Individuals in Time

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

Series Editors Werner Abraham
University of Vienna

Elly van Gelderen
Arizona State University

Advisory Editorial Board Cedric Boeckx
Harvard University

Ian Roberts
Cambridge University

Guglielmo Cinque
University of Venice

Ken Safir
Rutgers University, New Brunswick NJ

Günther Grewendorf
J.W. Goethe-University, Frankfurt

Lisa deMena Travis
McGill University

Liliane Haegeman
University of Lille, France

Sten Vikner
University of Aarhus

Hubert Haider
University of Salzburg

C. Jan-Wouter Zwart
University of Groningen

Christer Platzack
University of Lund

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

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

María J. Arche

John Benjamins Publishing Company
Amsterdam / Philadelphia

8

TM

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

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

2006042929

© 2006 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

..............2 The IL/SL Dichotomy as a Syntactic Distinction..1............................................. Kratzer (1988...........................................2.........1.......33 2...........53 3....1 .....................................51 3..1.....................15 2.............2.......................1 Proposals about the Individual-Level/Stage-Level Dichotomy ..... Raposo and Uriagereka (1995)........................... 1995) ...4 The IL/SL Contrast as a Categorical/Thetic Distinction.............1....................38 Chapter 3 Event Classes and Individual-Level Predicates ..............3 Summary of Section 2.......................... Higginbotham and Ramchand (1996) ............ When There Is More Than One Copular Verb: Spanish Ser and Estar.....2............3 Summary of Section 3.........12 2.......39 3..................................................1 The IL/SL Distinction as a Semantic Distinction...........61 3.............................8 2........................Table of Contents Acknowledgments...... A Brief Stop at “Agentivity”...................... The Structure of Copular Constructions ..1 Chapter 2 Individual-Level Predicates .....5 2................ Inner Aspect and Event Types .....................................39 3..........1.......................1 Distribution of Copular Verbs in Spanish ....2 .......................2...........59 3...................39 3........40 3....................................2.........................................................3....................2............................1...........................................................25 2..........................................4 Summary of the Chapter ...............3 Summary of Section 3........................................61 .................................2..1 A Cluster of Notions ..................................3 IL Predicates as Inherently Generic Predicates.................32 2.........2 Ser and Estar as an Aspectual Differentiation ......1............................................... Chierchia (1995) .14 2...........................................xi Chapter 1 Presentation of the Study .......................1...................... Introduction to the Tests Discerning the IL/SL Status.....2................................. Milsark (1974) and Carlson (1977)................................................2 .2 Agents in Event Structure ........................53 3................................16 2........ by Tim Stowell ............11 2.................5 2.5 Summary of Section 2.....................2 Event Types and Event Structure...........................1 ..5 2..........3 Aspectual Differences among Individual-Level Predicates .........................................................................1..................1 Inner Aspect..ix Foreword......................

.......3......................84 4................4..........................135 4......80 3..............91 4...................................2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor123 4........6.................133 4.............7........ The Cruel-type as a Small Clause Integrant.................................3 .............................................62 3.....86 4..............90 4.......................................................8..........................................96 4.1 Aspectual Tests on IL Predicate .........116 4..105 4.......1 The Two-Copulas Hypothesis ......2 Be as a Copula Shifting a State into an Activity ..108 4.........2 Differences between Activities and States ...117 4................1...........3......7 ....117 4.........3 Summary of Section 4....................76 3.5..........96 4.................................1 On the Interpretation of the Relational PP ........4...............85 4...................... Summary of the Chapter ..............83 4..............................69 3..................6 An Account Based on the Relational PP Complement ......................5...143 4..1 Prepositions as Aspect Encoders ....................4............5.........................................2 Summary of Section 4...................2 Summary of Section 4...................4 Summary of Section 3........98 4........................................2..........138 4............................................................................................3 The Relational PP with Other APs ........3 Adjectives Predicated of an Implicit Event Argument ........................4.......3 Relational Mental Properties......................72 3.........2 On the Optionality of the (Affected) Goal PP ..............................................1 Cruel-type Small Clauses Taken by Verbs Other Than the Copula .....................2 Syntactic Approaches ..........................................................................................2 Summary of Section 3............................................1................1 Lexicalist and Logico-Semantic Approaches ........7.4 The Issue: Aspectual Alternation in IL Copular Clauses ..4 Summary of Section 4.......4 .....................110 4......3..5.....68 3....6..........4.1 Similarities between Activities and States ........................................................................................ Summary of the Chapter .............1.....3.................. States and Activities: A Grammatically Relevant Difference?........106 4.........2 Adjectival Predicates Showing an Activity-like Behavior .......108 4.............................6...................2 ............2........................................................................4 Summary of Section 4..........95 4............ The Relational PP Complement.1 Previous Explanations of Nonstative Copular Clauses ...............................3 Some Confusing Arguments in Defense of a State/Activity Distinction...........vi Individuals in Time 3......................5 ...1 Lexical Semantic Classes of Adjectives ..... 69 3.................................3.....5 Justifying the Approach ........81 Chapter 4 Aspectual Alternations in Individual-Level Predicates .....143 .............107 4...............6......84 4......................................................................6 ........3 .........................137 4.....3 The DP Subject of Relational MPs ......................3...7...............................

....158 5........ 1996) .........................................................5 The Lifetime Reading in Compound Sentences .......... 1997) .3 Articulating the Account...................................................................... A Brief Summary of Aspect Notions .....1 Permanent IL Predicates ............................................3. Aspect as an Ordering Predicate ........................................... 176 5................2 Nonpermanent IL Predicates.........................2 Perfective Homogeneous Predications ....................................205 6.....3................................2 Aspect as a Quantifier over Occasions .....2....... Summary of the Chapter .......................2...........................215 6...............................................................199 6......220 6..197 6..................4 Summary of Section 6....148 5.......205 6............................................2 The Internal Argument of Tense: The External Argument of Aspect...............4....................................................... 186 5.... Proportion...........3 The Progressive and Adjectival IL Predicates ...... 192 Chapter 6 Tense and Individual-Level Predicates ...............................................2 ....157 5................. 179 5...........2 Differentiating Temporal Extensions of IL Predicates ........................205 6...213 6..........................4.................152 5.....................................................................1 Tense and Aspect as Ordering Predicates .......... 174 5..........1 Habitual Heterogeneous Predications...............1 Temporal Interpretation as a Consequence of Argument Structure....2 The Perfective and Adjectival IL Predicates ...............2 The Habitual Interpretation: Iteration......5....3 The Content of the TT in the Arising of Lifetime Readings .1 Quantifying over Occasions .........1 The Imperfect and Adjectival IL Predicates ......... 180 5...............1 Tense.................6............................... Stowell (1993.147 5..........................2..............3 Inner and Outer Aspect .................................2.........................2..2.. 1995) ..............................3..........................4................Table of Contents vii Chapter 5 Outer Aspect and Individual-Level Predicates .....................212 6............4 ......1...........................................199 6.................3.....4 Summary of Section 6....1.............210 6.................................3 On the Relation between the TT and the Habitual Q<occ> ....... Kratzer (1988....194 6...........................1 When the Subject Is a QDP ............................3 Brief Notes about the Complement of Mental Properties APs ................................. 187 5...........................................5 Adjectival Individual-Level Predicates and Viewpoint Aspect ........................................................1 Necessary Conditions for Lifetime Effects .........................2.....5......................163 5....................................3 The Arising of Lifetime Effects ....3................... 177 5.......................................202 6........................173 5..............................4..................2 Context Associated to Individuals ..148 5...........5................................193 6...................................208 6........218 6.....4 The Determination of the TT Content and Lifetime Effects ....................222 ............... Musan (1995...........171 5...................... and Systematicity ..........2 Introducing the Determining Role of Contextual Factors......4....

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

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

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

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. and shows that it is ultimately . and the status of the subject is confined to that of a participant in the reported event or situation. but no clear consensus has emerged on how it should be formalized. 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. on the other hand. and a third type (possibly the most numerous of the three) can have either type of interpretation. The SL/IL distinction is grounded in a core semantic intuition that IL predicates involve essential. This intuition is often associated with a related intuition about the discourse function of sentences whose main predicates belong to either type. At first glance. Sentences containing IL predicates have been described as categorical. adjectives are a diverse crowd. some are (usually) SL. or even immutable properties. Noun phrases are (almost) always IL. whereas verb phrases and prepositional phrases are (almost) always SL.Foreword This book provides a compelling account of the distinction between stage-level (SL) and individual-level (IL) predicates. others are (usually) IL. combining adjectives with various complements and superordinate aspectual and adverbial categories. 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. Adjectival predicates have constituted the central domain of study in the literature. they serve to report an event or situation. permanent. 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. but Arche systematically examines the evidence that seems to argue in favor of a lexical distinction between IL and SL adjectives. have been described as thetic. In this study. whereas SL predicates involve inessential or transient properties. In contrast. the subject of the IL predicate is the topic of the sentence. The SL/IL distinction has been the focus of an important body of work in the literature on formal semantic and syntactic theory over the past 30 years. with greater or lesser degrees of naturalness. depending on the surrounding context. because they alone among syntactic categories exhibit a robust internal division defined by the SL/IL divide. this claim is contradicted by many direct empirical observations. 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.

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

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

.

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

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

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

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

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. IL and SL. indicating some of the points on which my proposal differs. I concentrate on the most influential proposals about the contrast between individual-level (IL) and stagelevel (SL) predicates. syntactic. the one offered by Stowell (1978. since I am chiefly concerned with copular cases.1 Proposals about the Individual-Level/Stage-Level Dichotomy 2.1. *There are several policemen insane . specifically discussing arguments for semantic.4 summarizes the chapter. according to most authors. There are several policemen in the corner a. arguing for an aspectual distinction. In section 2. but those in (2) are excluded. the presence of functional projections) and its relationship with the IL/SL distinction. NP). Section 2. and introduce some of the concrete aspects for which this work makes alternative proposals. I also introduce proposals by Heycock (1994) and Becker (2000). I will introduce the syntactic analysis that I will assume for copular sentences. 1981). Likewise. For example. a part of this chapter presents the differences between the two Spanish copulas (ser and estar).1 The IL/SL Distinction as a Semantic Distinction. which.3. I present the syntactic analysis I assume for copular cases—namely. I present other two aspects basic for my purposes in this work. Since the data in the work are from Spanish. Section 2.2 is devoted to the differences between the two Spanish copulas. the predicates of (1) are allowed. In section 2. lexically distinguish the two types of predication. After that. There are several policemen available b. and pragmatic distinctions.1. Milsark (1974) and Carlson (1977). ser and estar. according to which the copular verb selects for a small clause (SC) containing the nominal argument and the predicate (AP. (1) (2) a. *There are several policemen intelligent b. who discuss the constituency of the SCs that copular verbs take (in particular. PP. This chapter is organized as follows.Chapter 2 Individual-Level Predicates In this chapter I introduce the concept of the individual-level predicate. I will first present the main descriptions of this notion found in the literature. 2. I will critically revise those analyses.

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

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

as Diesing 1992 proposes. from Davidson 1967. I will not discuss Kratzer’s perspective about where the NP subject should be assumed to generate: either in a specifier of the VP (along the lines of the VP internal subject hypothesis argued by Kitagawa 1986 and Koopman & Sportiche 1991) or in the specifier if the IP.3 In Kratzer’s proposal.3. denoting nonactions but performing actions and proposed that events could be treated as individuals. (∃e) (buttered (Jones. This gives rise to a difference in their argument structure which is the stem of the difference between the two types of predicates. e) & (with a knife. as in (i). with a knife. Whereas IL predicates are predicated of the individual directly. The Davidsonian argument is the external argument of the predicate. the toast. NP). (i) Jones buttered the toast slowly. the event variable is existentially quantified). 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. The trees in (13) just aim to capture Kratzer’s idea on the location of the external argument in the VP predicate. PP. 1995) In the spirit of Carlson (1977). and nouns.2 The eventive variable is considered a primitive element in the logical semantics of events. which.8 Individuals in Time This view has been adopted by some authors. SL predicates are predicated of something else: the eventive variable proposed by Davidson (1967). e)).1. to mention just a few. 1995) claims that IL and SL predicates differ with respect to what they are predicated of. 1995) and Diesing (1992) but disputed by others. Kratzer’s proposal looks like (13). in the bathroom. based on Williams’s (1981) Argument Linking description.2 The IL/SL Dichotomy as a Syntactic Distinction. Kratzer (1988.4 2 Davidson (1967) took the classical differentiation from Plato between verbs. IL predicates do not. e) & (slowly. 2. whereas SL predicates project this extra argument (which can be understood as a spatiotemporal argument). Kratzer (1988. at midnight. When it is not projected. e) & (at midnight. 3 4 I will add a few more words about the eventive variable in chapter 4. it is the DP referring to the individual itself which functions as external argument. can be modified and quantified over by adjuncts. (In the default case. . denoting actions. As will be shown in section 2. like nominal variables. such as Kratzer (1988. e) & (in the bathroom. He proposed that action sentences include an event variable as a primitive element in their logical semantics. such as Jäger (1999) and Becker (2000). Schematically.

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

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

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

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

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

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

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

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.” as cited previously (see section 2.” The copula ser is used to refer to properties that are considered essential of the individual. the copula used can be estar.12 10 Note that these descriptions fall in line with the ones given by Milsark (1974) for “properties” and “states. In the case of Spanish. How can the distribution of the two copular verbs be described? As traditionally observed. or ser. Este animal es/*está that animal ser/estar-PRES-3SG a goat “That animal is a goat” professor c. if not a mistake. as . 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. insufficient to characterize the copular alternation. if the subject is a physical entity (43).1. However. when the predicate is an NP. if. Pedro es/*está Pedro ser/estar-PRES-3SG teacher “Pedro is a teacher” When the predicate is a PP. the copular verb must be ser (42).10 Essential properties are understood as permanent.11 which. un ordenador de 1960 (42) a. Ese artefacto es/*está that artifact ser/estar-PRES-3SG a computer from 1960 “That artifact is a computer from 1960” una cabra b. despite being ungrammatical with ser.1). obviously designates a nontemporary property.1. I would like to note Demonte’s (1979) observation that this distinction poses an asymmetry in the account of copula distribution. Roldán 1974) describe the distribution of the two copulas ser and estar on the basis of the Aristotelian dichotomy “essence” versus “accident. other authors observed that the distinction between permanent and temporary was. discussed in section 2. 2.2. 11 See also comments by Milsark (1974) and Carlson (1977) about the same predicate. The classic counterexample in this respect is estar muerto ‘be dead’. there is no general agreement in the literature on how the copula alternation should be characterized. and accidental properties as temporary.1 Distribution of Copular Verbs in Spanish Some traditional grammarians (Bello 1847.1. 12 Although I will not work out this difference here. if the subject denotes an event (44).

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

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

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

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

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

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. it is not the case that any perception verb allows for any type of adjectives as its complement. which. short. it rejects closed-scale ones. correlates with an “absolute” (i. The following sentences where they are used in comparatives and are modified by proportional degree modifiers such as parcialmente ‘partially’. whereas it allows open-scale adjectives.22 Individuals in Time Other perception verbs.. context dependent) standard of comparison: (v) partially/half/completely {empty. One existing difference between the adjectives of (68) and (69) is that only the latter can be modified by muy ‘very’: Noté a Juan muy {alterado/preocupado/cansado} noted Juan very {agitated/worried/tired} (ii) *Noté a Juan muy {desnudo/descalzo} noted Juan very {naked/barefoot} (i) However. then. correspondingly. Notar seems sensitive to other properties of the predicates appearing in the small clause selected by it.e. which. interesting. correlates with a “relative” (i. For example.e. full. although cansado ‘tired’ and preocupado ‘worried’ seem to behave as proper closed-scale adjectives (cf. This and other issues concerning the semantics of the predicates allowed by perception verbs need further examination. adjectives compatible with proportional degree modifiers are those gradable predicates with a closed structure scale (with minimal/maximal elements). (vii) and (viii)). verbs such as percibir ‘perceive’ or advertir . they argue. In turn. completamente ‘completely’.18 such as ver ‘see’ seem to discriminate IL versus SL at first sight. non-context-dependent) standard of comparison. open. this does not mean that predicates such as desnudo ‘naked’ or descalzo ‘barefoot’ are not gradable. gradable adjectives incompatible with such modifiers involve an open scale structure (without minimal/maximal elements).) notes. (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. 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). However. closed} (vi) ??partially/half/completely {long.c. alterado ‘agitated’ sounds quite natural when combined with a proportional degree modifier.. inexpensive} We could.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.

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

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

they must bear the feature “–[–perfective]. there are many other factors intervening in the semantics of these constructions. as contento ‘glad’. She establishes a parallelism between estar and predicates like write a letter. Although. When the predicates are used to refer to imperfective states—that is. basically in the sense of ‘durative’ versus ‘punctual’. 2. the work is from 1945. Examples (75) and (76) are Luján’s (1981:177). whose beginning and end are assumed (75). How can we explain those cases where an adjective can combine with both copulas? Luján argues that. 2.” As I understand Luján’s proposal. Thus. In turn.2.2. ser expresses that a predicate applies to an individual during a stretch of time with no beginning or end assumed (76). states where no beginning or end is assumed—they select ser. Ser is conceived as a predicate parallel to write or admire. is already found in Gili Gaya 1961. When they refer to perfective states.1 Luján (1981). Following Querido (1976). but can be either [+perfective] or [–perfective] (Luján 1981:175). in general terms. I will discuss work by Luján (1981).2 Ser and Estar as an Aspectual Differentiation The idea that the distinction between the two copulas ser and estar should be described in aspectual terms rather than as a permanent/transitory dichotomy.2. where a beginning or end is assumed. and the [±perfective] feature in the copula. Among those who have worked along these lines. a delimited process. (75) Perfective predicate A(x) at time tj 23 The cited date corresponds to the ninth edition. it is not that the copula ser is atemporal. both undelimited predicates. Schmitt (1992). her account of the distribution of copular verbs (concretely in combination with adjectives) relies on two factors: the [±perfective] feature in the adjective. they select estar. if they combine with both. According to Luján. or at least one of them is. Luján (1981) claims that all adjectives are stative.Individual-Level Predicates 25 In sum. .23 This author argued that the property distinguishing ser and estar is imperfective versus perfective. 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). but rather that it makes a different contribution to temporal reference. the two copulas differ in the nature of the temporal reference they make. the results emerging from them are consistent with the distribution of the copulas in Spanish. Estar expresses that a predicate is true of an individual for a delimited time period.

. PPs (when the subject is a mobile object) Spanish NPs. PPs (when the subject is an event (party) or an immobile object) APs (tired. scenarios incompatible with a result state are excluded for estar.1]. since there is no reason to consider in the garden as a delimited predicate. female. PPs (when the subject is a mobile and immobile object) Estar + Table 2. It is not a state. which suggests. Estar corresponds to a result state of an accomplishment verb. (More on this in chapter 7.2. APs (human. States are eventualities with no duration or endpoint (be tall). Portuguese and Spanish copular verbs The concrete claim of Schmitt (1992) is that estar involves aspectual properties. PPs (when the subject is an event (party)) APs (tired. its wider flexibility in combining with adjectives indicates that it does not involve internal temporal characterization. and processes or activities are those events with duration but without an inherent culminating point (walk. I take it that her conclusions about the topic can be applied to Spanish. whereas ser is underspecified with respect to aspect. 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. Since estar refers to a result state. In the spirit of Luján. round. APs (human.1. Schmitt (1992) argues that the distinction between ser and estar is aspectual in nature. round. Among the facts that Schmitt cites to demonstrate the internal structure of estar and the underspecification of ser are those enumerated in (a)–(d). its underspecification in aspect. (Schmitt works on Portuguese.2). female. happy).26 Individuals in Time (76) Imperfective predicate A(x) at times tj… tj+k However. That is. old). since the distribution of Portuguese copulas nearly parallels those of Spanish [see Table 2.24 In a nutshell.) Copula Ser + Portuguese NPs. However. ser manifests a wider flexibility. her reasoning is as follows. swim). 24 All these concepts will be adequately presented in chapter 3 (see section 3. ser has no inherent temporal structure. Accomplishments are those predicates that refer to a process with a inherent endpoint (write a letter or build a house).1. for this author. nor is it an event or a process. happy). old). In contrast. the description of SL predicates as perfective (in the sense of “delimited”) does not capture the intuition about the whole class of SL predicates.2 Schmitt (1992).) 2. but.2. Predicates like be in the garden are not covered by Luján’s generalization.

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

1. However. they may just show that estar predicates behave as states. Chapter 4 develops a proposal about the strict correlation between the type of APs and this syntactic behavior. (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. is not allowed as a perception verb complement. 1995). such as when(ever)-clauses: . Schmitt also notes that estar. the reason why estar and the perfect are excluded as complements of perception verbs may be because both behave as states. (c) Inability of estar to appear as a perception verb complement In relation to its inability to appear in the progressive. In other words. the fact (also noted by Schmitt) that not all adjectival predicates combine with the progressive (cf. (See section 2. (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.2.28 Individuals in Time aspectual head. 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. as I already suggested. and (infinitive) states are not accepted as perception verb complements.” (I will introduce these notions properly in chapter 3).) (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. it is not clear that tests along the lines of (82) and (83) show that estar designates a result state. like all verbs in the perfect. (81)) remains unaccounted for. 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.

that is. 25 . whenever-clauses are allowed only in the presence of an eventive argument.2. she proposes that when(ever)-clauses do not select for SL-hood but for “certain aspectual properties. from Kratzer’s perspective. since the copula ser perfectly fits in such contexts. I do not share this judgment in Spanish. That when(ever)-clauses select for properties that can be iterated in time seems clear. and find (85) and (86) grammatical. however. That is. Recall that. yo le atendí whenever he come-PAST-PERF-3SG to see-me I him attended “Whenever he came to see me.Individual-Level Predicates 29 cruel.” Specifically. predicates in perfect aspect are excluded in when(ever)-clauses. no matter what type of predicate is involved. as Schmitt seems to intimate. ha estado allí para escucharme “Whenever I have needed him. Furthermore. only with SL predicates. among others). I also agree that this does not have to be equated with SL-hood. As the glosses indicate. I would not say that only some adjectival predicates are allowed in these contexts. the situational variable linked to SL predicates is just one of the possible variables susceptible of “referential variability. but. Consider first the contrast in (87) and (88). se arrepiente después (84) Siempre que Juan es whenever Juan ser-PRES-3SG cruel he-regrets right-after “Whenever Juan is cruel. rather. On my view. First. if ser is the copula designating IL-hood. See section 2.25 Schmitt offers an explanation for the acceptability of cases like (84) based on two points. I attended him” (86) Siempre que lo he necesitado.1. which does not mean that it selects for predicates denoting properties of stages. when(ever) can refer to predicates that can be iterated in time.” which is the crucial factor licensing distributivity (Beghelli 1995. he has been there to listen to me” Schmitt claims that when(ever) selects for predicates that can be distributed over time. he regrets it right after” In effect. the fact that it is allowed in when(ever)-contexts is completely unexpected under Kratzer’s hypothesis. that different adjectival predicates need different conditions to license their appearance in when(ever)clauses.” According to her judgment. Beghelli & Stowell 1996. Szabolcsi 1996. Some specific examples will help explicate this point. (85) Siempre que vino a verme. (85) is a simple past-perfective form and (86) a compound perfect form. she argues that when(ever)-clauses select for “some sort of durative aspect.

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

which are conveyed by a type of adjectives in a specific way. ser lacks any inherent temporality and is described as “inert with respect to aspect. Ser predicates are considered parallel to write or admire. in her view. Luján (1981) and Schmitt (1992) defend the idea that the distinction between ser and estar should be understood in aspectual terms. by “inert with respect to aspect. Another author arguing in similar terms as Luján (1981) and Schmitt (1992) is Fernández Leborans (1999). and estar as a predicate that is true of an individual for a delimited period of time. Whereas estar possesses internal temporal structure. I argue that the availability of the progressive is related to the internal temporal properties of the construction. For this reason. 2. Although Luján refers to this differentiation as an imperfective/perfective contrast. 2.2. which do involve internal temporal structure (estar in her own proposal and many others). In her study on copulative clauses of Spanish.” she means that serpredicates are not tied to temporal limits—the reason why she argues such predicates can be described as “durative. However. Fernández Leborans considers that in contexts such as the progressive (91) or temporal forms suggesting that the property does not hold any longer (92).2. I show that the perfective preterit can be licensed under the appropriate contextual conditions. I will not have to propose that when the copula ser combines with them it becomes an SL predicate. I argue that the oddity (or not) of such aspectual forms is not related to the IL/SL distinction.2. Likewise. as she explicitly points out (Luján 1981:206). she argues that the ser/estar distinction can be described as an IL/SL dichotomy. ser predicates work as SL predicates. Luján conceives of ser as a predicate that applies to an individual during a stretch of time with no beginning or end assumed. where no . which. just a partial answer.2.” As I understand her proposal. (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. alludes to the lexical content of the predicates.2.3 Fernández Leborans (1999).Individual-Level Predicates 31 parallel to when(ever)-cases. In sum.” As chapter 7 discusses in greater detail. are also perfectly grammatical in such scenarios.2.4 Summary of Section 2. is founded on aspectual properties. to say that ser is allowed because it does not involve internal temporal structure is. since other predicates. I consider any appearance of ser to denote an instance of an IL predicate. at best. the opposition.

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

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

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

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

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

1995) in that SL predicates do not involve the projection of just an event argument (here an Event Phrase). Thus. Because there is not a thematic subject. Felser concludes. Pollock 1989). but the projection of two nodes. the contribution of this book will be twofold. (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. In the first place. Heycock (1994). I argue that IL-hood 29 See also Guéron and Hoekstra 1995.1. However. Whereas SL predicates are linked to aspectual content. . for Becker. As can be appreciated from (105). assuming that to occurs in the T head (Emonds 1976. argued for the existence of two homophonous verbs be. unlike Schmitt. in particular to copular sentences. as pointed out in section 2. see (105). Becker’s (2000) idea differs with Kratzer’s (1988. Felser proposes the extra layer to be an Aspect projection. perception verbs take Aspect Phrases. Event Phrase and Aspect Phrase. Chomsky 1986. differing in this respect from Carlson (1977). other authors have argued for the existence of functional projections inside them too. the difference between SL and IL copular constructions resides in the complement of the copula. Becker establishes a correspondence between the presence of Aspect and SL-hood.Individual-Level Predicates 37 (108) I wouldn’t like to see [there be so many mistakes] Her argumentation goes as follows.1. Regarding the internal constitution of the SCs taken by the copular verb. The reasons to discard a CP or TP projections come from the ungrammaticality of sentences like (109) and (110). Summarizing. and Becker (2000)29 have put forth the presence of a functional node (Aspect). whereas Stowell (1981) originally conceived SCs as lexical projections. In particular. Thus. the exclusion of a to-infinitive. Furthermore. suggests that the perception-verb complement is not a TP either. such an aspectual content is absent in IL predicates. The proposal defended by Becker (2000) shares the spirit of Schmitt’s (1992). who. since both establish a relationship between copular distinction and aspectual content. which are presented as codependent. it cannot be said to have been generated inside the VP but inserted at a higher level within the perception-verb complement. 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. authors such as Schmitt (1992). She just needs to assume one copular verb. Becker considers that an aspectual distinction does not exclude in and of itself a split conceived as IL versus SL. Likewise.

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

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

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

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

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

made clear that internal arguments and certain types of locative phrases affect the category a predicate can be classified into before entering into syntax. others test out not aspectual properties strictly speaking. ??Juan está dándose cuenta de que su madre tiene razón Juan is realizing that his mother is right c. Ryle (1949). the possible or impossible occurrence in the progressive form. and Dowty (1979). I will be adding some discussion about them. I then proceed with the tests that differentiate among different types of events.Event Classes and Individual-Level Predicates 43 affected by other elements present in the sentence (cf. The test in (9). Some distinguish between states versus nonstates. 1994).) 3. Identifying which type an eventuality behaves like may be helpful just to pinpoint the determining inner features of a predicate. That is. Finally. has been taken (since Lakoff 1966. 1970) as a proof distinguishing statives versus nonstatives. (3)). (The judgments I give are based on native intuitions from Spanish. I start out by discussing some tests that distinguish statives versus nonstatives in (9)–(11). Next. (9) Occurrence in the progressive form a. 1991). rather than a lexical one. put it.1. *Juan está siendo alto Juan is being tall b. Others make visible the features that two or more event types have in common. Verkuyl (1972.2. As I present them. I present a summary of the tests to diagnose event-type membership from Kenny (1963). like agency. Tenny (1987. I will follow this second option throughout this work. aspect is a crucially compositional issue.1 Events versus States. 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. 1993). and Ritter and Rosen (1996). Juan está paseando Juan is walking d. Each test establishes a distinction of a different kind. 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. Pustejovsky (1988. as Pustejovsky (1988). among others. (2) vs. Put in other words. Only the latter can appear in such a form. proposes. among others. “event templates” do not have a grammatical reality.. for example. 1989. but properties that seem to depend on aspectual ones. as Verkuyl (1993) and Borer (2005). Juan está trazando un círculo Juan is drawing a circle . The other (more radical) option is to think that there are no such event type primitives. Vendler (1967).

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

(11) Interpretation as habitual in present tense a.2. sentences like Juan is usually sick or Juan is usually at his best friend’s place. it might seem that this test is a diagnostic for IL-hood rather than for stativity. I start out with the tests in (13) and (14). 3.1. However. The eventualities whose interpretation in the present is not habitual are states. Juan realizes that his mother is right c. according to the temporal adverbials that refer to the time an event lasts they admit. A durative adverbial. Normalmente. whereas Juan walks does. nonstatives are interpreted as habitual. Bertinetto 1986. It seems clear that Juan is sick or Juan is at his best friend’s place does not have the habitual reading as the only (and preferred) reading. Normalmente. Normalmente. . I show that this is a possibility for every type of predicate (including IL predicates) if a suitable context is created. it is worthwhile to notice that when no (habitual) adverb is present. are completely grammatical. I argue that the crucial condition for a predicate to appear in habitual form is its possibility of restarting. Juan pasea Usually.2).Event Classes and Individual-Level Predicates 45 habitual (Comrie 1976. Juan draws a circle Nevertheless. whereas (SL) statives can be said to be ambiguous. such as “for + x time” goes well with those predicates that refer to eventualities that extend over time with no involvement of an 3 I make a few more remarks on this in chapter 5 (section 5. Juan traza un círculo Usually. is excluded in combination with a habitual adverb (which proves it cannot be interpreted as habitual). Cinque 1999). *Normalmente. which divide activities and states. Juan knows mathematics b. Such compatibility cuts the pie of eventualities in two: those that have a habitual interpretation. and those that do not.3.2 Activities and States versus Achievements and Accomplishments.4 Whereas (11a). 4 I thank Tim Stowell for interesting discussion on this topic. on the other. on the one hand. with stative though SL predicates. involving an IL stative predicate. 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.” as their reaction against combination with habitual adverbs proves. Juan walks d. Juan se da cuenta de que su madre tiene razón Usually. Juan sabe matemáticas Usually. and achievements and accomplishments.

see Piñón 1999.5 (12) For + x time a. which is the reason why they are compatible just with eventualities entailing a delimit point. 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. in-adverbials refer to the period of time it takes to complete a certain event. Pablo estuvo enfermo durante tres semanas Pablo was ill for three weeks b. *Pablo construyó una casa durante tres semanas 6 Pablo built a house for three weeks d. 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. IL predicates can also combine with for-adverbials: (i) Juan fue camarero durante tres semanas Juan was a waiter for three weeks 6 Not all accomplishments are excluded with a for-adverbial: (i) The police jailed John for four weeks The for-adverbial refers to the period of time that the result state from jailing lasts. *Pablo estuvo enfermo en tres semanas Pablo was ill in three weeks b. As the ungrammaticality of (14) shows. (13) In + x time a. Pablo construyó una casa en tres semanas Pablo built a house in three weeks d. For a recent account about the different readings available in such sentences. Pablo viajó durante tres semanas Pablo traveled for three weeks c. . *Pablo se dio cuenta de que su madre tenía razón durante tres semanas Pablo realized that his mother was right for three weeks In turn.46 Individuals in Time ending point. the entailment of a delimit 5 Note that this test does not distinguish between IL and SL predicates. *Pablo viajó en tres semanas Pablo traveled in three weeks c. as (13) shows.

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

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

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

states and achievements. as van Voorst (1988) and others point out (see also section 3. In the other. I want to introduce another group detecting the presence of an agentive subject (Ryle 1949. However. *¡Estate enfermo! Be sick! b.2 gives a more precise notion of “agent. 3.2. ¡Construye la casa! Build the house! c.”) The roles that DPs play in the event and the analysis of the event itself usually overlap. an accomplishment. the subject did not get to walk or realize. which are grammatical. ??¡Date cuenta de que estás confundido! Realize that you are wrong! . ¡Pasea! Take a walk! d. (21) Interpretative entailments from the adverb almost a. the tests in (22)–(25) can be used to diagnose stativity. can have two interpretations.5 in this regard).50 Individuals in Time The possible interpretations of almost put in contrast accomplishments with other event types. In one of them. To conclude this survey of aspectual tests. too. (22) Occurrence in command imperative form a. the subject was involved in the process of building up a house but did not finish it. To the extent that an agent is conceivable only if a proper event (a nonstate) is at hand. an activity. Pablo casi caminaba con ocho meses Pablo almost walked when he was eight months c. the subject did not even start out the process of building. As can be appreciated.1. in (22)–(25). get differentiated from activities and accomplishments. (Section 3. Pablo casi se dio cuenta de que su madre tenía razón Pablo almost realized that his mother was right Sentence (21a). can only have the second interpretation.5 Agentivity Tests. (21b). an achievement. 1970). that is. Lakoff 1966. Pablo casi construyó la casa Pablo almost built the house b. yielding ungrammatical results. and (21c).

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

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

although volition usually entails control. (26) (27) El portazo rompió el cristal The bang broke the glass Juan rompió el cristal Juan broke the glass . but just holds.Event Classes and Individual-Level Predicates 53 properties. 3. not all controllers involve volition. which diagnose agency. it is important to distinguish the semantic features it contains and to establish whether there is some feature more basic than the other. I consider two aspects related to this.” Then. stuff that is not performed. I present a brief discussion about the agency tests above and about the notion of “agent” that I will observe in this work. In this section. Further discussion about the tests and about some concrete event types will come in section 3. Vendler (1967) himself considered states and achievements a natural class. may be used to discern properties of eventualities as well.2. To the extent that just some types of events can be undertaken and commanded.” and therefore “commanded. when I discuss the aspectual behavior of copular IL predicates. agency tests work as event-type tests. while volitionality and control are properties that just animate causers can embrace. As will become clear. three notions involved: causation. the term “agent” clearly denotes ‘the argument that brings about an event’. which gives bad results in contexts with adverbials such a deliberately. agency tests separate states and achievements from activities and accomplishments.” Also. volitionality. at least. In fact. there are. The results of (17). and (22)–(25) also suggest that states and achievements have characteristics in common. cannot hold with any willfulness on the subject’s behalf. As mentioned before. In the following section. (20). To begin this investigation.” The cleft with “do” and the test of command imperative detect eventualities that can be “done. 3.1 A Cluster of Notions Although. in principle. activities and accomplishments may be the most persistent co-patterning of the table. and control. Actually. I will deal with the different notions that get blended in the term “agent. consider the classical minimal pair in (26) and (27).2 A Brief Stop at “Agentivity” The set of tests in (22)–(25). causation is the most basic notion involved in agency contexts. In particular. I will briefly introduce the relation between agents and event structure. Likewise. agency tests detect eventualities that denote stuff that can be “brought about. First. In agency.3.

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

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

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

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

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

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

In this respect. With activities and achievements. where there is no agency or responsibility over the event involved. 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. when he saw he did not trust him A crucial difference between states and activities and accomplishments is that. *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. the beginning is understood as the moment at which the subject starts undertaking the process. *Juan empezó a estar en Madrid cuando consiguió trabajo Juan started being in Madrid when he got a job b. but they contrast with activities and accomplishments. Some stative predicates yield sentences more acceptable than others: (41) a. however. there is an interesting difference with parar de. Empezar can have both [+animate] as well as [–animate] subjects. whose subjects can only be [+animate] and agentive. 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. 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.60 Individuals in Time (39) a. too. States (40a) are not completely excluded. ??Juan empezó a estar enfermo con 50 años Juan started being sick in his fifties b. . Only in dynamic events is the subject is responsible for the beginning of the event. Juan/*la rueda paró de girar Juan/the wheel stopped rotating The interpretation of the subjects of activities and achievements in combination with start is interesting. Juan/*la rueda paró de moverse Juan/the wheel stopped moving b. both of which give completely grammatical results. Juan empezó a estar a disgusto con su jefe cuando vio que desconfiaba de él Juan started being uncomfortable with his boss.

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

*Lo que sucedió fue que Juan fue esquimal What happened was that Juan was an Eskimo b. states are eventualities that do not “happen. In (43)–(45) I reproduce the tests marking for stativity versus eventivity. taken to belong to the group of states.” do not “take place.3. (43) Occurrence in the progressive form a.” but just “hold.2.1 Aspectual Tests on IL Predicates The aim of this subsection is. Juan es esquimal Usually Juan is Eskimo b. in general. *Normalmente. *Juan estaba siendo rubio Juan was being blond c.1 Events versus States. To do that.1. Normalmente. 3.62 Individuals in Time 3. *Juan estaba siendo esquimal Juan was being an Eskimo b. I will make use of the tests introduced above in the same order. *Normalmente.” They are considered to be eventualities that lack any bounding and are inherently durative. Juan es rubio Usually Juan is blond c. as has been widely agreed on in the literature. Juan estaba siendo muy cruel con Pedro Juan was being very cruel with Pedro (44) Pseudocleft with happen (‘take place’) a. As described above. simply. IL predicates are. ?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. The predicates I will test are a classificatory AP (Eskimo) and two qualifying APs (blond and cruel). The eventualities yielding ungrammatical results are taken as stative.1 regarding the present-tense interpretation as habitual (see also chapter 5 for more important . to identify the aspectual properties of IL predicates.3. I will concentrate on copular clauses and test out whether all IL predicates are appropriately grouped in the aspectual set of states. *Lo que sucedió fue que Juan era rubio What happened was that Juan was blond c. Juan es cruel con Pedro Usually Juan is cruel with Pedro Despite all the observations introduced in section 3.

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

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. the combination of ser and an AP such as cruel ‘cruel’ looks good. Juan ha sido cruel con Pedro Juan has been cruel to Pedro 3. only activities. and not accomplishments. *Juan fue esquimal en dos horas Juan was an Eskimo in two hours b.3. *Juan fue esquimal durante tres semanas Juan was an Eskimo for three weeks b. As mentioned in section 3.1. where there is no endpoint. such a predicate shows a nonstate patterning. (48b) is a licit entailment from (48a). As shown in section 3.1. Juan está siendo cruel con Pedro Juan is being cruel to Pedro b.2. which argues in favor of considering the predicate (ser cruel) as an activity. In principle. a predicate like ser + AP would be unexpected as a complement of parar de ‘stop’. yield a perfect entailment from the progressive form. However.2. since this is only possible with eventive predicates. (48) Perfect entailment from the progressive form a.1.3 Activities and Accomplishments versus States and Achievements. tests like (49) divide activities and accomplishments from states and achievements.64 Individuals in Time (46) For + x time 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). Juan fue rubio durante tres años Juan was blond for three years c. *Juan fue rubio en dos horas Juan was blond in two hours c.3. . Once again.2.

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

5 Agentivity Tests. Juan no empezó a ser cruel con su adversario Juan did not start being cruel to his opponent c. *Juan casi fue esquimal Juan almost was an Eskimo 3.66 Individuals in Time 3.1. which also test accomplishments versus all the other eventualities. ?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. Juan finished being an Eskimo b. as mentioned in section 3. the latter is with activities. The former reading is available with accomplishments. confirm be cruel as an activity. Only ser cruel is possible under a command imperative form (54) or in combination with adverbs like deliberately (55). None of the AP predicates with ser. Tests (52) and (53) show some more contrasts consistent with the results from (49). 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. *Juan terminó de ser rubio Juan finished being blond c. Although be cruel predicates do not give perfect results after finish. (52) As a complement of finish a. they clearly contrast with the other two. *Juan terminó de ser esquimal.2. 20 Recall that. ?Juan casi fue cruel con su adversario Juan almost was cruel to his opponent b. and (b) that the subject did not even start out the process. except for cruel. like (53c). since the unique licit entailment from (53a) is (53b).4 Accomplishments versus All the Rest. which suggests that dynamic properties are somehow involved in these predicates.20 Although (53a) does not sound extremely natural. The next set of contrasts aim at testing whether the process that be cruel can be considered is agentive or not.3.1. whose combination with almost is impossible.1. . All tests give the same results. fit in canonical agentive contexts.3. (53) Interpretative entailments from the adverb almost a. it contrasts with other APs with ser.4.

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

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

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

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

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

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

Juan está trazando un círculo Juan is drawing a circle d. Verkuyl 1993. and. van Voorst 1988. *Pablo paró de amar a María Pablo stopped loving María b. Following Pustejovsky (1988). these contexts separate states together with one type of events—namely. As a number of authors have pointed out (Smith 1991. achievements (76) can get . ?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. this can be legitimately taken as evidence for classifying them separately. *Pablo paró de darse cuenta de que su madre tenía razón Pablo stopped realizing that his mother was right (75) Pseudocleft with do a. (73) Occurrence in the progressive form a. Pablo paró de pasear Pablo stopped walking d. However. Juan está paseando Juan is walking c. 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. achievements. due to the subprocess that can be adduced to be involved in them. Pablo paró de construir la casa Pablo stopped building the house c. Lo que hizo fue pasear What he did was walk d.Event Classes and Individual-Level Predicates 73 contexts. *Juan está siendo alto Juan is being tall b. in my opinion. which differentiates them from purely dynamic events: activities and accomplishments. I assumed that the reason for such similar patterning is that achievements involve a (resultant) state in their structure. Lo que hizo fue construir la casa What he did was build the house c. *Lo que hizo fue estar enfermo What he did was be sick b.

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

the speaker expresses a guess about a particular activity she thinks the subject. (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. 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’ . Roughly described. the different interpretation of predicates with modal verbs diagnoses agency versus lack of agency. Martha. however.26 I have stopped knowing her/getting to know her I have nothing to say. epistemic and deontic. the speaker makes a guess about Martha’s state. On the latter the speaker expresses a command that Martha has to effectuate. The same can be said for accomplishments (80) and achievements (81). the oddness of (xiv)) or an inchoative one (xv). however. However. is usually involved in. the simple present form being ungrammatical. (xiv) *This table is missing a leg more and more (xv) *This table is getting to miss a leg 26 On my view. the modal has just an epistemic reading. with an activity such as walk around the park. rather than stativity versus dynamicity. (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. On the former. about examples from English such as those pointed out by van Voorst (1988).Event Classes and Individual-Level Predicates 75 (76) Pablo se estaba dando cuenta de que Pedro tenía razón Pablo was realizing that Pedro was right (77) Pablo was about to realize that Pedro was right Consider now two more arguments in favor of distinguishing states and activities. the modal has two meanings— namely. Piñón (1995) notes that modal verbs have different interpretations with dynamic and stative predicates. The examples below show where stative verbs have to appear in the progressive form in English.

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

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

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

Event Classes and Individual-Level Predicates

79

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

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

80

Individuals in Time

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

Event Classes and Individual-Level Predicates

81

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

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

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

2 an agentive interpretation is the expected one given the event type of the outcome (an activity). a temporary reading. In a nutshell. it is left unexplained why the stative reading is also available in such cases. the subject of sleep. with be present. 2 Rothstein (1999) also notes this is not the case for all activities. where Rothstein explicitly acknowledges that activity and agentive properties are not obvious. That is. Rothstein (1999) resorts to the conversational principles (Grice 1975) of Quantity (“do not say more than required”) and Manner (“avoid obscurity”). it is understood as a property predicated of the whole person. since subjects of activities and accomplishments are typically understood as agents.2 Be as a Copula Shifting a State into an Activity Rothstein (1999) proposes an alternative account to that based on two copular entries. the analysis of the two copulas does not capture the fact that such “agentive properties” are activated with a specific set of adjectives. cannot have an agentive reading. an activity. the copula shifts a state into an eventuality with activity-like properties. Rothstein argues that. other adjectives (blond.1 Regarding the interpretation of the subject. Eskimo) do not appear in activity-agentive contexts. To cover this question. 4.3). The proposal by Rothstein leaves the following points unanswered. if it is the copula that shifts a state into an activity. depending on the context. as a state. for example. would violate these maxims of Actually. (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). She offers a semantic investigation about the behavior of the predicates showing an agentive behavior as a complement of causative verbs like make. (8) Dafna is polite Also. it is not clear how the account can be extended to simpler inflected cases like (8). Rothstein argues that.1. 1 . whereas when the copula is overt (7). Rothstein (1999) argues (quite unspecifically) that “be + AP” constructions move quite freely among event types. leaving the eventuality in (7). given the role of be to shift an eventuality from a state into an activity. if the small clause (SC) of (7) be polite is eventive by virtue of the presence of be. associated to a particular eventuality.Aspectual Alternations in IL Predicates 85 And finally. Rothstein claims that when the adjective comes naked (6). As shown in chapter 3 (section 3. First. locatable in time and space. is more salient.

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

only activities and accomplishments fit. (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. This is expected. and the tree in (24) depicts the dyadic one. 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. As the following contrasts show. the event. The structures Stowell (1991) proposes for these potentially dyadic APs are the following.Aspectual Alternations in IL Predicates 87 Third. given the fact that only activities and accomplishments refer to “actions. (17) To invest in the bonus market is a cunning/mean action (18) *To invest in the bonus market is a cunning/mean fact Finally. represented by the infinitive clause.” (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. from a finer grained typology of eventualities. refers to “actions” rather than “facts. The tree in (23) represents the syntax of the adjective in its monadic usage. the infinitival arguments can be neither states nor achievements. (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).

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

2). According to Stowell.4 he takes it that the presence of such an argument is proper of SL predicates. Following Kratzer (1988. However. estar.3). 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. which is the lexical marking for SL-hood. In Spanish. Note that the contrast between ser and estar in this case falls in line with the contrasts studied in chapter 2 (section 2.Aspectual Alternations in IL Predicates 89 Stowell (1991) establishes a correspondence between the monadic version of these adjectives and individual-level (IL) predication.1) regarding the APs that can appear with either copula (cf. as (26) illustrates. he argues that the action denoting argument is an overt version of the (Davidsonian) eventive argument. 1995). in his account. 1995) account.2. (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. I take up the discussion related to the interpretation of the “infinitival argument” in Spanish and its status (as argument or adjunct). and the contrast between the examples with ser and those with estar (27) can be captured by paraphrases such as the one in (28). Specifically. the event argument represented by the infinitive is implicit in simpler copular cases such as the ones under investigation here. Ser is completely grammatical. (28)). I will make two observations. First. . that is not case. (25) John was cruel to Peter Regarding Stowell’s analysis. 4 For details of Kratzer’s (1988. see chapter 2 (section 2.1. this proposal would entail that in all instances of APs followed by a relational PP. but he is not cruel to him usually” Later. to admit there is an eventive argument that is implicit amounts to admitting that (25) is a SL predication. necessarily. I repeat one of the examples here in (29). in chapter 6 (section 6. and between the dyadic one and stage-level (SL) predication. the copular verb should be.

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. my proposal will not be based on properties of the copular verb. to account for nonstative IL copular clauses. 4.90 Individuals in Time guapo. since they are not the same. More specific- . Second. These two facts lead me to discard this hypothesis based on the existence of implicit event arguments. As shown in the previous chapter. (8) above).2 Adjectival Predicates Showing an Activity-like Behavior In chapter 3 (section 3. but he looks very handsome in that suit” Given the grammaticality of ser. as shown above. I will consider these sentences as instances of IL predication. My purpose in the remainder of the chapter is twofold. these nonstative copular clauses pattern with activities and not with accomplishments. I have argued in the previous section that the activity-like behavior should not be attributed to properties of the copular verb. it cannot be said it is the copula that induces the active features. the event argument Stowell proposes corresponds to an accomplishment (21) or activity (22). 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. we observed that there were certain copular combinations that behave like states and others that behave like activities. I will then offer an explanation regarding the stative reading also available with such predicates (cf. Thus. which is proved by their acceptation of during-adverbials and the rejection of in-adverbials. as a consequence.3). Since the activity-like behavior depends on the adjectival predicate combining with the copula. I will first elaborate an explanation as to what specific kind of adjectives are the ones that show an activity-agentive patterning and why. since it is not the copula that varies in the minimal pairs that can be construed (be Eskimo/be cruel) and. 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. 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.

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

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

A subset of them pattern with states and the two other subsets pattern with activities. this does not make them agentive. cunning c. Apt. inside the group patterning with activities only a subset of them involves agentive properties.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. More specifically.6 These are adjectives referring to properties that 5 Observe that although APs such as intelligent can refer to agentive events in constructions of the type discussed by Stowell (1991) (John was intelligent to close the window). kind Inner aspect patterning States States States States States Agentive activities States Activities Agentive activities Most of them show a uniform patterning with states. the group of MPs manifests a variety of behavior. However. Furthermore. capable b. Intelligent. those APs referring to velocity and those referring to aptitudes and human dispositions or mental properties (MPs). . there are two groups that pattern with activities: namely. Cruel.5.

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

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

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

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

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

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

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

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

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” . Stowell (1991) was right in considering the simple versions of relational MPs as IL and the ones accompanied by the PP as SL. This could in principle lead us to think that. The first scenario where cruel cannot appear followed by a complement PP is small clause (SC) complements of epistemic verbs such as consider. 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. (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. and rejects SL predicates. (81) Juan considera a Pedro cruel Juan considers Pedro cruel (82) ??Juan considera a Pedro cruel con Maria Juan considers Pedro cruel to Maria To elucidate the kind of predicates licensed in the SCs of consider is quite delicate. such as Demonte and Masullo (1999). in fact. Sentences (83) and (84) are based on examples from Demonte and Masullo (1999). Some authors have claimed that consider selects for SCs containing IL predicates. In the remainder of the chapter I will leave aside these (somewhat confusing) interpretive contrasts. without any apparent contradiction. However. at the same time. 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. I will conclude that cruel-type APs are not inherently relational. SCs complement of consider can perfectly have SL predicates. be claiming that the person is not cruel. From these cases. To show that the PP is optional. as other authors.102 Individuals in Time (80) Juan no es cruel. have pointed out.

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

¡sé cruel! “Image. (94) a. Examples (93)– (95) demonstrate that the possibility of the PP to appear depends on the nature (animate/inanimate) of the DP subject. such as the progressive or as a complement of parar de ‘stop’. En Canadá el frío es muy cruel In Canada the cold is really cruel *En Canadá. b. and. such as the imperative form. more interestingly. b. siendo cruel (96) *Esa imagen está that image estar-PRES-INDIC-3SG ser-ing cruel “That image is being cruel” (97) *Esa imagen ha parado de ser cruel that image has stopped ser-ing cruel “That image has stopped being cruel” As can be expected.104 Individuals in Time (93) a. they show a correlation between the relational PP and the properties that the subject has to possess. 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]. when the DP subject is inanimate. the construction cannot appear in contexts proper of dynamic eventualities. 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” . the construction cannot appear in agency scenarios either. el frío es muy cruel con los habitantes In Canada. (95) a. b. as a complement of verbs like force or in combination with volition adverbials: (98) Occurrence in command Imperative *Imagen. When the subject is inanimate.

Consider (101) and (102). since. as Violeta Demonte makes me notice.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. since it needs to bear the required properties for that (and animacy is the most basic one). namely. (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.3.3 The Relational PP with Other APs There are other MPs. although these adjectives referring to human (or animate) aptitudes can appear in combination with a relational PP. I consider these cases with inanimate subjects as evidence that the PP complement cannot be an obligatory complement of cruel. they do not have the same relationship with the PP. but just a “theme. while the subject of cruel is not an agent. I therefore. with particular characteristics of the construction. can also take a relational PP complement. as it is possible with the other adjectives (cruel. This is because they describe a property that can be understood in relation to another animate entity.). In the next sections. 4. I account for the relation between the two inner aspect forms. conclude that the structure of cruel-type APs does not always involve a PP complement and aspectually behave as states in principle.” if we use traditional vocabulary. not included in the group of relational MPs in the table above which. (104) is not the interpretation of (103). which are not totally excluded. as mentioned before. I will consider that the presence of the PP correlates with a subject involving particular properties and. If the PP complement is not overt. dynamicity. otherwise. its impossibility to appear with certain subjects would remain unaccounted for. etc. (103) Juan es inteligente Juan is intelligent . Summarizing. it is not inferred from the context or interpreted as generic. kind. also. which enables agency. In the first place.

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

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

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

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

1 Event Roles. since.1). among others. First. in principle. and the presence or absence of other elements in the sentence (the PP complement in our cases) that.” given the fact that both are atelic and do not “culminate. 1994).2 Syntactic Approaches 4. 1989.5. 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. As discussed earlier (see section 3. this approach does not establish any relationship between the aspectual term (Hold). 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. and van Voorst (1988). authors such as Verkuyl (1972). both would contain the term “hold. Tenny (1987. as captured in the Event Correspondence Rule (116). showed the relevance of arguments in the inner aspect properties of a sentence.1. there is no obvious way to distinguish between the two constructions in point (stative and dynamic copular clauses). In particular.” Second. 1994) and van Voorst (1988) elaborate both on the observation that arguments grammaticize the points that give the temporal contour of an event. As van Voorst puts it.110 Individuals in Time This approach leaves several questions unanswered. Tenny (1987. arguably. each entity corresponds to an aspectual notion. its syntactic position and its role in the determination of the aspectual properties of the sentence. 1989. Dowty (1979). (113) •---------------------------------------object of origin/actualization (114) •----------------------------------------• object of origin object of termination (115) ----------------------------------------• object of termination Activities Accomplishments Achievements (116) Event Correspondence Rule (Van Voorst 1988) Subject NP Direct object NP •----------------------------------------• event-object of event-object of termination origin/actualization . these authors emphasized the role of the internal object as an event delimiter. correlate with the aspectual outfits of the construction. 4. which opens the possibility of establishing a direct relationship between the thematic properties of the arguments present in a sentence. the dynamic or stative properties.2.5.

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

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

In other words. I will refer to her book (Borer 2005). More accurately. 2000) argue that functional projections responsible for Case processes (Agreement projections) assign the event-roles of delimitation and initiation. This head has the property of “quantity” (AspQMAX or Quantity). whereas a DP headed by a numeral. Likewise. Borer proposes that the core functional projection of the clause is Aspect. Mostly along the lines of Verkuyl (1972.15 In what follows.14.17 Borer proposes that the core feature of aspectual distinctions is [±quantity]. Borer (2005) elaborates on the intimate relationship between quantification and telicity or lack thereof. 2001b). she argues that aspectual properties of a determined event All the authors cited make proposals in a similar spirit in this regard. 2000). Thus. 2000. 2001a. Borer (1994. When considering that functional projections are in charge of event-roles too. where functional projections are in charge of Case assignment and agreement processes. that is. a telic one. that gives the type of aspectual interpretation. Event-roles are not assigned by a particular head but are simply derived as an entailment from the aktionsart of the whole event. In turn. It is the syntactic structure. A [+quantity] object. in turn. events are not lexically specified for any type of aktionsart. 1998. 2005) defends the idea that Case-checking processes are satisfied through aspectual projections. which I take as a frame of reference for the remaining discussion. events are not states or accomplishments from the lexicon. I introduce some points of the work by Borer (2005). 15 For further discussion about the relationship between case and event structrure. argument structure is computed on the basis of the syntactic structure. with telicity.16 For Borer. 14 . 1999. whose core is constituted by aspectual projections. since it presents the latest versions of her ideas and proposals.Aspectual Alternations in IL Predicates 113 and functional structure is interpreted as event structure. an object that can be marked as [–quantity] correlates with atelicity. a verb projects an unordered number of arguments whose interpretation depends just on the functional projection they move to. 17 This was introduced in chapter 3 (section 3. Ritter and Rosen (1996. As mentioned in chapter 3. 2001.4). all the relevant grammatical processes get a unified account. 1998. Assuming the parallel distinctions between mass/count nouns and atelic/telic verb interpretation. Compare: (124) John drank beer (125) John drank two beers A bare mass noun gives an atelic interpretation. 1989. 16 The work by Borer has appeared progressively in a number of articles. These works are framed in the Minimalist Program (Chomsky 1995. aspectual properties can be recast into quantity properties. the quantification properties of the DP direct object make a direct difference regarding telicity. see Kratzer 2004 and Kiparsky 1998. That is.

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

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

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

To capture (a)—that is. the following two properties: (a) When there is a relational PP. As I develop the proposal.1 Prepositions as Aspect Encoders I have shown that cruel-type APs. (b) When an animate DP subject and the PP are present. among others. based on Hale (1984). I argue. the aspectual alternation (state/activity) is not unpredictable at all. behave as activities. Stowell (1993). that prepositions can be conceived of as heads encoding aspect properties. which gains dynamic properties from the preposition heading the PP denoting the DP goal. That is. all of them very likely emerging from animacy itself: the DP has to be able to behave as a controller agent. 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.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 will take such a correlation seriously and I will propose that the source of the aspectual properties (specifically. 2000. 2004). and I will also be concerned with the aspectual treatment of the arguments expressed via a PP. and Demirdache and UribeEtxebarría (1997. 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). I will address two theoretical points. As I intimated before. the dynamic aspectual behavior) actually resides in the preposition itself. Point (a) is discussed in section 4. I will discuss the syntactic treatment of the opposition state/activity. it has become clear that the account for the cruel-type constructions has to capture. I propose that the dynamic properties become a part of the construction through the preposition heading the PP. To capture (b). From the description in the previous sections. 4. In support of this proposal.3. but it correlates with the existence of the PP relational complement. the PP makes them a dynamic predicate.6.) in the active cruel-type constructions is a projection (the PP) that combines with the “basic” structure of the AP. at least.Aspectual Alternations in IL Predicates 117 4.6. . I propose that the source of dynamicity and its related properties (agency etc. As I advanced. the construction has characteristics proper of processes. Let us begin by discussing point (b) concerning the aspectual contribution of the preposition. I will argue that the stative version of the cruel-type APs corresponds to their most “basic” structure. as soon as they combine with a (relational) PP.

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

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

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

in contrast to stative locative verbs participating in progressive forms. in Spanish and English. they notice that. the combination of a motion verb with an elative component and the preposition from yields a perfect interpretation: (146) Max COME FROM home ‘Max has come from home’ As Demirdache and Uribe-Etxebarría (1999) notice. in Dutch. such as from. . 2004) conclude that Aspect can be analyzed as a head establishing relations between two terms. in Nigerian Margi and Palaung (from the AustroAsiatic family of languages). (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. Demirdache and UribeEtxebarría (1997. to (143). where “come from” can be used with the aspectual meaning of the perfect. indicating movement toward (centripetal movement). appear in the form used to express close future (going to). Also. are present in perfect aspectual forms in some languages. 2000. the prepositions a (142). similar cases are found in Romance languages. (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). as prepositions do.23 Prepositions like the Italian per (‘for’. For example. prepositions denoting ‘movement from’ (centrifugal movement). For example. prepositions of noncentral coincidence are present in the auxiliaries expressing perfect and prospective aspect across languages. In turn. Specifically. 23 Note also the presence of a verb denoting movement. ‘to’) (144) or the English about to (145) are constituent parts of prospective forms.Aspectual Alternations in IL Predicates 121 ‘(with)in/on/at’. the locative preposition at directly combines with the verb in infinitival form to construe the progressive.

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

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.1 Con: A Locative and a Directional Preposition.Aspectual Alternations in IL Predicates 123 Assuming that aspectual meaning is reducible to prepositions (and. like in examples such as the following: (153) Pablito. that prepositions have aspectual meaning).2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor In this section I am going to describe the preposition heading the PP that refers to the (affected) goal in cruel-type construals in Spanish. and assuming with Demirdache and Uribe-Etxebarría that all temporal relations can be reduced to the same set of primitives. con ‘with’. it is conceivable that prepositions can also convey inner aspect properties. 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. as seen in the examples above. I am going to argue that the preposition introducing the (affected) goal carries inner aspectual properties affecting the aspectual nature of the whole construction. and. Preposition con establishes a central coincidence relation between the figure and the ground (in the terms introduced before). 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: . 4. denoting the goal. This preposition is associated with stative locational meanings. conversely. I propose that this is the case. run with your daddy! (154) Me volví con mis padres I went-back with my parents “I went back to my parent’s place” Importantly.6.6.2. the preposition con. ¡corre con tu papa! Pablito. The preposition introducing the PP in Spanish is. con-PPs can also appear with motion verbs. paying special attention to the APs I am taking care of in this chapter. 4.

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

Aspectual Alternations in IL Predicates

125

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

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

26

126

Individuals in Time

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

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

Aspectual Alternations in IL Predicates

127

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

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

30

128

Individuals in Time

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

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

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

Aspectual Alternations in IL Predicates

129

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

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

the discontinuous dotted line for the trajectory. therefore. the DP subject can be considered as the figure and the DP inside the PP as the ground.) (182) (183) TOWARD: preposition that establishes a centripetal noncentral coincidence relation between the figure and the ground. it locates the figure in its trajectory toward the ground.6. we saw that Demirdache and Uribe-Etxebarría (1997.1.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. 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). 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 . and (183) represents it graphically. The two terms are put under a specific relationship according to the meaning of the preposition. (The bullet stands for the figure and the little house for the ground. They are. …● ⌂ In section 4. associated with ‘future’ in the realm of Tense and with ‘prospective’ in the domain of outer aspect. 2000. it designates the movement of a figure toward a place or ground. 2004) proposed that centripetal noncentral prepositions translate as “before” in the temporal domain. (182) summarizes the description of the preposition. The two of them are related by a preposition meaning ‘toward’. (181) Juan went to Pedro In the terms of figure and ground presented before. As a directional preposition. that is.

an incomplete process. by their lexical meaning. in other words. 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.e. but not in past. consider the temporal contribution that motion verbs make in Lillooet Salish. we can say that such prepositions convey also the meaning of ‘before’ in the inner aspect domain. there are other adjectives (stupid. That is. sentences containing this verb and lacking an explicit temporal marker.) shows motion verbs in the mentioned language. the PP is not contextually inferred if it is not overtly present. they gain dynamic and agentive properties.132 Individuals in Time reached (“the destination has not been reached”). illustrate this point. shrewd) that. a concluded process) are interpreted as past tense sentences. Finally. in inner aspect terms. indicating a reached destination (i. Interestingly. admit a relational PP complement. (i) Motion toward speaker Motion away from speaker Destination not reached ts7as Nas Destination reached t’iq Tsicw 33 Verbs ts7as and nas share the property of referring to “movement” whose destination has not been reached. which translates as ‘nonaccomplishment’. although. in prep. In principle. to some extent. (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. Since the figure is captured “before” the ground.). Due to this reason.3. no process has been fulfilled33.3. sentences with t’iq or tsicw. The following table (Davis. cunning. then. (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’ .. it cannot be said it has reached its destination. are interpreted in present tense (or future). when these adjectival predicates combine with the PP. but merely as a ‘process’ or ‘activity’. as mentioned in section 4. both refer to an ongoing process. Because of the same reasoning.

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

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

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

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

a `centripetal noncentral coincidence preposition.Aspectual Alternations in IL Predicates 137 structure to emerge. That is. the paraphrases of both being ‘to’ (The Latin preposition in could be followed by a nominal either in ablative case. Both cases are similar regarding the elements that participate in them: a preposition. 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. since it is by virtue of a preposition that an adjectival construction becomes dynamic. Etymologically. contra Borer’s (2005) idea that statives are due to the presence of specialized functional structure (a sort of a “stativizer”). they come from the Latin “in” and “ad”. 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. an adjective and a verbal piece. If followed by a nominal in ablative case. or in accusative case. 4. black.7 The Cruel-type as a Small Clause Integrant The account I have developed thus far differentiates from previous explanations according to which the peculiar properties (which I have described as an activity-aspectual behavior) displayed by predicates of the type of cruel were due to the properties of the copular verb. whereas if followed by an accusative. it had a directional meaning (‘to’). The structure I have proposed for deadjectival verbs exploit the participation of a preposition to account for their dynamic properties. red) and refer to dynamic processes. The prepositions participating in the cases above are en (en-negreció) and a (a-largó). (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. the preposition had a locative (stative) meaning (English ‘in’).

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

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

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

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

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

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

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

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

.

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

as Tense and Aspect are considered ordering predicates. Stowell (1993.148 Individuals in Time In section 5. Lillooet Salish). Speech Time. but in dynamicity versus lack of dynamicity.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.e. as something that took place in the past). this being why the eventuality is computed as “dynamic and unbounded” (i. and simultaneity).1. like to. in Spanish at least. I tried to show that all aspectual properties could be reduced to the same set of primitives. Since the line of this work is not formal semantics I do not discuss them here. and inclusion are semantically defined as partial order relations fulfilling a number of properties.6. Therefore. At that point. 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. and Eventuality Time). I argued that dynamicity properties (that activities involve) can be described through prepositions. which I mentioned briefly in chapter 4 (section 4.. specifically. ter Meulen.g. I show that. I turn my attention to adjectival IL predicates and show the different interpretations available with each type of viewpoint mentioned.2 One of the facts I alluded to was that.3 Authors such as Zagona (1990) and Stowell (1993. posteriority. and Wall 1993.1). I argued that inner aspectual properties could also be conceived in ordering terms. It turns out that this cannot be described in terms of presence or absence of quantity. based on Zagona 1990. In particular. atelic). For an introduction of such notions. by the same token. I show that states are compatible even with the perfective. However. 5. 2 . that establish centripetal noncentral coincidence relations between a figure and a ground. 3 Anteriority. The idea is depicted in (1). the progressive seems to impose restrictions on the type of eventuality.5. These predicates translate into the temporal domain as ‘before’.. posteriority. Consistent with the conclusion that viewpoints do not interfere with quantity properties. 5. Reichenbach gave a formalization of tenses in natural languages based on three intervals (Reference Time. and the possible relationships that can be established among them (anteriority. inner-aspect characterization may work as an aspect indicator (an accomplishment can be taken as a perfect action. 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. in some languages (e. for example) and. therefore. inner aspect can be as well. the interested reader is referred to Partee. aspect properties can serve as indicators of temporal properties (a perfect action is computed as finished and.1 Tense.

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

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

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

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

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

The progressive is construed by the verb estar (locative be) plus the ing-form of the verb (Spanish -ndo). That is. the event has already taken place at the time denoted by the adverbial. I follow Klein’s (1994) proposal that there is an ordering relationship between the TT and the ET. I will thus use “perfective preterit” and “imperfect preterit” to refer to the two simple past forms. 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 . More strictly speaking. with the perfective. whereas with the perfect. Tense and Aspect consist in the same mechanism—namely. The verb estar acts as an auxiliary and can appear in any tense and aspect form. as Demirdache and Uribe-Etxebarría (2000) put it. I am not going to analyze all these aspect viewpoints.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.3SG Trabaja-ba he-work-IMPF-PRET. that of ordering temporal arguments. Aspect. the adverbial refers to the time at which the event takes place. the next question is how the singled-out portion of the situation (TT) relates to the rest of the event. the difference between the composed forms of “have” corresponding to the perfective and the perfect is that. In the glosses for my Spanish examples. 9 As many authors have pointed out. is an ordering predicate. 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.1.3SG Había salido a las tres he-had left at 39 Había salido ya a las tres he-had left already (at 3) Trabajará he-will-work Trabajaría he-would-work Table 5. Aspect nomenclature (García 1999) As indicated earlier. Returning to the internal working of Aspect. I mention them here only as clarifications.3SG Trabajó he-work-PERF-PRET. which is the element in charge of establishing a relationship between these two intervals. Thus.

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

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

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

3. As I will repeat later in section 5. the three forms differ in the number of occasions they allude to. Q<occ> can also have an existential value. also contain information about the number of occasions an eventuality occurs. I propose that aspectual viewpoints involve. aspectual viewpoints. (25) ----------------------x---------------------(26) ------x------x--------x-------x---------x-In the spirit of Verkuyl (1999). It appears . 5. inner aspect.2.5. besides the ordering component. The number of occasions the progressive and the perfective form allude to is just |1|.158 Individuals in Time of Quantity. besides ordering. who observed that the difference between the progressive and the habitual rests on the number of occasions they refer to. the imperfect in (24) refers to a plural number of occasions that can be described as greater than one (|>1|). a quantificational component. and the progressive. which I analyze as complex phrases containing a quantifier over occasions and ordering information. the nonhabitual or “continuous” imperfect. the functional projection which makes the eventuality heterogeneous and whose absence leaves it homogeneous. as the reader may have conjectured already. The continuous imperfect does not count the number of occasions a determined predicate takes place.2 and next in section 5. I will consider two points—first. as in (26). the habitual imperfect. However. and second. In this latter regard. Let me start by considering sentences (22)–(24). I address these points in turn. as shown in (25).1 Quantifying over Occasions As just mentioned. can be either |1| or |>1|. here in section 5. giving rise to another interpretation. However.1. what the relationship is between the number of occasions alluded to by each viewpoint and the quantity nature of predicates (that is. how many occasions each viewpoint alludes to. The values of such a quantifier. whose head is a quantifier over occasions (Q<occ>). In this section I am chiefly concerned with the perfective. (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.

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

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

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

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

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

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

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

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

Outer Aspect and Individual-Level Predicates

167

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

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

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

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

168

Individuals in Time

(60)

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

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

(62)

(63)

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

Outer Aspect and Individual-Level Predicates

169

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

170

Individuals in Time

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

Outer Aspect and Individual-Level Predicates

171

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

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

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

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

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

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

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

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

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

|>1|. which does not make the sentence habitual. this is borne out.25 The existential quantifier corresponds to the continuous imperfect. the ordering component takes the quantified eventuality and orders it with respect to the TT interval.4. |>1|. and not only with SL ones.26 The quantifier over occasions binds the eventive variable (e) which. (98) ZP 2 Zi VP 2 ei VP (99) AspP 2 TT Asp′ 2 Asp0 ZP (after/before/in) 2 VP Z0 (|1|. if “one. When the quantifier denotes a number roughly describable as “greater than one. This unifies the working of Aspect. following Stowell (1993. 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). Elaborating on Stowell’s (1993) proposal. 1996).” habitual imperfect is at stake. I used Spanish habitual and perfective data to show that aspectual viewpoints do not have any impact on the inner-aspect properties of predicates. as Kratzer (1988) originally proposed.178 Individuals in Time cation) is depicted in (99).5. As will be shown in section 5. ∃) should be possible with any type of predicate. In the last section.” perfective or progressive. independently from the kind of predicate. where there are no occasions to enumerate (think of stative homogeneous predicates such as Peter is tall). ∃) 2 e VP Thus. and makes the prediction that the quantifiers (|1|. See section 5.1 for further discussion. but these predicates are also compatible with the habitual quantifier (She was sick usually). presented in (3) above (repeated here as (98)) about the internal constituency of ZPs. 26 This aspect form is involved with stative SL predicates (She was sick). Specifically. I argued that. I assume to be present with every type of predicate (SL and IL). 25 . in Spanish. 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).

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

(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. 5. imperfect.5. *(En aquella ocasión) Juan estaba siendo esquimal that time Juan was ser-ing Eskimo “(That time) Juan was being Eskimo” b.3 (about the relation between quantity and perfective and imperfect viewpoint). I will explore the interpretation that adjectival IL predicates have under the aspect forms mentioned thus far (perfective. This will lead me to discuss the restriction existing between the progressive and the property of stativity mentioned in chapter 4. 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” . Juan está siendo cruel con Pedro Juan is ser-ing cruel to Pedro “Juan is being cruel to Pedro” b.180 Individuals in Time (101) Progressive form a. and progressive). This restriction differs from the one discussed in section 5. *(En aquella ocasión) Juan estaba siendo rubio that time Juan was ser-ing blond “(That time) Juan was being blond” c.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. 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.

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

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

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

when a relational complement is added (123). whereas activities in imperfect (117) can also have a progressive paraphrase (118). although the progressive form is correct with cruel. rather than the habitual (cf. in Spanish.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.2.3.30 Finally. as noted in section 5. when a complement that can act as a “distributee” is added (125). which. a habitual interpretation emerges. However. (122)). Likewise.1. they acquire activity-like properties. 30 For the reasons why the progressive is excluded with Eskimo and blond.184 Individuals in Time It is worth noticing that.5. 29 . as noted above. this is just marginally available with be cruel (cf. (119) and (120)). 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. the progressive paraphrase of the imperfect is more natural in these contexts. see section 5. make the habitual reading available. accordingly. the progressive is not a paraphrase saliently available from the imperfect form. the preferred interpretation of adjectives denoting intellectual aptitudes (121) is the continuous reading.

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

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

the progressive in English works differently in some contexts.31 The progressive viewpoint does not seem to work like the imperfect and the perfective. cruel/inteligente toda su vida (140) Juan fue Juan ser-PRET-PERF-3SG cruel/ intelligent all his life “Juan was cruel his entire life” 5.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. As can be seen. and therefore reject the expression of progression in time.5. whereas processes take time and progress in time (although this progression in time does not mean they advance toward a culminating point). 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).3 The Progressive and Adjectival IL Predicates I now turn my attention to the progressive viewpoint in Spanish. in the sense that the progressive is not compatible with every kind of predicate. according to which it is compatible with basically any type of predicate. it is also possible to have the stative version of these adjectives together with the perfective. states hold in time but do not take time. . consistent with the description of the perfective above. 31 As mentioned in chapter 3. SL (141) and IL (142) statives are excluded with the progressive. (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. the progressive is possible with other class of homogeneous predicates.

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

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

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

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

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

Chapter 6 Tense and Individual-Level Predicates In this chapter I study another temporal realm—namely. These two features suppose a difference with respect to previous accounts of IL predicates in general and. Throughout the discussion I will be chiefly concerned with the interpretation of IL sentences in the past tense. it does not influence the ultimate temporal interpretation of the sentence. it is not the case that Tense has the same interpretive impact on all individual-level (IL) sentences. by exploring the interpretation of IL predicates in compound sentences. Likewise. This chapter is organized as follows. The two main points I would like to make here are the following.4 I show how contextual factors intervene in the rising of this temporal interpretation. those dubbed “lifetime effects” (Musan 1995).3 presents the discussion of lifetime effects and describes the properties the predicate has to involve to trigger them. the domain of Tense. As it appears. As mentioned in chapter 2. In section 6. I assume tenses are syntactically represented in the node of Tense. . in particular. which has consequences for their temporal interpretation. Second. As I have shown. formerly discussed by Kratzer (1988. which. structurally. 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’. Section 6. In this regard. I will be concerned with the interpretive contribution that tenses (present. in section 6. I propose a division among (adjectival) IL predicates that correlates with different temporal interpretations. past.5 discusses the nature of the TT in greater detail. to previous accounts of the temporal interpretation of IL predicates. although it is not only the type of predicate that counts. The last section summarizes the conclusions. 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. 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. 1995). Section 6. IL predicates offer different inner-aspect properties. First. it is the content of the Topic Time (TT) that ultimately decides the temporal interpretation of the sentence. is located higher than the Aspect node. as a consequence. since this tense form gives rise to one of the most important interpretive effects with IL predicates—namely. and future) have on copular sentences with adjectival predicates. In the first section I introduce one previous account for the particularities of the temporal interpretation of IL predicates.2.

1. Kratzer proposes that the differentiation between SL and IL predicates resides in their argument structure.”2 Specifically. she proposed that IL predicates applied directly to the subject.1). see Anderson 1973. as Diesing 1992 proposes. the representation in (1). in the same vein as Carlson (1977). The configurational difference between IL and SL predicates can be depicted roughly as in (1). whereas permanent ones lack such an argument. Kratzer argues that SL predicates involve an extra argument in their argument structure—namely. putting in direct correlation the presence of a spatiotemporal 1 2 For an earlier brief discussion on this phenomenon. SL predicates IP 1 I′ 1 I VP 1 <e> b. That is. IL predicates IP 1 I′ 1 I VP 1 θ-subj Regarding temporal interpretation.3 (1) a. just aims to capture Kratzer’s idea on the location of the external argument with respect to the maximal projection of the predicate. 1995) Kratzer (1988.” See chapter 2 (section 2.1. she argues that the eventive argument is the external argument of (SL) predicates. temporariness and permanency is encoded in the argument structure of predicates: temporary predicates involve a spatiotemporal argument. For Carlson (1977). Kratzer proposes that the spatiotemporal argument is syntactically represented. First. Furthermore.194 Individuals in Time 6. a spatiotemporal or eventive (<e>) argument (Davidson 1967). 1995) is the first in establishing an explicit correlation between the temporal interpretation later dubbed “lifetime effects” and IL predicates. the “realization function. In particular. .2). I will mention two points of this proposal. As said before in chapter 2 (section 2. whereas SL predicates needed something “extra.1 Temporal Interpretation as a Consequence of Argument Structure: Kratzer (1988.1 I will briefly summarize her line of reasoning and show that there are cases it cannot account for. SL predicates involve an extra semantic operation. This author assumes that the factor that draws the line between the two types of predicates is “being temporary” versus “being permanent. 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.” Likewise. 3 As I also said in chapter 2.

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

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

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

198

Individuals in Time

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

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

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

Tense and Individual-Level Predicates

199

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

200

Individuals in Time

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

Tense and Individual-Level Predicates

201

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

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

202

Individuals in Time

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

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

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

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

but can give what can be called a “forward-lifetime effect” (46). 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 . does not activate the reading in (43).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). A sentence like (44). In (40). In the remainder of the discussion I will not make any other allusion to future tense cases. (43) Pedro is no longer alive The second necessary condition is that the predicate must be tensed in the past. in the present. the predicate is understood as referring to school time. The future tense does not yield the interpretation of (43). an interpretation like (43) is available when the predicate is a lifetime IL predicate. However. either. 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.

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

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

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

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

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

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

a set of propositions that form the shared background of the participants in the conversation.3 I will show how my proposal (to account for all the cases uniformly) works in a systematic way. a “context” contains a “common ground”—that is. I will show that the TT is influenced by the discourse topic.4. In section 6. 1981).” Following widespread proposals in semantics and pragmatics. Grice (1975). Based on Keenan-Ochs and Schieffelin 1976.4. 1979). 6. that it can be described as an anaphoric relationship. where. 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. Among the elements constituting the discursive common ground. 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.Tense and Individual-Level Predicates 213 interpretation of these two types (with and without context) varies (i. they are anaphoric to distinct “discourse topics.e. the content of their TTs is different) because their TTs get their values from distinct “discourse topics”. In the next two sections. I define “discourse topic” as the subject of immediate concern in the conversation.. lifetime effects do not arise. (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 . The common ground determines the options that are live at any point in the conversation. following von Fintel (1994). nevertheless. Kratzer (1977. and von Fintel (1994). Consider the following sentences. As defined by Stalnaker (1972. among many others. I assume that sentences are uttered against this background. I assume. 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).1 When the Subject Is a QDP I am going to examine a set of examples without any previous explicit context.

Longobardi 1994). in the world. As is classically argued (Heim 1982. Thus. Then. as mentioned in chapter 5. in these cases we cannot allude to the presence of a previously mentioned past interval to account for the mitigation of the interpretation. Lewis 1986). the meaning of (70) is that of (71). but just over the contextually relevant set of plates. constitutes the quantificational restriction. In essence. Higginbotham 1983. where the reading can be defended to exist as a possibility. say. Stump (1981). the simplest hypothesis is to suppose that the difference in the availability of the reading resides in the different nature of the DPs. I am going to argue that the very nature and working of quantifiers is what precludes the arising of the reading. propose that such a restriction is formalized through a free variable in the determiner. and this. The contextual relevance gets into play through the contextual variable appearing as a subindex of the determiner. among others. 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 context provides a set that the NP intersects with. In (70). That set is called the “restrictive clause. which the noun is a predicate of (69a). However. all together. Partee (1984a). D x [ x is N] g restrictive clause Dx [x(N)] Furthermore. . determiners/quantifiers are interpreted with respect to the set contextually provided.” (69) The boys a. and von Fintel (1994). the set that determiners quantify over is restricted by the context (Westerståhl 1984. every is not quantifying over the whole set of plates. Since the difference between (52) (Gregory was from America). and these examples resides in the nature of the subject DPs. QP/DP 2 Di NP the 2 xi N′ g boy b. quantifiers bind the open argument (variable x). Stowell 1989.214 Individuals in Time (68) Ningún chico era esquimal No guy was Eskimo In none of these examples is the lifetime reading available. boys in (69).

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

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

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

4. more systematically. Elaborating on Musan 1995. along the same lines that specific DPs are generally understood (Pesetsky 1987. following von Fintel (1994). The relationship between the contextual background and the TT or the free variable C has been taken as anaphoric. The occurrence of lifetime effects depends on the content of TT of the sentence. which contains contextual information associated to the individual referred to by the DP. and. 2. it becomes the source for antecedents. so that all the surveyed cases can be accounted for in a uniform way. how this is articulated. I want to show now. The relevant context can be brought to the scenario by the DP subject itself. That is. In other words. Beghelli & Stowell 1996).218 Individuals in Time situation. Conversely. 6. The TT refers to a specific interval. Enç 1991a. The leading ideas introduced thus far are the following: 1. the context acting as background constitutes the set where anaphoric elements and free variables find their antecedents. where the individual referred to by João is present. The context relevant to find the “antecedent” for the TT does not have to be another past interval necessarily. 4. as Musan (1995) suggests. 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. This seems to suffice to convert the utterance situation as the background that counts as relevant for the antecedent of the TT. I also argued that proper name DPs involve a similar context variable. 5.3 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. I propose that the mechanical interpretive procedure determining whether a lifetime effect arises is the same: the content of the TT is determined by the elements in the salient discourse. the utterance situation is the most salient context. 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. as a consequence. A specific interval is an interval that is linked to a previous context. 3. I have argued that this is brought about through the context variable restricting quantifiers (free variable C). (77) Salient context " TT .

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

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

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

and its eventuality will be located with respect to the eventuality of the main clause. This way. since the RT would be controlled by the main ET rather than by the main TT. According to him. the content of the RT is.15 However. by default. namely. as in (i). subordinate clauses (complement and relative). the ordering between the TTs should depend on the specification of the embedded RT. As Hornstein (1990) and Demirdache and Uribe-Etxebarría (2004). starting by the former ones. instead of with respect to the TT. as the subindexes of the ZPs in (i) suggest. . In this work. However. I argued that what Tense locates is not the ET as such. Following the logic of Stowell’s proposal discussed above.M.5. the prototypical case of this is the modification of the past perfect. Specifically. giving rise to different temporal readings.M. in compound sentences. if the RT of the subordinate clause is in the c-command domain of the ET. I have just dealt with IL sentences which establish a link with a previous one paratactically. The time adverb at 3 can modify the TT. to understand the temporal interpretation of a compound sentence is to understand the relationship holding between the two (or more) TTs. Peter’s leaving occurs at 3 P. Thus. In simple sentences.M. the subordinate clause will be temporally dependent on the main one. TT and ET can be modified by time adverbs. I focus now on the interpretation of lifetime IL predicates in another type of syntactic environment. (Thus far. As mentioned in chapter 5. so that it can be controlled by the ET of the main clause. this does not mean that ETs always remain in the dark for interpretation.1 Complement Clauses In chapter 5 (section 5. have noted.) I will deal with complement clauses and relatives in turns.1.5 The Lifetime Reading in Compound Sentences For the sake of a more complete survey of cases.14 but rather the TT. given that the ET is the closest ZP. 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. 6. yielding the reading in (ii) or the ET.1) I described the way temporal interpretation is accounted for following Stowell’s (1993. the subordinate RT can be affected by the ET of the main clause. yielding the interpretation in (iii). yielding different interpretations. the embedded TT would become ordered with respect to the main ET.222 Individuals in Time 6. among others. this could lead us to a situation where the TTs are not ordered between each other strictly speaking. temporal interpretation is derived from the ordering relationship Tense establishes between the RT and the ET. 1996) work. whose nature I argue to be also a ZP. the UT. (i) (ii) (iii) 15 14 Peter had left at 3 P. Peter’s leaving occurs prior to 3 P.

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

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

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

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

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

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

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

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

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

(102) Juan estaba buscando (a) una niña John was looking for a girl The two interpretations of such DPs have been traditionally accounted for in scopal terms. a girl > look for When the main verb is extensional. Abusch demonstrated that the different temporal readings of RCs correlate with the interpretation of the DP the relative depends on. Secondly.2 Relative Clauses In this section. The reading roughly described as [–specific] would correspond to a narrow scope position with respect to the intensional verb. which may be determined by that of the DP . I will summarize general points concerning the temporal interpretation of RCs. Rivero 1975. I will make some considerations about the interpretation of lifetime predicates in another syntactic environment: Relative Clauses (RCs). Donnellan 1966. 1996) work. Firstly. a lifetime interpretation is not expected from a syntactic situation such as (101). Adriana. it can have two interpretations. as I have been doing thus far.232 Individuals in Time the TT is located in the past. or Juan was looking for a particular girl. and the [+specific] interpretation to a wide scope position. Stowell (1993. If we take an example like (102). when the DP is indefinite and acts as a complement of an intensional verb. 1996) builds on Ladusaw (1977) and Abusch (1988) to elaborate his proposal about the temporal interpretation of RCs. 6. namely. I will divide the task in two steps. we see that the interpretation of the DP a girl. which can be described as a [–specific] and a [+specific] (Russell 1905.5. Summarizing very much. I will pose what this analysis predicts with respect to the lifetime predicates and show whether it is borne out or not. the DP can be considered as unambiguously [+specific]. look for > a girl b. (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. following Stowell’s (1993. among others). (103) a. whoever she might be. can be paraphrased in two different ways: either Juan was looking for any girl.

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

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

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

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

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

.

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

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

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

aspect. However. I understand that the difference consists of something describable along the lines of the rightmost column.2. To understand a property as permanent means either that the property holds for the entire lifetime span of the individual (intelligent) or that. Thus. Chierchia 1995.” or “temporally anchored” (in contrast to IL predicates. 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). Different number of occasions eventualities occur and different orderings between the Eventuality Time (ET) and the Topic Time (TT) intervals. nonstable predicates. 1995. tense). among others). In this work. Table 7. I distinguished three temporal levels (inner aspect. I have considered that each of them is syntactically represented (quantity. I have argued that to attribute permanency to a property amounts to assigning it a determined temporal interval that the property extends over. tense) and have described the working of each of these heads as well as the differentiations we can establish among predicates depending on the distinctions made by each temporal head. Temporal unit descriptions and definitions In what follows I summarize the conclusions I have drawn from the study of each temporal realm and discuss whether the IL/SL dichotomy can be ultimately established in temporal terms.” “temporally bound. outer aspect.242 Individuals in Time of them describe SL predicates as “involving aspectual properties. which lack all such characteristics). once “acquired.2 summarizes the temporal units. In contrast.D. and gave concrete definitions to each. One of the difficulties in these descriptions is the vagueness as to what exactly the authors mean by such terms. Differentiation between homogeneous and heterogeneous predicates Outer aspect Inner aspect Table 7.3 The IL/SL Dichotomy Is Not a Permanent/Episodic Distinction IL predicates have been defined in the literature as permanent predicates (Kratzer 1988. SL predicates are conceived as episodic. and the distinctions that can be made according to each one. their definitions.). if we say that SL and IL predicates differ in certain temporal terms. 7.” it lasts for the remainder of the individual’s lifetime (Ph. if we .

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

.

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

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

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

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

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

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

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

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

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

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

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

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

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

.

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

207. 116. 191 Picallo. 55 Rosen. 40–44. 120 Parsons. J. 245 Martin. 37 Pustejovsky. de 30. 36–38. 44. 189. 81. 50 Landman. 3. 119. 143. F. 40. 70. 226. 165. 37. 25 Quine. 115. S. 84. 113. L. N. 71 Ramchand. T. 251 Sportiche. 1. 84. 147. 156 Tenny. 37. S. 114 Kuroda. 117. 94 Morera. 116. G. 237. W. M. 127 Swart. 56. 5. 250 Ladusaw. 250 Rapoport. W. A. 75. H. 191 Ogihara. 191 Pollock. S. P. 156. 225. 148–150.276 Individuals in Time Kearns. R. S. 240 Rivero. 112 Tungseth. 107. E. 109. 178. M. de 48 Milsark. 29 Talmy. 38 Mithun. M. C. 12–14. 205. 215. 127 Uriagereka. 218. 252. B. 79. S. 144. P. 148. 98. C. 143 Russell. 122. 220. 176. 15. 112. 208. 16 Rosen. 211. 214 Longobardi. 150. 162. M. 88 Lemmon. 218 Phillips. 61. 113 Kitagawa. T.-Y. 226 Pagliuca. 96. G. C. 6. 148. 164 Miguel. 240 Rothstein. 8–12. 73. 240 Kondrashova. 22 Meulen. 145. 43. 94 Musan. 216 Reichenbach. 113. H. 138. 214 Perkins. ter 148. 61. 214 Luján. 11. 240. 232. 113. 3.-Y. 145. 235. 80 Ritter. 15 McConnell-Ginet. 12–14. 240 . 55. W. 5. 41. E. 163. 61. 150 Reinhart. M. 205. J. C. 213 Stowell. 40–43. 43. 172. 45. 126. 77. P. 135 Kiparsky. R. 153. 176 Szabolcsi. 119. 251 Quer. 191 Partee. 77 Roldán. 130 Mourelatos. V. R. 118 Matushansky. T. T. P. 217–220. 151. 140. 211. 152. 110– 112 Torii. G. J. 29. J. 178. 22 Kenny. 167. 114 Schieffelin. 102. 16. 25. 109. 192. 10. 136. 242–244. M. 89. 33. 155. 136. C. 31. L. 176 Munro.-L. 12–14. 46. 85. 251. 106. M. 70. 122. M. 71. D. 93. 40. 94. 55. A. 144. 122. 18 Schmitt. 43. L. W. C. 78. R. 232 Lakoff. M. 188 Keenan-Ochs. 86. S. D. A. 3. 194 Klein. 8. 222. 38. 50. 49. 26. 147. 120 Pesetsky. 60. E. 3. J. 3. 15 Raposo. 80. 97. 164. 216 Uribe-Etxebarría. 8. 40. 143. 70. 189. 109 McNally. Y. J. 126 Morimoto. A. 202. 236 Naumann. 112 Schein. 244 Stump. E. L. 29. 213. 193–196. C. 33. 47 Keyser. 232–235. 34. 214. 194 Kratzer. 116. E. J. 127. 74. 191 Larson. 233 Rohrer. 144. 57 Masullo. 12. 240. 41. 56 Sanz. O. 243 Travis. 225. 59. 195 Leonetti. 38. 208–212. 145. 251 Krifka. 188. M. 138. 14. 214 Svenonius. 215 Stalnaker. T. G. 112. T. 192. 190. 124. B. 131. G. 18 Piñón. B. 254 Lewis. A. 144. 117. Y. 43. 38. 55. 11. M. 61. 120– 123. 213 Schmidt. 154. 119. J. 84. R. 25–32. 208. 32. 38. 194. 15. 96. 86–90. 152. 71. S. 15 Koopman. 106. 97. 162. 3. 43. G. 193. 196. 213 Kennedy. 222. B. 127. H. 8. 32. 245. 232 Ryle. 102 Matthewson. 152. 58. 225. 154. R. P. 207. C. 233 Querido. 14. A. 28. D.

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

.

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

257 Perform(er) 53. 72. 156 Focus 13. 195. 212. 152. 227–229. 159 Quantification restriction 11 Quantity 113–116. 227 Simultaneous reading 76–79. 141. 132. 189. 207. 158. 122. 94 Pomo 94 Prepositions Directional prepositions 127. 244. 118. 174–178. 112 Juba Arabic 120 Kinesis 40 Krio 120 Lakhota 55. 147. 178. 186. 35. 248 Swahili 120 Temporal interpretation of Complement clauses 76. 12. 252 Realization function 6. 122. 80. 158–160. 201. see Temporal interpretation Past-shifted reading 76–78. 157. 131. 221 Habitual 77. 249 Relative clauses temporal interpretation of. 191 Event(uality) Time (ET) 3. 248. 187–192. 176. 26. 186. 106. 162. 233 Perfect entailment 47. 215 Goal(Path) 96–98. 222–225. 112. 148 Sequence of Tense 226. 160. 37 Thetic (judgments) 12–14. 172. 246. 181 Existential reading 6. 207. 94. 180. 134–136. 130. 175. 155. 31 Imperfective paradox see Progressive paradox Initiation. 257. 14 Generic reading 6. 57. 230. 72. 158. 212. 79. 142. 181. initiator 61. 249. 80. 116.280 Individuals in Time Event-object 110. 180 Perfective (preterit) 31. 176–178. 150. 258 Progressive paradox 191. 100. 184. see also Utterance Time Stativizer 115. 240. 224. 245. 230. 14. see Temporal interpretation Russian 15 Salish 118. 158. 148. 158–160. 194 Reference Time (RT) 3. 76. 250 . 148. 14. 184. 178. 165. 129. 233. 224. 244. 136 Place/situational prepositions 120. 240. 192 (Im)perfective contrast 25. 144. 171– 175. 258 Hindi 120 Homogeneous (predicates) 70. 187. 132. see also Subinterval property and Additivity property Narration temporal interpretation of. 208. 148. 148–152. 128. 226. 196. 248. 67. 123. 88. 226. 144. 159. 179. 172. 186. 236 Theme 105. 249 Existential coda 12 Existential quantification 8. 186. 134 There-sentences 5. 123– 126. 258 Imperfect Continuous 76. 131. 165. 176. 242. 83. 172–174. 242. 189– 192. 11. 241. see also Event(ive) argument Speech Time 122. 64. 229. 173. 148–152. 153–160. 234 Situation aspect 39 Small v 55 Spatiotemporal argument/variable 8–10. 184. 192. 11. 178. 212 Mereological properties 3. 35 Figure 119. 202 Ground 119. 136. 221–223. 189. 227. 186. 55. 156 Hawaii-Creole 120 Hebrew 15 Heterogeneous (predicates) 72. 161. 135. 131. 227. 136. 116. 154. 105. 158. 246–248. 147. 233–235. 246. 187. 189. 242. 173–180. 58. 258. 209. 231. 117. 155–157. 192 Prospective 120–122. 232. 147. 211. 194. 168 French 121 Generic operator 11. 123. 94 Margi 121 Maxim of Quantity 85. 191. 132. 6. 115. 147. 246. 162. 148. 189. 113. 125. 192. 79. 167. 148. 183. 177–180. 129 Progressive 153–163. 183. 12. 154–157. 86. 229. 211. 111. 134–136. 189. 137 Subinterval property 70. 222. 35. 175–177. 72. 144. 230 of Narration 77–80 of Relative clauses 223. 226. 141. 129. 128–131. 134 Instigator 55.

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

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

Sign up to vote on this title
UsefulNot useful