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

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

...72 3...1 Aspectual Tests on IL Predicate ..3 The Relational PP with Other APs ...............................................3..................................................................96 4.................................2 Differences between Activities and States .....6 An Account Based on the Relational PP Complement .........................5.............................................7...................86 4..................1 The Two-Copulas Hypothesis ...........................84 4.........4 Summary of Section 4............................3 Summary of Section 4....3..............................................2.....143 4............133 4............8........105 4........4.........................................91 4....................5.....80 3.........2 On the Optionality of the (Affected) Goal PP .1..............................137 4................4 Summary of Section 4............................................. The Relational PP Complement..84 4............................1 Lexicalist and Logico-Semantic Approaches ............................81 Chapter 4 Aspectual Alternations in Individual-Level Predicates .1..1 Cruel-type Small Clauses Taken by Verbs Other Than the Copula ........4......6........7......90 4..............................107 4. The Cruel-type as a Small Clause Integrant.......4 Summary of Section 3......4.............5.................................3 Adjectives Predicated of an Implicit Event Argument .............................3....vi Individuals in Time 3.............................3.......... 69 3............110 4..................................6..........................................96 4..68 3..............................................1...............2 Syntactic Approaches .................................62 3.......2........138 4.......................................................3 Relational Mental Properties.........................1 Prepositions as Aspect Encoders .............2 Adjectival Predicates Showing an Activity-like Behavior ....3 ..................98 4.....4............3........................................83 4...................5 Justifying the Approach .........1 On the Interpretation of the Relational PP ............................85 4....6....................143 .........2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor123 4.....117 4.. States and Activities: A Grammatically Relevant Difference?..............7........4 ............................3 Some Confusing Arguments in Defense of a State/Activity Distinction.....................2 .................................................3.....116 4.............1 Lexical Semantic Classes of Adjectives ....1 Similarities between Activities and States .4 The Issue: Aspectual Alternation in IL Copular Clauses . Summary of the Chapter .........5 .....95 4............108 4..............................108 4..........2 Be as a Copula Shifting a State into an Activity ...............1 Previous Explanations of Nonstative Copular Clauses ....106 4.....3 The DP Subject of Relational MPs .......6 ..4..7 .....................................................2 Summary of Section 4........................................................................... Summary of the Chapter ........135 4..........76 3................117 4.69 3.....................2 Summary of Section 3..................6.........................3 .............................2 Summary of Section 4............................5.................................................

.......5 Adjectival Individual-Level Predicates and Viewpoint Aspect .........5..199 6.158 5................. 179 5........... 180 5.......................................................194 6.....4...................................................3 The Arising of Lifetime Effects ..163 5.............................................215 6......................................................................................................................................................................................2 ..............................................5........... 1995) .........3...................5 The Lifetime Reading in Compound Sentences .................3 Brief Notes about the Complement of Mental Properties APs .....205 6...........................................2 The Internal Argument of Tense: The External Argument of Aspect.........1 The Imperfect and Adjectival IL Predicates ..........210 6.....212 6....4........................2...2....................1 Permanent IL Predicates ....................218 6............................................213 6...3 Inner and Outer Aspect . Aspect as an Ordering Predicate ...............1 Temporal Interpretation as a Consequence of Argument Structure...................................................................1 Habitual Heterogeneous Predications.............................................................171 5.1...........3...6.............................................. 1996) ......4 The Determination of the TT Content and Lifetime Effects .......4............................................2.....................................2 The Habitual Interpretation: Iteration....................3 On the Relation between the TT and the Habitual Q<occ> .199 6............... Stowell (1993...193 6...................................................2 Differentiating Temporal Extensions of IL Predicates .................................1 Tense and Aspect as Ordering Predicates .......2 Nonpermanent IL Predicates............2 Aspect as a Quantifier over Occasions .........4.....152 5.4 Summary of Section 6.157 5.......................173 5..................... A Brief Summary of Aspect Notions .....................2 Perfective Homogeneous Predications .......... Kratzer (1988.........202 6.............. 177 5..................2..................................................3 The Content of the TT in the Arising of Lifetime Readings .......4.....3 Articulating the Account.........................148 5.5....................208 6.........................1 When the Subject Is a QDP .................3..........2 Context Associated to Individuals .................222 ..................................................... 1997) .................................Table of Contents vii Chapter 5 Outer Aspect and Individual-Level Predicates ..........................197 6............................................4 ......................147 5..... Summary of the Chapter ......................3......................................2...........220 6.........1 Quantifying over Occasions ............3..........2... 192 Chapter 6 Tense and Individual-Level Predicates ......................................... 186 5...... Proportion...................... 174 5...........1 Tense.....205 6.....3 The Progressive and Adjectival IL Predicates .....1.......148 5............ 187 5...........................4 Summary of Section 6.....1 Necessary Conditions for Lifetime Effects ......................................2.....205 6........ and Systematicity ....... 176 5.2 Introducing the Determining Role of Contextual Factors.2 The Perfective and Adjectival IL Predicates ............. Musan (1995............

.................................3........................235 Chapter 7 Conclusions and Final Remarks .......2.......................1.....8.............................................................................. Summary of the Chapter ........................ Summary ... The IL/SL Dichotomy Is Not a Permanent/Episodic Distinction ............................................. Recasting the IL/SL Distinction ............242 7......................2 Relative Clauses .....................................261 Name Index ................................................6................................... Summary of the Conclusions ..5............1 Complement Clauses ..........................................279 ....................................239 7..........5. The IL/SL Dichotomy and the Ser/Estar Contrast ......................7...........................249 7..................................249 7............................. Outer Aspect Does Not Affect the IL/SL Distinction ....5............254 7.............257 References ...........................222 6....................viii Individuals in Time 6.................245 7...................................................................275 Subject Index....................................241 7..........................................................232 6................................................... Some Remaining Questions .......................... The IL/SL Distinction Is Not a Matter of Inner Aspect…Completely ..............................4.239 7..................................................6..........................................

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

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

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

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

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

.

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

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

I will show that prepositions can contribute an inner aspect shift from states to activities. I will propose a formal constitution of Aspect as a complex functional projection including an ordering predicate (which locates the so-called Topic Time [Klein 1994. This book is organized as follows. 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’). 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). along the lines of Demirdache and Uribe-Etxebarría 2000. based on hypotheses that attribute aspectual content to prepositions (Hale 1984). 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). Focusing on copular clauses. which is described as a viewpoint referring to a plural number of occasions. dynamicity is proposed to correlate with the presence of the relational complement and argued to be rooted in the prepositional phrase. the dynamic properties of clauses where adjectives referring to mental properties appear (such as cruel. mean) are proposed to correlate with the presence of the relational complement of these adjectives (cruel to Mary).Presentation of the Study 3 tense. 1996). I start the discussion of the temporal domain of inner aspect by first presenting the different event types acknowledged in the literature and the tests to diagnose them. In particular. Ramchand 2003. I argue that the IL/SL contrast should be rooted in the association to a particular circumstance (SL) or its lack thereof (IL). Chapter 2 introduces the notion of IL predicate and presents the distribution of the two copular verbs in Spanish. I study the set of IL predicates patterning with activities and argue that this behavior is rooted in the nature of the syntactic frame intro- . I present some reflections about habituality. Thus. my proposal will concern the role of prepositions in event structure. kind. The discussion of inner aspect is framed in the theoretical debate regarding the effects of syntactic structure on the semantics of predicates. I will then discuss the theoretical and empirical consequences for the IL/ SL dichotomy. Specifically. 2004) and a quantifier over occasions. giving the number of times a certain eventuality holds (along the lines of Verkuyl 1999). this being understood in inner aspect terms (Borer 2005. the ideas defended in this monograph are sympathetic to the conception of syntactic structure as event structure. In this respect. In chapter 4. 1995] with respect to the totality of the Eventuality Time. Ritter & Rosen 2000). In chapter 3. 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. Outer aspect and tense are conceived as dyadic ordering predicates that take two time-denoting arguments (Stowell 1993.

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

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

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

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

e) & (slowly. such as Kratzer (1988. the event variable is existentially quantified). NP). denoting actions. 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. 1995) In the spirit of Carlson (1977). it is the DP referring to the individual itself which functions as external argument. with a knife. e) & (with a knife. Kratzer (1988.3. denoting nonactions but performing actions and proposed that events could be treated as individuals. He proposed that action sentences include an event variable as a primitive element in their logical semantics. (∃e) (buttered (Jones. 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.3 In Kratzer’s proposal. which.4 2 Davidson (1967) took the classical differentiation from Plato between verbs. IL predicates do not. When it is not projected. based on Williams’s (1981) Argument Linking description. PP. The Davidsonian argument is the external argument of the predicate. Kratzer (1988. 3 4 I will add a few more words about the eventive variable in chapter 4. and nouns. . As will be shown in section 2. e) & (at midnight. like nominal variables.2 The eventive variable is considered a primitive element in the logical semantics of events. 2. whereas SL predicates project this extra argument (which can be understood as a spatiotemporal argument).8 Individuals in Time This view has been adopted by some authors. at midnight. e) & (in the bathroom.2 The IL/SL Dichotomy as a Syntactic Distinction. This gives rise to a difference in their argument structure which is the stem of the difference between the two types of predicates. e)).1. from Davidson 1967. as in (i). 1995) and Diesing (1992) but disputed by others. as Diesing 1992 proposes. such as Jäger (1999) and Becker (2000). Schematically. Kratzer’s proposal looks like (13). (i) Jones buttered the toast slowly. can be modified and quantified over by adjuncts. Whereas IL predicates are predicated of the individual directly. (In the default case. the toast. The trees in (13) just aim to capture Kratzer’s idea on the location of the external argument in the VP predicate. SL predicates are predicated of something else: the eventive variable proposed by Davidson (1967). to mention just a few. in the bathroom. 1995) claims that IL and SL predicates differ with respect to what they are predicated of.

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

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

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

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

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

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. 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. which is not found in the argument structure of the predicates per se. Chierchia (1995). which makes the predicate apply to the subject in all circumstances—that is.5 Summary of Section 2. the theoretical framework of Kratzer’s work. I have discarded those aspects of Kratzer’s and Chierchia’s approaches based on an assumption I consider to be incorrect—namely. for these authors. and the Davidsonian argument for Kratzer. On my view. is the subject of predication. that IL predicates are necessarily permanent predicates.3.” the scoping analysis does not add anything we did not have beforehand (although it may be compatible with it).1.1. Raposo and Uriagereka (1995) and Higginbotham and Ramchand (1996) argue that the distinction between both types of predicates is a matter of information structure.14 Individuals in Time I believe that having estar instead of ser in (35) is not a small detail. considers that all predicates involve such a Davidsonian argument. the semantic contribution of each copula cannot be ignored. to look like an inalterable property. Higginbotham and Ramchand (1996) also appeal to Kuroda’s (1972) analysis to account for the difference between SL and IL predicates. according to which all predicates involve a Davidsonian argument. If we end up concluding that estar itself denotes “linking to a particular situation. Very similarly to Raposo and Uriagereka (1995). 1995) argue that SL predicates are more complex predicates since they involve something extra: the realization function for Carlson. 2. Carlson (1977) and Kratzer (1988. at Logical Form. Framed in the neo-Davidsonian approach introduced in section 2. Higginbotham and Ramchand associate this situational variable with the SL constructions. Such an assumption leads Kratzer to suppose that the spatiotemporal argument can appear in the last step of the syntactic derivation.1 In this section I reviewed the most influential descriptions of IL predicates and their differentiation from SL predicates. He proposes that what distinguishes IL from SL predicates is that the latter involve an obligatory generic operator. what are usually called “SL predicates” are cases of “thetic” sentences. when interpretation is decided. these authors consider that SL predicates correspond to predications where an independent external situational variable. Cases where an external situation is the topic of the clause . Finally. however. In sum.

see also Rapoport 1987 and Rothstein 1995. I expand the discussion regarding what seems to be an adequate description for the IL/SL dichotomy.9 In Russian. A large part of the population uses Bambara as its mother tongue. I take up this issue also in chapter 7. I have introduced the idea that equating permanency of a property and ILhood is inaccurate. where the semantics of the two Spanish copulas are introduced. found across most of western Africa. When the predicate is an AP → copula “ka” b. 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. The Hebrew copula alternates between overt or null. Matushansky 2000).Individual-Level Predicates 15 are instances of SL predications. Spanish and Portuguese have two lexically different copular verbs. the postcopular NP (the predicate) alternates in case (nominative or instrumental) depending on whether the property is permanent (40) or nonpermanent or noninherent (41). the copula is null in the present tense but overt in the past tense (Kondrashova 1995. Bambara is one of the officially declared languages of Mali. Other languages. When the predicate is an NP → copula “don” c. In turn. . In the past-tense cases. 1996.2 When There Is More than One Copular Verb: Spanish Ser and Estar A good number of languages show copular alternations. hinging on whether the predicate expresses an inherent or definitional property of the subject (38) or not (39) (Greenberg 1994). ser and estar. (37) a. In the next section. such as Bambara (Koopman 1992). 2. those clauses where the topic is an individual are instances of IL predications. 9 For further discussion on the overtness of Hebrew copula.8 distinguish three types of copular verbs depending on the category the predicate belongs to (37). and as secondary language it is employed to communicate nationwide.

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

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

Further discussion is presented in chapter 7 (see section 7.6). (b) impossible modification by a degree adverb (48). and Demonte 1999. and those in (51) combine with both copular verbs. it is not gradable. Are origin adjectives qualifying or classifying? I will treat them as classifying because they behave as canonical classifying adjectives in the other three tests that Schmidt and Bache motivate.13. If the subject is a resultative nominal. Bosque and Picallo 1996. 14 Note. When the adjective is referred to the place where one is born. also.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. For more details about classifying adjectives. the copular clause sounds degraded: (i) the Cuban guy → the guy is Cuban (ii) the Cuban fight against the embargo → ??the fight is Cuban The reading where African is understood as a particular way of behaving is not relevant here. that classifying adjectives are excluded as copulative predicates (The presidential trip vs. However. that their acceptability as predicates in copular constructions seems to be confined to those cases where the subject is a physical entity. These tests are (a) impossible participation in comparatives (47). precisely. and (c) impossible participation in binary correlations (49).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. 15 13 . see Bosque 1993. *The trip was presidential). one of the properties that Schmidt (1972) and Bache (1978) provide to distinguish classifying adjectives from qualifying ones is.

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

all of the inflected versions in (55)–(58) that we can construe as paraphrases have to be done by using estar. Interestingly.1. Correspondingly. 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. Consider (59)–(62). Consider the next group of examples. Thus. the copula designing SL-hood. see section 2. pero está moreno (54) Pablo es Pablo ser-PRES-3SG very light-skinned but estar-PRES-3SG dark-skinned “Pablo is very light-skinned. (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. With the copular distinction in Spanish in mind. not ser— that is. where a perception verb (note) has a SC (NP + AP) as a complement. pero está muy guapo (53) Pablo no es Pablo not ser-PRES-3SG handsome but estar-PRES-3SG very handsome “Pablo is not handsome.1.20 Individuals in Time nada gracioso. throughout this work. I am dealing with an IL predicate. I will consider that.1). 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. where the pound symbol indicates when the paraphrases are not appropriate. whenever the copular verb is ser. (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” . as we already know from (51). 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. the copula estar will be considered a lexical sign of an SL predicate. but he looks very handsome” muy pálido.

which only combine with estar. and would therefore be expected to be grammatical as complements of notar. as in (63) and (64). as in (65) and (66).17 Adjectives such as desnudo or descalzo are described as “cut-short adjectives” (Bosque 1990). are not so. since the participial forms corresponding to the verbs they are related to are desnudado and descalzado respectively. 16 . Observe the following contrasts: Juan estaba/*era {desnudo/descalzo/alterado/preocupado/cansado} Juan estar/ser-PRET-IMPF-3SG {naked/agitated/worried/tired} “Juan was naked/agitated/worried/tired” (68) *Noté a Juan desnudo/descalzo I-noted Juan naked/barefoot (69) Noté a Juan {alterado/preocupado/cansado} I-noted Juan agitated/worried/tired (67) Although alterado/preocupado/cansado (strict participial adjectives) work as expected.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. whereas those combining with ser are excluded in such a position. since they combine with estar. (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. desnudo and descalzo16 do not. Chapter 7 contains more discussion on these adjectives. some adjectives.

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

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

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

the work is from 1945. Ser is conceived as a predicate parallel to write or admire. the results emerging from them are consistent with the distribution of the copulas in Spanish.23 This author argued that the property distinguishing ser and estar is imperfective versus perfective. How can we explain those cases where an adjective can combine with both copulas? Luján argues that.2. In turn. basically in the sense of ‘durative’ versus ‘punctual’. whose beginning and end are assumed (75). When the predicates are used to refer to imperfective states—that is. they select estar. and the [±perfective] feature in the copula. they must bear the feature “–[–perfective]. there are many other factors intervening in the semantics of these constructions. her account of the distribution of copular verbs (concretely in combination with adjectives) relies on two factors: the [±perfective] feature in the adjective. but rather that it makes a different contribution to temporal reference.2. ser expresses that a predicate applies to an individual during a stretch of time with no beginning or end assumed (76). According to Luján. . or at least one of them is. Following Querido (1976). When they refer to perfective states. Schmitt (1992). Examples (75) and (76) are Luján’s (1981:177).2. the two copulas differ in the nature of the temporal reference they make. 2.1 Luján (1981). tests based on perception verbs should not be taken as totally safe auxiliary tests to distinguish between IL and SL predicates.” As I understand Luján’s proposal. as contento ‘glad’. in general terms.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. a delimited process. I will discuss work by Luján (1981). Although. is already found in Gili Gaya 1961. She establishes a parallelism between estar and predicates like write a letter. 2. both undelimited predicates. Among those who have worked along these lines. and Fernández Leborans (1999). states where no beginning or end is assumed—they select ser. but can be either [+perfective] or [–perfective] (Luján 1981:175). if they combine with both. Estar expresses that a predicate is true of an individual for a delimited time period.Individual-Level Predicates 25 In sum. Thus. where a beginning or end is assumed. it is not that the copula ser is atemporal. (75) Perfective predicate A(x) at time tj 23 The cited date corresponds to the ninth edition. Luján (1981) claims that all adjectives are stative.

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Event types Sentences (1)–(8) illustrate the event classification as proposed by Vendler (1957. the latter two refer to a habit that does not have to end. walk around the park be green. belong. fall asleep denotes an endpoint. has neither duration nor endpoint. That is. push a cart. be born. awaken. Both are conceived as habitual happenings. 1994 and Pustejovsky 1988). However. In (6). that is why. collapse. be tall. but. the subject is involved in an action that is instantaneous and lacks an intrinsic endpoint. Instantaneous and nonculminating events swim. write novels.42 Individuals in Time like (4) works as an accomplishment. importantly. with an endpoint but no duration Actions with a culminating point that take duration to be completed. This predicate does not take time or involve any kind of “process”. so I will not discuss them further. and it is the habit that is not understood as delimited in time. Events with duration but no endpoint die. each building-a-house or reading-a-novel event is conceivable to reach an endpoint. Table 3. we also see that such templates cannot be a pure lexical matter. there is a slight difference between activity-typed events like walk or push a cart and others like build houses or read novels. However. the same with a bare plural as an internal argument instead (build houses) behaves as an activity (see especially Tenny 1987.1. Whereas the former two clearly refer to processes that need not ever end. 1989. knock Table 3. read a chapter Accomplishments Activities States Semelfactives No-actions that hold in time but do not take time. 1967) (in addition to the class proposed by Smith). Eventualities like this are named “states.” In (8). it is not an action in any sense. recognize. it lacks duration. in itself. be sick. explode walk to the beach.” The predicate in (7). Achievements Instantaneous events. love. They lack any kind of internal structure. build houses and read novels are conceived as different building-houses and reading-novels events distributed over time. Verbs involving such a property are called “achievements. fall sleep. build a house. Semelfactives are not of great relevance to this work. since they can be . know. They can be considered as complex events with internal steps toward an end. hate cough. sneeze. As noted earlier. arrive. such events are called “semelfactives” (Smith 1991).1 briefly characterizes each event type and provides some more examples. I return to the meaning of habituality in chapter 5.

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

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

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

which is the reason why they are compatible just with eventualities entailing a delimit point. *Pablo viajó en tres semanas Pablo traveled in three weeks c. 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.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 construyó una casa durante tres semanas 6 Pablo built a house for three weeks d. as (13) shows. As the ungrammaticality of (14) shows. Pablo viajó durante tres semanas Pablo traveled for three weeks c. *Pablo estuvo enfermo en tres semanas Pablo was ill in three weeks b. Pablo se dio cuenta de que su madre tenía razón en tres semanas Pablo realized that his mother was right in three weeks The following contexts also group achievements and accomplishments together. IL predicates can also combine with for-adverbials: (i) Juan fue camarero durante tres semanas Juan was a waiter for three weeks 6 Not all accomplishments are excluded with a for-adverbial: (i) The police jailed John for four weeks The for-adverbial refers to the period of time that the result state from jailing lasts. see Piñón 1999. (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. 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.5 (12) For + x time a. *Pablo se dio cuenta de que su madre tenía razón durante tres semanas Pablo realized that his mother was right for three weeks In turn.

accomplishments can be asserted to have taken place just when the process toward a culminating endpoint is over. Juan salió a pasear por el parque y todavía sigue paseando Juan went for a walk. *Juan llegó.1.3 Activities and Accomplishments versus States and Achievements.7 as (15) proves. this test and the for-adverbial one distinguish between permanent and nonpermanent properties. since there is no endpoint privileged. Pablo ha construido una casa Pablo has built a house 3. unlike activities and states.2. it is legitimate to assert that the event has taken place at any point of its process. With activities. *Juan arregló la lámpara. The test in (17) reveals one property that activities and accomplishments share. Pablo está paseando Pablo is walking b. and he is still walking b. (14) a. it is not licit to draw (16d) from (16c). rather than between IL and SL predicates. Juan estaba enfermo y todavía lo está Juan was sick and he still is The next test (Kenny 1963). and he is still arriving (15) a. . which is that they can be interrupted in their process and their subjects may have the control over the stopping of the event. (16) Perfect entailment from the progressive form a. Whereas it is licit to establish (16b) as an entailment from (16a). and he is still fixing it b. y todavía sigue arreglándola John fixed the lamp. y todavía sigue llegando Juan arrived. However. Pablo está construyendo una casa Pablo is building a house d. in (16).Event Classes and Individual-Level Predicates 47 point in accomplishments and achievements makes them be incompatible with an assertion of noncompletion. but not when it is ongoing. constitutes another diagnostic to differentiate accomplishments from 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. Pablo ha paseado Pablo has walked c.

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

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

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

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

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

may be used to discern properties of eventualities as well. the term “agent” clearly denotes ‘the argument that brings about an event’.” Also.Event Classes and Individual-Level Predicates 53 properties. and (22)–(25) also suggest that states and achievements have characteristics in common. and control. which gives bad results in contexts with adverbials such a deliberately. while volitionality and control are properties that just animate causers can embrace. but just holds. (26) (27) El portazo rompió el cristal The bang broke the glass Juan rompió el cristal Juan broke the glass . at least. it is important to distinguish the semantic features it contains and to establish whether there is some feature more basic than the other. three notions involved: causation. In this section.” and therefore “commanded. which diagnose agency. As mentioned before. agency tests work as event-type tests. I present a brief discussion about the agency tests above and about the notion of “agent” that I will observe in this work. Likewise. when I discuss the aspectual behavior of copular IL predicates. In agency.1 A Cluster of Notions Although. Actually. In particular. causation is the most basic notion involved in agency contexts. 3. agency tests detect eventualities that denote stuff that can be “brought about. In fact. not all controllers involve volition. I will deal with the different notions that get blended in the term “agent. stuff that is not performed. Vendler (1967) himself considered states and achievements a natural class. (20). agency tests separate states and achievements from activities and accomplishments.3. The results of (17). 3.2 A Brief Stop at “Agentivity” The set of tests in (22)–(25). First. although volition usually entails control. To begin this investigation. in principle. I will briefly introduce the relation between agents and event structure. To the extent that just some types of events can be undertaken and commanded. consider the classical minimal pair in (26) and (27). volitionality. I consider two aspects related to this.” Then. As will become clear.2. activities and accomplishments may be the most persistent co-patterning of the table.” The cleft with “do” and the test of command imperative detect eventualities that can be “done. In the following section. cannot hold with any willfulness on the subject’s behalf. Further discussion about the tests and about some concrete event types will come in section 3. there are.

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

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

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

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

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

*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. where the verb parar de does not have a subject that can be said to involve “control.” yet nevertheless is grammatical. Pablo paró de construir la casa Pablo stopped building the house c.c. the syntactic contexts reviewed in section 3. dynamic events yield grammatical sentences when they appear as complements of parar de ‘stop’ (see (17). show a certain correspondence between events that are agentive and those that can appear in the progressive.15 15 Tim Stowell (p. thus cannot be understood as agentive.Event Classes and Individual-Level Predicates 59 3.2. which suggests that only with dynamic events the subject has control over its ending. Assuming that progression in time defines dynamic events. repeated here). I will not consider this kind of case as a severe counterexample to my suggestion. used as tests to identify the different event types. In the second place. that is.1. (17) As a complement of parar de ‘stop’ a. Pablo paró de pasear Pablo stopped walking d.) points out the interesting case of Paró de llover ‘It stopped raining’. At this point I just want to notice the consonance of the results of the tests diagnosing dynamicity and of those diagnosing agency. Latin. . where their subject was a [+animate] DP such as Jupiter). Given the special status of the subject of meteorological verbs (cf. As the following contrasts show. which I take up later in the work. the verb parar de implies dynamic control by its subject. I will assume that the ability of an eventive verb to combine with the progressive is evidence that the event involves dynamicity. the infinitive is grammatical after parar de only when the subject can be agentive. We can prove that an event involves “internal movement” if it can express its progression in time. these events are typically activities and accomplishments. When the subject is [–animate] and. As is traditional. the sentences are ungrammatical. In the first place. *Pablo paró de amar a María Pablo stopped loving María b.2 Agents in Event Structure In this subsection I offer a few introductory words about the possible role of agents in event structure. we can conclude that only dynamic events can have agentive subjects.

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

Throughout the tests. given the correspondence between certain types of events and certain types of subjects. it seems that three properties can be connected: dynamicity. In the case of accomplishments. I examine the behavior of copular cases with ser. 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. which overlaps with aspectual notions. as Ritter and Rosen (2003) put it. We saw that not all arguments that could be labeled “agents” embrace all such properties. In turn. the ending is arbitrary and subject to the subject’s willingness. in the event structure. as mentioned earlier. as argued in chapter 2.2. through the battery of tests introduced above to diagnose event types. the end is inherent. In particular. Now.Event Classes and Individual-Level Predicates 61 (42) La rueda/Juan empezó a girar The wheel/Juan started rotating In sum. I will ignore the differences between imperfect and perfective aspect that exist in Spanish. and (with the considerations just made) control over the beginning of the event. although all volition agents involve control. (which. In the last part of the section. 1994) and Pustejovsky (1988) observed the role that objects have in the structure of events. the initiator of the event. not all controller agents involve volition. I looked briefly at some of the notions involved with the concept “agent”: causation. I introduced the issue of the role of agents in the eventive structure itself as “initiators” of the event. which has been very often linked to conscious volition only. Objects grammaticize telicity. is the lexical expression of IL-hood). 16 . 3. but I leave the discussion about their interaction with aktionsart to chapter 5. This will constitute the first step in the account of such predicates offered here.3 Summary of Section 3. but only animates can involve volition and control. 1989.2 This section includes remarks about the tests regarding the thematic property of agency. They are relevant to some points. it is natural to wonder whether the agentive argument plays any role in aspectual properties. 3. Specifically. This supposes a remark on the notion of agency. and control. in the case of activities.3 Aspectual Differences among Individual-Level Predicates This section explores a group of IL predicates from an aspectual perspective. volition. control over the ending. all must involve causation. Tenny (1987. Agentive subjects can grammaticize an aspectual role—namely.

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

which is why they are compatible only with eventualities entailing a delimit point.Event Classes and Individual-Level Predicates 63 remarks). on the other. it seems that these tests consistently separate the APs in two groups: Eskimo and blond. Since it appears that the AP in (c) behaves like an event.1. its acceptability in (46b) is expected. as reasoned in the literature. for-adverbials express how long someone has been engaged in a particular activity or a state has held of an individual. we see that one group comprises APs showing a state nature (a. It does so as well with IL predicates. pseudoclefts with happen are acceptable with states when present tense is involved. but the oddness in (46a) calls for further explanation. Even in the absence of the habitual adverbial. 3. As I introduced in chapter 2 (and will discuss further in chapter 6).2 Activities and States versus Achievements and Accomplishments. If we compare these results with the ones obtained in the previous section.2.1. In turn. on the one hand. when they hold of an individual. The test in (46) distinguishes. Consider a scenario like the following: (i) Lo que sucede/pasa es que su primo es esquimal. If. . Actually. the reasons for its ungrammaticality are independent from the test itself. in-adverbials refer to the period of time it takes to complete a certain event. 17 Much like do-pseudoclefts (see footnote 10). that the period that a stative IL adjectival predicate such as blond holds of the subject can be restricted with an adverbial (46b). which is why the limitation of its duration with a for + x time complement sounds weird. although it gives ungrammaticality for one of the predicates that above tested out as a state (Eskimo).18 states and activities from accomplishments and achievements.b)17 and the other comprises one kind of AP manifesting eventive properties (c). predicates like Eskimo refer to properties that.2. and cruel. These three tests show that not all APs in combination with ser behave the same way. as shown earlier. the next step is to find out exactly what kind of event it is. once again. Rather. (45c) gets a habitual interpretation. por eso sabe hablar Inuit “What happens is that her cousin is Eskimo. states are durative and for-adverbials are durative adverbials. a clear contrast emerges among different types of APs in combination with the copula ser. that is why she can speak Inuit” 18 As mentioned in section 3. It is interesting to note. as the contrasts between (46c) and (47c) show. 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.3. rather than a state. hold for that person’s entire lifetime.

2. 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.1. *Juan fue esquimal en dos horas Juan was an Eskimo in two hours b. which argues in favor of considering the predicate (ser cruel) as an activity.1. Once again. . In principle. a predicate like ser + AP would be unexpected as a complement of parar de ‘stop’. *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). (48) Perfect entailment from the progressive form a.3 Activities and Accomplishments versus States and Achievements.2.3.1. *Juan fue esquimal durante tres semanas Juan was an Eskimo for three weeks b. where there is no endpoint. However. (48b) is a licit entailment from (48a). yield a perfect entailment from the progressive form. such a predicate shows a nonstate patterning. Juan está siendo cruel con Pedro Juan is being cruel to Pedro b. the combination of ser and an AP such as cruel ‘cruel’ looks good. Juan fue rubio durante tres años Juan was blond for three years c. Juan ha sido cruel con Pedro Juan has been cruel to Pedro 3.3. *Juan fue rubio en dos horas Juan was blond in two hours c. As mentioned in section 3. since this is only possible with eventive predicates. and not accomplishments.64 Individuals in Time (46) For + x time a.2. As shown in section 3. only activities. tests like (49) divide activities and accomplishments from states and achievements.

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

?Juan casi fue cruel con su adversario Juan almost was cruel to his opponent b. *Juan casi fue esquimal Juan almost was an Eskimo 3.5 Agentivity Tests.3. whose combination with almost is impossible. 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 esquimal. since the unique licit entailment from (53a) is (53b). the latter is with activities. The next set of contrasts aim at testing whether the process that be cruel can be considered is agentive or not.20 Although (53a) does not sound extremely natural.1. 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. *Juan terminó de ser rubio Juan finished being blond c. like (53c). .1.3. (52) As a complement of finish a. All tests give the same results. except for cruel. The former reading is available with accomplishments.4 Accomplishments versus All the Rest. Juan no empezó a ser cruel con su adversario Juan did not start being cruel to his opponent c. Juan finished being an Eskimo b.4. confirm be cruel as an activity. fit in canonical agentive contexts. (53) Interpretative entailments from the adverb almost a.2. they clearly contrast with the other two. and (b) that the subject did not even start out the process. which also test accomplishments versus all the other eventualities. which suggests that dynamic properties are somehow involved in these predicates. it contrasts with other APs with ser. ?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. Although be cruel predicates do not give perfect results after finish. Tests (52) and (53) show some more contrasts consistent with the results from (49).66 Individuals in Time 3. as mentioned in section 3.1. 20 Recall that.

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

as shown by a number of tests.2. The leaf was falling down from the tree). we have learned that. there seems to be a group whose behavior patterns with canonical activities. *Lo que Juan hizo fue ser rubio What Juan did was be blond c. Among them. agency is notable. *Juan forzó a Pedro a ser rubio Juan forced Pedro to be blond c. This amounts to claiming that there is a group of IL cases that involve typical eventive properties. not all IL predicates belong to the category of states. ?Lo que Juan hizo fue ser cruel con su adversario What Juan did was be cruel to his opponent 3. *Juan lamentó haber sido esquimal Juan regretted to have been an Eskimo b. taking into account the properties of this group introduced in section 3.1.3 In this section we have learned two things.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). 21 .e.3.68 Individuals in Time (59) As a complement of persuade or force a.. I will attribute such an eventive-like behavior to the properties of the AP itself. among state be-predicates. differing from other authors’ account.2 Summary of Section 3. *Juan forzó a Pedro a ser esquimal Juan forced Pedro to be Eskimo b. not all of them can be considered alike. where it is attributed to the copular verb be instead. 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. only processes (i. as shown in the set of aspectual tests. Rather. Juan lamentó haber sido cruel con su adversario Juan regretted to have been cruel to his opponent (61) Pseudocleft with do a. There seems to be a group of APs that can be taken as Examples (60a. First.22 I dedicate detailed attention to these predicates that do not behave as states in chapter 4. activities and accomplishments) can be agentive. 22 Recall that although not all processes are necessary agentive (cf. *Juan lamentó haber sido rubio Juan regretted to have been blond c. *Lo que Juan hizo fue ser esquimal What Juan did was be Eskimo b. Second.

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

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

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

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

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

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

(xiv) *This table is missing a leg more and more (xv) *This table is getting to miss a leg 26 On my view. rather than stativity versus dynamicity. Consider the interpretation of the following example: (i) The leaf could fall → ‘the speaker believes it was possible the leaf fell’ → #‘the leaf had the capability of falling’ . the speaker expresses a guess about a particular activity she thinks the subject. Martha.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 just an epistemic reading. Piñón (1995) notes that modal verbs have different interpretations with dynamic and stative predicates. the simple present form being ungrammatical.26 I have stopped knowing her/getting to know her I have nothing to say. about examples from English such as those pointed out by van Voorst (1988). The examples below show where stative verbs have to appear in the progressive form in English. however. (xii) This table is missing a leg (xiii) *This table misses a leg This pair cannot be explained either by alluding to a scalar interpretation (cf. On the former. epistemic and deontic. however. Roughly described. The same can be said for accomplishments (80) and achievements (81). However. the modal has two meanings— namely. the oddness of (xiv)) or an inchoative one (xv). with an activity such as walk around the park. the different interpretation of predicates with modal verbs diagnoses agency versus lack of agency. (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. is usually involved in. On the latter the speaker expresses a command that Martha has to effectuate. the speaker makes a guess about Martha’s state.

------------walked-------said------------Utterance Time----However.3 Some Confusing Arguments in Defense of a State/Activity Distinction Finally. rather than due to inner-aspect properties. but they differ in their outer-aspect properties (imperfect/perfective)..76 Individuals in Time 3.27 whereas in the second one (perfective). -------------------///////said///////---------Utterance Time-------? ? be sick (83) a.4. 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. which gives the same results as stative predicates regarding its interpretation with modal verbs. they get the so-called past-shifted reading) (83). at least according to the Spanish data. 27 Imperfect forms can also yield a past-shifted interpretation (see chapter 5 for more discussion). It has been argued. (84) a. the temporal interpretation is simultaneous. Since I have not introduced outer-aspect notions yet.e. outer-aspect differences make a difference regarding the overlapping/past-shifted temporal contrast. In (i) there is a dynamic verb with a [–animate] subject (therefore incompatible with the existence of agency). 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. that events and states get a different temporal interpretation in (past) complement clauses. Whereas stative predicates in the subordinate clause are interpreted as overlapping the event of the main one (82). 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. This leads us to conclude that the different interpretation in combination with modals diagnoses agency versus lack thereof. in Spanish. Juan said that Peter walked around the park b. 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. this can be considered as an effect due to the properties of (outer) aspect. (82) a. whereas the perfective cannot yield a simultaneous reading. I will not focus on the technical treatment but limit myself to show that. John said that Peter was sick b. the stative predicate gets located in the past with respect to the main predicate. . In the first case. mainly for English.

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

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

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

different from the “regular copular be”.84 Individuals in Time presence of a constituent acting as a dynamicity inductor. based on Hale (1984). Specifically. Recall pairs like (4) and (5). this is an uneconomical move (and. They proposed a theta-assigning active copula for cases such as (1).1. 2004). I will defend the preposition heading the PP complement as an actual aspect head. 4. of cases such as (3). among others. as shown in chapter 3.1 The Two-Copulas Hypothesis Partee (1977) and Dowty (1979). 4. properties like agency cannot be attributed to the lexical entry of the verb but to the whole configuration. and empirically. it finds the obstacle that there is no evidence proving the actual existence of such two homophonous copulas. where the verb is the same but it is observed that agency is a property just of the animate DP subject. as I discussed in chapter 3. the two-copulas hypothesis is unappealing because it multiplies lexical entries with no independent evidence to support it.1 Previous Explanations of Nonstative Copular Clauses In this section I summarize previous proposals concerning the active behavior observed in certain copular clauses. (4) (5) John broke the window on purpose The ball broke the window (*on purpose) . 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. with a meaning close to act (2). Second. First. undesirable in itself). therefore. and Demirdache and Uribe-Etxebarría (2000. Stowell (1993). (1) (2) (3) John was cruel on purpose John acted cruelly John is tall (*on purpose) Among the limitations of this analysis. In section 4.7 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. Theoretically. rather than to the properties of the copular verb itself. semantically null. as has been maintained thus far in the literature. The last section summarizes the chapter. attributed the agentive properties displayed by adjectives such as cruel to a different lexical entry of the copula. there are the following three.

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

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

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

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

Note that the contrast between ser and estar in this case falls in line with the contrasts studied in chapter 2 (section 2. I take up the discussion related to the interpretation of the “infinitival argument” in Spanish and its status (as argument or adjunct).2. in chapter 6 (section 6.3). 4 For details of Kratzer’s (1988. I will make two observations. he argues that the action denoting argument is an overt version of the (Davidsonian) eventive argument.2).Aspectual Alternations in IL Predicates 89 Stowell (1991) establishes a correspondence between the monadic version of these adjectives and individual-level (IL) predication. (26) Juan es muy cruel con Pedro Juan ser-PRES-3SG to Pedro (27) Juan está muy cruel con Pedro Juan estar-PRES-3SG to Pedro (28) Juan está muy cruel con Pedro esta tarde.2. to admit there is an eventive argument that is implicit amounts to admitting that (25) is a SL predication. Following Kratzer (1988. Ser is completely grammatical. this proposal would entail that in all instances of APs followed by a relational PP. see chapter 2 (section 2. pero normalmente Juan estar-PRES-3SG very cruel to Pedro this evening but usually no es cruel con él not ser-PRES-3SG cruel to him “Juan is being very cruel to Pedro this evening. but he is not cruel to him usually” Later. the copular verb should be. 1995) account. which is the lexical marking for SL-hood. in his account. First. Specifically. (25) John was cruel to Peter Regarding Stowell’s analysis. I repeat one of the examples here in (29). and the contrast between the examples with ser and those with estar (27) can be captured by paraphrases such as the one in (28). However. (28)). necessarily. as (26) illustrates.1. 1995). the event argument represented by the infinitive is implicit in simpler copular cases such as the ones under investigation here.1) regarding the APs that can appear with either copula (cf. that is not case. . In Spanish. and between the dyadic one and stage-level (SL) predication.4 he takes it that the presence of such an argument is proper of SL predicates. According to Stowell. estar.

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

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

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

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

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

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

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”). whether or not the PP expresses an “affected argument” depends on the concrete type of action undertaken by the subject. offend someone else either by saying something unpleasant. 1994] and .3 Relational Mental Properties: The Relational PP Complement As just described. Likewise. 4. although it seems intuitively clear that the PP specifies the goal addressed by the subject’s action. to name just a few. 1988.3. focusing on two aspects: its interpretation and its optionality or obligatoriness. 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). can be considered as an “affected goal. abuse. harass. Thus. and such an action is qualified as cruel. Incidentally. and discuss all the consequences derivable from this. I investigate the nature of the PP complement. it is understood that Juan was the agent of an action that had Pedro as its affected argument.” This captures his intuition that such a DP refers to the individual that gets affected by the (underlying) action undertaken by the subject DP. 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. or by acting in a certain way. One can. even if we finally concluded that the PP is an “affected argument”. in italics in (58). for example. set on fire and bother. we would have still to discuss the aspectual role of such a PP regarding the delimitation of the event (bearing in mind the correlation between affectedness and event delimitation pointed out by Tenny [1987. In this section. I will show that its syntactic presence correlates with the aspectual patterning of the construction.96 Individuals in Time correlation with their aspectual characteristics. (58) Juan fue muy cruel con Pedro Juan was very cruel to Pedro The subject is understood as the agent of an action that has the DP inside the PP as the affected argument. offend or regale. 4.1 On the Interpretation of the Relational PP Stowell (1991) suggests that the DP inside the relational PP.

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

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

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

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

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

Some authors have claimed that consider selects for SCs containing IL predicates. at the same time.102 Individuals in Time (80) Juan no es cruel. The first scenario where cruel cannot appear followed by a complement PP is small clause (SC) complements of epistemic verbs such as consider. such as Demonte and Masullo (1999). without any apparent contradiction. as other authors. in fact. However. Sentences (83) and (84) are based on examples from Demonte and Masullo (1999). as their obligatory combination in Spanish with the copula estar proves: de vacaciones (85) Estoy/*soy estar/ser-PRES-1SG on vacation “I am on vacation” . (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. 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. From these cases. (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. SCs complement of consider can perfectly have SL predicates. have pointed out. 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. Stowell (1991) was right in considering the simple versions of relational MPs as IL and the ones accompanied by the PP as SL. and rejects SL predicates. This could in principle lead us to think that. To show that the PP is optional. In the remainder of the chapter I will leave aside these (somewhat confusing) interpretive contrasts. be claiming that the person is not cruel. I will concentrate on some syntactic scenarios where an overt relational PP cannot appear.

Consider now the following examples from English. (87) *Considero a Pedro padre I consider Pedro a father (88) Considero a Pedro un buen padre I consider Pedro a good father However. the predicate should refer to something which can be a matter of subjective opinion. a PP involving an arbitrary nominal could be syntactically noticed. this does not seem to give us the reason of the contrast in (81) and (82). even in the case where it appears alone. since. The contrast found in (81) and (82) suggests that the presence of the PP makes a difference. there should not be any difference between cruel by itself and “cruel + PP”. I conclude. then. In the first place. which leads to the conclusion that it is not always there. if cruel-type APs always involved a PP complement. (89) *I consider John to build houses (90) *I consider John to walk around the park (91) I consider John to know mathematics (92) I consider John to be on vacation I take the contrast between (89) and (90) versus (91) and (92) to mean that consider selects for stative predicates and rejects eventive ones. Surely. such as activities or accomplishments. one can agree as a matter of opinion about whether someone is cruel to someone else or not. when the DP is inanimate the relational PP is excluded. for example. 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. As observed in the examples below. Consider the contrasts between (87) and (88). where just some infinitival predicates are possible as predicates of consider SCs. In sum. 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.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. .

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

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

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

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

whose contrast has been debated in the literature (see section 3. we would be forced to assume that each aspectual behavior of adjectival copular cases (stative and nonstative) corresponds to different (but homophonous) lexical items. María tocó la verja (dos veces) María touched the fence (twice) The alternation here is of states versus activities.4). That is. by definition. This perspective cannot capture any of the other properties (interpretation of the subject.1 Lexicalist and Logico-Semantic Approaches The observed correlations among the presence of the PP relational complement.5.5 Justifying the Approach 4.). 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. La pared tocaba la verja The wall touched the fence b. From a lexical point of view. 4. according to which aspect properties are fully encoded in the lexical items. 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. kind. 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. it does not allude to the other elements present in the syntactic configuration. I begin by discussing which approach to event representation can make sense of the observed correlations among the syntactic presence of a constituent. However. mean. In the remainder of the chapter I will be concerned with such a contrast in the realm of adjectival copular clauses. (107) El delantero avanzó hasta la portería The foremost advanced until the goal (108) a. etc. two different entries (stative and dynamic) for each adjective (cruel.2) the contrast between animate and inanimate DPs in pairs like (109) and (110).” Recall in this regard from chapter 3 (section 3. syntactic presence of a PP) as correlations since.108 Individuals in Time tioned thus far. I argued that such a contrast is grammatically relevant. since there are syntactic and semantic facts suggesting its actual substantiation. the thematic interpretation of the subject and particular aspectual properties. . In (107) it is a PP which is exerting a crucial aspectual role in making the event an accomplishment. both atelic eventualities.

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

1989.2 Syntactic Approaches 4. and the presence or absence of other elements in the sentence (the PP complement in our cases) that. correlate with the aspectual outfits of the construction. among others. each entity corresponds to an aspectual notion. 1994) and van Voorst (1988) elaborate both on the observation that arguments grammaticize the points that give the temporal contour of an event. its syntactic position and its role in the determination of the aspectual properties of the sentence.” Second. as captured in the Event Correspondence Rule (116). Dowty (1979).5. As discussed earlier (see section 3. showed the relevance of arguments in the inner aspect properties of a sentence. 1994). since. Tenny (1987. Tenny (1987.1). 1989.2. First. this approach does not establish any relationship between the aspectual term (Hold). authors such as Verkuyl (1972). This perspective does not make any direct prediction about the correlation between the aspectual properties of a VP and the thematic properties of the DPs.5. there is no obvious way to distinguish between the two constructions in point (stative and dynamic copular clauses).” given the fact that both are atelic and do not “culminate. which opens the possibility of establishing a direct relationship between the thematic properties of the arguments present in a sentence.110 Individuals in Time This approach leaves several questions unanswered. As van Voorst puts it. the dynamic or stative properties. (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 . in principle.1. both would contain the term “hold. these authors emphasized the role of the internal object as an event delimiter. 4.1 Event Roles. and van Voorst (1988). 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. In particular. arguably.

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

which I have argued are in strict correspondence. the event-role perspective predicts the mapping of different arguments to particular syntactic positions more accurately than the UTAH.112 Individuals in Time In these two sentences the theta-roles are the same. van Voorst’s and Tenny’s proposals do not provide a direct way to capture the correlation between the presence of a PP and the aspectual behavior as an activity. 1998. the systematic correlation between object and delimitation and subject and origin). 2005). and Sanz (2000) elaborates on the idea that argument structure is licensed by functional syntactic structure. different lexical items have different ways to project. 2000. In a sense. but not so their syntactic positions. although both Tenny and van Voorst explicitly acknowledge the compositionality of events. Compare (120) and (121) with (122) and (123). Benua and Borer (1996). this assumption makes them similar to lexical approaches. 2000). (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). therefore. . However. the work of authors such as Borer (1994. occupies the subject position. As Rosen (1999) points out. appears in the subject position. Furthermore. for them. Although these event-role approaches provide accurate insights in important respects (the fact that arguments can affect the event type. For our present concerns.5. and. she does nothing that can be called buying in (122). The car is the theme in both cases. Note further that the NPs play different event roles in each one of them. The position of thematic roles is not restricted. However. and. 4. he does nothing that can be called selling in (123). they cannot capture aspectual alternations such as the one we are interested here in a natural way. these approaches would have to assume that each aspectual behavior corresponds to different adjectival projections. Incorporating the insights from the lexical argument-oriented approaches just reviewed. a natural account of such compositionality cannot be derived from their lexicalbased approaches. Ritter and Rosen (1996.2. In other words. In sum.2 Syntactic Structure as Event Structure. but the position of event-roles seems to be so. as such. regarding the stative/dynamic contrast in IL copular clauses. The old lady instigates the selling in (123). 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. Travis (1994. 2003). 1998. correspondingly. since.

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

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

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

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

3. To capture (a)—that is. the dynamic aspectual behavior) actually resides in the preposition itself. In support of this proposal. 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. As I intimated before. based on Hale (1984). the following two properties: (a) When there is a relational PP. the DP subject has to involve a particular set of properties. I will address two theoretical points. To capture (b). 4. among others. and I will also be concerned with the aspectual treatment of the arguments expressed via a PP. 2004). that prepositions can be conceived of as heads encoding aspect properties. the construction has characteristics proper of processes. As I advanced. Stowell (1993). That is. as soon as they combine with a (relational) PP. since their intervention in the event structure is not obvious (given that they are oblique arguments). I argue.6.6.) in the active cruel-type constructions is a projection (the PP) that combines with the “basic” structure of the AP. I will discuss the syntactic treatment of the opposition state/activity.1 Prepositions as Aspect Encoders I have shown that cruel-type APs. which gains dynamic properties from the preposition heading the PP denoting the DP goal. behave as activities. and Demirdache and UribeEtxebarría (1997. I will take such a correlation seriously and I will propose that the source of the aspectual properties (specifically. all of them very likely emerging from animacy itself: the DP has to be able to behave as a controller agent. the PP makes them a dynamic predicate. Let us begin by discussing point (b) concerning the aspectual contribution of the preposition. but it correlates with the existence of the PP relational complement. As I develop the proposal. I propose that the dynamic properties become a part of the construction through the preposition heading the PP. the aspectual alternation (state/activity) is not unpredictable at all. it has become clear that the account for the cruel-type constructions has to capture. 2000. Point (a) is discussed in section 4. I propose that the source of dynamicity and its related properties (agency etc. From the description in the previous sections.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. at least. I will argue that the stative version of the cruel-type APs corresponds to their most “basic” structure. .Aspectual Alternations in IL Predicates 117 4. (b) When an animate DP subject and the PP are present.

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

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

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

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

In (149). he characterizes tense distinctions as relations of priority. When I entered the room. when he entered the room. When I entered the room. Some clarifying examples appear below. Maria was cradling the baby [----------////////------------b. the kind of the relations being either of central coincidence (progressive aspect) or noncentral coincidence (perfect and prospective aspect). and in (149c). Maria had cradled the baby [----------------]///////// c. Progressive AspP 2 b.” The discontinuous line represents the event and the slashes the TT. Following Reichenbach (1947). Demirdache and Uribe-Etxebarría (1997. 1996) proposal that the semantics of tenses can be described in analogy to prepositions like before or after. which establish a temporal ordering between time-denoting arguments. It is conceived as “contained” in the span of time the event of cradling takes place in (149a). In (149b) the TT is located “after” the span of time the event of cradling takes place. 2000. Maria was going to cradle the baby //////////[------------The trees below schematize Demirdache and Uribe-Etxebarría’s (1997. When I entered the room. . 2004) argue that Aspect can be conceived as a head establishing a temporal ordering relationship between two times. Demirdache and Uribe-Etxebarría argue that Aspect can be conceived in parallel terms. This time is the Topic Time (TT). 2000) proposal inspired in Hale 198424 and Stowell 1993.122 Individuals in Time Demirdache and Uribe-Etxebarría take as a point of departure Stowell’s (1993. the time the speaker is talking about is a concrete time. (150) a. Based on Klein’s (1994. namely. (149) a. the TT is captured “before. 1995) idea that Aspect orders the event time with respect to the time the speaker refers to in the sentence (the “Topic Time” or “Assertion Time”). Perfect AspP 2 c. simultaneity and posteriority describable in terms of central/noncentral coincidence between the speech time and the Reference Time and aspect distinctions in terms of central/noncentral coincidence between the Reference Time and the Event Time. 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.

6. and.1 Con: A Locative and a Directional Preposition. ¡corre con tu papa! Pablito. con-PPs can also appear with motion verbs. con ‘with’. I am going to argue that the preposition introducing the (affected) goal carries inner aspectual properties affecting the aspectual nature of the whole construction.6. and assuming with Demirdache and Uribe-Etxebarría that all temporal relations can be reduced to the same set of primitives. in these cases the preposition con is in complementary distribution with the directional preposition a ‘to’ when the DP complement is [– animate] and is interpreted as a location: . I am going to argue that the aspectual shift from state to activity found in the cruel-type constructions is due to the presence of the PP relational complement. The preposition introducing the PP in Spanish is. the preposition con. 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. This preposition is associated with stative locational meanings.Aspectual Alternations in IL Predicates 123 Assuming that aspectual meaning is reducible to prepositions (and. conversely. as seen in the examples above. 4. run with your daddy! (154) Me volví con mis padres I went-back with my parents “I went back to my parent’s place” Importantly. 4. paying special attention to the APs I am taking care of in this chapter. I propose that this is the case.2.2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor In this section I am going to describe the preposition heading the PP that refers to the (affected) goal in cruel-type construals in Spanish. denoting the goal. that prepositions have aspectual meaning). like in examples such as the following: (153) Pablito. Preposition con establishes a central coincidence relation between the figure and the ground (in the terms introduced before).

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

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.

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

(The bullet stands for the figure and the little house for the ground. 2004) proposed that centripetal noncentral prepositions translate as “before” in the temporal domain. that is.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. associated with ‘future’ in the realm of Tense and with ‘prospective’ in the domain of outer aspect. it designates the movement of a figure toward a place or ground. The two terms are put under a specific relationship according to the meaning of the preposition. (181) Juan went to Pedro In the terms of figure and ground presented before.6.1. 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. (182) summarizes the description of the preposition. it locates the figure in its trajectory toward the ground. we saw that Demirdache and Uribe-Etxebarría (1997. As a directional preposition. I consider that the relation between the DP subject Juan and the goal Pedro is not very different from the one existing between Juan and Pedro in (181). 2000. and (183) represents it graphically. the discontinuous dotted line for the trajectory. …● ⌂ In section 4. The two of them are related by a preposition meaning ‘toward’. They are. 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 . therefore.

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

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

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

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

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

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

First. In the set of cases above. a stative SC (with the bare AP) is accepted. The aim of this section is to analyze whether this is borne out. when the PP is present. and Rothstein (1999) defends that the copular verb maps states onto eventuality-denoting predicates. can take . volverse ‘become’. as a complement of hacer ‘make’ and as a complement of volver(se) ‘become’. the constructions get degraded in some cases. but when the SC contains the adjective plus the PP (197) it gets degraded. (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. Partee (1977) and Dowty (1979) argue for a homophonous agentive copula with a meaning close to act. three things. in all the examples. however. we observe.7.1 “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’. it is not the case that “cruel + PP” fits in all cases. However. 4.138 Individuals in Time chapter. traditionally considered very close to be. In my proposal. Second. at least. a verb traditionally considered “pseudocopulative. the verb seem. In the following paragraphs I attempt to explain the contrast between allowing and disallowing an “AP + PP” SC. This hypothesis makes the prediction that such active properties should be retained when the taking verb is other than the copula. can take an SC where the AP appears by itself (196).” with an intrinsic inchoative meaning. As the examples show. the activity-like properties are derived from the properties of the (adjectival) predicate heading the small clause (SC) taken by the copular verb.

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

(iii) a. I argue. 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.140 Individuals in Time reason.35 Regarding inchoatives like become. (The judgments are from Spanish). the aspectual property they involve is. it cannot “lower” at Logical Form. “seem to me” in (iiib). Therefore. Although I do not have an explanation to it now. (190) with the plain adjective. Tim Stowell (p. the same one I propose for the PP in these cases (both involve process properties). whereas the simpler one is more defective (looking very close to a modal). for example. and the scopal ambiguity disappears. precisely. for example. they are aspectually compatible. in English. whereas the AP with the PP (an activity SC) gets degraded. When a QP rises from the lower subject position across. therefore behaving as a state. b. (208) √ cruel 2 volverse cruel 2 cruel con 2 Juan con 35 con 2 Pedro It is worthwhile to notice that parecer ‘seem’ in its combination with a clitic in Spanish (with a PP in English) gives different results. María me parece cruel con Juan Maria me CL-seems cruel with Juan “Maria seems cruel to me with Juan” (ii) ?María parece cruel con Pablo Maria seems cruel to Pablo (i) In this respect.c. the contrast between (i) and (ii) is a piece of evidence suggesting that parecer has different properties than “clitic-parecer”. I will bring up two remarks suggesting that seem and “seem + PP” are actually different. is accepted. quantifier lowering is blocked across “seem + PP”. (iv) ??María pareció cruel Maria seemed cruel (v) ??Maria ha parecido cruel Maria has seemed cruel . This hypothesis could explain. and does not sound natural in other tenses other than the present. even in the simple “parecer + AP” cases. why the clitic forms can be found in all tenses.) observes that.

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

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

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

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

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

.

I show that the perfective form does not contribute quantity (i. Since.Chapter 5 Outer Aspect and Individual-Level Predicates In the two previous chapters I dealt with inner-aspect properties. telicity) to the construction and the imperfect is compatible with quantity predicates (telic predicates).e. the external argument of Aspect is the internal argument of Tense.1). and the progressive.2 that aspect viewpoints also involve quantification over occasions. I will first describe the meaning of different aspectual viewpoints and the way outer aspect works (section 5. I argue that Aspect is a predicate (located structurally higher than Quantity) that establishes an ordering relationship between two interval arguments. and its absence amounts to leaving the predicate homogeneous. the perfective.1 Besides the ordering component just mentioned. I am going to deal now with another aspectual realm. Following Klein (1994. 1995) and Demirdache and Uribe-Etxebarría (2000). in Spanish.. Its presence makes the predicate heterogeneous. Focusing on the imperfect and perfective. through the functional projection of Quantity. 1 The particulars of temporal interpretation are the topic of the next chapter. for example. Following Borer (2005).. Throughout the discussion I will focus on three particular viewpoints: the imperfect. I assumed that different event types are decided through functional structure. I conclude that. I propose in section 5. outer-aspect properties do not fulfill quantity (i. namely. In section 5.3. according to Demirdache and Uribe-Etxebarría’s proposal. I will also present how Tense works here. I also introduced some discussion about the difference between the two types of homogeneous predicates—states and activities. outer aspect. which I will capture through a quantifier inside the Aspect composition. 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. 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. . Centered on individual-level predicates.e. inner aspect) properties. The main points I focus on are: (a) the description of a number of outer-aspect forms (“viewpoints”) and (b) the relations and restrictions that can be argued to exist or not exist between inner and outer aspect. I discuss the relationship between inner and outer aspect. specifically. so that the complete articulation of Tense and Aspect and their parallel behavior can be easily contemplated.

atelic). aspect properties can serve as indicators of temporal properties (a perfect action is computed as finished and. It turns out that this cannot be described in terms of presence or absence of quantity. I show that. therefore. The idea is depicted in (1). Therefore. which I mentioned briefly in chapter 4 (section 4.148 Individuals in Time In section 5. 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. However. Since the line of this work is not formal semantics I do not discuss them here.3 Authors such as Zagona (1990) and Stowell (1993. In particular. Reichenbach gave a formalization of tenses in natural languages based on three intervals (Reference Time. 3 Anteriority. and inclusion are semantically defined as partial order relations fulfilling a number of properties. Stowell (1993.1 Tense.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. specifically. Lillooet Salish).1. as Tense and Aspect are considered ordering predicates.5. posteriority.. by the same token. and Wall 1993. I show that states are compatible even with the perfective. inner aspect can be as well. in some languages (e. this being why the eventuality is computed as “dynamic and unbounded” (i. as something that took place in the past). that establish centripetal noncentral coincidence relations between a figure and a ground.6. For an introduction of such notions.. the interested reader is referred to Partee.1). 5. These predicates translate into the temporal domain as ‘before’. At that point. in Spanish at least. posteriority. and Eventuality Time). ter Meulen.e. the progressive seems to impose restrictions on the type of eventuality. Consistent with the conclusion that viewpoints do not interfere with quantity properties. Speech Time. but in dynamicity versus lack of dynamicity. and simultaneity). I tried to show that all aspectual properties could be reduced to the same set of primitives. I turn my attention to adjectival IL predicates and show the different interpretations available with each type of viewpoint mentioned.g. 2 . I argued that dynamicity properties (that activities involve) can be described through prepositions.2 One of the facts I alluded to was that. for example) and. 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. I argued that inner aspectual properties could also be conceived in ordering terms. inner-aspect characterization may work as an aspect indicator (an accomplishment can be taken as a perfect action. based on Zagona 1990. like to. and the possible relationships that can be established among them (anteriority. 5.

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

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

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

5. for example. makes them visible for the receiver to be able to catch them in a photograph. According to (9). as Klein (1994) pointed out: Tense does not order the ET with respect to RT. However. This leads to a big conclusion.152 Individuals in Time as an ordering predicate. The former would correspond with readings of the sort in (10). I assume that the element that does this job is Aspect. 1996) work. by focusing objects. he intends (10) rather than (9). That time is in the past. (8) María estaba lavando el coche Maria was washing the car (9) Maria was involved in the task of washing the car at some point in the past. The observation that Tense can refer to a concrete interval traces back to Partee (1973. The lens of a camera.1.2 The Internal Argument of Tense: The External Argument of Aspect. the total interval during which María has been washing the car is left in the dark. but the interval the speaker is referring to. (10) There is a specific time at which Maria was involved in the task of washing the car. 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. at that time. be paraphrased as (9) or (10). María was involved in the task of washing the car. The analogy between DPs and ZPs enables Stowell to distinguish between “definite” and “indefinite” (with just pure existential force) ZPs. Following Klein (1994). Now. Smith (1991) makes an insightful comparison between aspect and the lens of a camera. (10) seems the most accurate paraphrase of (8). when someone utters (8). Aspect as an Ordering Predicate The temporal interpretation of a sentence like (8) can. the speaker is asserting that María has been washing the car some time in her life. That is. Depending on the focus . 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. Between the two. (10) wants to capture a different assertion: the speaker has a concrete time in his mind and he is asserting that. 1984b) and was also noted in Stowell’s (1993. in principle. It orders the time the speaker talks about with respect to a time of reference. In (8). I call the interval the sentence is about (or “the interval the speaker talks about”) “Topic Time” (TT). whereas the latter with those along the lines of (9). the obvious question is: what tool picks out the interval the sentence makes an assertion about? Following Smith (1991) and Klein (1994).

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

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

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

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

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

besides ordering. aspectual viewpoints. also contain information about the number of occasions an eventuality occurs.2. 5. The number of occasions the progressive and the perfective form allude to is just |1|. and the progressive.158 Individuals in Time of Quantity. The values of such a quantifier. which I analyze as complex phrases containing a quantifier over occasions and ordering information. the functional projection which makes the eventuality heterogeneous and whose absence leaves it homogeneous. the nonhabitual or “continuous” imperfect. As I will repeat later in section 5. Q<occ> can also have an existential value. here in section 5. However. what the relationship is between the number of occasions alluded to by each viewpoint and the quantity nature of predicates (that is. besides the ordering component. a quantificational component. as the reader may have conjectured already. as shown in (25). I will consider two points—first. can be either |1| or |>1|.5. how many occasions each viewpoint alludes to. The continuous imperfect does not count the number of occasions a determined predicate takes place. (25) ----------------------x---------------------(26) ------x------x--------x-------x---------x-In the spirit of Verkuyl (1999). In this latter regard. It appears .1 Quantifying over Occasions As just mentioned. the three forms differ in the number of occasions they allude to. (22) Pablo estaba paseando por el parque Pablo was walk-ing around the park (23) Pablo paseó por el parque Pablo walk-PRET-PERF-3SG around the park (24) Pablo paseaba por el parque Pablo walk-PRET-IMPF-3SG around the park These sentences all refer to a (homogeneous) eventuality that took place in the past. However. who observed that the difference between the progressive and the habitual rests on the number of occasions they refer to. In this section I am chiefly concerned with the perfective. giving rise to another interpretation. inner aspect. I address these points in turn. and second. whose head is a quantifier over occasions (Q<occ>). as in (26). the imperfect in (24) refers to a plural number of occasions that can be described as greater than one (|>1|).2 and next in section 5.3. I propose that aspectual viewpoints involve.1. the habitual imperfect. Let me start by considering sentences (22)–(24).

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

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

the TT is contained in the period of time when Pablo was walking. Likewise. (36) -----------xxxxx[/]xxxxxx---------------↑ when I saw him (37) ------x------[x--------x-------x]---------x-ie his teens In (36). Although the habitual viewpoint will be more extensively discussed shortly. the progressive and the habitual share the same ordering component. That is. that of “containing.Outer Aspect and Individual-Level Predicates 161 una casa María tenía Maria have-PRET-IMPF-3SG a house (32) *María estaba teniendo una casa Maria was having a house (33) *Habitualmente María tenía una casa habitually Maria have-PRET-IMPF-3SG a house “Maria used to have a house” (31) As indicated in (27). the period of his teens is included in the period of time during which Pablo used.” The TT in (34) is identified by the temporal clause when I saw him. Pablo was walking around the park” por el parque (35) En su adolescencia. Pablo paseaba in his teens Pablo walk-PRET-IMPF-3SG around the park “In his teens. the TT interval in his teens of (37) is included in the period of time during which Pablo used to walk in the park. . to go walking in the park. consider the following contrast to illustrate this shared characteristic: (34) Cuando lo vi. and maybe still use in the present. Pablo used to walk around the park” The relationship established between the TT and the ET is the same in both cases—namely. Pablo estaba paseando por el parque when him saw Pablo was walk-PROG around the park “When I saw him. 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. for example). the TT in (35) is specified in the phrase in his teens.

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

as described in (27c). Pablo estaba llamando a la puerta tres veces In the two occasions I heard him. whereas the progressive and the perfective (in the absence of adverbial modification) refer to a singular occasion. However. define habituality as an iteration of instances of a given eventuality distributed in time. In this subsection I deal with the properties of one such quantifier that gives rise to habitual interpretation.17 among others. Specifically. Proportion. and propose that it is expressed through functional quantificational structure. The speaker’s assumptions about the actual number of instances of a habitually quantified event depends on heterogeneous (pragmatic) factors. I assume that the meaning of habituality is lexically expressed in quantifier adverbials such as habitually or usually.1 Iteration. 5. 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. the exact number of event instances is not specified.Outer Aspect and Individual-Level Predicates 163 progressive.2. (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.2 The Habitual Interpretation: Iteration. Pablo was knocking at the door three times. Consider the contrast between (40) and (41) as a starting point. I will consider that habituality is encoded in the “quantificational floor” of Aspect. 5. (42) Juan smokes four times a year. Now suppose we interpret (40) and (41) against the following scenarios in (42) and (43). Following the quantificational-based perspective defended by Verkuyl. I consider that habituality corresponds to a quantifier over occasions denoting a plurality of occasions of a particular eventuality. . the habitual viewpoint refers to a plural number of occasions by itself. and Systematicity Verkuyl (1999) and Stowell (2000).2.2. As just mentioned. (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.

(45) does not properly describe the situation. as such.164 Individuals in Time Although the implicated number of times is the same (four). is affected by contextual parameters (Westerståhl 1984). In particular. In other words. we conclude that the crucial notion in describing many resides in the “average ratio. (41) is very likely to be judged as appropriate. If the average ratio of right-handed people is (statistically) higher than that.2.2 Proportion. . two intervening factors: (a) iteration. it seems that. (40) does not reach the category of “habitual. Specifically. (44) should be judged as a sentence that properly describes the situation. Whereas (40) is very likely to be judged as inappropriate. given the situations of (42) and (43). and Wall (1993). The number of times considered “average” is established taking into account extralinguistic information.2. Sentence (44) would be judged inappropriate for such a scenario. note. As Westerståhl (1984) and Partee. ter Meulen. a habitual sentence is considered appropriate for depicting such a scenario. Suppose now that the number of right-handed people in the class is also five. when judging whether a habitual form is appropriate. what counts as “many” depends on different contextual parameters. ter Meulen. (40) and (41) are judged differently.” which is established by external information. (46) expresses this formally. that in the determination of “habituality” there are. whose definition heavily depends on contextual basis. 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. How does such pragmatic information get integrated into the aspectual encoding? I propose that aspect Q<occ> behaves like regular quantifiers and.” In turn. (44) Many students got an A (45) Many students are right-handed Suppose that the set under consideration is a class of 20 students. among others. at least. Since the number of times people usually smoke tobacco in our culture highly exceeds four times a year. and Wall. It seems. Thus. and (b) a certain proportion with respect to the number of times the action at stake is usually performed. therefore. since the eventualities take place more than once. I propose that habitual Q<occ> behaves similarly to canonical quantifiers such as many. If five is more than the average ratio of students that get an A. The following pair is from Partee. 5. some notion of “average” regarding the number of occasions that an action is performed is taken into account. I will review the behavior of this quantifier in the nominal realm briefly. (41) is judged as appropriate for the situation in (43). Suppose further that the number of students who got an A is five.

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

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

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

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

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

(84) and (85). In contrast. Their telic nature is tested by the appropriateness of the modifier in x time. either in the perfective or imperfect form. 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” . 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. 5. which points to the conclusion that the viewpoint does not affect in any sense the nature of the predicate. and offer a general overview of the relationships in the temporal system. Bearing in mind this telicity proof. which tests the predicate swim as atelic. atelic) predicates. To begin.174 Individuals in Time relation between inner and outer aspect. and (b) the co-occurrence of perfective viewpoint and homogeneous (nonquantity. are odd in the presence of such a modifier. I will discuss the following points: (a) the compatibility of telic (quantity) predicates and habituality.3.1 Habitual Heterogeneous Predications The Spanish data show that the imperfect-habitual viewpoint is compatible with telic predicates. prepare the meal and write the report.

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

2 Perfective Homogeneous Predications I turn my attention now to homogeneous predicates appearing in perfective form. I argue that a perfective form does not trigger telicity effects on predicates. As discussed back in chapter 3 (section 3. The meaning of these sentences is that Pablo was once involved in the task of swimming (91) and walking (92). de Swart (1998). I will argue that. who conceive the perfective as a modifier that makes an eventuality telic. a perfective form does not make a homogeneous predicate heterogeneous. he has had that car at every single moment between 1974 and 1985. (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. a perfective form does not interfere with the quantity nature of the predicate. It is precisely this fact that has led many authors to argue that perfective is an “event type coercer” in the sense that it has the power to turn homogeneous predicates into telic ones. In other words. sentences such as these show that the presence of the perfective viewpoint does not contribute telicity. at least in Spanish.176 Individuals in Time 5.3. Contrary to Mourelatos (1978). In the same vein as before. it does not mean that he has been walking at every single moment of that interval.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.” as mentioned in the previous section. With the perfective. unlike states. Piñón (1995). In contrast. activities are not 100 percent homogeneous at a micro-level. draw a conclusion in the same direction. If we say John walked from 2 to 3. 24 23 . However.24 as subinterval and additivity proofs show. Bach (1986). such as Bertinetto (2000). which leads to the inference of “completeness” and “boundedness.4). if we say John had that car from 1974 to 1985. (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. and others.

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

(98) ZP 2 Zi VP 2 ei VP (99) AspP 2 TT Asp′ 2 Asp0 ZP (after/before/in) 2 VP Z0 (|1|. 26 This aspect form is involved with stative SL predicates (She was sick). ∃) 2 e VP Thus. 25 . in Spanish. independently from the kind of predicate. I would like to propose that the quantifier over occasions of the eventuality is the ZP itself (the internal argument for T in Stowell’s work). I argued that.” perfective or progressive.4. In the last section. following Stowell (1993. and makes the prediction that the quantifiers (|1|.1 for further discussion. I assume to be present with every type of predicate (SL and IL). the ordering component takes the quantified eventuality and orders it with respect to the TT interval. as Kratzer (1988) originally proposed. 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). When the quantifier denotes a number roughly describable as “greater than one.26 The quantifier over occasions binds the eventive variable (e) which. Specifically. 1996). This unifies the working of Aspect. but these predicates are also compatible with the habitual quantifier (She was sick usually).178 Individuals in Time cation) is depicted in (99). where there are no occasions to enumerate (think of stative homogeneous predicates such as Peter is tall). |>1|. if “one. I used Spanish habitual and perfective data to show that aspectual viewpoints do not have any impact on the inner-aspect properties of predicates.” habitual imperfect is at stake. ∃) should be possible with any type of predicate. |>1|. this is borne out.25 The existential quantifier corresponds to the continuous imperfect. As will be shown in section 5.5. which does not make the sentence habitual. Elaborating on Stowell’s (1993) proposal. and not only with SL ones. See section 5. presented in (3) above (repeated here as (98)) about the internal constituency of ZPs.

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

Juan está siendo cruel con Pedro Juan is ser-ing cruel to Pedro “Juan is being cruel to Pedro” b. Consider the following cases: esquimal (103) Pablo era Pablo ser-IMPF-PRET-3SG Eskimo “Pablo was Eskimo” rubio (104) Pablo era Pablo ser-IMPF-PRET-3SG blond “Pablo was blond” . This will lead me to discuss the restriction existing between the progressive and the property of stativity mentioned in chapter 4. and progressive). This restriction differs from the one discussed in section 5. I will explore the interpretation that adjectival IL predicates have under the aspect forms mentioned thus far (perfective. 5. 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.180 Individuals in Time (101) Progressive form a.5.3 (about the relation between quantity and perfective and imperfect viewpoint).1 The Imperfect and Adjectival IL Predicates In this section I discuss the different interpretations (continuous or habitual) that an imperfect form can have with copular adjectival constructions depending on the adjectival predicate at stake. (En aquella entrevista) Juan estaba siendo muy cruel in that interview Juan was ser-ing very cruel con el entrevistador to the interviewer “(In that interview) Juan was being very cruel to the interviewer” (102) Perfect entailment from the progressive form a. *(En aquella ocasión) Juan estaba siendo rubio that time Juan was ser-ing blond “(That time) Juan was being blond” c. imperfect. *(En aquella ocasión) Juan estaba siendo esquimal that time Juan was ser-ing Eskimo “(That time) Juan was being Eskimo” b.

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

182 Individuals in Time definition (that is. Juan was a very cruel person” I am assuming that every time one is born counts as a different “individual. that the structure of (108) is not the only one they can have). this is not the only meaning that the imperfect can have with be cruel. since it refers to a cyclic. inside the existence of an individual. I am aware. However. though.28 For a habitual interpretation to be (somewhat) acceptable. Consider the following scenario: chino habitualmente (109) En sus encarnaciones humanas. they permanently hold. As I will amplify in the next chapter. the whole big interval containing (or overlapping) with the interval of the eventuality has to be repeated. 28 Note that IL predicates not referring to necessarily permanent properties. Juan era before becoming pacifist Juan ser-PRET-IMPF-3SG a person very cruel “Before becoming a pacifist. which suffices to license the habitual paraphrase (107) ‘Juan used to be cruel in his interactions with people’. Li used to be Chinese” In principle. However. The predicate in (105) be cruel is not permanent and can restart multiple times. If we create a situation where permanent properties can re-hold because their subject can be reborn (in possibly different circumstances). I cannot undertake here. such as blond. 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. even predicates covering an entire lifetime can be treated as iterative. as a consequence. unfortunately.” That is. When. where the adverbial in summer activates a habitual interpretation. the sentences become less odd. time. As can be seen. are quite natural in contexts such as (i). Li era in his incarnations human Li ser-IMPF-PRET-3SG Chinese usually “In his human incarnations. of each permanent predicate. in (109) there is an individual who has lived more than once. Other suitable contexts for (105) can be the following: (una persona) muy cruel (110) Antes de hacerse pacifista. the interval over which permanent predicates hold totally overlaps with the interval of the life of the individual they are predicated of. rubio en verano (i) Juan era Juan ser-IMPF-3SG blond in summer 27 . that this point may deserve more discussion. which. these properties hold. since we can establish the temporal bounds of each “existence” and. hence repeatable.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 in (113). Habitual AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 >1 b. when the complement inducing the activity-like properties is entered into the structure. in a similar way as we say John was blond or John was Eskimo. cruel con Pedro (113) Juan era Juan ser-PRET-IMPF-3SG cruel to Pedro “Juan was cruel to Pedro” cruel con Pedro (114) Normalmente Juan era usually Juan ser-PRET-IMPF-3SG cruel to Pedro “Juan used to be cruel to Pedro” This result dovetails with the fact that the most salient interpretation of activities in imperfect is the habitual reading shown in (116). The two types of paraphrases (habitual and continuous) are possible with be cruel and each corresponds to a different viewpoint structure. When the PP complement is present. the habitual reading is the most salient. the only reading is the habitual reading (114).Outer Aspect and Individual-Level Predicates 183 (una persona) muy cruel. That is. 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. The imperfect form can involve different quantifiers over occasions (Q<occ>): (112) a. (115) Pablo paseaba Pablo walk-IMPF-PRET-3SG (116) “Pablo used to walk” .

the preferred interpretation of adjectives denoting intellectual aptitudes (121) is the continuous reading. in Spanish.2. Likewise. see section 5. as noted above. the progressive is not a paraphrase saliently available from the imperfect form. this is just marginally available with be cruel (cf.3. 30 For the reasons why the progressive is excluded with Eskimo and blond. although the progressive form is correct with cruel. accordingly. make the habitual reading available. (122)). a habitual interpretation emerges. However. whereas activities in imperfect (117) can also have a progressive paraphrase (118). rather than the habitual (cf.1.30 Finally. 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. they acquire activity-like properties. which.5. as noted in section 5. 29 . the progressive paraphrase of the imperfect is more natural in these contexts. (119) and (120)). when a complement that can act as a “distributee” is added (125).184 Individuals in Time It is worth noticing that. when a relational complement is added (123).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.

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

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

.5.31 The progressive viewpoint does not seem to work like the imperfect and the perfective. As can be seen. the progressive is possible with other class of homogeneous predicates. according to which it is compatible with basically any type of predicate. (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. 31 As mentioned in chapter 3. consistent with the description of the perfective above.Outer Aspect and Individual-Level Predicates 187 muy cruel/inteligente (137) Juan fue Juan ser-PRET-PERF-3SG very cruel/intelligent “Juan was very cruel/intelligent” (138) ‘Juan was very cruel to Peter’ (139) ‘Juan was very intelligent in that business’ However. 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. it is also possible to have the stative version of these adjectives together with the perfective. the progressive in English works differently in some contexts. SL (141) and IL (142) statives are excluded with the progressive. in the sense that the progressive is not compatible with every kind of predicate. 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. whereas processes take time and progress in time (although this progression in time does not mean they advance toward a culminating point). states hold in time but do not take time. and therefore reject the expression of progression in time.

eventualities with such a temporal structure that between every two temporal points.” related to the concept of “movement. a third point can be established. progression in time is not possible for those eventualities that are “dense”—that is. However. nondynamic eventualities do not. where the verb empezar ‘start’ refers to the onset of the predicate remain. the idea that states lack onsets argues in the same direction than Comrie’s (1976) description that states are those situations which are not subject to any input of energy to hold.”34 alludes to “temporal stages” to describe the distribution of the progressive. an event is a stage of another event if the second can be regarded as a more developed version of the first. However. nevertheless. The input of energy correlates with the onset of the predicate. According to Landman. footnote 4.1) for the introduction of this concept. that is. According to these authors. for example. strictly speaking.32. nevertheless. if we can point at it and say ‘it’s the same event in a further stage of development. argue that progression in time is not possible for those eventualities lacking internal development. . accordingly. An eventuality can appear in the progressive if it is divisible in stages.” Dynamic eventualities progress in time. does not allow us to use the progressive. we can distinguish different stages in. since estar enfermo ‘be sick’ and estar preocupado ‘be worried’ are states and. it is difficult to find an explanation in more technical terms. In some sense. such as Landman (1991).33 Landman (1992). See also chapter 6. a sickness. as a result. based on Carlson’s (1977) notion of “stage.188 Individuals in Time The property usually invoked to explain the differences regarding progression (advancement) in time is “dynamism. and. unexplained under this view. In a similar vein. followed by Bertinetto (2000). In fact. can be argued to lack any input of energy. examples such as the following. 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. 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. they are excluded in the progressive form. which. as the ungrammaticality of cases like *Juan estaba estando enfermo ‘Juan was being sick’ shows. Aside from the intuitive understanding of the correlation between dynamism and progression in time. Some authors. Kearns (1991) explains the incompatibility between the progressive and states by alluding to their property of essentially lacking onsets and culminations.

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

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

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

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

3 presents the discussion of lifetime effects and describes the properties the predicate has to involve to trigger them. as a consequence. those dubbed “lifetime effects” (Musan 1995). Section 6. As mentioned in chapter 2. although it is not only the type of predicate that counts. IL predicates offer different inner-aspect properties.2.4 I show how contextual factors intervene in the rising of this temporal interpretation. 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. The two main points I would like to make here are the following. I will be concerned with the interpretive contribution that tenses (present. and future) have on copular sentences with adjectival predicates. . lifetime effects refer to the interpretation whereby the referent of the DP subject can be interpreted as ‘no longer alive’. I will make special reference to the work by Musan (1995.5 discusses the nature of the TT in greater detail. in section 6.Chapter 6 Tense and Individual-Level Predicates In this chapter I study another temporal realm—namely. Likewise. Throughout the discussion I will be chiefly concerned with the interpretation of IL sentences in the past tense. These two features suppose a difference with respect to previous accounts of IL predicates in general and. which has consequences for their temporal interpretation. In the first section I introduce one previous account for the particularities of the temporal interpretation of IL predicates. by exploring the interpretation of IL predicates in compound sentences. the domain of Tense. which. 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. is located higher than the Aspect node. The last section summarizes the conclusions. In section 6. First. in particular. structurally. I propose a division among (adjectival) IL predicates that correlates with different temporal interpretations. it does not influence the ultimate temporal interpretation of the sentence. past. Section 6. In this regard. formerly discussed by Kratzer (1988. it is the content of the Topic Time (TT) that ultimately decides the temporal interpretation of the sentence. since this tense form gives rise to one of the most important interpretive effects with IL predicates—namely. to previous accounts of the temporal interpretation of IL predicates. 1995). As I have shown. it is not the case that Tense has the same interpretive impact on all individual-level (IL) sentences. I assume tenses are syntactically represented in the node of Tense. Second. This chapter is organized as follows. As it appears.

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. 1995) Kratzer (1988. IL predicates IP 1 I′ 1 I VP 1 θ-subj Regarding temporal interpretation. she proposed that IL predicates applied directly to the subject. As said before in chapter 2 (section 2.3 (1) a. a spatiotemporal or eventive (<e>) argument (Davidson 1967).1). whereas SL predicates needed something “extra.1. Kratzer proposes that the spatiotemporal argument is syntactically represented.2).1. she argues that the eventive argument is the external argument of (SL) predicates. This author assumes that the factor that draws the line between the two types of predicates is “being temporary” versus “being permanent. Furthermore.1 I will briefly summarize her line of reasoning and show that there are cases it cannot account for. temporariness and permanency is encoded in the argument structure of predicates: temporary predicates involve a spatiotemporal argument.194 Individuals in Time 6. the “realization function. SL predicates involve an extra semantic operation. SL predicates IP 1 I′ 1 I VP 1 <e> b.” Likewise.”2 Specifically. 3 As I also said in chapter 2. as Diesing 1992 proposes. see Anderson 1973. whereas permanent ones lack such an argument. First. Kratzer proposes that the differentiation between SL and IL predicates resides in their argument structure. I will mention two points of this proposal. in the same vein as Carlson (1977). 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. In particular.1 Temporal Interpretation as a Consequence of Argument Structure: Kratzer (1988. . just aims to capture Kratzer’s idea on the location of the external argument with respect to the maximal projection of the predicate. the representation in (1). 1995) is the first in establishing an explicit correlation between the temporal interpretation later dubbed “lifetime effects” and IL predicates. That is. The configurational difference between IL and SL predicates can be depicted roughly as in (1).” See chapter 2 (section 2. For Carlson (1977).

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

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

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

198

Individuals in Time

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

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

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

Tense and Individual-Level Predicates

199

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

200

Individuals in Time

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

Tense and Individual-Level Predicates

201

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

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

202

Individuals in Time

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

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

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

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

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

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

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

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

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

1). In this case.Tense and Individual-Level Predicates 211 sentence—namely. this interval is called “Topic Time” (TT) (Klein 1994).e.. since we have proved that it is the content of the TT that matters. as Stowell (1993) proposes for the other two (see chapter 5. I assume that the nature of the TT is not different from other intervals like Reference Time (RT) and Eventuality Time (ET). 1997) talks about “the neutralization of lifetime effects. Thus. the interval at which Maria and Harry arrived in the USA.1. here assumed to be the Utterance Time (UT) (i.2). no lifetime reading is available. but the interval the speaker is referring to. I assume that the TT is a time-denoting phrase (ZP). I propose that the lifetime reading arises when the TT is defined as the interval of the individual’s existence: 9 Musan (1995.1. is located within the span of time at which being from California holds and after the RT. I thank Tim Stowell for this remark. In other words. In the spirit of Musan (1995). the interval corresponding to the arrival in the USA. From this.9 What needs to be defined is which content gives rise to the lifetime interpretation. section 5. the temporal structure of Harry was from California of (59) can be depicted as follows: (60) (UT) RT TP 3 T′ 2 T AspP (after) 2 TT Asp′ (arrival in the USA) 2 Asp VP (within) 2 ET VP (being from California) The structure in (60) says that the TT. Likewise. Tense is locating in the past not the whole interval the eventuality (be from California) holds. lifetime effects can be reasonably considered a phenomenon of temporal interpretation. since from the perspective I propose here it is not the case that the reading arises and then it becomes neutralized. . it can be observed that the value of the TT is decisive in the arising of such readings. in the past). it seems more accurate to talk about the arising or not of the interpretation.” However. As I introduced in chapter 5 (section 5. The occurrence of the lifetime effect is the opposite side of the coin of their absence.

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

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

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

a lifetime reading does not arise. I would like to argue that via such an obligatory contextual restriction. during which these examples arose.10 This hypothesis establishes the previous context as the context that counts for the sentence in the conversation. since the reference to a previous scenario obtains. no lifetime effect arises. it concurs with Musan’s (1995) observation that an element from previous context may make the reading not to appear. I want to argue that this contextual information establishes the link with the background where the TT finds its content. as explained by von Fintel (1994) and mentioned before. the lack of a (free) restricting variable can be argued to be the difference between generic and nongeneric DPs. and. I turn to more subtle situations. The point I would like to introduce here is that DPs referring to individuals also have contextual information associated with them. and. that lifetime effects arise in sentences such as (i). The contextual restriction of the quantifiers directs us to a previous context where the free variable C finds its antecedent. too. and makes it the common ground where the TT finds its antecedent. the value of the TT is not a lifetime span.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). In essence. Compare these two situations. 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. This absence of a link to a previous context has an outcome in the temporal realm—namely. as a result. Actually. Los dinosaurios eran del Polo Norte (ii) Dinosaurs are extinct 11 I thank Dominique Sportiche and Daniel Büring for the discussions around these topics. a background is built up. (i) a. even though we have a lifetime predicate in past tense. As a consequence. 6.4. where the preferred interpretation is (ii). 10 . Let me explain what I mean with an example. the availability of the lifetime interpretation does not arise. Along similar lines as before.11 It is interesting to note that English bare plurals (ia) or generic Romance DPs (ib) do not seem to be contextually restricted.2 Context Associated to Individuals At this point. If the TT content is other than the lifetime span of the individual. Dinosaurs were from the North Pole b.

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

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

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

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

by default. then. the determiners) have. when there is no specific context the free variable is to be linked to. Gregory himself TT → the whole interval overlapping Gregory’s lifetime span. no lifetime effect arises. I have argued that the establishment of the relevant context is brought about through the free variable DPs (more specifically. Result: a past form is allowed in the copular clause and.4 Summary of Section 6. As Kratzer (1988. the value it takes by default is the individual himself and the interval associated to him.4 In determining the context relevant for the specification of the TT content (which is what finally decides the temporal interpretation of a sentence). there is no reason to consider (59) different from (72) and (73). the ZP . 1995) and Musan (1995) assume. (…) Topic of the sentence that may establish the connections to a previous background = Harry. (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. 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. since it refers to the time of the arrival. it finds its antecedents in the most salient context. The interpretive procedure is always the same and works the following way. (82) Gregory was from America (83) Topic of the sentence that may establish the connections to a previous background = Gregory (84) Gregory c= undefined. let us consider a sentence where a lifetime reading can arise. Once a context is fixed. 1997). these typically are out of the blue sentences. since they are the same in all relevant respects—all of them are copular sentences in context. Bearing in mind that the TT is a ZP context sensitive. (80) Harry was from California.” they establish the relevant context.4. Since the DP subjects of copular sentences can be considered “topical elements. Finally. I propose. Actually. Result: lifetime arises With Musan (1995. 6.220 Individuals in Time crucial importance in the licensing and interpretation of a past form. it is the information associated to the DP subject that counts. I assume that. a lifetime effect does not arise.

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

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

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

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

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

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

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

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

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

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

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

1996) work. whoever she might be. Stowell (1993. can be paraphrased in two different ways: either Juan was looking for any girl.5. as I have been doing thus far. among others). or Juan was looking for a particular girl. Adriana. it can have two interpretations. a lifetime interpretation is not expected from a syntactic situation such as (101). (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. Abusch demonstrated that the different temporal readings of RCs correlate with the interpretation of the DP the relative depends on. I will divide the task in two steps.232 Individuals in Time the TT is located in the past. 6. look for > a girl b. I will pose what this analysis predicts with respect to the lifetime predicates and show whether it is borne out or not. I will make some considerations about the interpretation of lifetime predicates in another syntactic environment: Relative Clauses (RCs). a girl > look for When the main verb is extensional. we see that the interpretation of the DP a girl. Rivero 1975.2 Relative Clauses In this section. which can be described as a [–specific] and a [+specific] (Russell 1905. If we take an example like (102). 1996) builds on Ladusaw (1977) and Abusch (1988) to elaborate his proposal about the temporal interpretation of RCs. (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. namely. Donnellan 1966. Secondly. and the [+specific] interpretation to a wide scope position. when the DP is indefinite and acts as a complement of an intensional verb. which may be determined by that of the DP . The reading roughly described as [–specific] would correspond to a narrow scope position with respect to the intensional verb. Firstly. (103) a. the DP can be considered as unambiguously [+specific]. I will summarize general points concerning the temporal interpretation of RCs. following Stowell’s (1993. Summarizing very much.

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

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

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

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

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

.

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

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

these notions. the notion involved. 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. By answering this question. a number of authors have proposed that the IL/SL distinction should be cast in “aspectual” terms. Although I will not be able to offer a definitive answer to this.Conclusions and Final Remarks 241 which the Topic Time can take its value. most . most authors dealing with the ser/estar contrast concur with the idea that estar-predicates put the property in relation to a situation. and I have given an account for the different temporal interpretations that IL sentences can have without yielding inaccurate generalizations. Claim (in previous literature) IL predicates express permanent properties When an IL predicate is at stake. One of the important questions here is whether the difference between ser/estar clauses (or. we will be proposing something about the terms in which the IL/SL dichotomy is represented in the grammar. more in general. In examining this question we address the key difference between IL/SL predicates. Table 7. As mentioned in chapter 2. classically alluded to in the description of IL predicates.” leaving the property without any association to a determined situation. while ser is more “innocuous. One of the crucial questions is how the association to a situation is brought about.1 summarizes the claims previously made. and the semantic domain they belong to. These conclusions differ from most previous literature. are temporal concepts in nature. where the definition of IL predicates has been mainly associated to the notions of permanent property and stativity. As was surveyed. Ser simply classifies the subject into the category denoted by the predicate that appears in combination with it. between IL/SL predicates) can be cast in temporal terms.1.2 The IL/SL Dichotomy and the Ser/Estar Contrast With major or minor differences. 7. Previous claims regarding IL predicates Throughout this work I have analyzed IL predicates in their relation with the temporal realms cited. I will introduce some discussion in the following sections. As noted in the beginning of the book.

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

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

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

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

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

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

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

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

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

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

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

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

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

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

the obligatory predication with estar is explained. the ungrammaticality of Juan es en Brasil is explained. although episodicity can be one of the consequences of the linking to a particular situation. In chapter 4. among other things. but does not account for why they can appear just with ser and cannot be conceived as linked to a particular situation. the cruel-type. I proposed that the peculiar properties to the constructions these APs participate in emanate from the particular properties the APs themselves involve. the description of estar as ‘episodic’. The definition of ser given explains why nouns combine with ser (40) (since nouns denote classes). Consistently with this hypothesis. Ser classifies the subject by categorizing it into the class denoted by the predicate following the copula. en España (41) Madrid está Madrid estar-PRES-3SG in Spain Since a location is not a property of any individual. 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. unlike adjectives.256 Individuals in Time en Brasil (39) Juan está/*es Juan estar/ser-PRES-3SG in Brazil ‘Juan is in Brazil’ profesor (40) Pedro es/*está Pedro ser/estar-PRES-3SG teacher ‘Pedro is a teacher’ We can make sense of (39) by alluding to the description of ser given. As we already know. This perspective also allows us to avoid. Since a location is not a class. 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 . but something external to it. such as agency. I would like to add a few words about the combination of estar and the type of APs I have devoted special attention to in this work. Thus. it seems that both notions can be conceived apart from each other. those peculiar properties.

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

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

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

.

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

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

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

Linguistic Inquiry 22:1–25. Word Meaning and Montague Grammar. 2000. The semantics of specificity.. Revista Española de Lingüística 9:133–171. D. The stative in the progressive and other essence/accident contrasts. M. Uriagereka (eds. 129–215.). 1999. Bosque and V. 1976. Philosophical Review 75:281–304. 1987. 1991. Structure-Preserving. V. D. J. 1999. J. Emonds. Dixon. The ingressive as a new category of verbal Aktionsart. Guéron and J. D. V. Depraetere. I.). Egg. 1986.264 Individuals in Time Demirdache. Enç. 1979. 1979. 2004.). Uribe-Etxebarría. Bosque and V. Enç. 157–186. 1995. V. Dowty. In Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik. Linguistic Inquiry 18:633–657. I. Journal of Semantics 12:311–356. I. and Local Transformation. The syntax of time adverbs. H. Cambridge. Dowty. Madrid: Espasa Calpe. The primitives of temporal relations. Language 67:547–619. Demonte (eds. H and M. and Masullo. Demirdache. New York: Academic Press. In The Syntax of Time. Dordrecht: Reidel. 1995. Michaels. Dowty. Donnellan. M. R. 2461–2524. Uribe-Etxebarría. . M. On the absence of the present tense morpheme in English. The effects of aspectual class on the temporal structure of discourse: Semantics or pragmatics? Linguistics and Philosophy 9:37–61. Dowty. 1991b. M. Lecarme (eds. D. 1991a. La predicación: Los complementos predicativos. Demonte. Enç. 1992. Clases de adjetivos: La posición del adjetivo en el SN. 1966. In Gramática Descriptiva de la Lengua Española. Demonte. MA: MIT Press. Cambridge. Madrid: Espasa Calpe. and M. M. Indefinites. Anchoring conditions for tense. Demonte. D. Linguistic Inquiry 6:579–588. In Gramática Descriptiva de la Lengua Española. Martin. 1975. H. Demirdache. Linguistics and Philosophy 18:1–19. Cambridge.. Demonte (eds. Uribe-Etxebarría. Madison. Ms. and M. University of Wisconsin. Sintaxis y semántica de las construcciones con ser y estar. 1999. MA: MIT Press. and J. 1977. Diesing. K. R. J. 143–179. M. Reference and definite descriptions. Thematic proto-roles and argument selection. MA: MIT Press. A Transformational Approach to English Syntax: Root.). University of Nantes and University of the Basque Country/Basque Center for Language Research. Ms. On the necessity of distinguishing between (un)boundedness and (a)telicity. M. Where have all the adjectives gone? Studies in Language 1:19–80. Towards a theory of the diversity of temporal systems.

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

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

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

Pittsburgh. Lecarme (eds. R. 1967. Stative Verbs and Adjectives in English [Harvard Computational Laboratory Report NSF-17]. Thematic relations as links between nominal reference and temporal constitution. Linguistic Inquiry 25:609– 666. 1996. The categorical and the thetic judgment: Evidence from Japanese syntax. F. 1981. Stage-level and individual-level predicates. On the double object construction. Zaring (eds. 1994. Davidson’s “The Logical Form of action sentences. Landman. J. 1991.. Lakoff. Lewis. 1995. Kratzer.). 29–53. On the Plurality of Worlds. J. The Determination of Grammatical Relations in Syntax. A. 1994. In The Syntax of Time. Cambridge. Stanford. A.).” In The Logic of Decision and Action. D. M. J. Dordrecht: Kluwer. 1966. Paper given at the Berkeley Linguistic Society. . Kratzer. Comments on D. The Spanish copulas as aspect indicators. M. A. Texas Linguistic Forum 6:89–102. Ms. In Phrase Structure and The Lexicon [Studies in Natural Language and Linguistic Theory]. Ph. 389–423. 1988. Carlson and F. Lemmon. 1992. S.).). 125–175. Tübingen: University of Tübingen. 2004. Guéron and J. 1970. Telicity and the meaning of objective Case. Rooryck and L. A. Chicago: University of Chicago Press. Severing the external argument from the verb. 1972. I. Kratzer. Kratzer. The event argument. Krifka (ed.). Natural Language Semantics 1:1–32.268 Individuals in Time Kratzer. CA: CSLI Publications. Martin. Szabolcsi (eds. Los Angeles. MA: Harvard University. DC. Cambridge. The progressive. G. Rinehart and Wilson. Kuroda. 1988. Hensey (eds. 109–137. W. In Genericity in Natural Language [SNS-Bericht]. 1991. Luján. Larson. University of Massachusetts. 2000. G. February 2000. In The Generic Book. Current Studies in Romance Linguistics.). Longobardi. 1986. Washington. Building statives. MA: MIT Press. Landman. PA: University of Pittsburgh Press. Some problems with tense in PTQ. M. F. Sag and A. Structures for semantics. Oxford: Blackwell. In Lexical Matters. New York: Holt. 1977. G. Linguistic Inquiry 19:335– 391. Lakoff.D. Amherst. 96–103. and F. Lingua 54:165–210. Foundations of Language 9:153–185. dissertation. 1992. Ladusaw. 1976. E. 247–284. Irregularity in syntax. Georgetown University Press. Luján. M. Stage-level and individual-level predicates. A. University of California. Dordrecht: Kluwer.). J. Pelletier (eds.-Y. Rescher (ed. Reference and proper names. A. Kratzer. A. Krifka. N. G.

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

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

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

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

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

.

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

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

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

.

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

186. 134–136. 246. 246–248. 172–174. 240. 224. 12. 148. 167. 184. 222–225. 131. 94 Pomo 94 Prepositions Directional prepositions 127. 86. 227 Simultaneous reading 76–79. see Temporal interpretation Russian 15 Salish 118. 113. initiator 61. 72. 231. 160. 134 There-sentences 5. 147. 148. 246. 67. 31 Imperfective paradox see Progressive paradox Initiation. 242. 132. 221–223. 128–131. 158–160. 191 Event(uality) Time (ET) 3. 106. 187–192. 186. 175–177. 229. 212 Mereological properties 3. 14. see also Event(ive) argument Speech Time 122. 154–157. 176–178. 147. 105. 123– 126. 162. 148. 233. 94 Margi 121 Maxim of Quantity 85. see also Subinterval property and Additivity property Narration temporal interpretation of. 189. 35 Figure 119. 189. 215 Goal(Path) 96–98. 132. 154. 159 Quantification restriction 11 Quantity 113–116. 144. 230. 173. 134–136. 100. 14 Generic reading 6. 148. 158–160. 183. 189. 176. 111. 207. 211. 208. 6. 212. 192. 234 Situation aspect 39 Small v 55 Spatiotemporal argument/variable 8–10. 252 Realization function 6. 178. 175. 248 Swahili 120 Temporal interpretation of Complement clauses 76. 178. 221 Habitual 77. 165. 258 Progressive paradox 191. 11. 192. 141. 156 Hawaii-Creole 120 Hebrew 15 Heterogeneous (predicates) 72. 236 Theme 105. 244. 116. 148–152. 136. 116. 258. 192 Prospective 120–122. 209. 186. 172. 233 Perfect entailment 47. 222. 80. 58. 184. 230.280 Individuals in Time Event-object 110. 181 Existential reading 6. 242. 224. 186. 11. 12. 258 Hindi 120 Homogeneous (predicates) 70. 122. 233–235. 211. 136 Place/situational prepositions 120. 158. 176. 76. 180 Perfective (preterit) 31. 165. 241. 226. 177–180. 129 Progressive 153–163. 249 Existential coda 12 Existential quantification 8. 184. 162. 257. 123. 148–152. 131. 194 Reference Time (RT) 3. 122. 129. 240. 156 Focus 13. 155–157. 26. 168 French 121 Generic operator 11. 201. 152. 148. 141. 147. 202 Ground 119. 130. 196. 37 Thetic (judgments) 12–14. 183. 64. 179. 147. 250 . see also Utterance Time Stativizer 115. 248. 117. 257 Perform(er) 53. 134 Instigator 55. 226. 150. 258 Imperfect Continuous 76. 249. 79. 148 Sequence of Tense 226. 14. 192 (Im)perfective contrast 25. 57. 131. 159. 161. 181. 158. 137 Subinterval property 70. 123. 94. 72. see Temporal interpretation Past-shifted reading 76–78. 189– 192. 128. 246. 158. 226. 83. 187. 142. 212. 35. 115. 227–229. 79. 249 Relative clauses temporal interpretation of. 172. 227. 144. 195. 245. 118. 178. 80. 136. 153–160. 189. 112 Juba Arabic 120 Kinesis 40 Krio 120 Lakhota 55. 187. 232. 144. 180. 55. 242. 229. 155. 88. 248. 194. 72. 191. 135. 189. 186. 171– 175. 227. 125. 230 of Narration 77–80 of Relative clauses 223. 132. 157. 207. 173–180. 174–178. 158. 112. 35. 244. 129.

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

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

Sign up to vote on this title
UsefulNot useful