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

...........3 IL Predicates as Inherently Generic Predicates........2........................2..........................................................2...................................................................................................................................53 3................................................................. Chierchia (1995) ..........59 3.......25 2......1............3 Aspectual Differences among Individual-Level Predicates ............ Higginbotham and Ramchand (1996) ..........2 Event Types and Event Structure......................................................................................2... Raposo and Uriagereka (1995)....................................1 ......................................................53 3...14 2..............2 ...................................39 3..............39 3..................5 2...........................................................51 3..61 3..2..............3 Summary of Section 3.. A Brief Stop at “Agentivity”................ Milsark (1974) and Carlson (1977).....1 A Cluster of Notions ..........5 2......2..........1.............5 2............1............4 The IL/SL Contrast as a Categorical/Thetic Distinction........................................................ Kratzer (1988.......39 3...........................................1.......................2 ....2 Agents in Event Structure .. Inner Aspect and Event Types ..............2 The IL/SL Dichotomy as a Syntactic Distinction......................................................1 Distribution of Copular Verbs in Spanish ......11 2..................1 .......................8 2.....2.38 Chapter 3 Event Classes and Individual-Level Predicates ....2 Ser and Estar as an Aspectual Differentiation ....................1........ 1995) ....................1 The IL/SL Distinction as a Semantic Distinction... Introduction to the Tests Discerning the IL/SL Status................................40 3....................... by Tim Stowell .........3 Summary of Section 2....................................1................ The Structure of Copular Constructions ....................1 Proposals about the Individual-Level/Stage-Level Dichotomy ...3...1 Inner Aspect..........................................1 Chapter 2 Individual-Level Predicates .....Table of Contents Acknowledgments....61 ................................................1.....32 2.....................1.....33 2...............2................16 2.......4 Summary of the Chapter .......5 Summary of Section 2................................................................ix Foreword..xi Chapter 1 Presentation of the Study ......1................12 2................15 2.... When There Is More Than One Copular Verb: Spanish Ser and Estar...3 Summary of Section 3.......................................................

..................108 4..................117 4............5...............................96 4.........5 Justifying the Approach .................................72 3................ Summary of the Chapter ..138 4...............90 4.......117 4..............................................96 4.....4 Summary of Section 4....4.........................................................................81 Chapter 4 Aspectual Alternations in Individual-Level Predicates ...3......4..91 4..........3 The DP Subject of Relational MPs .143 ...........116 4............86 4..............2......................135 4............. The Relational PP Complement...............2 Summary of Section 3..........2 On the Optionality of the (Affected) Goal PP .................. States and Activities: A Grammatically Relevant Difference?.........1 Lexical Semantic Classes of Adjectives ..............................6 An Account Based on the Relational PP Complement ...........137 4...1..................107 4........7.......4........1 Prepositions as Aspect Encoders .84 4...........................................6.5........69 3...........................7 ..................6............3 Relational Mental Properties.................................................. 69 3.......4 Summary of Section 3............3 Summary of Section 4...............2 Summary of Section 4..............................3 The Relational PP with Other APs ........................................................5...........................1 Previous Explanations of Nonstative Copular Clauses ........76 3........2 Differences between Activities and States ................................ Summary of the Chapter .................62 3...2 Be as a Copula Shifting a State into an Activity ............................133 4................................................3.......8..............................2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor123 4..98 4.....106 4.......................3..............................1 The Two-Copulas Hypothesis ....................................................................3 ....................3 ....4 The Issue: Aspectual Alternation in IL Copular Clauses .......vi Individuals in Time 3..1 Cruel-type Small Clauses Taken by Verbs Other Than the Copula .............................5..........6 ..3.................84 4.143 4.......................................2 Adjectival Predicates Showing an Activity-like Behavior ..........3 Some Confusing Arguments in Defense of a State/Activity Distinction.......85 4.....................................................4....95 4..............1 Similarities between Activities and States ......68 3.....6..........................2 Syntactic Approaches ........1.2 Summary of Section 4..............................................110 4.........2 ...................1 Aspectual Tests on IL Predicate .................6.................5 ..........................................3 Adjectives Predicated of an Implicit Event Argument ....4 ............................ The Cruel-type as a Small Clause Integrant..................1.....7.3....................83 4...................3........108 4..............80 3......7.......1 Lexicalist and Logico-Semantic Approaches .......................................................105 4....................4 Summary of Section 4.......................2..........4..................................................1 On the Interpretation of the Relational PP .................................................

...................................... 1995) .....148 5.................157 5.......3.4.......................1 Temporal Interpretation as a Consequence of Argument Structure.......................2 Nonpermanent IL Predicates.......173 5..........2 ...............................................................2 The Perfective and Adjectival IL Predicates .4...212 6.............3 Articulating the Account.5..3 On the Relation between the TT and the Habitual Q<occ> .................3 The Progressive and Adjectival IL Predicates ..... A Brief Summary of Aspect Notions ..........4 Summary of Section 6.......210 6...3................202 6...................................4.........2 The Habitual Interpretation: Iteration....205 6...171 5..............................4 Summary of Section 6...........3..........2.........1 Habitual Heterogeneous Predications...1........................................................ Aspect as an Ordering Predicate ..3 The Content of the TT in the Arising of Lifetime Readings .................................................6.................... 187 5.........3.................................................4 ..............................222 .......208 6..................................................................152 5..........................194 6..4..............205 6...........................................213 6....................................................2 Aspect as a Quantifier over Occasions ....................................2..............................1 When the Subject Is a QDP ..............163 5.............193 6....................................3.............................199 6....................... 176 5.....1 Permanent IL Predicates ......5....................................................................... 180 5......................................5 The Lifetime Reading in Compound Sentences .................205 6.... 1997) ....... 192 Chapter 6 Tense and Individual-Level Predicates .................................3 Brief Notes about the Complement of Mental Properties APs .................2.......................2 Context Associated to Individuals ...220 6.........................................218 6. Stowell (1993.1 The Imperfect and Adjectival IL Predicates ............1..................147 5....148 5.........3 Inner and Outer Aspect ................................................... 1996) ...... 174 5...... Musan (1995............................2..5 Adjectival Individual-Level Predicates and Viewpoint Aspect ... Proportion............3 The Arising of Lifetime Effects ...........................................2 Introducing the Determining Role of Contextual Factors........................215 6..............2 Differentiating Temporal Extensions of IL Predicates ...............1 Tense.............. 186 5.....................................Table of Contents vii Chapter 5 Outer Aspect and Individual-Level Predicates ......................4 The Determination of the TT Content and Lifetime Effects ....................2 Perfective Homogeneous Predications .... Summary of the Chapter .2........................................2.2 The Internal Argument of Tense: The External Argument of Aspect... 179 5....199 6.............................197 6.................158 5....... Kratzer (1988......................4...1 Quantifying over Occasions ....................2............................1 Necessary Conditions for Lifetime Effects ......1 Tense and Aspect as Ordering Predicates ......................................................................... 177 5................... and Systematicity ................................................................5........

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

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

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

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

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

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

.

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

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

After showing that the ser/estar dichotomy cannot be recast in terms of tense nor in terms of outer aspect (by alluding to the completion or perfection of the property) nor in terms of inner aspect (mereological properties). In this respect. Focusing on copular clauses. Thus. along the lines of Demirdache and Uribe-Etxebarría 2000. The discussion of inner aspect is framed in the theoretical debate regarding the effects of syntactic structure on the semantics of predicates. Ritter & Rosen 2000). In this vein. 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. In particular. Outer aspect and tense are conceived as dyadic ordering predicates that take two time-denoting arguments (Stowell 1993. I will show that prepositions can contribute an inner aspect shift from states to activities. Tense is proposed to order the Topic Time with respect to a Reference Time and is argued to work uniformly with all kinds of predicates (SL and IL). giving the number of times a certain eventuality holds (along the lines of Verkuyl 1999). I argue that the IL/SL contrast should be rooted in the association to a particular circumstance (SL) or its lack thereof (IL). Ramchand 2003. which is described as a viewpoint referring to a plural number of occasions. the dynamic properties of clauses where adjectives referring to mental properties appear (such as cruel. my proposal will concern the role of prepositions in event structure. In chapter 4. 2004) and a quantifier over occasions. kind. 1995] with respect to the totality of the Eventuality Time. I will then discuss the theoretical and empirical consequences for the IL/ SL dichotomy. The study of tense focuses on the interpretive phenomenon known as ‘Lifetime Effects’ (whereby the individual referred to by an IL predicate in past can be understood as ‘no longer alive’). I present some reflections about habituality. In chapter 3. I study the set of IL predicates patterning with activities and argue that this behavior is rooted in the nature of the syntactic frame intro- . Chapter 2 introduces the notion of IL predicate and presents the distribution of the two copular verbs in Spanish. mean) are proposed to correlate with the presence of the relational complement of these adjectives (cruel to Mary). dynamicity is proposed to correlate with the presence of the relational complement and argued to be rooted in the prepositional phrase. 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. Specifically. 1996). based on hypotheses that attribute aspectual content to prepositions (Hale 1984). This book is organized as follows. the ideas defended in this monograph are sympathetic to the conception of syntactic structure as event structure.Presentation of the Study 3 tense. 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. this being understood in inner aspect terms (Borer 2005.

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

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

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

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

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

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

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

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

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

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

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

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

obviously designates a nontemporary property.11 which. if the subject is a physical entity (43). un ordenador de 1960 (42) a. However. Este animal es/*está that animal ser/estar-PRES-3SG a goat “That animal is a goat” professor c. Pedro es/*está Pedro ser/estar-PRES-3SG teacher “Pedro is a teacher” When the predicate is a PP. the copular verb must be ser (42). and accidental properties as temporary.2. the copula used can be estar. The classic counterexample in this respect is estar muerto ‘be dead’. 11 See also comments by Milsark (1974) and Carlson (1977) about the same predicate. In the case of Spanish.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. which led to the claim that ser corresponds to permanent properties and estar to temporary ones.1. other authors observed that the distinction between permanent and temporary was. despite being ungrammatical with ser. discussed in section 2.10 Essential properties are understood as permanent. I would like to note Demonte’s (1979) observation that this distinction poses an asymmetry in the account of copula distribution.” as cited previously (see section 2. 12 Although I will not work out this difference here.1). How can the distribution of the two copular verbs be described? As traditionally observed.1.1. as . Ese artefacto es/*está that artifact ser/estar-PRES-3SG a computer from 1960 “That artifact is a computer from 1960” una cabra b. if not a mistake. if. 2. or ser. there is no general agreement in the literature on how the copula alternation should be characterized.1 Distribution of Copular Verbs in Spanish Some traditional grammarians (Bello 1847. Roldán 1974) describe the distribution of the two copulas ser and estar on the basis of the Aristotelian dichotomy “essence” versus “accident. if the subject denotes an event (44). whereas estar is used to refer to those properties considered accidental. when the predicate is an NP.” The copula ser is used to refer to properties that are considered essential of the individual.12 10 Note that these descriptions fall in line with the ones given by Milsark (1974) for “properties” and “states. insufficient to characterize the copular alternation.

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

Bosque and Picallo 1996. For more details about classifying adjectives. and Demonte 1999. 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. (b) impossible modification by a degree adverb (48).13. also. that classifying adjectives are excluded as copulative predicates (The presidential trip vs. When the adjective is referred to the place where one is born. that their acceptability as predicates in copular constructions seems to be confined to those cases where the subject is a physical entity. 14 Note. 15 13 . precisely. If the subject is a resultative nominal.6). and those in (51) combine with both copular verbs. However. These tests are (a) impossible participation in comparatives (47). Further discussion is presented in chapter 7 (see section 7.14 (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.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. see Bosque 1993. it is not gradable. *The trip was presidential). and (c) impossible participation in binary correlations (49). one of the properties that Schmidt (1972) and Bache (1978) provide to distinguish classifying adjectives from qualifying ones is. Are origin adjectives qualifying or classifying? I will treat them as classifying because they behave as canonical classifying adjectives in the other three tests that Schmidt and Bache motivate.

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

(55) Noté a Juan muy guapo I-noted Juan very handsome (56) Noté a Juan muy pálido I-noted Juan very pale (57) Noté a Juan muy moreno I-noted Juan very dark-skinned (58) Noté a Juan muy gracioso I-noted Juan very funny All of the adjectives in (55)–(58) can combine with either ser or estar. pero está muy guapo (53) Pablo no es Pablo not ser-PRES-3SG handsome but estar-PRES-3SG very handsome “Pablo is not handsome. Thus. Correspondingly. as we already know from (51). throughout this work. where the pound symbol indicates when the paraphrases are not appropriate. (59) Noté que Juan estaba/#era muy guapo noted that Juan estar/(#ser)-PRET-IMPF-3SG very handsome “I noted that Juan was looking/was very handsome” . where a perception verb (note) has a SC (NP + AP) as a complement. but he is tanned” The terms in which the contrast between ser and estar has been described fall in line with the IL/SL distinction introduced in section 2. all of the inflected versions in (55)–(58) that we can construe as paraphrases have to be done by using estar. Consider the next group of examples. Interestingly. I am dealing with an IL predicate. the copula designing SL-hood. I would like to bring up some discussion about the contribution of perception verbs in distinguishing an IL predicate from an SL one (a test used by Carlson 1977. not ser— that is. but he looks very handsome” muy pálido. pero está moreno (54) Pablo es Pablo ser-PRES-3SG very light-skinned but estar-PRES-3SG dark-skinned “Pablo is very light-skinned.1).20 Individuals in Time nada gracioso. Consider (59)–(62). 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. whenever the copular verb is ser. the copula estar will be considered a lexical sign of an SL predicate.1. see section 2. With the copular distinction in Spanish in mind.1. I will consider that.

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

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

improve the sentence. 20 . when the adjective is combinable with the two copulas. however. as in (ii). 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. but you stay’). and so on that contribute to the acceptability of the sentence. or gracioso. In this regard. 50 above) and. Other adjectives. Roughly speaking. physical. the semantics of perception verbs deserves a deeper investigation than what I can offer here. rather than IL/SL-hood. 19 With infinitive complements. where a state such as know languages is excluded. Consider the following contrast. subject pronouns need not be overt in Spanish. the corresponding paraphrases contain estar. pálido.20 When they are overt. (70) Te he visto desnudo “I have seen you naked” (71) *Te he visto ignorante “I have seen you ignorant” (72) Te he visto muy pálido → He visto que estabas/#eras muy pálido you have-seen very pale → have-seen that estar/ser-IMPF-PRET-2SG… “I have seen you looking/being very pale” However. 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. and disambiguate between verbal forms (cantaba ‘sing-IMPF’ can refer to the first and the third person). they appear to contrast (Tu hermano puede venir al cine. not all the results are so clear cut. bastante ‘quite’. the verb see distinguishes between states and events. pero tú te quedas ‘Your brother can come to the theater. Juan {guapo/pálido/moreno/gracioso} (i) *Percibí/advertí a I-perceived/noticed Juan {handsome/pale/tanned/funny} al contrincante {muy tranquilo/nervioso/apagado} (ii) Percibí I-perceived the opponent {very quiet/nervous/spiritless} As mentioned in footnote 16. the sense of the verb is that of sub- ‘notice’ are odd with predicates such as guapo. emphasize (Quiero que vengas tú ‘I want you [and not other] to come’).Individual-Level Predicates 23 such as (70) and (71) are those that combine with estar (cf. moreno. or both types of perception) and the role of modifiers like muy ‘very’. rather than ser (72). it is interesting to note the role that overt subjects (pronouns) play in the meaning of this perception verb in Spanish.

. when subjects are overt. the past tense (ii) favors an SL interpretation.21 Consistent with this fact. 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). Te veo muy guapa (hoy. te sientan muy bien esos pantalones) you-CL I-see very pretty (today. you look very good in those pants)’ b. for example. you look very good in those pants) ‘I see you very pretty (today. Yo te veo muy guapa (no sé por qué tienes tantos complejos) I you-CL see very pretty (not know why have so-many complexes) ‘I see you very pretty (I don’t know why you have so many complexes)’ Interestingly. Yo 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. Yo lo veo muy ignorante I him see very ignorant “I think/for me he is very ignorant” 21 When the subject is a third-person DP.24 Individuals in Time jective perception and the predicate of the SC is understood as IL. Consider the following contrast: (73) a.22 (74) a. (i) Mario te ve muy guapa Mario you sees very pretty “Mario sees you very pretty” (‘Mario thinks that you are very pretty’) (ii) Mario te vio muy guapa Mario you saw very pretty “Mario saw you very pretty” muy guapa (iii) A Mario le pareció que estabas to Mario it seemed that estar-PAST-IMPF-2SG very pretty “Mario thinks/thought you looked very pretty” 22 I owe such an interesting observation to Olga Fernández Soriano. whereas without the overt subject (73a) the sentence should be paraphrased by using the copula estar (74a). those adjectives that go with ser are allowed (74c) . 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). with an overt subject the sentence is paraphrased by using ser (74b). with estar. Veo que estás muy guapa I-see that you-estar-PRES-3SG very pretty “I see that you look very pretty” b. Whereas.

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

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

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

In other words. as I already suggested. such as when(ever)-clauses: . (c) Inability of estar to appear as a perception verb complement In relation to its inability to appear in the progressive. the fact (also noted by Schmitt) that not all adjectival predicates combine with the progressive (cf. (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. like all verbs in the perfect. the reason why estar and the perfect are excluded as complements of perception verbs may be because both behave as states. (81)) remains unaccounted for.) (d) Acceptability of ser in whenever-clauses Schmitt (1992) raises an important issue regarding the acceptability of ser in scenarios considered typical for SL predicates by Kratzer (1988. I will argue that a set of APs (to be delimited) involves particular properties that make the constructions they are in pattern as “activities” rather than “states. is not allowed as a perception verb complement. they may just show that estar predicates behave as states. Schmitt also notes that estar.1. However. 1995).” (I will introduce these notions properly in chapter 3). Chapter 4 develops a proposal about the strict correlation between the type of APs and this syntactic behavior. (81) *Juan estaba siendo esquimal Juan was ser-PROG Eskimo “Juan was being Eskimo” One of the central goals of this book is to establish a link between the type of adjectives and the peculiar aspectual behavior just mentioned.2. (See section 2.28 Individuals in Time aspectual head. it is not clear that tests along the lines of (82) and (83) show that estar designates a result state. 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. and (infinitive) states are not accepted as perception verb complements.

however. I do not share this judgment in Spanish. On my view. That is. ha estado allí para escucharme “Whenever I have needed him. Furthermore. whenever-clauses are allowed only in the presence of an eventive argument. Szabolcsi 1996. the situational variable linked to SL predicates is just one of the possible variables susceptible of “referential variability. I also agree that this does not have to be equated with SL-hood. Some specific examples will help explicate this point. which does not mean that it selects for predicates denoting properties of stages. Consider first the contrast in (87) and (88).25 Schmitt offers an explanation for the acceptability of cases like (84) based on two points. that different adjectival predicates need different conditions to license their appearance in when(ever)clauses. only with SL predicates. First. I would not say that only some adjectival predicates are allowed in these contexts.” which is the crucial factor licensing distributivity (Beghelli 1995.2. Recall that. no matter what type of predicate is involved. as Schmitt seems to intimate. and find (85) and (86) grammatical. 25 . yo le atendí whenever he come-PAST-PERF-3SG to see-me I him attended “Whenever he came to see me. since the copula ser perfectly fits in such contexts. (85) is a simple past-perfective form and (86) a compound perfect form. That when(ever)-clauses select for properties that can be iterated in time seems clear. See section 2. among others). she proposes that when(ever)-clauses do not select for SL-hood but for “certain aspectual properties. from Kratzer’s perspective. that is.1.” According to her judgment. the fact that it is allowed in when(ever)-contexts is completely unexpected under Kratzer’s hypothesis.Individual-Level Predicates 29 cruel. I attended him” (86) Siempre que lo he necesitado. when(ever) can refer to predicates that can be iterated in time. he has been there to listen to me” Schmitt claims that when(ever) selects for predicates that can be distributed over time. if ser is the copula designating IL-hood.” Specifically. (85) Siempre que vino a verme. predicates in perfect aspect are excluded in when(ever)-clauses. he regrets it right after” In effect. rather. se arrepiente después (84) Siempre que Juan es whenever Juan ser-PRES-3SG cruel he-regrets right-after “Whenever Juan is cruel. As the glosses indicate. but. Beghelli & Stowell 1996. she argues that when(ever)-clauses select for “some sort of durative aspect.

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

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

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

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

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

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

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.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. can host an argument (a subject) in its specifier. I saw John draw a circle b. in the presence of a functional projection (Aspect) and an Event Phrase (in the spirit of Kratzer 1988. 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.1 for a discussion in which opposition (eventive/stative or IL/SL) is at stake. besides hosting the event argument. as in (107). *I saw John know the answer As for the existence of an extra projection. Felser argues for the existence of an eventive argument and for the existence of a functional projection that.” rather than stative. 1995) or its lack thereof. .

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

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

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

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

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

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

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

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

with stative though SL predicates. it is worthwhile to notice that when no (habitual) adverb is present. whereas Juan walks does. sentences like Juan is usually sick or Juan is usually at his best friend’s place. (11) Interpretation as habitual in present tense a. Juan draws a circle Nevertheless. Juan knows mathematics b. Juan pasea Usually. Cinque 1999).” as their reaction against combination with habitual adverbs proves. I show that this is a possibility for every type of predicate (including IL predicates) if a suitable context is created. .2). I argue that the crucial condition for a predicate to appear in habitual form is its possibility of restarting.1.2 Activities and States versus Achievements and Accomplishments. are completely grammatical. according to the temporal adverbials that refer to the time an event lasts they admit. Normalmente. 3. It seems clear that Juan is sick or Juan is at his best friend’s place does not have the habitual reading as the only (and preferred) reading. involving an IL stative predicate. whereas (SL) statives can be said to be ambiguous.Event Classes and Individual-Level Predicates 45 habitual (Comrie 1976. I start out with the tests in (13) and (14). Normalmente. Such compatibility cuts the pie of eventualities in two: those that have a habitual interpretation. 4 I thank Tim Stowell for interesting discussion on this topic. Juan walks d. on the one hand. Juan realizes that his mother is right c. Juan traza un círculo Usually. 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. States are argued to be the only type that cannot be interpreted as “habitual. A durative adverbial. However. which divide activities and states. it might seem that this test is a diagnostic for IL-hood rather than for stativity. and those that do not. nonstatives are interpreted as habitual. Bertinetto 1986. Normalmente. is excluded in combination with a habitual adverb (which proves it cannot be interpreted as habitual). The eventualities whose interpretation in the present is not habitual are states.4 Whereas (11a). *Normalmente.2. and achievements and accomplishments. on the other.3. Juan sabe matemáticas Usually. In this subsection I go through a number of tests that separate activities and states from achievements and accomplishments. Juan se da cuenta de que su madre tiene razón Usually.

They express for how long someone has been engaged in a particular activity or for how long a particular state has held of an individual. *Pablo estuvo enfermo en tres semanas Pablo was ill in three weeks b. see Piñón 1999. 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. as (13) shows. which is the reason why they are compatible just with eventualities entailing a delimit point. . 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. As the ungrammaticality of (14) shows. *Pablo viajó en tres semanas Pablo traveled in three weeks c. For a recent account about the different readings available in such sentences. (13) In + x time a.46 Individuals in Time ending point. Pablo se dio cuenta de que su madre tenía razón en tres semanas Pablo realized that his mother was right in three weeks The following contexts also group achievements and accomplishments together. the entailment of a delimit 5 Note that this test does not distinguish between IL and SL predicates. Pablo viajó durante tres semanas Pablo traveled for three weeks c. Pablo construyó una casa en tres semanas Pablo built a house in three weeks d. *Pablo se dio cuenta de que su madre tenía razón durante tres semanas Pablo realized that his mother was right for three weeks In turn. *Pablo construyó una casa durante tres semanas 6 Pablo built a house for three weeks d.

which is that they can be interrupted in their process and their subjects may have the control over the stopping of the event.Event Classes and Individual-Level Predicates 47 point in accomplishments and achievements makes them be incompatible with an assertion of noncompletion. . y todavía sigue llegando Juan arrived. it is not licit to draw (16d) from (16c). accomplishments can be asserted to have taken place just when the process toward a culminating endpoint is over. rather than between IL and SL predicates. However. *Juan arregló la lámpara. in (16). (14) a.1. Juan estaba enfermo y todavía lo está Juan was sick and he still is The next test (Kenny 1963). y todavía sigue arreglándola John fixed the lamp. Pablo ha construido una casa Pablo has built a house 3. Pablo está construyendo una casa Pablo is building a house d.3 Activities and Accomplishments versus States and Achievements. (16) Perfect entailment from the progressive form a. Whereas it is licit to establish (16b) as an entailment from (16a). *Juan llegó. constitutes another diagnostic to differentiate accomplishments from activities. Pablo ha paseado Pablo has walked c. and he is still walking b. and he is still fixing it b. (17) separates predicates that 7 Note that an IL predicate gives the same results as a SL (stative) one like (15b): (i) Juan era camarero y todavía lo es Juan was a waiter and he still is As discussed in chapter 6. since there is no endpoint privileged.7 as (15) proves. and he is still arriving (15) a. Pablo está paseando Pablo is walking b. unlike activities and states. Juan salió a pasear por el parque y todavía sigue paseando Juan went for a walk. it is legitimate to assert that the event has taken place at any point of its process. With activities. but not when it is ongoing. The test in (17) reveals one property that activities and accomplishments share.2. this test and the for-adverbial one distinguish between permanent and nonpermanent properties.

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

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

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

3 Summary of Section 3. which pattern together because of the stative feature they share. but it is a compositional matter. Pablo forzó a Juan a pasear Pablo forced John to walk d. 3. I also introduced the idea that inner aspect does not depend on the inner properties of the verb itself.2. Lo que hizo fue pasear What he did was walking d. Pablo construyó la casa deliberadamente Pablo built the house deliberately c. 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. In the second part. Pablo paseó deliberadamente Pablo walked deliberately d. *Pablo forzó a Juan a estar enfermo Pablo forced John to be sick b. where other elements like the object play a determining role. This accounts for the ungrammaticality of both states and achievements. A “+” .1. *Pablo se dio cuenta de que estaba confundido deliberadamente Pablo realized that he was wrong deliberately (25) As a complement of force 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. Table 3.1 The first part of this section presented the notion of inner aspect.2 summarizes the tests and the behavior of event types. *Pablo estuvo enfermo deliberadamente Pablo was sick deliberately b. *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. Lo que hizo fue construir la casa What he did was build the house c. the good behavior of activities and accomplishments in agency contexts is due to the property of dynamicity the two of them involve. Pablo forzó a Juan a construir la casa Pablo forced John to build the house c. *Lo que hizo fue estar enfermo What he did was be sick b.

as pointed out above in the description of each test. The bracketed numbers in the leftmost column refers to the number of the test as they appear in the text above. Tests for event types Among other things. a “–” when it cannot. (20).2. 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 3.52 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. and a “%” if the event type shows a positive behavior in the test but just under a certain interpretation. (17). and (22)–(25) reveal that activities and accomplishments share an important part of their . the results of the tests (9)–(11).

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

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

12 (30) vP 3 deliberately vP → volition causer 2 DP[+animate] v (vi) B: Sí. where agreement markers switch according to the relative animacy of the arguments. 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. The first guess can be that there are actually two types of small v. 2000. 11 See Chomsky 1995. Now. As a result. Collins 1997.11 any uneconomical proposal becomes automatically undesirable. lo que hizo la insulina fue restablecer los niveles de azúcar. such as Algonquian languages (Berardo 1999) and Southern Tiwa (Rosen 1984). or be tired). rather than the head (small v) by itself. pero no pudo controlar los de acetona. thus far. There are languages with animacy splits. Assuming the generative premises that economy principles are observed by the system of language. but it could not control the acetone levels. A possible alternative is the following. What the insulin did was reestablish sugar levels. but. I am grateful to Olga Fernández Soriano for discussion on this point. 2001a. . There are many other kinds of splits regarding subjects (Ritter & Rosen 2003). a possible question might be how volitional cause is distinguished from nonvolitional cause. 12 It would be crucial to investigate whether there are languages that distinguish overtly between volition or animate causers and nonvolition ones. languages exist that show agent/patient splits. 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.” However. multiplying elements in the lexicon is an uneconomical move. and references therein about this issue. die. Since volition seems to be tied to properties of the argument in the specifier. As is known. such as Lakhota (Mithun 1991). One conveying something like “animate cause” and another one involving “inanimate cause. the DP in its specifier gets interpreted as the ‘causer’. I have not found any clear example showing that an inanimate causer is treated morphosyntactically different from an animate one. given that not all causers are understood the same way.Event Classes and Individual-Level Predicates 55 In the framework of the generative Minimalist Program (Chomsky 1995 and subsequent work). it is usually considered that a functional projection (socalled small v) contributes the content of ‘cause’. It can very well be the case that it is the whole set “specifier + head” that licenses intentional adverbs (or not).

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

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

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

these events are typically activities and accomplishments. that is. As is traditional. In the first place. dynamic events yield grammatical sentences when they appear as complements of parar de ‘stop’ (see (17).2 Agents in Event Structure In this subsection I offer a few introductory words about the possible role of agents in event structure. show a certain correspondence between events that are agentive and those that can appear in the progressive.Event Classes and Individual-Level Predicates 59 3. *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. repeated here). which suggests that only with dynamic events the subject has control over its ending. As the following contrasts show. We can prove that an event involves “internal movement” if it can express its progression in time. I will not consider this kind of case as a severe counterexample to my suggestion.2.1. used as tests to identify the different event types. the verb parar de implies dynamic control by its subject.c. Pablo paró de pasear Pablo stopped walking d.) points out the interesting case of Paró de llover ‘It stopped raining’. where their subject was a [+animate] DP such as Jupiter). . the sentences are ungrammatical. Pablo paró de construir la casa Pablo stopped building the house c. thus cannot be understood as agentive. At this point I just want to notice the consonance of the results of the tests diagnosing dynamicity and of those diagnosing agency. 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. Latin. In the second place. When the subject is [–animate] and.” yet nevertheless is grammatical. Assuming that progression in time defines dynamic events. *Pablo paró de amar a María Pablo stopped loving María b. which I take up later in the work. the syntactic contexts reviewed in section 3. (17) As a complement of parar de ‘stop’ a. Given the special status of the subject of meteorological verbs (cf.15 15 Tim Stowell (p. where the verb parar de does not have a subject that can be said to involve “control. we can conclude that only dynamic events can have agentive subjects.

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

in the case of activities. it seems that three properties can be connected: dynamicity. through the battery of tests introduced above to diagnose event types. Tenny (1987. I introduced the issue of the role of agents in the eventive structure itself as “initiators” of the event. The fact that agents only appear with events that may have an end and whose subjects have the control over its beginning16 leads us to think that agentive subjects are responsible for the initiation and durativity (the sustaining) of the event. They are relevant to some points. Specifically. 1989. 3. is the lexical expression of IL-hood). given the correspondence between certain types of events and certain types of subjects. as argued in chapter 2. (which. the end is inherent. Now. control over the ending. and control. 16 . although all volition agents involve control. I will ignore the differences between imperfect and perfective aspect that exist in Spanish. Agentive subjects can grammaticize an aspectual role—namely. the ending is arbitrary and subject to the subject’s willingness. 1994) and Pustejovsky (1988) observed the role that objects have in the structure of events.3 Aspectual Differences among Individual-Level Predicates This section explores a group of IL predicates from an aspectual perspective. as mentioned earlier. Objects grammaticize telicity. all must involve causation. volition. but only animates can involve volition and control. which has been very often linked to conscious volition only. as Ritter and Rosen (2003) put it. not all controller agents involve volition. Throughout the tests. In the last part of the section. 3. This supposes a remark on the notion of agency. I examine the behavior of copular cases with ser. I looked briefly at some of the notions involved with the concept “agent”: causation. In turn.Event Classes and Individual-Level Predicates 61 (42) La rueda/Juan empezó a girar The wheel/Juan started rotating In sum. In the case of accomplishments. it is natural to wonder whether the agentive argument plays any role in aspectual properties. which overlaps with aspectual notions.3 Summary of Section 3.2. In particular. the initiator of the event. We saw that not all arguments that could be labeled “agents” embrace all such properties. This will constitute the first step in the account of such predicates offered here. and (with the considerations just made) control over the beginning of the event. in the event structure. but I leave the discussion about their interaction with aktionsart to chapter 5.2 This section includes remarks about the tests regarding the thematic property of agency.

(43) Occurrence in the progressive form a.2. states are eventualities that do not “happen. Juan es esquimal Usually Juan is Eskimo b. *Juan estaba siendo rubio Juan was being blond c. as has been widely agreed on in the literature.1 regarding the present-tense interpretation as habitual (see also chapter 5 for more important . 3. to identify the aspectual properties of IL predicates.62 Individuals in Time 3.1 Aspectual Tests on IL Predicates The aim of this subsection is.1. *Normalmente. taken to belong to the group of states.1 Events versus States. I will make use of the tests introduced above in the same order. ?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 estaba siendo muy cruel con Pedro Juan was being very cruel with Pedro (44) Pseudocleft with happen (‘take place’) a.” They are considered to be eventualities that lack any bounding and are inherently durative.” but just “hold. Juan es rubio Usually Juan is blond c.1.3. *Juan estaba siendo esquimal Juan was being an Eskimo b. *Lo que sucedió fue que Juan era rubio What happened was that Juan was blond c. *Normalmente. To do that.3. *Lo que sucedió fue que Juan fue esquimal What happened was that Juan was an Eskimo b. I will concentrate on copular clauses and test out whether all IL predicates are appropriately grouped in the aspectual set of states. Normalmente.” do not “take place. As described above. Juan es cruel con Pedro Usually Juan is cruel with Pedro Despite all the observations introduced in section 3. IL predicates are. In (43)–(45) I reproduce the tests marking for stativity versus eventivity. The eventualities yielding ungrammatical results are taken as stative. in general. The predicates I will test are a classificatory AP (Eskimo) and two qualifying APs (blond and cruel). simply.

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

*Juan fue cruel con su adversario en dos horas Juan was cruel with his opponent in two hours The test in (48) corroborates the results of (46) and (47). (48) Perfect entailment from the progressive form a.3. Once again.1. such a predicate shows a nonstate patterning. In principle. 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. As shown in section 3. *Juan fue esquimal durante tres semanas Juan was an Eskimo for three weeks b.1. *Juan fue esquimal en dos horas Juan was an Eskimo in two hours b.3. *Juan fue rubio en dos horas Juan was blond in two hours c. which argues in favor of considering the predicate (ser cruel) as an activity. tests like (49) divide activities and accomplishments from states and achievements. a predicate like ser + AP would be unexpected as a complement of parar de ‘stop’. As mentioned in section 3. (48b) is a licit entailment from (48a). the combination of ser and an AP such as cruel ‘cruel’ looks good.64 Individuals in Time (46) For + x time a. only activities. Juan está siendo cruel con Pedro Juan is being cruel to Pedro b. yield a perfect entailment from the progressive form. Juan ha sido cruel con Pedro Juan has been cruel to Pedro 3. However. where there is no endpoint.3 Activities and Accomplishments versus States and Achievements.1.2. and not accomplishments. . Juan fue rubio durante tres años Juan was blond for three years c.2.2. since this is only possible with eventive predicates.

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

(52) As a complement of finish a.4.5 Agentivity Tests. which suggests that dynamic properties are somehow involved in these predicates. Juan no empezó a ser cruel con su adversario Juan did not start being cruel to his opponent c.3.3.1. whose combination with almost is impossible. ?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. which also test accomplishments versus all the other eventualities. *Juan terminó de ser esquimal. the latter is with activities. and (b) that the subject did not even start out the process.1. None of the AP predicates with ser. All tests give the same results. Only ser cruel is possible under a command imperative form (54) or in combination with adverbs like deliberately (55). since the unique licit entailment from (53a) is (53b).2. The next set of contrasts aim at testing whether the process that be cruel can be considered is agentive or not. ?Juan casi fue cruel con su adversario Juan almost was cruel to his opponent b. fit in canonical agentive contexts. (53) Interpretative entailments from the adverb almost a. Juan finished being an Eskimo b. *Juan casi fue esquimal Juan almost was an Eskimo 3. The former reading is available with accomplishments. except for cruel.4 Accomplishments versus All the Rest. Although be cruel predicates do not give perfect results after finish. they clearly contrast with the other two. Tests (52) and (53) show some more contrasts consistent with the results from (49). *Juan terminó de ser rubio Juan finished being blond c. . as mentioned in section 3. 20 Recall that. confirm be cruel as an activity. it contrasts with other APs with ser. 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.66 Individuals in Time 3. like (53c).1.20 Although (53a) does not sound extremely natural.

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

2 Summary of Section 3. taking into account the properties of this group introduced in section 3. 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. as shown by a number of tests.2.68 Individuals in Time (59) As a complement of persuade or force a. Rather. differing from other authors’ account. *Lo que Juan hizo fue ser rubio What Juan did was be blond c. among state be-predicates. there seems to be a group whose behavior patterns with canonical activities. *Juan forzó a Pedro a ser esquimal Juan forced Pedro to be Eskimo b. ?Lo que Juan hizo fue ser cruel con su adversario What Juan did was be cruel to his opponent 3. only processes (i. *Juan lamentó haber sido rubio Juan regretted to have been blond c. *Juan lamentó haber sido esquimal Juan regretted to have been an Eskimo b.e. *Lo que Juan hizo fue ser esquimal What Juan did was be Eskimo b. First. 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. we have learned that. as shown in the set of aspectual tests. I will attribute such an eventive-like behavior to the properties of the AP itself.3 In this section we have learned two things.3.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). not all IL predicates belong to the category of states. activities and accomplishments) can be agentive. The leaf was falling down from the tree). Second. This amounts to claiming that there is a group of IL cases that involve typical eventive properties. not all of them can be considered alike. Among them.22 I dedicate detailed attention to these predicates that do not behave as states in chapter 4. 22 Recall that although not all processes are necessary agentive (cf. *Juan forzó a Pedro a ser rubio Juan forced Pedro to be blond c.1.. agency is notable. 21 . where it is attributed to the copular verb be instead.

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

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

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

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

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

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

however. the different interpretation of predicates with modal verbs diagnoses agency versus lack of agency. the modal has two meanings— namely. 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 oddness of (xiv)) or an inchoative one (xv). On the latter the speaker expresses a command that Martha has to effectuate. epistemic and deontic. Roughly described. with an activity such as walk around the park. however. about examples from English such as those pointed out by van Voorst (1988). the speaker expresses a guess about a particular activity she thinks the subject. On the former. the modal has just an epistemic reading.26 I have stopped knowing her/getting to know her I have nothing to say. is usually involved in. The same can be said for accomplishments (80) and achievements (81). (78) Marta debe estar enferma Martha must be sick (79) Marta debe pasear todos los días por el parque Martha must walk every day around the park (80) Marta debe trazar un círculo ahora mismo Martha must draw a circle right now (81) Marta debe llegar al final de la carrera Marta must arrive at the end of the race Note that with the stative verb be sick. the speaker makes a guess about Martha’s state. (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. 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 examples below show where stative verbs have to appear in the progressive form in English. However. Piñón (1995) notes that modal verbs have different interpretations with dynamic and stative predicates. rather than stativity versus dynamicity. the simple present form being ungrammatical. (xiv) *This table is missing a leg more and more (xv) *This table is getting to miss a leg 26 On my view.

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

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

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

Event Classes and Individual-Level Predicates

79

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

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

80

Individuals in Time

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

Event Classes and Individual-Level Predicates

81

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

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

1 The Two-Copulas Hypothesis Partee (1977) and Dowty (1979). attributed the agentive properties displayed by adjectives such as cruel to a different lexical entry of the copula. as has been maintained thus far in the literature. Theoretically. Second. undesirable in itself).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. First. In section 4.1 Previous Explanations of Nonstative Copular Clauses In this section I summarize previous proposals concerning the active behavior observed in certain copular clauses. 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. I will defend the preposition heading the PP complement as an actual aspect head. as I discussed in chapter 3. as shown in chapter 3. different from the “regular copular be”. it finds the obstacle that there is no evidence proving the actual existence of such two homophonous copulas. among others. this is an uneconomical move (and. Recall pairs like (4) and (5). and Demirdache and Uribe-Etxebarría (2000. of cases such as (3). 4. rather than to the properties of the copular verb itself.1. Specifically. with a meaning close to act (2). 2004). therefore. semantically null. Stowell (1993). and empirically. (1) (2) (3) John was cruel on purpose John acted cruelly John is tall (*on purpose) Among the limitations of this analysis. there are the following three. 4. They proposed a theta-assigning active copula for cases such as (1). (4) (5) John broke the window on purpose The ball broke the window (*on purpose) . based on Hale (1984). the two-copulas hypothesis is unappealing because it multiplies lexical entries with no independent evidence to support it. The last section summarizes the chapter. properties like agency cannot be attributed to the lexical entry of the verb but to the whole configuration.84 Individuals in Time presence of a constituent acting as a dynamicity inductor. where the verb is the same but it is observed that agency is a property just of the animate DP subject.

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

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

(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.Aspectual Alternations in IL Predicates 87 Third. refers to “actions” rather than “facts. (23) John/To tell the truth is intelligent AP 5 Spec A′ | | | A John intelligent To tell the truth . (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. from a finer grained typology of eventualities. given the fact that only activities and accomplishments refer to “actions. and the tree in (24) depicts the dyadic one.” (19) States *It was very kind of John to know mathematics/to own a house/to be an African/to want that coat (20) Achievements *It was very cunning of John to reach the top/to recognize the thief/to find the needle. only activities and accomplishments fit. The tree in (23) represents the syntax of the adjective in its monadic usage. the infinitival arguments can be neither states nor achievements. This is expected. represented by the infinitive clause. The structures Stowell (1991) proposes for these potentially dyadic APs are the following.” Compare (17) and (18). As the following contrasts show. the event. it is very imprudent of you to run every day/to swim in the ocean In sum. 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.

whereas the event denoting argument in English is conceived as an infinitive. the subject of the predication is in the specifier of the adjective. For the cases where the adjective is understood as simultaneously predicated of the individual and the action. The controller of this PRO is the DP subject (John). in Spanish there is no such a restriction. which is the DP (John). whose PRO subject is understood as an agent (recall that the only predicates possible here are. (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 . Thus. in Spanish they are not entirely parallel to those in English. in the dyadic cases the adjective is predicated of the action denoting argument. (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. precisely. which allows predicates to project additional maximal projections to accommodate all of their arguments. the performer of such an action. In the first place. As Stowell notes. the dyadic interpretation is obtained this way: the adjective is predicated of the action denoting argument and. whereas in English the event denoting argument cannot co-occur with the complement some MP adjectives can have. Compare (i) and (ii). based on Larson’s (1988) proposal of phrase structure.88 Individuals in Time (24) John was intelligent to tell the truth AP wp Event A′ | 5 A AP | | | 5 | Spec A′ | | | | | | | A 6 [e] John intelligent to tell the truth In the two monadic examples in (23). those that can be agentive. Stowell proposes (24). The adjective takes two arguments via a double-shell structure. gets also qualified as intelligent for having performed the action. states and achievements are excluded). formed by the preposition (a) (‘to’) plus the definite article. by the same token.3 I will not undertake an analysis of these dyadic constructions in this work. in Spanish it has to be introduced by the contraction al. since.

2. However. the event argument represented by the infinitive is implicit in simpler copular cases such as the ones under investigation here. (25) John was cruel to Peter Regarding Stowell’s analysis. in chapter 6 (section 6. the copular verb should be.4 he takes it that the presence of such an argument is proper of SL predicates. he argues that the action denoting argument is an overt version of the (Davidsonian) eventive argument. First. but he is not cruel to him usually” Later. Specifically.2). Ser is completely grammatical. to admit there is an eventive argument that is implicit amounts to admitting that (25) is a SL predication. (26) Juan es muy cruel con Pedro Juan ser-PRES-3SG to Pedro (27) Juan está muy cruel con Pedro Juan estar-PRES-3SG to Pedro (28) Juan está muy cruel con Pedro esta tarde. . I will make two observations.1. necessarily. in his account. see chapter 2 (section 2. 1995) account. Note that the contrast between ser and estar in this case falls in line with the contrasts studied in chapter 2 (section 2. Following Kratzer (1988. this proposal would entail that in all instances of APs followed by a relational PP. I take up the discussion related to the interpretation of the “infinitival argument” in Spanish and its status (as argument or adjunct). (28)). which is the lexical marking for SL-hood. and the contrast between the examples with ser and those with estar (27) can be captured by paraphrases such as the one in (28). According to Stowell. estar.Aspectual Alternations in IL Predicates 89 Stowell (1991) establishes a correspondence between the monadic version of these adjectives and individual-level (IL) predication. pero normalmente Juan estar-PRES-3SG very cruel to Pedro this evening but usually no es cruel con él not ser-PRES-3SG cruel to him “Juan is being very cruel to Pedro this evening. In Spanish. 4 For details of Kratzer’s (1988. that is not case.1) regarding the APs that can appear with either copula (cf. I repeat one of the examples here in (29). as (26) illustrates. and between the dyadic one and stage-level (SL) predication.2.3). 1995).

Second. 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. my proposal will not be based on properties of the copular verb. Thus. it cannot be said it is the copula that induces the active features. cruel con Pedro (*en una hora/durante (30) Juan fue Juan ser-PRET-PERF-3SG cruel to Pedro in an hour/during toda la entrevista) all the interview “Juan was cruel to Pedro (*in an hour/during the whole interview)” However. The idea I will defend here is that nonstative properties of copular clauses are due to the characteristics of the APs heading the SC taken by the copula.90 Individuals in Time guapo. we observed that there were certain copular combinations that behave like states and others that behave like activities. since they are not the same.3). I will then offer an explanation regarding the stative reading also available with such predicates (cf. pero está muy (29) Pablo no es Pablo not ser-PRES-3SG handsome but estar-PRES-3SG very guapo con ese traje handsome in that suit “Pablo is not handsome. My purpose in the remainder of the chapter is twofold. these nonstative copular clauses pattern with activities and not with accomplishments. as shown above. to account for nonstative IL copular clauses. 4. but he looks very handsome in that suit” Given the grammaticality of ser. I will consider these sentences as instances of IL predication. since it is not the copula that varies in the minimal pairs that can be construed (be Eskimo/be cruel) and. which is proved by their acceptation of during-adverbials and the rejection of in-adverbials. Since the activity-like behavior depends on the adjectival predicate combining with the copula. More specific- . I have argued in the previous section that the activity-like behavior should not be attributed to properties of the copular verb.2 Adjectival Predicates Showing an Activity-like Behavior In chapter 3 (section 3. the event argument Stowell proposes corresponds to an accomplishment (21) or activity (22). These two facts lead me to discard this hypothesis based on the existence of implicit event arguments. As shown in the previous chapter. (8) above). as a consequence. I will first elaborate an explanation as to what specific kind of adjectives are the ones that show an activity-agentive patterning and why.

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

*La mesa era ancha/pequeña a propósito The table was wide/small on purpose (40) *El mueble era ligero/pesado a propósito The piece of furniture was light/heavy on purpose (41) *La mesa era blanca/azul/marrón/redonda/cuadrada a propósito The table was white/blue/brown/round/squared on purpose (42) a. *Juan era alto/bajo a propósito Juan was tall/short on purpose b. *La noticia estaba siendo nueva/reciente The piece of news was being new/recent (36) *La mesa estaba siendo preciosa/horrible The table was being beautiful/horrible (37) Juan estaba siendo rápido/lento Juan was being quick/slow (38) Juan estaba siendo *apto/*capaz/inteligente/ingenioso/cruel/ amable Juan was being apt/capable/intelligent/cunning/cruel/kind Combination with on purpose (39) a. *Juan era joven/viejo a propósito Juan was young/old on purpose b. *La noticia era nueva/reciente a propósito The piece of news was new/recent on purpose (43) *La mesa era preciosa/horrible a propósito The table was beautiful/horrible on purpose (44) Juan era rápido/lento a propósito Juan was quick/slow on purpose (45) Juanera*apto/*capaz/*inteligente/*ingenioso/cruel/amable a propósito Juan was apt/capable/intelligent/cunning/cruel/kind on purpose .92 Individuals in Time (35) a. *Juan estaba siendo joven/viejo Juan was being young/old b.

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

and ma as a marker for ‘patients’. I do not mean that each type of the adjectives of the table lexically belongs to states or to activities. who. it takes a different marker than when it is stative (iv). and ?a· reflects semantic agency. I consider that intelligent or cunning in examples such as (54) are states. Mithun describes Lakhota prefix wa as a marker for ‘agents. whereas the interpretation of stative verbs can be habitual but need not. somehow. I do not mean that the inner aspect patterning is lexically given. (i) Lakhota a. their interpretation as processes is easily obtained (55). to· mká·t’ “I am careful” “I am mean” “I am cold” “I’m surprised” In Chichesaw (a Western Muskogean language. that when a complement (involving nouns that can be said to. Very similarly. to· kasíla d. others do because they describe properties that can have certain types of complements. ?a· ?eč·baya c. Central Pomo to· goes with semantic patients. The data are from Mithun (1991). The data are from Munro and Gordon (1982). such as business or jokes) is added. Chichesaw chokma-LI SA-chokma (iii) (iv) 7 “I act good” “I am good” Recall that the interpretation of eventive verbs in present tense is ‘habitual’ (John drives to work. ?a· yá · qač’in b. “take place”. ‘John usually drives to work’). Following Stowell (1991). instigators’. however.7 Note. waksápa b. malákhota “I am prudent” “I am Sioux” (ii) Central Pomo a. 6 In Lakhota (a Sioux language) and Central Pomo (spoken in the Clear Lake in Northern California). . spoken in the American Southwest) there is also evidence that when an adjective has an agentive sense (iii). I will call them “relational MPs”. describe the fact as no systematic. however. That is. performers.94 Individuals in Time can be understood in relation to another individual: be cruel/kind to someone. the core of the following sections is dedicated to show that such complements are the source of the dynamic behavior. See chapter 3 for discussion. In fact. as their interpretation in present tense (not necessarily ‘habitual’) proves. The point I want to make with the classification of (53) is that whereas some adjectives do not seem to easily have the possibility of behaving like activities. As I will amplify later in the chapter. adjectives of this type morphologically show agentive markers.

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

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

(60) The soldiers destroyed the city That the event in (60) is. it . our cases are not like the ones in (62) or (63) since it is not a direct internal argument (a naked DP) the (arguably) affected one. 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. as the suitability of in x time adverbials show. it is not the case that all affected internal arguments actually delimit the event. event delimitation and affectedness has been discussed by other authors. 1988.Aspectual Alternations in IL Predicates 97 Pustejovsky [1988] mentioned in the previous chapter). Getting back to the relational PP of our adjectival cases. In turn. Jackendoff (1996) challenges Tenny’s correlations and points out that it is not always an internal affected argument which delimits the event and. (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. however. but a PP. I will deal with all this in turns. where. 1994) describes “affected argument” as the direct internal argument which undergoes some change. nor do they delimit the event. there are internal arguments that can be considered as “affected. also. distinct PPs (into the house. and delimits the event. like the city in (60).” since they undergo a change. after this brief discussion. Both examples below are from Jackendoff (1996). in effect. nevertheless. The first point I will address is the meaning of the notion of “affected argument. it has become clear that the correlation among internal argumentaffectedness-delimitation as conceived by Tenny is too strict. Humiliate has direct internal arguments and.” Tenny (1987. there is no delimitation of the event in (63). as the ungrammaticality of (64) shows. (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). in what sense can we say that the relational PP is an “affected” goal? Although. over the bridge) delimit the event.

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

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

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

habituals of non state verbs keep their dynamic and agency properties in the habitual forms. as already shown. as has been claimed in the literature. I will keep the idea that the habitual interpretation should not be considered stative in inner aspect terms. While we can have habitual predicates (go walking every day) combined with volition adverbials (iii) and as complements of verbs like force (iv). which invites us to take the parallelism between habituals and statives with a bit of caution. whereas (74) above was considered a contradiction.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. . states are excluded from such contexts (v) and (vi). (i) Juan ha dejado de tener sed Juan has given up being thirsty (ii) Juan ha dejado de pasear todos los días Juan has given up going walking every day 12 Even though habituals of nonstate verbs behave as statives in some contexts. as a consequence. In fact. To begin. habituals are stative. it is true that states and habituals behave the same way in certain contexts. These contrasts suggest that dynamic predicates keep as such even in the contexts they share with states. (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. which suggest that habituals and statives should not be considered to be exactly the same thing. note that. so that there is no crucial difference between (78) and (79). Examples (i) and (ii) illustrate this. One context is after dejar de ‘≈give up’. the interpretation as habitual in present tense is not the unique reading available. the opposite assertion does not sound that strange: One could argue that.

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

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

Examples (93)– (95) demonstrate that the possibility of the PP to appear depends on the nature (animate/inanimate) of the DP subject. (95) a. when the DP subject is inanimate. b. be cruel!” (99) As a complement of forzar ‘force’ *El director forzó a la imagen a ser cruel “The director forced the image to be cruel” . more interestingly. the construction cannot appear in contexts proper of dynamic eventualities. 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]. En Canadá el frío es muy cruel In Canada the cold is really cruel *En Canadá. el frío es muy cruel con los habitantes In Canada. as a complement of verbs like force or in combination with volition adverbials: (98) Occurrence in command Imperative *Imagen. (94) a.104 Individuals in Time (93) a. b. the construction cannot appear in agency scenarios either. such as the imperative form. they show a correlation between the relational PP and the properties that the subject has to possess. such as the progressive or as a complement of parar de ‘stop’. ¡sé cruel! “Image. and. When the subject is inanimate. siendo cruel (96) *Esa imagen está that image estar-PRES-INDIC-3SG ser-ing cruel “That image is being cruel” (97) *Esa imagen ha parado de ser cruel that image has stopped ser-ing cruel “That image has stopped being cruel” As can be expected. b.

it is not inferred from the context or interpreted as generic. dynamicity. (103) Juan es inteligente Juan is intelligent . while the subject of cruel is not an agent. since. 4. with particular characteristics of the construction. I will consider that the presence of the PP correlates with a subject involving particular properties and. I consider these cases with inanimate subjects as evidence that the PP complement cannot be an obligatory complement of cruel.” if we use traditional vocabulary. (104) is not the interpretation of (103). namely. kind. (101) ?Juan fue estúpido con el entrevistador (y no le contestó a ninguna pregunta) “Juan was very stupid to the interviewer (and did not answer any question)” (102) ?Juan fue muy inteligente con su jefe (y consiguió lo que quería) “Juan was very intelligent to his boss (and got what he wanted)” I will make two brief considerations in this respect.). If the PP complement is not overt. In the next sections. its impossibility to appear with certain subjects would remain unaccounted for. as it is possible with the other adjectives (cruel. also. since it needs to bear the required properties for that (and animacy is the most basic one). Consider (101) and (102). I account for the relation between the two inner aspect forms. In the first place.3.3 The Relational PP with Other APs There are other MPs. I therefore. they do not have the same relationship with the PP. but just a “theme. which are not totally excluded. otherwise. Summarizing. which enables agency. as mentioned before. although these adjectives referring to human (or animate) aptitudes can appear in combination with a relational PP. as Violeta Demonte makes me notice. This is because they describe a property that can be understood in relation to another animate entity. can also take a relational PP complement. etc.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. conclude that the structure of cruel-type APs does not always involve a PP complement and aspectually behave as states in principle. not included in the group of relational MPs in the table above which.

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

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

since there are syntactic and semantic facts suggesting its actual substantiation. From a lexical point of view.5 Justifying the Approach 4.5. María tocó la verja (dos veces) María touched the fence (twice) The alternation here is of states versus activities.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.108 Individuals in Time tioned thus far.” Recall in this regard from chapter 3 (section 3. (107) El delantero avanzó hasta la portería The foremost advanced until the goal (108) a. 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. mean. the thematic interpretation of the subject and particular aspectual properties.2) the contrast between animate and inanimate DPs in pairs like (109) and (110). 4. In (107) it is a PP which is exerting a crucial aspectual role in making the event an accomplishment.). I begin by discussing which approach to event representation can make sense of the observed correlations among the syntactic presence of a constituent. Along the lines of Borer (2005) (among others). etc.1 Lexicalist and Logico-Semantic Approaches The observed correlations among the presence of the PP relational complement. we would be forced to assume that each aspectual behavior of adjectival copular cases (stative and nonstative) corresponds to different (but homophonous) lexical items. . it does not allude to the other elements present in the syntactic configuration. according to which aspect properties are fully encoded in the lexical items. That is. However. kind. In the remainder of the chapter I will be concerned with such a contrast in the realm of adjectival copular clauses. by definition. whose contrast has been debated in the literature (see section 3. This perspective cannot capture any of the other properties (interpretation of the subject. two different entries (stative and dynamic) for each adjective (cruel. both atelic eventualities. syntactic presence of a PP) as correlations since. I argued that such a contrast is grammatically relevant. The pair in (108) from Borer (2005) shows an alternation stative/dynamic which correlates with the [±animate] properties of the DP subject. La pared tocaba la verja The wall touched the fence b.

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

As discussed earlier (see section 3.1 Event Roles. 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.” Second.” given the fact that both are atelic and do not “culminate.5. This perspective does not make any direct prediction about the correlation between the aspectual properties of a VP and the thematic properties of the DPs. as captured in the Event Correspondence Rule (116). in principle. each entity corresponds to an aspectual notion. and the presence or absence of other elements in the sentence (the PP complement in our cases) that.1). there is no obvious way to distinguish between the two constructions in point (stative and dynamic copular clauses). 1994). both would contain the term “hold. First. its syntactic position and its role in the determination of the aspectual properties of the sentence.2 Syntactic Approaches 4. arguably.5. 1989. Tenny (1987. which opens the possibility of establishing a direct relationship between the thematic properties of the arguments present in a sentence.1. this approach does not establish any relationship between the aspectual term (Hold). since.2. As van Voorst puts it. and van Voorst (1988). (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 . 1989. 4. authors such as Verkuyl (1972). these authors emphasized the role of the internal object as an event delimiter. Tenny (1987.110 Individuals in Time This approach leaves several questions unanswered. among others. correlate with the aspectual outfits of the construction. 1994) and van Voorst (1988) elaborate both on the observation that arguments grammaticize the points that give the temporal contour of an event. In particular. showed the relevance of arguments in the inner aspect properties of a sentence. the dynamic or stative properties. Dowty (1979).

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

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

16 For Borer. As mentioned in chapter 3. 17 This was introduced in chapter 3 (section 3. When considering that functional projections are in charge of event-roles too. that gives the type of aspectual interpretation. This head has the property of “quantity” (AspQMAX or Quantity). Mostly along the lines of Verkuyl (1972.Aspectual Alternations in IL Predicates 113 and functional structure is interpreted as event structure. 1989.15 In what follows. 2000) argue that functional projections responsible for Case processes (Agreement projections) assign the event-roles of delimitation and initiation. It is the syntactic structure. These works are framed in the Minimalist Program (Chomsky 1995. 2001a. 2005) defends the idea that Case-checking processes are satisfied through aspectual projections. she argues that aspectual properties of a determined event All the authors cited make proposals in a similar spirit in this regard. Ritter and Rosen (1996. Borer (1994. where functional projections are in charge of Case assignment and agreement processes. whereas a DP headed by a numeral. events are not lexically specified for any type of aktionsart.14. In other words. Compare: (124) John drank beer (125) John drank two beers A bare mass noun gives an atelic interpretation. Likewise. Thus. with telicity. 15 For further discussion about the relationship between case and event structrure. whose core is constituted by aspectual projections. 1998. I will refer to her book (Borer 2005). a verb projects an unordered number of arguments whose interpretation depends just on the functional projection they move to. see Kratzer 2004 and Kiparsky 1998. 1998. all the relevant grammatical processes get a unified account. events are not states or accomplishments from the lexicon. That is. 1999. that is. argument structure is computed on the basis of the syntactic structure.4). 16 The work by Borer has appeared progressively in a number of articles. Borer proposes that the core functional projection of the clause is Aspect. in turn. an object that can be marked as [–quantity] correlates with atelicity. A [+quantity] object. since it presents the latest versions of her ideas and proposals. More accurately. the quantification properties of the DP direct object make a direct difference regarding telicity. 2001b). a telic one. 2001. Event-roles are not assigned by a particular head but are simply derived as an entailment from the aktionsart of the whole event. I introduce some points of the work by Borer (2005). Borer (2005) elaborates on the intimate relationship between quantification and telicity or lack thereof.17 Borer proposes that the core feature of aspectual distinctions is [±quantity]. which I take as a frame of reference for the remaining discussion. In turn. 14 . 2000). aspectual properties can be recast into quantity properties. 2000. Assuming the parallel distinctions between mass/count nouns and atelic/telic verb interpretation.

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

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

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

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

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

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

these authors also note the important role of verbs as aspectual sources. “be there” (Krio). María está en casa Maria is at home (ii) María está estudiando Maria is studying (i) This situation is found across languages. In Spanish. and spatial and temporal locations conflated. “sleep”/“spend the night” (Bemba [Bantu]). the use of locative (central coincidence) prepositions to express ongoing activities can be nicely illustrated with examples such as (141B). I am at it Demirdache and Uribe-Etxebarría (1997.22 For example. estoy en ello No. As can be appreciated. The following examples are from these two works respectively: (139) John is on/at hunting/building a house (140) He was a. where the auxiliary is the locative copula estar. “lie down/stand”. “sit” (Siluyana [Bantu]. Sentence (i) illustrates a pure locative context of estar.coming home Other pieces of (synchronic) evidence argue in the same direction. “be with” (Swahili). Observe the verbs used as progressive-durative markers. Juba Arabic). The purity of the temporal meaning emerges when the locative sense is lost. this preposition in combination with a gerund noun phrase is the origin of such an aspectual form. Perkins. they show that prepositions of central coincidence are a very common source for progressive aspect. Such an idea is nicely illustrated by the progressive forms in Spanish. “live/reside” (Hindi). They suggest that locative prepositions were first used to indicate that the subject was involved in an activity at a certain location. 2000) point out that the progressive form in Basque is construed by combining the verb ari ‘engage’ with the nominalized form of the verb suffixed with the locative postposition -n 22 Although I am focusing on the aspectual value of prepositions. . the auxiliary in progressives is a verb with a locative component. (ii) is an example of progressive usage. the English locative preposition on or at (shortened to a-. whereas those of noncentral coincidence are on the basis of prospective aspect.120 Individuals in Time historic and cross-linguistic study. (141) A: ¿Has hecho la cama? Have you made your bed? B: No. as in asleep) is considered the historical antecedent for the English progressive. The data are from Givón (1982): “stay” (Hawaii-Creole). According to Vlach (1981) and Bybee. and Pagliuca (1994).

similar cases are found in Romance languages. in Nigerian Margi and Palaung (from the AustroAsiatic family of languages). such as from. 23 Note also the presence of a verb denoting movement. 2004) conclude that Aspect can be analyzed as a head establishing relations between two terms. where “come from” can be used with the aspectual meaning of the perfect. the locative preposition at directly combines with the verb in infinitival form to construe the progressive. prepositions denoting ‘movement from’ (centrifugal movement). ‘to’) (144) or the English about to (145) are constituent parts of prospective forms. the combination of a motion verb with an elative component and the preposition from yields a perfect interpretation: (146) Max COME FROM home ‘Max has come from home’ As Demirdache and Uribe-Etxebarría (1999) notice. as prepositions do. prepositions of noncentral coincidence are present in the auxiliaries expressing perfect and prospective aspect across languages. to (143). the prepositions a (142). in Spanish and English. (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). 2000. are present in perfect aspectual forms in some languages. For example.23 Prepositions like the Italian per (‘for’. indicating movement toward (centripetal movement). in contrast to stative locative verbs participating in progressive forms. In turn.Aspectual Alternations in IL Predicates 121 ‘(with)in/on/at’. . For example. (147) French Je viens d’être malade I come from be sick “I have been sick” (148) Spanish Vengo de matricularme en la Universidad come-I from register-me in the university “I have just registered at the university” Since prepositions express aspect relations. Demirdache and UribeEtxebarría (1997. Also. Specifically. they notice that. appear in the form used to express close future (going to). in Dutch.

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

I propose that this is the case. 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. and.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. con-PPs can also appear with motion verbs. 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.2. it is conceivable that prepositions can also convey inner aspect properties. 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. and assuming with Demirdache and Uribe-Etxebarría that all temporal relations can be reduced to the same set of primitives.1 Con: A Locative and a Directional Preposition. 4. The preposition introducing the PP in Spanish is. ¡corre con tu papa! Pablito.6. as seen in the examples above. I am going to argue that the preposition introducing the (affected) goal carries inner aspectual properties affecting the aspectual nature of the whole construction. the preposition con. 4. Preposition con establishes a central coincidence relation between the figure and the ground (in the terms introduced before). in these cases the preposition con is in complementary distribution with the directional preposition a ‘to’ when the DP complement is [– animate] and is interpreted as a location: .6. denoting the goal. This preposition is associated with stative locational meanings.Aspectual Alternations in IL Predicates 123 Assuming that aspectual meaning is reducible to prepositions (and. that prepositions have aspectual meaning). conversely. paying special attention to the APs I am taking care of in this chapter. con ‘with’. like in examples such as the following: (153) Pablito.

in previous centuries. XIII] (159) (…)& guardandosse segunt su aluedrio fue muy cruel contra los que lo non merecían & keeping himself according to his will. in previous periods of Spanish. 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. as a path with a goal (a Goal Path. (157) Con + [+animate] DPs A + [–animate] DPs Therefore. as (157) summarizes. XIII] (160) Paulo orosio (…) que era muy cruel sobre los ciudadanos Paulo orosio (…) who was very cruel over the citizens [Alfonso X: General Estoria IV. the only difference between con ‘with’ and a ‘to’ is the nature of the DP acting as their complements.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] 25 I would like to thank Isabel Pérez for bringing this source into my knowledge. It is interesting to note that. or hurt them… [Alfonso X : Siete partidas. Svenonius 2004). Consider first the following examples taken from the corpus of Spanish from the Illinois State University (Davies 1999). the prepositions introducing the relational complement in cruel-type cases were of same characteristics as con. 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’). he was very cruel against whom did not deserve it [Alfonso X: General Estoria V. con (‘with’) and a (‘to’) can be said to share properties and can be both analyzed as prepositions denoting a trajectory with a destination. 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. with a goal. contra (‘against’). S.25 As reported in the corpus. 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.

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

that is. and (183) represents it graphically. associated with ‘future’ in the realm of Tense and with ‘prospective’ in the domain of outer aspect. They are. it designates the movement of a figure toward a place or ground. (181) Juan went to Pedro In the terms of figure and ground presented before.Aspectual Alternations in IL Predicates 131 (180) Juan fue cruel con Pedro ser 3 ser cruel 3 cruel GoalPathP 3 Juan GoalPathP 3 GoalPath PlaceP (para-Ø) 3 Place DP (con) → dynamicity Bearing in mind this discussion. I consider that the relation between the DP subject Juan and the goal Pedro is not very different from the one existing between Juan and Pedro in (181). 2004) proposed that centripetal noncentral prepositions translate as “before” in the temporal domain.1. The two terms are put under a specific relationship according to the meaning of the preposition. 2000. we saw that Demirdache and Uribe-Etxebarría (1997. therefore. it locates the figure in its trajectory toward the ground.) (182) (183) TOWARD: preposition that establishes a centripetal noncentral coincidence relation between the figure and the ground. The two of them are related by a preposition meaning ‘toward’. (182) summarizes the description of the preposition. the discontinuous dotted line for the trajectory. As a directional preposition.6. (The bullet stands for the figure and the little house for the ground. …● ⌂ In section 4. the DP subject can be considered as the figure and the DP inside the PP as the ground. 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 .

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

it suggests that the properties of the noun in the DP subject and the PP stand in a codependent relationship. 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” . (188) Juan/*esa imagen está siendo cruel Juan/that image estar-PRES-INDIC-3SG ser-ing cruel (189) Juan/*esa imagen ha parado de ser cruel Juan/that image has stopped ser-ing cruel Likewise. as a complement of verbs like force or in combination with volition adverbials: (190) Occurrence in command imperative Juan/*imagen. such as the progressive or as a complement of parar de ‘stop’. 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.3 The DP Subject of Relational MPs As I mentioned in the beginning of the section. when the subject is inanimate. And second. If that were the case. 4. First. the construction cannot appear in contexts proper of dynamic eventualities. when the DP subject is inanimate. the PP complement would be possible independently from the properties of the DP subject. it suggests that the PP is not a plain complement of the adjective. eventuality properties are obtained from the very syntactic structure. 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.2.6. As shown in section 4. This fact suggests two things. the construction cannot appear in agency scenarios either. when adjectives such as cruel are predicated of an inanimate subject.Aspectual Alternations in IL Predicates 133 construction is decided. In sum. ¡sé cruel! “Juan/image. and we could not explain why it cannot appear in sentences like (186). I repeat the contrasts below. the relational PP complement cannot appear.3. such as the imperative form.

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

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

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

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

we observe. it is not the case that “cruel + PP” fits in all cases. This hypothesis makes the prediction that such active properties should be retained when the taking verb is other than the copula. can take . however. In my proposal. traditionally considered very close to be.1 “Cruel-type” Small Clauses Taken by Verbs Other Than the Copula I will start by considering the existing contrast between cruel and “cruel + PP” as a complement of parecer ‘seem’. as a complement of hacer ‘make’ and as a complement of volver(se) ‘become’. 4. Partee (1977) and Dowty (1979) argue for a homophonous agentive copula with a meaning close to act. First. In the set of cases above. in all the examples.138 Individuals in Time chapter. three things. a verb traditionally considered “pseudocopulative. a stative SC (with the bare AP) is accepted. However. volverse ‘become’. the activity-like properties are derived from the properties of the (adjectival) predicate heading the small clause (SC) taken by the copular verb. when the PP is present. the constructions get degraded in some cases. Second. at least.” with an intrinsic inchoative meaning. In the following paragraphs I attempt to explain the contrast between allowing and disallowing an “AP + PP” SC.7. the verb seem. but when the SC contains the adjective plus the PP (197) it gets degraded. As the examples show. The aim of this section is to analyze whether this is borne out. (196) María parece cruel Maria seems cruel (197) ?María parece cruel con Pablo Maria seems cruel to Pablo (198) María se volvió cruel Maria CL-REFL become cruel (199) María se volvió cruel con Pablo Maria CL-REFL become cruel to Pablo (200) María volvió cruel a Juan Maria turned Juan cruel (201) ??María volvió cruel a Juan con Pablo Maria turned Juan cruel to Pablo (202) María hizo cruel a Juan Maria made Juan cruel (203) ??María hizo cruel a Juan con Pablo Maria made Juan cruel to Pablo Note that. can take an SC where the AP appears by itself (196). and Rothstein (1999) defends that the copular verb maps states onto eventuality-denoting predicates.

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

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

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

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

I have shown cases where the PP cannot appear overtly. the better it bears a command imperative. Although null. 37 In imperative forms with ser. unlike previous accounts (Partee 1977. Dowty 1979.Aspectual Alternations in IL Predicates 143 verb is. become. I also discussed Stowell’s (1991) work.2 Summary of Section 4.1 previous accounts by Partee (1977). I have shown that this is not necessarily the case. In order to support the hypothesis that cruel-type APs can also appear without the PP complement. corresponds to the presence of a (relational-directional) PP. but it is only present if the PP complement is also present. in the cases where an explicit PP is odd. no null PP would be proposed. I have shown that the dynamic behavior is not inherent to this type of APs. and make. the better it accepts the active SC. I argued that to attribute the dynamic properties to pieces inside the SC.7 I have analyzed the behavior of relational MP SCs as a complement of verbs other than be.8 Summary of the Chapter In this chapter I have accounted for those copular constructions that behave as activities from an inner aspect perspective. Rothstein 1999). the P brings the dynamic properties enabling forms such as imperatives. as the full grammaticality of Spanish copular clauses with ser suggests. which can make sense of the preferences of different predicates such as seem. I have shown that this aspectual behavior is not idiosyncratic or unpredictable but. Taking into consideration the aspectual properties of the SC. This fact led me to reject in section 4. rather. which proves its empirical superiority. makes more predictions. mean. Dowty (1979). 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 explained as aspectual (in)-compatibilities the appropriateness of the relational MP SCs with each verb. Correspondingly. I assume a null PP when it is not overt. where they were attributed to the copula. Such a hypothesis has to admit that in a number of cases it is not possible to say at first sight whether the sentence is an activity (with a PP null) or a state. I have 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). .37 4. more suitable with those adjectives describing properties that can be understood in relation to another animate entity (cruel. accordingly. 4. and Rothstein (1999). I showed that the properties attributed to the SC headed by the adjective are “active” independently of the taking verb.7. according to which such dynamic properties are in correlation with the status of the predicate as SL. kind to someone).

I have considered that the (most) basic structure corresponds . Any adjective (e. 2004). Davis (in prep. following a strong configurational perspective such as that defended by Borer (2005) or Ritter and Rosen (2000). Regarding the broader theoretical issue about the real grammatical difference between states and activities. 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). Based on the adjectival constructions I have examined.). it becomes an activity because the adjective and the PP are met in the structure. I have treated the alternation state/activity as an instance of aspect shift. this amounts to saying that aspectual properties are not lexically encoded but are decided through the syntactic construction of the clause. The interpretation of the subject (as an agent) has been explained as an entailment from the aktionsart properties. While it is due to the lexical properties of the adjectives that they can easily take a relational complement or not. I have argued that the difference between these two types is dynamicity. all temporal relationships: Tense. and (inner) Aspect. but the syntactic structure corresponding to a relational meaning they are inserted in that yields such particular aspect properties.144 Individuals in Time Examples of this are sentences with inanimate DP subjects and clauses where the adjectival SC is a complement of the verb consider.g. As a consequence. (outer) Aspect. Aspect-Quantity being the unique relevant head and it being absent in both cases. as a consequence. Rather. Once in such a structure. those adjectives that more naturally allow for such a relational complement are the ones that manifest active aspect properties. and others that prepositions can bear aspect information. it would involve dynamic properties. it is not because of the lexical properties of the adjective that the construction is an activity. triggered by the preposition. I have concluded that they are not to be structurally distinguished through a purely functional projection. Thus. Therefore. have been reduced to the same set of temporal primitives. the account exposed here highlights the relationship between prepositions and event structure. Stowell (1993). Theoretically. which can be brought onto the construction through lexical indirect or oblique arguments (PPs). 2000. it is not adjectives themselves. the hearer would have to come up with a suitable interpretation for the construction. Based on evidence brought up by Hale (1984). Likewise. taking seriously the correlation between the PP complement and the active behavior. both homogeneous predicates. bald) would acquire a relational meaning if inserted in a relational structure and. Demirdache and Uribe-Etxebarría (1997. That is..

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

.

through the functional projection of Quantity. I propose in section 5. I discuss the relationship between inner and outer aspect. 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.1). and its absence amounts to leaving the predicate homogeneous. Since. so that the complete articulation of Tense and Aspect and their parallel behavior can be easily contemplated. I show that the perfective form does not contribute quantity (i. the external argument of Aspect is the internal argument of Tense. I am going to deal now with another aspectual realm. the perfective.e. I assumed that different event types are decided through functional structure. 1995) and Demirdache and Uribe-Etxebarría (2000).2 that aspect viewpoints also involve quantification over occasions. 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. . outer-aspect properties do not fulfill quantity (i. 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. In section 5. Following Borer (2005). specifically. 1 The particulars of temporal interpretation are the topic of the next chapter. Centered on individual-level predicates. Throughout the discussion I will focus on three particular viewpoints: the imperfect. Focusing on the imperfect and perfective. according to Demirdache and Uribe-Etxebarría’s proposal. I will also present how Tense works here.. for example.Chapter 5 Outer Aspect and Individual-Level Predicates In the two previous chapters I dealt with inner-aspect properties. which I will capture through a quantifier inside the Aspect composition. namely.3. outer aspect. Its presence makes the predicate heterogeneous.e. inner aspect) properties. and the progressive. 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.. Following Klein (1994.1 Besides the ordering component just mentioned. I conclude that. in Spanish. telicity) to the construction and the imperfect is compatible with quantity predicates (telic predicates). I also introduced some discussion about the difference between the two types of homogeneous predicates—states and activities.

atelic). In particular. I show that states are compatible even with the perfective. Since the line of this work is not formal semantics I do not discuss them here. I argued that dynamicity properties (that activities involve) can be described through prepositions. in some languages (e.g. and inclusion are semantically defined as partial order relations fulfilling a number of properties. as something that took place in the past). 2 . as Tense and Aspect are considered ordering predicates. and simultaneity). and Eventuality Time). that establish centripetal noncentral coincidence relations between a figure and a ground. 5. For an introduction of such notions. posteriority. These predicates translate into the temporal domain as ‘before’.1). inner-aspect characterization may work as an aspect indicator (an accomplishment can be taken as a perfect action.148 Individuals in Time In section 5. but in dynamicity versus lack of dynamicity. The idea is depicted in (1). 1996) The idea that temporal interpretation can be conceived as a relationship between the time of the event the sentence is about and a given time traces back at least to Reichenbach 1947. and Wall 1993. I argued that inner aspectual properties could also be conceived in ordering terms.. based on Zagona 1990. At that point. in Spanish at least. for example) and.3 Authors such as Zagona (1990) and Stowell (1993. and the possible relationships that can be established among them (anteriority. Consistent with the conclusion that viewpoints do not interfere with quantity properties.e.2 One of the facts I alluded to was that. I turn my attention to adjectival IL predicates and show the different interpretations available with each type of viewpoint mentioned. posteriority. Speech Time. I show that. like to.6. by the same token. which I mentioned briefly in chapter 4 (section 4. therefore. I tried to show that all aspectual properties could be reduced to the same set of primitives. aspect properties can serve as indicators of temporal properties (a perfect action is computed as finished and.1 Tense and Aspect as Ordering Predicates In this section I deal in greater detail with the conception of Tense and Aspect as ordering predicates. Stowell (1993. 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.1 Tense.. 5.5. Lillooet Salish). the interested reader is referred to Partee. Reichenbach gave a formalization of tenses in natural languages based on three intervals (Reference Time. 3 Anteriority.1. Therefore. It turns out that this cannot be described in terms of presence or absence of quantity. inner aspect can be as well. the progressive seems to impose restrictions on the type of eventuality. However. specifically. ter Meulen. this being why the eventuality is computed as “dynamic and unbounded” (i.

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

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

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

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

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

(i) Perfective → the leaving occurs at 3 (ii) Perfect → the leaving occurs prior to 3 .3SG Había salido a las tres he-had left at 39 Había salido ya a las tres he-had left already (at 3) Trabajará he-will-work Trabajaría he-would-work Table 5. which is the element in charge of establishing a relationship between these two intervals. Tense and Aspect consist in the same mechanism—namely. 9 As many authors have pointed out.3SG Trabaja-ba he-work-IMPF-PRET. is an ordering predicate. Aspect nomenclature (García 1999) As indicated earlier. the adverbial refers to the time at which the event takes place. I restrict my attention to the forms called “imperfect” and “perfective” (and concretely to the perfective simple form) as well as with the progressive. with the perfective. Thus. that of ordering temporal arguments. More strictly speaking. whereas with the perfect. the next question is how the singled-out portion of the situation (TT) relates to the rest of the event. I follow Klein’s (1994) proposal that there is an ordering relationship between the TT and the ET. In the glosses for my Spanish examples. the difference between the composed forms of “have” corresponding to the perfective and the perfect is that. how does the interval corresponding to the singled out portion of the situation (the TT) relate to the rest of the ET? As mentioned earlier. Aspect. the event has already taken place at the time denoted by the adverbial. The progressive is construed by the verb estar (locative be) plus the ing-form of the verb (Spanish -ndo). I mention them here only as clarifications.3SG Trabajó he-work-PERF-PRET.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. The verb estar acts as an auxiliary and can appear in any tense and aspect form. That is. as Demirdache and Uribe-Etxebarría (2000) put it. I am not going to analyze all these aspect viewpoints.1. Returning to the internal working of Aspect. I will thus use “perfective preterit” and “imperfect preterit” to refer to the two simple past forms.

10 Actually. ‘before’. with the perfective. Tense orders the TT with respect to an RT. Likewise. the asserted part is within the situation. la canción a las 3 (i) Juan cantó Juan sing-PERF-PRET-3SG the song at 3 “Juan sang the song at 3” Accomplishment . or to the end of the event (iii). Although the working of aspect was also described in chapter 4. ‘within’). Tense and Aspect only differ with respect to which arguments these predicates take to order. However. the TT is partly included in the interval at which the event takes place. As Klein (1994:109) points out. (14) RT T TP 2 T′ 2 AspP 2 TT Asp′ 2 Asp VP 2 ET VP 2 e VP In the previous section I showed the precise ways the different values of T (‘before’. for the sake of completeness I repeat it here. I will not make distinctions among the different types of perfective. depending on the inner-aspect properties of the predicates. the perfective can refer either to the time at which the event started.6. ‘after’. with the perfective. I will assume that the ordering predicate for the perfective is ‘after’. while with the perfect the TT is conceived ‘after’ the ET. aspectual relationships can be described through the same type of prepositions as Tense’s relationships (‘after’. (i) and (ii). it appears before. As Bertinetto (1986) points out. and the dotted line the entire situation. The slashes indicate the part of the situation asserted (TT). the content of both predicates can be reduced to the same set of primitives. ‘[with]in’) work in the tense realm. this is an oversimplification. since nothing of what I will comment on hinges on these distinctions. without entering into the differences it displays with the perfect. The structure in (14) captures all this.Outer Aspect and Individual-Level Predicates 155 As these authors also point out. With the progressive (15). the assertion time is after the situation. As shown in chapter 4 (section 4. Aspect orders the TT with respect to the situation time or ET.10 and when the prospective is involved.1).

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

” as mentioned in the previous section. such as Bertinetto (2000). As discussed back in chapter 3 (section 3. I will argue that. In contrast. it does not mean that he has been walking at every single moment of that interval.4).24 as subinterval and additivity proofs show. In the same vein as before. and others. who conceive the perfective as a modifier that makes an eventuality telic. If we say John walked from 2 to 3.2 Perfective Homogeneous Predications I turn my attention now to homogeneous predicates appearing in perfective form. a perfective form does not interfere with the quantity nature of the predicate. The meaning of these sentences is that Pablo was once involved in the task of swimming (91) and walking (92). I argue that a perfective form does not trigger telicity effects on predicates. activities are not 100 percent homogeneous at a micro-level. he has had that car at every single moment between 1974 and 1985. 24 23 . 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. Bach (1986). which leads to the inference of “completeness” and “boundedness. Contrary to Mourelatos (1978). (95) *Pablo nadó en una hora Pablo swim-PRET-PERF-3SG in an hour “Pablo swam in an hour” Other authors. unlike states. at least in Spanish. if we say John had that car from 1974 to 1985. In other words. draw a conclusion in the same direction.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. a perfective form does not make a homogeneous predicate heterogeneous. sentences such as these show that the presence of the perfective viewpoint does not contribute telicity. the TT is located ‘after’ the ET.3. However. (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.176 Individuals in Time 5. Piñón (1995). de Swart (1998). With the perfective.

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

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

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

and progressive). Juan ha sido cruel con Pedro Juan has ser-PAST-PART cruel to Pedro “Juan has been cruel to Pedro” I will concentrate now on the analysis of these adjectival predicates in copular constructions regarding outer aspectual information. I will explore the interpretation that adjectival IL predicates have under the aspect forms mentioned thus far (perfective. *(En aquella ocasión) Juan estaba siendo rubio that time Juan was ser-ing blond “(That time) Juan was being blond” c. (En aquella entrevista) Juan estaba siendo muy cruel in that interview Juan was ser-ing very cruel con el entrevistador to the interviewer “(In that interview) Juan was being very cruel to the interviewer” (102) Perfect entailment from the progressive form a.3 (about the relation between quantity and perfective and imperfect viewpoint). This restriction differs from the one discussed in section 5. Consider the following cases: esquimal (103) Pablo era Pablo ser-IMPF-PRET-3SG Eskimo “Pablo was Eskimo” rubio (104) Pablo era Pablo ser-IMPF-PRET-3SG blond “Pablo was blond” . 5.5. Juan está siendo cruel con Pedro Juan is ser-ing cruel to Pedro “Juan is being cruel to Pedro” b. imperfect.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. This will lead me to discuss the restriction existing between the progressive and the property of stativity mentioned in chapter 4. *(En aquella ocasión) Juan estaba siendo esquimal that time Juan was ser-ing Eskimo “(That time) Juan was being Eskimo” b.180 Individuals in Time (101) Progressive form a.

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

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

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

(119) and (120)). make the habitual reading available. as noted in section 5. this is just marginally available with be cruel (cf. Likewise. rather than the habitual (cf. see section 5. when a relational complement is added (123). the preferred interpretation of adjectives denoting intellectual aptitudes (121) is the continuous reading. as noted above. 29 .5.3.2.184 Individuals in Time It is worth noticing that. estúpido/inteligente (121) Juan era Juan ser-IMPF-PRET-3SG stupid/intelligent “Juan was stupid/intelligent” (122) ??Juan used to be stupid/intelligent muy estúpido con su hermano (123) Juan era Juan ser-IMPF-PRET-3SG very stupid to his brother “Juan was stupid to his brother” (124) “Juan used to be stupid to his brother” The examples are based on temporal subordinate clauses because. when a complement that can act as a “distributee” is added (125). although the progressive form is correct with cruel. (122)). 30 For the reasons why the progressive is excluded with Eskimo and blond. in Spanish. the progressive is not a paraphrase saliently available from the imperfect form. the progressive paraphrase of the imperfect is more natural in these contexts.30 Finally. accordingly. However. a habitual interpretation emerges. they acquire activity-like properties.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. whereas activities in imperfect (117) can also have a progressive paraphrase (118). which.1.

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

Regarding nonpermanent predicates. 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. The perfective form does not make a homogeneous predicate heterogeneous or telic. Li was Eskimo” una serpiente venenosa (En una vida anterior). the following cases sound quite acceptable: (133) esquimal (En una vida anterior). or Juan was very intelligent in doing something (139). Li was a poisonous snake” (134) Again. in principle. Be Eskimo and be a poisonous snake are homogeneous predicates. Permanent predicates are said not to combine with the perfective form. compatible even with this kind of predicate. 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). regardless of the aspectual viewpoint at stake. Li fue una serpiente venenosa en una hora in a previous life Li ser-PERF-PRET-3SG a poisonous snake in an hour “In a previous life. Li era habitualmente chino in his human incarnations Li ser-IMPF-PRET-3SG usually Chinese en una hora in an hour “In his human incarnations. If they are true of an interval of time. Note that an in + time adverbial is excluded with the perfective as well as with the habitual imperfect: (135) *(En una vida anterior). despite the overt absence of the PP.5. With respect to be cruel or be intelligent. However. as argued before. if an appropriate context is created. Li fue in a previous life Li ser-PERF-PRET-3SG Eskimo “In a previous life.186 Individuals in Time 5.2 The Perfective and Adjectival IL Predicates I will start by discussing the interpretation of the perfective with permanent predicates. Li was a poisonous snake in an hour” (136) *En sus encarnaciones humanas. they are true of every subinterval of that time. 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. a viewpoint based on a Q<occ> = |1| is. .

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

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

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

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

Landman 1992. Vlach 1981. do not come from the properties encoded in the aspectual functional projection of Quantity. it is compatible with an in + time adverbial (152). before getting the endpoint) are incompatible because both . as a result. Delfitto & Bertinetto 1995. the expression of progression and the expression of telicity cannot cooccur. This is known as the “imperfective paradox” (Dowty 1979) or the “progressive paradox” (see.. makes possible their conception in progress. I have argued that the properties enabling the use of the progressive are not of mereological nature and. which. However. among others. Specifically. Asher 1991. the progressive is incompatible with adverbials denoting telicity. such as in + time.e. it has to be noticed that the progressive interferes in a particular way with the mereological properties of the predicates. therefore. (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. as a telic predicate.Outer Aspect and Individual-Level Predicates 191 siendo muy ingenioso (147) ?Juan estaba Juan estar-PRET IMPF-3SG ser-ing very cunning “Juan was being very cunning” siendo muy inteligente en la entrevista (148) Juan estaba Juan estar-PRET-IMPF-3SG ser-ing very intelligent in the interview “Juan was being very intelligent in the interview’ siendo muy ingenioso en sus bromas (149) Juan estaba Juan estar-PRET-IMPF-3SG ser-ing very cunning in his jokes “Juan was being very cunning in his jokes” The interview or his jokes denote “event-objects” (Dowty 1979) that take time. Bertinetto 2000. Parsons 1990. (150) *Juan estaba arreglando dos sillas en una hora Juan was fixing two chairs in an hour As a dynamic predicate. to fix two chairs is compatible with the progressive form (151) and. Naumann & Piñón 1997. Summarizing thus far. Although it seems intuitively clear that an adverbial expressing the amount of time invested in the whole event and the expression of the event in its progression (i. the progressive viewpoint (in Spanish) is not a possible option with states. Kazanina & Phillips 2003).

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

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

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

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

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

each of them involves a different semantics that can properly be described as corresponding to what Carlson (1977) denominated “individual level” and “stage level. since I argue that permanency is not a necessary characteristic of IL predicates. I am expected to offer an alternative definition for IL predicates. after having introduced the relevant temporal distinctions. In the remainder of the chapter.Tense and Individual-Level Predicates 197 by the existential limitation of the referent. I have consistently considered that every (copulative) predication with ser is IL. I will construct an analysis of the temporal properties of IL predicates. why the Spanish copular verb does not change. if the predication with ser is all right. permanency or. without being able to explain. more accurately. “lifetime permanency” is a property of just a subset of IL predicates. I propose that being “nonpermanent” does not amount to being “stage level. As before.” In other words. a result that proposals equating IL-hood and permanency are pushed to. This perspective leads me to two outcomes. the effects are not borne out. Since be blond or be cruel does not have to overlap with the whole span of time that an individual’s existence extends over. The following paragraphs show examples of copular sentences with ser and different adjectival predicates. Throughout this work. despite the fact that the predicate is not understood as permanent. among other facts. I am not forced to say that the nature of the predicate has mutated from IL into SL. I dedicate a part of chapter 7 to such discussion. 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. I take the strong native intuitions about the different interpretation of adjectives that can combine with ser and estar to suggest that. in fact. I consider ser as the copula marking IL-hood. as has classically been upheld.” Consider. Pablo es muy gracioso Pablo ser-PRES-3SG very funny “Pablo is very funny” muy gracioso b. Thus. Regarding the temporal interpretation that gets the existence of the individual implicated (the lifetime reading). 6.2 Differentiating Temporal Extensions of IL Predicates In this section I argue that the span of time over which an IL predicate extends can vary depending on the type of property. I take it to mean that the predicate is IL. First. Second. As mentioned in chapter 2. when an IL predicate comes temporally or spatially restricted (some way or other). Pablo está Pablo estar-PRES-3SG very funny “Pablo is being very funny” . the next contrasts: (11) a.

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

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. that of ‘simultaneity’) is by asking “when the eventuality designed by the predicate took place. Al rodar por las escaleras. Pedro fue muy amable al acompañarme a casa Pedro ser-PERF-PRET-3SG very kind in walking me home b. Juan se rompió el pie In rolling down the stairs Juan broke his foot (35) a. it is proved it acts as a temporal modifier. According to García (1999). similar to the during + DP or while-clauses above.” If the al + infinitive clause constitutes a satisfactory answer. However. . at least under this examination. as the # symbol in (31) aims to capture. Hernanz (1999) points out that temporal infinitives can freely appear in either initial or final position (see (34)). Pedro fue muy amable In walking me home. does not work as a temporal adjunct. (34) a.Tense and Individual-Level Predicates 203 and I want to discuss whether such “infinitive clauses” act as a temporal adjunct restricting the interval of time the property denoted by the adjective holds of the individual. as (35b) shows. the status of the clause is not temporal. that the infinitival clause (in Spanish) spells out the interval of time the property holds. (30) Pedro fue muy amable al acompañarme a casa Pedro ser-PERF-PRET-3SG very kind in walking me home (31) ¿Cuándo fue amable Pedro? #Al acompañarme a casa When ser-PERF-PRET-3SG Pedro kind? In walking me home (32) Juan se rompió el pie al rodar por las escaleras Juan broke his foot in rolling down the stairs (33) ¿Cuándo se rompió el pie Juan? Al rodar por las escaleras When did Juan break his foot? In rolling down the stairs Whereas the al + infinitive clause is a natural answer of a when-clause in (32) and (33). if it does not. 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. this is not a possibility for al + infinitive clauses in the copular sentences at stake. it is not so in the adjectival cases. Juan se rompió el pie al rodar por las escaleras Juan broke his foot in rolling down the stairs b. it cannot be concluded. one way to test whether an al + infinitive clause has a temporal meaning (concretely. #Al acompañarme a casa.

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

As already mentioned. In the following section.3.3 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.2. in their lexical entry. according to independent grounds (such as their combination with a particular lexical choice of the copula in languages like Spanish). 6. holds of the individual for all his lifetime (mainly those referring to the origin and genetic nature of beings and to those properties that once acquired cannot be lost.). Musan (1995. Musan (1995) does. I argue that lifetime effects are a salient option in those cases where the predicate. these three conditions have to be met: (a) the predicate has to be an IL lifetime predicate. 1997).Tense and Individual-Level Predicates 205 6. whether they denote a lifetime property. I want to briefly enumerate the conditions that are necessary (not sufficient) for the lifetime reading to appear. 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 do not consider that those predicates encode. In support of the first claim. For the interpretation of the DP subject as ‘no longer alive’ to be an option (in languages like Spanish). nevertheless. according to which the temporal interpretation known as lifetime effects derives from the argument structure of IL predicates. for example. More accurately. 1997) observed that contextual factors.2 I have made two main points in this section. do not have to be permanent properties.g. and (c) the past form should be in the imperfect form. literally. in the line of Musan (1995. Ph. such as the presence of another past tense around. (b) the predicate has to appear in the past form. 6.4 Summary of Section 6. can be argued to be IL and. as. First.D. Second. for lexical reasons. can neutralize the lifetime effects.. I focus on another restriction for this temporal interpretation. since there is a large number of predicates that. I argued that those accounts. Differing from Kratzer (1988. equating IL-hood to property stability is inadequate. That is.1 Necessary Conditions for Lifetime Effects First. compare the following sentences: . I am going to argue. 1995). e. 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.

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

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

6.4. Both examples here are from Musan (1995). Then. This was first noted by Musan (1995.208 Individuals in Time other verbs appearing in lifetime predications. In the framework she assumes. such as have brown eyes or have long legs. Musan argues as follows. whereas a lifetime reading is available in (52). In this section I show that even when the predicate denotes a lifetime property. I would like to emphasize that these are not sufficient conditions for the lifetime reading to arise. Musan’s idea can be informally depicted as in (54). it can be restricted by the context through a variable C. and others with a previous context. 1997) As noted earlier. “out of the blue” cases. is exactly the same. ‘the day I was introduced to Gregory and Eva-Lotta’. as Kratzer’s examples were. in the next sections. as any other operator. Kratzer’s (1995) account predicts a lifetime reading to be available with IL predicates in all circumstances. Gregory was from America and Eva-Lotta was from Switzerland. the lifetime reading does not necessarily come out. despite the fact that the sentence. since it was a consequence of their argument structure. Westerståhl (1984) and Lewis (1986) observed that domains of quantification are contextually restricted. 8 As mentioned in the previous chapter.3.2 Introducing the Determining Role of Contextual Factors. the temporal operator of Gregory was from America in the second case is restricted by the previous context. Although I am abstracting away from the mechanics of the semantic formalization. I was introduced to Gregory and Eva-Lotta. I have already shown that the equation this author establishes between permanency and IL-hood is not accurate and I have suggested that the lifetime reading is expected to be salient when a lifetime property is at stake. The third condition is that the viewpoint form must be the imperfect. The sentence Gregory was from America is given a specific temporal context in which it is interpreted temporally. Musan (1995. As Musan observes. I return to this point in section 6. roughly. This author proposed to consider in contrast sentences like (52). .1. To account for this fact. like (53).8 Thus. 1997). I will show a number of situations where such a reading does not arise. (52) Gregory was from America (53) On that day. tense is taken as a sentence operator. Precisely. Gregory was from America. The second condition is that the tense form has to be the past. it is not in (53).

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

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

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

also in the line of Musan. a lifetime reading arises (or not) depending on the content of the TT. The leading idea in this regard will be that the TT is topic sensitive—that is. the reading that arises is a lifetime reading. I conceive the lifetime reading as an interpretive outcome derived from the location in the past of a TT with a specific content—namely. For sentences inserted in a context with a previous past interval mentioned. I claimed. as I said before. focusing on sentences with lifetime predicates. based on Musan (1995). In what follows. and Tense restriction for contextual ones). I argue that the temporal TP 2 . For “out of the blue” examples. that the TT takes its content from the previous interval mentioned (I will refine this account in section 6.3). 6. it is the whole span of time of Harry’s existence that gets located in the past. we have seen two types of cases: “out of the blue” cases and sentences inserted in a context. Thus. 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). that referring to the lifetime span of the individual referred to by the DP subject. it gets its content from salient elements in the conversation. in this way differing from Musan.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 lifetime interval of the individual referred to by the DP subject.4 The Determination of the TT Content and Lifetime Effects Thus far. in (61). In sum. I argue that these two cases can be explained uniformly. I will discuss the role of the contextual information conveyed by the DP subject itself in particular. who. In this sense. Put in these terms. The next sections discuss the sources for the TT.4. I proposed. the lifetime issue is reduced to the working of temporal interpretation in general. accounts for them by appealing to different resources (Gricean Maxima of Quantity for “out of the blue” cases.

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

among others. propose that such a restriction is formalized through a free variable in the determiner. every is not quantifying over the whole set of plates. In (70). in the world. As is classically argued (Heim 1982. That set is called the “restrictive clause. I am going to argue that the very nature and working of quantifiers is what precludes the arising of the reading. QP/DP 2 Di NP the 2 xi N′ g boy b. Barwise and Cooper (1981) pointed out that determiners do not quantify over the total domain of entities but just over the set specified by the noun. The context provides a set that the NP intersects with. and this. where the reading can be defended to exist as a possibility. Longobardi 1994). The contextual relevance gets into play through the contextual variable appearing as a subindex of the determiner. D x [ x is N] g restrictive clause Dx [x(N)] Furthermore.” (69) The boys a. and von Fintel (1994). Stowell 1989. However. say. Thus. but just over the contextually relevant set of plates. . boys in (69). the set that determiners quantify over is restricted by the context (Westerståhl 1984.214 Individuals in Time (68) Ningún chico era esquimal No guy was Eskimo In none of these examples is the lifetime reading available. which the noun is a predicate of (69a). as mentioned in chapter 5. Higginbotham 1983. In essence. the simplest hypothesis is to suppose that the difference in the availability of the reading resides in the different nature of the DPs. Stump (1981). determiners/quantifiers are interpreted with respect to the set contextually provided. the meaning of (70) is that of (71). Then. Lewis 1986). quantifiers bind the open argument (variable x). in these cases we cannot allude to the presence of a previously mentioned past interval to account for the mitigation of the interpretation. constitutes the quantificational restriction. all together. Partee (1984a). Since the difference between (52) (Gregory was from America). and these examples resides in the nature of the subject DPs.

the availability of the lifetime interpretation does not arise. no lifetime effect arises. I want to argue that this contextual information establishes the link with the background where the TT finds its content. where the preferred interpretation is (ii).10 This hypothesis establishes the previous context as the context that counts for the sentence in the conversation. Let me explain what I mean with an example. that lifetime effects arise in sentences such as (i). This absence of a link to a previous context has an outcome in the temporal realm—namely. Compare these two situations.Tense and Individual-Level Predicates 215 (70) Every plate (71) Everyc x [x(plate)] The point I want to make here is that it is because this universal property of quantifiers of being contextually restricted that a lifetime reading is absent in the sentences in (62)–(68). the lack of a (free) restricting variable can be argued to be the difference between generic and nongeneric DPs. and. and makes it the common ground where the TT finds its antecedent. a lifetime reading does not arise. it concurs with Musan’s (1995) observation that an element from previous context may make the reading not to appear. In essence. too. the value of the TT is not a lifetime span. during which these examples arose. Actually. I turn to more subtle situations. as a result. As a consequence. Along similar lines as before.4. since the reference to a previous scenario obtains. as explained by von Fintel (1994) and mentioned before. and. (i) a.2 Context Associated to Individuals At this point. These examples show that the factor triggering the nonappearance of the reading can be other than the very presence of a previously mentioned past interval. The point I would like to introduce here is that DPs referring to individuals also have contextual information associated with them. 6. The contextual restriction of the quantifiers directs us to a previous context where the free variable C finds its antecedent. 10 . Dinosaurs were from the North Pole b. I would like to argue that via such an obligatory contextual restriction. a background is built up. If the TT content is other than the lifetime span of the individual. Los dinosaurios eran del Polo Norte (ii) Dinosaurs are extinct 11 I thank Dominique Sportiche and Daniel Büring for the discussions around these topics.11 It is interesting to note that English bare plurals (ia) or generic Romance DPs (ib) do not seem to be contextually restricted. even though we have a lifetime predicate in past tense.

” 13 I thank Olga Fernández Soriano for this discussion . Fisher. whereas one of the individuals named Amàlia is a coworker. I propose that individuals also have contextual information associated and this becomes available through the contextual variable of the determiner.. since we got along with him. “is a predicate. 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). He happened to be moving to Lisbon. In few words. whereas in the first situation. for example. I suggest that even proper names may be subject to contextual information. This kind of information (“the Amàlia from work” vs. “the Amàlia from the For Raposo and Uriagereka (1995:191). the tourist guide we got in Brazil? (It turns out that) he grew up in Lisbon.” not the other one.12 A situation where it can be seen that context information matters in the interpretation of a proper name (i.13 Suppose the speaker and the hearer know two people with the same name. 12 . Amàlia. the other is a friend of theirs from the town where they usually spend the summer. really? So he is (#was) from Portugal! (that is why we did not notice so much difference between his accent and ours…) The first observation that can be made is that. This morning Eva ran into my office and told me: —Eva: You remember João. Suppose further that. This morning Eva came into my office and told me: —Eva: You know what? João grew up right here in Palmela! —Me: Oh. Following Higginbotham (1988) and Raposo and Uriagereka (1995). we used to hang out with him very often. and. the most likely and plausible interpretation will be that the girl they are talking about is “the Amàlia from work. —Me: Oh! So he was from Portugal! (that is why we did not notice so much difference between his accent and ours…) (73) Situation 2: My friend Eva and I are Portuguese. We went to Brazil three years ago and got a tourist guide for our excursion through the jungle. and so what makes Fisher Fisher is (perceived or imagined) ‘Fisherhood’. it does not seem so in the second situation. 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. We went to Brazil three years ago and got a tourist guide named João. the use of a past tense sounds natural (He was from Portugal). a name. to know who we are referring to by that name) is the following.e. If the speaker and the hearer are once talking about an issue from work and someone called “Amàlia” arises in the conversation.216 Individuals in Time (72) Situation 1: My friend Eva and I are Portuguese.

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

2. 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. I also argued that proper name DPs involve a similar context variable. I have argued that this is brought about through the context variable restricting quantifiers (free variable C).4. Beghelli & Stowell 1996). 3. it becomes the source for antecedents. along the same lines that specific DPs are generally understood (Pesetsky 1987. how this is articulated. The TT refers to a specific interval. as Musan (1995) suggests. the utterance situation is the most salient context. The relationship between the contextual background and the TT or the free variable C has been taken as anaphoric. and. so that all the surveyed cases can be accounted for in a uniform way. I want to show now. 4. In other words. 5. which contains contextual information associated to the individual referred to by the DP.3 Articulating the Account In the last two sections I have shown that the context that counts for the TT content can be that introduced by the DP subject itself. I 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. Conversely. the context acting as background constitutes the set where anaphoric elements and free variables find their antecedents. 6. The context relevant to find the “antecedent” for the TT does not have to be another past interval necessarily. The occurrence of lifetime effects depends on the content of TT of the sentence. (77) Salient context " TT . The leading ideas introduced thus far are the following: 1.218 Individuals in Time situation. The relevant context can be brought to the scenario by the DP subject itself. Elaborating on Musan 1995. lifetime effects do not arise when the interval relevant in the sentence is other than the lifetime span of the individual referred to by the DP subject. as a consequence. where the individual referred to by João is present. This seems to suffice to convert the utterance situation as the background that counts as relevant for the antecedent of the TT. Enç 1991a. more systematically. A specific interval is an interval that is linked to a previous context. following von Fintel (1994). That is.

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

can truthfully be captured by (105). (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. the lifetime reading does not arise either. . following Stowell’s suggestions. this case does not look different from (59) above and repeated below. where the TT of the IL predicate refers to the arrival time. Bearing all this in mind.234 Individuals in Time relative. where a simultaneous reading would be an option (making the lifetime reading unavailable as a result). equivalent to a forward-shifted reading. This would seem to pave the way for a lifetime reading to arise. and (107a). However. Furthermore. In fact. consider now the following example with a lifetime predicate in the RC. the indefinite DP a guy is interpreted as [+specific] and can be supposed to take wide scope (becoming an adjunct of TP). we can suppose that the RC moves along the DP. since the content of the subordinate TT is not in the domain of the higher one. (109) Selene besó a un chico que era de California Selene kissed a guy who was from California As a complement of an extensional verb (kiss).

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

although the syntactic dependence between the matrix and the subordinate tense complexes was different. in section 6. blond. I defended the lifetime reading as an interpretive outcome due to a specific content of the TT—namely. Actually. Among the issues I have not discussed in this work is the relationship between the TT content and time adverbials. 1997) insight that lifetime effects can be neutralized by the presence of another past interval in the context. most of the properties are like the latter ones. as. which prevents a lifetime reading from arising. 1997). and I argued that the arising of the lifetime reading depends on the content of the TT. Consider (113) in relation to (111). are not necessarily lifetime properties. when it refers to the whole interval of an individual’s lifetime. Harry was from California. I examined the arising of the lifetime reading in syntactic composition—specifically. I argued that the contextual restriction of DPs can intervene in building up a previous background. I concluded. that the content of the TT is not sensitive to syntactic distance but to discourse prominence. 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. even without the overt presence of another past tense. if the TT refers to the interval corresponding to the span of time an individual’s lifetime extends over. In sections 6. etc.3 and 6. their good combination with restrictive temporal adjuncts proves. complement and relative clauses. Others (kind. Thus. I showed that a lifetime reading is not available in all of the cases where a lifetime predicate appears in the past tense. I assimilated her observation to the theoretical terms assumed here.5. therefore. although it has become clear that the TT of an IL lifetime predicate can refer to an interval given by the discourse. for example. Finally. In both cases the lifetime reading does not arise. I also offered a number of examples of lifetime predicates in the past tense that. who defends the position that the lifetime characteristic comes specified in the lexical entry of the predicate. 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. based on independent grounds. I have shown that a good number of predicates that can be considered IL. In this respect. A large number of questions remain to which I have not even suggested a possible answer.4. that interval cannot be “spelled out” by an adverbial. For example. After introducing Musan’s (1995. he was from California . do not have a lifetime interpretation.) can hold of an individual just for a determined period of time shorter than the individual’s lifetime. a lifetime reading will be available. (111) Harry and Maria arrived in the USA. I differ from Musan (1995.236 Individuals in Time individual’s lifetime that is their subject.

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

.

correspondingly. In the following pages. In particular.Chapter 7 Conclusions and Final Remarks As set out in the presentation. and tense. in this book I sought to explore what lies behind the dichotomy traditionally known as individual-level (IL)/stage-level (SL). I examined whether this distinction is rooted in temporal terms. Only a subset of IL predicates are permanent predicates. Spanish distinguishes between two copulas. any instance of estar yields an SL one. the speaker predicates the properties of the subject on a particular occasion. outer aspect. most of them denoting properties related to the origin (ethnic or genetic) of the . whose semantic characterization I have argued corresponds to the IL/SL distinction. 7. or funny person. (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. linked to external reasons (maybe because he is wearing a nice suit. In the cases with estar (2). or is in a good mood). got tanned. dark-skinned. 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. contrary to widespread belief. First. by studying the temporal properties of IL predicates in copular sentences at three levels—inner aspect. 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. When ser is involved (1).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. The length of the interval an IL property extends over can be restricted in time. and. I have consistently taken it that any instance of ser yields an IL predication. ser and estar.

in Spanish. Third. mean) have been proposed to correlate with the presence of the relational complement of these adjectives (cruel to Mary). I have concluded that those proposals arguing that ser-predicates should be considered SL under aspectual forms like the perfective and the progressive cannot be correct.. however. Topic Time has been argued to be a time interval taking its value from the salient discursive context. Specifically. Rather. such as Eskimo. based on hypotheses that attribute aspectual content to prepositions (Hale 1984. this study is a contribution to a better understanding of the role of prepositions in the properties of event structure. but it is independent from that: cruel mean/kind to someone can appear with both copulas. contrary to general belief. in the case of IL predicates. The book proposes a uniform treatment of tense. the aspectual characteristics of the copular constructions are derived from their syntax. I have accounted for a variety of contextual situations illustrating the sources from . such as blond or young. are properties that can be true just of a limited period of time. Crucially also. which clearly pattern with activities in an inner-aspect classification. the temporal interpretation of IL predicates is derived. as is the case with any other type of predicate in Spanish (e. from the content of the so-called Topic Time (Klein 1994). I have argued that the dynamic nature of copular constructions does not correlate with an SL classification. tense takes and orders the nominal argument (the NP subject). In particular. IL predicates are not all stative. this book gives further empirical support to recent hypotheses according to which aspectual properties are syntactically encoded (Borer 2005. as is the case with any other type of predicate. Ramchand 2003. Demirdache & Uribe-Etxebarría 2000). kind. there are two types of IL predicates regarding inner aspect: stative and dynamic predicates. no stative verb can appear in the progressive). among others). differing from previous influential proposals like Kratzer 1988 where. Thus. Others.g. the dynamic properties of clauses where adjectives referring to mental properties appear (such as cruel. The restrictions between a ser-predicate and aspectual forms like the perfective or progressive are derived from the inneraspect properties of the predicate. gypsy.240 Individuals in Time individual. Finally. Second. ser and estar. unlike what is commonly assumed. As a result. I have proposed that the dynamic and active properties classically observed in certain copular constructions are due to the presence of a complement. In this respect. I have shown that ser-predicates can appear in any aspectual form. The investigation of Lifetime Effects (the interpretive phenomenon whereby the DP subject can be understood as ‘no longer alive’) in chapter 6 deepens our understanding of the concept of Topic Time. dynamicity is proposed to correlate with the presence of the relational complement and argued to be rooted in the Prepositional Phrase. or color-blind. Ritter & Rosen 2000.

Claim (in previous literature) IL predicates express permanent properties When an IL predicate is at stake. One of the crucial questions is how the association to a situation is brought about. between IL/SL predicates) can be cast in temporal terms. while ser is more “innocuous. most authors dealing with the ser/estar contrast concur with the idea that estar-predicates put the property in relation to a situation. In examining this question we address the key difference between IL/SL predicates. 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 7.2 The IL/SL Dichotomy and the Ser/Estar Contrast With major or minor differences. and I have given an account for the different temporal interpretations that IL sentences can have without yielding inaccurate generalizations. Table 7. As was surveyed. Although I will not be able to offer a definitive answer to this. By answering this question. where the definition of IL predicates has been mainly associated to the notions of permanent property and stativity. These conclusions differ from most previous literature. the notion involved. we will be proposing something about the terms in which the IL/SL dichotomy is represented in the grammar. Previous claims regarding IL predicates Throughout this work I have analyzed IL predicates in their relation with the temporal realms cited. are temporal concepts in nature. classically alluded to in the description of IL predicates. these notions. As noted in the beginning of the book.Conclusions and Final Remarks 241 which the Topic Time can take its value. As mentioned in chapter 2.1. more in general.” leaving the property without any association to a determined situation. One of the important questions here is whether the difference between ser/estar clauses (or. 7. and the semantic domain they belong to. a number of authors have proposed that the IL/SL distinction should be cast in “aspectual” terms. I will introduce some discussion in the following sections. Ser simply classifies the subject into the category denoted by the predicate that appears in combination with it.1 summarizes the claims previously made. most .

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

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

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

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

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

you get tanned in two weeks” Finally. estar-clauses admit in-adverbials with certain predicates (participles and perfective adjectives that have been described as “cut-short”2). . see Bosque 1990. which come from heterogeneous verbs. therefore. only if the adverbial in +x time is present. AspQMAX is projected. 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.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. rather than llenar > llenado > lleno (‘fill > filled > full’).e. from which participles derive. 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. 3 For participial and cut-short adjectives to yield a heterogeneous construction. ‘get sick’) which. participial adjectives. ‘sick-INF’. due to their atelic nature. Following the theoretical framework I have worked with here. There are other adjectives (enfermo ‘sick’) related to a verb (enfermar. estás with this cream estar-PRES-3SG dark-skinned in two weeks “With this cream. I would like to suggest the idea that in the cases of participles (14) and cut-short adjectives (15). but rather adjectives yielding verbs. the verb they are related to has to be of heterogeneous nature too. That is: lleno > llenar > llenado (‘full > fill > filled’). 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’. the predicate gains a telic interpretation: morena *(en dos semanas) (16) Con esta crema. 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.3 2 For a detailed discussion about participles. behaving. and cut-short adjectives.. Regarding cut-short adjectives. with predicates other than participles and cut-short adjectives (i. the ones of our minimal pairs in (1) and (2)).

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

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

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

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

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

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

cortado. according to what I said in section 7. When these participles are part of a passive form. falso ‘false’.10 Likewise. Escandell-Vidal and Leonetti (2002) argue that the reason is because these predicates take a propositional argument as their subject. in contrast to other authors who consider that adjectives of the type cited in (31) derive from participles. this is not the case. basically every adjective can appear in combination with 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. 12 This remark is restricted to (adjectival) copular constructions.12 (31) Estar/*ser descalzo. Being conceivable as an object of perception is. which is not an object of perception. which selects for [+perfect] SCs. conmovido. lleno. Nevertheless. necesario ‘necessary’. More specifically. It is the copular choice that makes the predicate either IL or SL. where a [+perfect] feature is difficult to defend.) 10 . any adjective is predicted to be able to combine with it.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). a necessary previous step for estar to induce coercion. 11 Recall from footnote 2 that Bosque (1990) argues that it is these adjectives that produce verbs and such verbs yield participles. would otherwise become un- Bosque (1999) points out that not all adjectives can combine with estar—among them. As Demonte (1999) points out. which in Spanish is formed by ser + past participle. and evidente ‘evident’. That estar (obligatorily) selects for [+perfect] SCs cannot be the case since its combination with adjectives such as handsome or pale.254 Individuals in Time (29) ??Juan está americano Juan estar-PRES-3SG American muy americano (30) Juan está Juan estar-PRES-3SG very American 7. (The [past] participle ending in Spanish is -ado/-ido. according to them. most of the adjectival predicates that only combine with estar are cut-short adjectives (31)11 and adjectival participles (32). llenado. contentado. the mentioned participles are allowed in combination with ser. contento. There are some adjectives that combine only with estar and reject ser. 13 The corresponding participles are: descalzado. given the innocuous characteristics attributed to ser. I leave aside here the study of the differences between the passive forms and the ser-copular constructions. harto13 full fed up estar/ser barefoot glad (32) Estar/*ser destrozado. hartado.6.

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

the description of estar as ‘episodic’. but something external to it. but does not account for why they can appear just with ser and cannot be conceived as linked to a particular situation. I proposed that the peculiar properties to the constructions these APs participate in emanate from the particular properties the APs themselves involve. unlike adjectives. The definition of ser given explains why nouns combine with ser (40) (since nouns denote classes). are retained in their combination with estar: cruel con Pedro a propósito (43) Juan estuvo Juan estar-PST-PERF-3SG cruel to Pedro on purpose . the cruel-type. Since a location is not a class. As we already know. those peculiar properties. 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. it seems that both notions can be conceived apart from each other. Ser classifies the subject by categorizing it into the class denoted by the predicate following the copula. In chapter 4. Consistently with this hypothesis. such as agency. among other things. Thus. which would leave (41) without explanation since in Spain is not a temporary property of Madrid. This perspective also allows us to avoid. although episodicity can be one of the consequences of the linking to a particular situation. the ungrammaticality of Juan es en Brasil is explained.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. en España (41) Madrid está Madrid estar-PRES-3SG in Spain Since a location is not a property of any individual. the obligatory predication with estar is explained. the linking to a particular situation is independent from the restriction of a predicate to a determined temporal interval: (42) Pedro fue camarero dos meses Pedro was a-waiter two months Finally.

the adjectival property is interpreted restricted in time and tied to a particular occasion. as I mentioned in chapter 5. poses some issues still unexplained. 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). which.Conclusions and Final Remarks 257 One important difference between the construction of these APs with ser or estar is the impossibility of estar to appear in the progressive. the fact that the progressive is incompatible with estar is not accounted for. waiting to be studied with the progressive form itself. since this does not happen with other adjectives: guapa (49) María fue Maria ser-PERF-PRET-3SG beautiful “Maria was a beautiful woman (in a period of her life)” guapa (50) María estuvo Maria estar-PERF-PRET-3SG beautiful “Maria was beautiful (as soon as she made herself up)’ . The closeness in interpretation cannot be attributed to an effect of the perfective. since it appears that estar behaves in this respect as if it were a stative predicate. where the contrast between the two copulas looks mitigated. contrary to ser. With cruel-type APs. One of them directly refers to the copula alternation in Spanish. (44) *Juan estaba estando cruel con Pedro Juan was estar-ing cruel to Pedro If the dynamic properties are kept no matter what copula is at stake. the difference between ser and estar seems slightly neutralized. which makes the meanings of ser and estar converge. Whereas there is a neat contrast between (45) and (46). the interpretive difference in (47) and (48) decreases significantly. There are other issues regarding these APs that deserve to be explored in deeper detail in future research. This is left unanswered here.

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

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

.

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

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

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

and J. Madison. and M.).. Word Meaning and Montague Grammar. Enç. Linguistics and Philosophy 18:1–19. M. 2000. In Gramática Descriptiva de la Lengua Española. R. La predicación: Los complementos predicativos. 1995. Thematic proto-roles and argument selection. Revista Española de Lingüística 9:133–171. 1991b. 1977. New York: Academic Press. The stative in the progressive and other essence/accident contrasts. Cambridge. 1979. Uribe-Etxebarría. Bosque and V. Diesing.). 1991a. Ms. MA: MIT Press. V. Egg. J. and Masullo. 143–179. MA: MIT Press. I. A Transformational Approach to English Syntax: Root. 1986. M. V. Uribe-Etxebarría. H. In The Syntax of Time. Demonte. D. 157–186.264 Individuals in Time Demirdache. University of Wisconsin. and M. Demirdache. Demonte (eds. I. 1999. Uriagereka (eds. 1999. Cambridge. Cambridge. Lecarme (eds. Journal of Semantics 12:311–356. D. Structure-Preserving. 1995. Linguistic Inquiry 22:1–25. 1979. and Local Transformation. Clases de adjetivos: La posición del adjetivo en el SN. 1975. On the necessity of distinguishing between (un)boundedness and (a)telicity. The effects of aspectual class on the temporal structure of discourse: Semantics or pragmatics? Linguistics and Philosophy 9:37–61. Dordrecht: Reidel.). D. Uribe-Etxebarría. Madrid: Espasa Calpe. The semantics of specificity. 1966. Demirdache. Donnellan. Sintaxis y semántica de las construcciones con ser y estar. Emonds. M. Linguistic Inquiry 18:633–657. Anchoring conditions for tense. 1976. In Gramática Descriptiva de la Lengua Española. K. Michaels. Demonte (eds. Demonte. J. J. R. Madrid: Espasa Calpe. . Linguistic Inquiry 6:579–588.). 2461–2524. Martin. M. Language 67:547–619. I. The ingressive as a new category of verbal Aktionsart. D. Ms. Bosque and V. MA: MIT Press. 1991. V. 2004. Enç. On the absence of the present tense morpheme in English. Demonte. Indefinites. 1999. Towards a theory of the diversity of temporal systems. Enç. Dowty. Philosophical Review 75:281–304. University of Nantes and University of the Basque Country/Basque Center for Language Research. M. M. 129–215. Dowty. H. Dowty. The syntax of time adverbs. M. 1992. Dixon. In Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik. 1987. H and M.. D. The primitives of temporal relations. Guéron and J. Where have all the adjectives gone? Studies in Language 1:19–80. Depraetere. Dowty. Reference and definite descriptions.

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

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

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

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

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

J. University of Utrecht. 1995. Ritter. G. In Small Clauses [Syntax and Semantics 28]. 1993. Cambridge. J. The theta system: syntactic realization of verbal concepts. Rapoport. C. Pustejovsky.). E. MA: MIT. Reuland and A. Linguistic Inquiry 20:365–424. 2001. Ritter. and Small Clauses: A study of Israeli Hebrew. Ramchand. Cambridge. Norquest (eds. dissertation. In The Projection of Arguments. 1995. A. J. V. Cambridge: Cambridge University Press. C. Proceedings of the Berkeley Linguistic Society 26:435. Piñón.D. MIT. Stanford. Ms. Probus 13:81–111. 1947. Linguistic Analysis 26:29–62. A. Hensey (eds. Stanford University. Interpreting mood. The semantics of copulative constructions in Portuguese.-Y. 2000. UiL OTS Working Papers. University of California. Bird. Delimiting events in Syntax. Rosen. In Lexical and Conceptual Structure. A. Mathematical Methods in Linguistics.). E. 1998. 1960. T.). MA: MIT Press. C. Pustejovsky. and the structure of IP. and S. . Washington. Uriagereka. 179–206. B. First phase syntax. and J. dissertation. 2003. CA: CSLI Publications. M. Strong and weak predicates: Reducing the lexical burden. Universal Grammar. University of Oxford. Pesetsky. Reichenbach. 1988. 1987. In Proceedings of WCCFL 18. J. 1996. Elements of Symbolic Logic. Ph. D. ter Meulen (eds. 1989.270 Individuals in Time Partee.). dissertation. 1999. Raposo. 1976. Cardinaletti and M-T. In Current Studies in Romance Linguistics. S. New York: The Free Press. Happening gradually. Rosen. Berkeley. Somerville. 1995. 19–39. J. Mood at the Interface. Cambridge. Ph. Dordrecht: Kluwer. Reinhart.D. Pinker (eds. E. 420–433. A Mereology for Aspectuality. T. 135–164. Levin and S. J.. 2000. In Studies in Generative Approaches to Aspect [Lexicon Project Working Papers 24]. 98– 129. Quer. A. Luján and F. ter Meulen and R. Tenny (ed. W. Pollock. The Generative Lexicon. E. Stanford.). Piñón. CA. Greuder and M. Verb movement. Durative adverbials for result states. 1991. Butt (eds. The geometry of events. Quine. 47–81. 1998. 343–366. Wh-in-situ: Movement and unselective binding. Wall. Nominal. Guasti (eds. W. Oxford: Blackwell. MA: Cascadilla Press. Quer. University of Utrecht. H.).. Querido. Copular. MA. B. C. Piñón. Ph. Word and Object. Haugen and P.D. In The Representation of (In)definites. and S. New York: Academic Press. Carnie. Cambridge. MA: MIT Press.). Pustejovsky. Two types of small clauses (Toward a syntax of theme/rheme relations. DC: Georgetown University Press. The syntax of event structure. 1987. J.

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

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

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

.

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

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

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

.

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

132. 224. 176. 227 Simultaneous reading 76–79. 136 Place/situational prepositions 120. 189. 194. 123. 144. 148. 128. 14. 177–180. 64. 201. 132. 116. 257. 233. 242. 184. 230 of Narration 77–80 of Relative clauses 223. 176. 184. 116. 167. 72. 176–178. 178. 171– 175. 37 Thetic (judgments) 12–14. 202 Ground 119. 135. 165. 180 Perfective (preterit) 31. 14. 136. 226. 80. 86. 227. 80. 229. see Temporal interpretation Past-shifted reading 76–78. 122. 211. 244. 252 Realization function 6. 178. 258 Imperfect Continuous 76. 147. 211. 141. 207. 180. 258 Hindi 120 Homogeneous (predicates) 70. 148. 186. 191. 158. 72. 14 Generic reading 6. see also Utterance Time Stativizer 115. 112 Juba Arabic 120 Kinesis 40 Krio 120 Lakhota 55. 129. 125. 142. 161. 148–152. 158. 226. 123– 126. 128–131. 156 Hawaii-Creole 120 Hebrew 15 Heterogeneous (predicates) 72. 181. 187. 12. 79. 186. 244. 55. 249 Relative clauses temporal interpretation of. initiator 61. 6. 94. 208. 117. 12. 249 Existential coda 12 Existential quantification 8. 130.280 Individuals in Time Event-object 110. 248 Swahili 120 Temporal interpretation of Complement clauses 76. 67. 118. 258 Progressive paradox 191. 57. 246. 221–223. 227–229. 72. 132. 248. 236 Theme 105. 160. 192. 134 There-sentences 5. 134 Instigator 55. 173–180. 155. 242. see Temporal interpretation Russian 15 Salish 118. 122. 147. 148 Sequence of Tense 226. 172–174. 76. 159 Quantification restriction 11 Quantity 113–116. 141. 246–248. 224. 26. 231. 175–177. 212 Mereological properties 3. 172. 134–136. 209. 150. 131. 131. 192 Prospective 120–122. 212. 137 Subinterval property 70. 207. 111. 148. 94 Margi 121 Maxim of Quantity 85. 162. 196. 115. 155–157. see also Subinterval property and Additivity property Narration temporal interpretation of. 215 Goal(Path) 96–98. 183. 257 Perform(er) 53. 156 Focus 13. 144. 226. 131. 246. 153–160. 242. 230. 158–160. 233–235. 129 Progressive 153–163. 147. 79. 258. 178. 105. 154. 241. 144. 147. 148–152. 162. 250 . 248. 94 Pomo 94 Prepositions Directional prepositions 127. 227. 246. 245. 194 Reference Time (RT) 3. 113. 186. 136. 58. 134–136. 175. 173. 192. 11. 183. 192 (Im)perfective contrast 25. 159. 232. 181 Existential reading 6. 234 Situation aspect 39 Small v 55 Spatiotemporal argument/variable 8–10. 31 Imperfective paradox see Progressive paradox Initiation. 112. 186. 189. 229. 148. 222. 189. 158. 172. 168 French 121 Generic operator 11. 157. 174–178. 195. 184. 154–157. 106. 249. 186. 189– 192. 165. 158. 11. 158–160. 148. 187. 189. 233 Perfect entailment 47. 35 Figure 119. see also Event(ive) argument Speech Time 122. 240. 230. 179. 152. 221 Habitual 77. 212. 222–225. 129. 35. 187–192. 123. 100. 240. 35. 88. 83. 189. 191 Event(uality) Time (ET) 3.

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

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

Sign up to vote on this title
UsefulNot useful