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

................. Raposo and Uriagereka (1995)...............3 Summary of Section 3.....................1. The Structure of Copular Constructions .....................................16 2.59 3...................................................2.......................33 2............................ 1995) ......2.........2 Ser and Estar as an Aspectual Differentiation ...1 ..........4 The IL/SL Contrast as a Categorical/Thetic Distinction....51 3...............5 2...............................1 Distribution of Copular Verbs in Spanish ........................................................................................ Introduction to the Tests Discerning the IL/SL Status.12 2.. Chierchia (1995) .................................................. Kratzer (1988............2 The IL/SL Dichotomy as a Syntactic Distinction..........................2................3 IL Predicates as Inherently Generic Predicates................................. When There Is More Than One Copular Verb: Spanish Ser and Estar.....................................................................1 ......................................11 2...............................1 Inner Aspect.....1 Chapter 2 Individual-Level Predicates ...............................................................................2 ....... Higginbotham and Ramchand (1996) ........3 Summary of Section 2....1.................................................1...........................8 2..xi Chapter 1 Presentation of the Study ............3 Summary of Section 3.................................................2 Event Types and Event Structure.....................................53 3.............. A Brief Stop at “Agentivity”....... Milsark (1974) and Carlson (1977)..........1.61 .....................1....................4 Summary of the Chapter ....3 Aspectual Differences among Individual-Level Predicates ............1........................61 3.......2.......2 .................... Inner Aspect and Event Types .........1.........................2..............32 2.39 3.............53 3.......5 Summary of Section 2........5 2..............................................................1 The IL/SL Distinction as a Semantic Distinction..............1 A Cluster of Notions ..2 Agents in Event Structure ...............................14 2........................................................2...3...........................1..38 Chapter 3 Event Classes and Individual-Level Predicates ........15 2.........5 2...................Table of Contents Acknowledgments.......2.............................40 3.......39 3....39 3................ix Foreword..........1......................25 2..................... by Tim Stowell .......................................................1 Proposals about the Individual-Level/Stage-Level Dichotomy ......................2.......

.......................................1 Previous Explanations of Nonstative Copular Clauses ........133 4.........3...................135 4.....85 4...........4 Summary of Section 4............4 The Issue: Aspectual Alternation in IL Copular Clauses ......................................................................3 ......117 4..........1 Cruel-type Small Clauses Taken by Verbs Other Than the Copula ...............................................72 3.........2 ...................................................1 Similarities between Activities and States ............................. The Cruel-type as a Small Clause Integrant.2 Syntactic Approaches .........................86 4......1....................1 Lexicalist and Logico-Semantic Approaches .....................................3 Adjectives Predicated of an Implicit Event Argument ..........3 Some Confusing Arguments in Defense of a State/Activity Distinction...................................................vi Individuals in Time 3................................................7 ..108 4..........4 Summary of Section 3...............83 4.....3....... 69 3......................8..........2 Summary of Section 4..3...............................................3......7.................. Summary of the Chapter ..............1..............6..................................4..................................................2........................................80 3.84 4......4........1...........................7..................6.....3 The DP Subject of Relational MPs ................62 3.............6 .................6......................................................2 Summary of Section 3....................... The Relational PP Complement........2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor123 4.......84 4....................................7.5....105 4.....5 Justifying the Approach ...............96 4................................................2......76 3..81 Chapter 4 Aspectual Alternations in Individual-Level Predicates ..............4 .............138 4....117 4................................................106 4...........116 4..................................................98 4...............................110 4...........................................3 .91 4..107 4.........6 An Account Based on the Relational PP Complement ....143 4.........2 On the Optionality of the (Affected) Goal PP ................................................108 4..............................................143 ....................2 Differences between Activities and States .6...............1 Aspectual Tests on IL Predicate ...3 Relational Mental Properties................5................68 3...4.4.....................................90 4................................................. States and Activities: A Grammatically Relevant Difference?.. Summary of the Chapter ...........5.95 4................1 The Two-Copulas Hypothesis ...........4...1 Lexical Semantic Classes of Adjectives ..........2 Be as a Copula Shifting a State into an Activity .........69 3.............3........................3 The Relational PP with Other APs ..........4 Summary of Section 4...............2 Summary of Section 4.137 4..............................................3 Summary of Section 4...........1 Prepositions as Aspect Encoders .............2 Adjectival Predicates Showing an Activity-like Behavior ............3....1 On the Interpretation of the Relational PP .....................96 4......5 .....5...........

......193 6.....5....3...............1.........3 On the Relation between the TT and the Habitual Q<occ> ....................1 Necessary Conditions for Lifetime Effects . Aspect as an Ordering Predicate .................1 Permanent IL Predicates .3 Brief Notes about the Complement of Mental Properties APs .................................3............5 The Lifetime Reading in Compound Sentences .210 6..........................215 6...2 The Perfective and Adjectival IL Predicates ................Table of Contents vii Chapter 5 Outer Aspect and Individual-Level Predicates ..... Kratzer (1988................3 The Content of the TT in the Arising of Lifetime Readings ........................4.....................................................3... 179 5.....3 Inner and Outer Aspect ................3 The Progressive and Adjectival IL Predicates ..............2...... 176 5......1 Habitual Heterogeneous Predications.................................................................2...3.................... 177 5.................................. 1996) ................171 5...........................................................2 The Internal Argument of Tense: The External Argument of Aspect.........1 Tense and Aspect as Ordering Predicates .................... 186 5................................1 When the Subject Is a QDP ........................................6.....3 The Arising of Lifetime Effects .........................................205 6....................148 5..........158 5.............................4..205 6.......4 Summary of Section 6............ A Brief Summary of Aspect Notions ....152 5...............5......................................... 174 5.......................202 6. Summary of the Chapter ...................197 6...............1...........2................................................ 192 Chapter 6 Tense and Individual-Level Predicates ................................................................. and Systematicity .....4...............................147 5...........2....................................................220 6..2 ..........1 Tense............2 The Habitual Interpretation: Iteration........................................ Stowell (1993........222 .......................1 The Imperfect and Adjectival IL Predicates ..............4 The Determination of the TT Content and Lifetime Effects ...4..............................................................................1 Quantifying over Occasions ..2.... 180 5...................1 Temporal Interpretation as a Consequence of Argument Structure.208 6...... 1995) ............................2 Context Associated to Individuals ...............148 5......199 6...............2...2 Introducing the Determining Role of Contextual Factors.................2 Aspect as a Quantifier over Occasions ........................................2 Perfective Homogeneous Predications .....5................212 6...199 6.205 6.............213 6........ 187 5..............................................................4.........2..........................5 Adjectival Individual-Level Predicates and Viewpoint Aspect ................................4 Summary of Section 6.........................................................................163 5..................3.............................2 Differentiating Temporal Extensions of IL Predicates ....................... 1997) .................................3 Articulating the Account..218 6.......... Musan (1995...................................194 6.....................173 5....2 Nonpermanent IL Predicates....................................................................................4 ....157 5..... Proportion....

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

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

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

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

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

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

.

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

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

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). After showing that the ser/estar dichotomy cannot be recast in terms of tense nor in terms of outer aspect (by alluding to the completion or perfection of the property) nor in terms of inner aspect (mereological properties). I argue that the IL/SL contrast should be rooted in the association to a particular circumstance (SL) or its lack thereof (IL). In particular. my proposal will concern the role of prepositions in event structure. Focusing on copular clauses. 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. Chapter 2 introduces the notion of IL predicate and presents the distribution of the two copular verbs in Spanish. 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. 1996). the ideas defended in this monograph are sympathetic to the conception of syntactic structure as event structure. In chapter 3. dynamicity is proposed to correlate with the presence of the relational complement and argued to be rooted in the prepositional phrase. giving the number of times a certain eventuality holds (along the lines of Verkuyl 1999). Ramchand 2003. 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- . 2004) and a quantifier over occasions. based on hypotheses that attribute aspectual content to prepositions (Hale 1984). mean) are proposed to correlate with the presence of the relational complement of these adjectives (cruel to Mary). The discussion of inner aspect is framed in the theoretical debate regarding the effects of syntactic structure on the semantics of predicates. I will show that prepositions can contribute an inner aspect shift from states to activities. In this vein. along the lines of Demirdache and Uribe-Etxebarría 2000.Presentation of the Study 3 tense. kind. 1995] with respect to the totality of the Eventuality Time. 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. which is described as a viewpoint referring to a plural number of occasions. 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. Specifically. In chapter 4. Outer aspect and tense are conceived as dyadic ordering predicates that take two time-denoting arguments (Stowell 1993. Ritter & Rosen 2000). This book is organized as follows. Thus. this being understood in inner aspect terms (Borer 2005. the dynamic properties of clauses where adjectives referring to mental properties appear (such as cruel. In this respect.

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

4 summarizes the chapter. Section 2.3. I present other two aspects basic for my purposes in this work. IL and SL. and pragmatic distinctions. I also introduce proposals by Heycock (1994) and Becker (2000).1 The IL/SL Distinction as a Semantic Distinction. I present the syntactic analysis I assume for copular cases—namely. I will introduce the syntactic analysis that I will assume for copular sentences. There are several policemen in the corner a. I concentrate on the most influential proposals about the contrast between individual-level (IL) and stagelevel (SL) predicates. the one offered by Stowell (1978.1 Proposals about the Individual-Level/Stage-Level Dichotomy 2. After that. a part of this chapter presents the differences between the two Spanish copulas (ser and estar). who discuss the constituency of the SCs that copular verbs take (in particular. 2. There are several policemen available b. *There are several policemen intelligent b. (1) (2) a. arguing for an aspectual distinction. In section 2. Likewise. I will first present the main descriptions of this notion found in the literature. Section 2. Introduction of the Tests Discerning the IL/SL Status Milsark (1974) noted that there is a predicate restriction in the coda of presentational there-sentences. PP. I will critically revise those analyses. the predicates of (1) are allowed. the presence of functional projections) and its relationship with the IL/SL distinction.1. specifically discussing arguments for semantic. which. For example.1. Since the data in the work are from Spanish. syntactic. according to most authors. according to which the copular verb selects for a small clause (SC) containing the nominal argument and the predicate (AP. lexically distinguish the two types of predication. NP). In section 2.Chapter 2 Individual-Level Predicates In this chapter I introduce the concept of the individual-level predicate. *There are several policemen insane . ser and estar. since I am chiefly concerned with copular cases.2 is devoted to the differences between the two Spanish copulas. This chapter is organized as follows. but those in (2) are excluded. Milsark (1974) and Carlson (1977). 1981). indicating some of the points on which my proposal differs. and introduce some of the concrete aspects for which this work makes alternative proposals.

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

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

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

This is precisely the case of the IL predicate in (19). Stage-level predicates IP 1 I′ 1 I VP 1 <e> VP b.3. it feeds its son According to Kratzer. as the ungrammaticality of (15) and (17) shows. In (18)–(20). she looks happy (19) *When(ever) Mary is a mammal. Other interesting contrasts are accounted for by this spatiotemporal variable: (18) When(ever) Mary is in Paris. and the spatiotemporal variable (18).Individual-Level Predicates 9 (13) a. Such an adverb unselectively binds all free variables it has in its scope. where the absence of indefinites and the absence of a spatiotemporal argument preclude the availability of a variable susceptible to binding. (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. when(ever)-clauses of the sort of (18)–(20) involve an implicit quantificational adverb. I discuss the semantics of when(ever)-clauses and. 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. Heim 1982) (an animal in (20)). When there is no variable to be bound. it is predicted to be a stable distinction. SL predicates (at your disposal) can. differing from Kratzer’s suggestion. the sentence results in ungrammaticality. potential variables to be bound are those that indefinites introduce (following Kamp 1981.5 Let me now introduce another important aspect of Kratzer’s proposal. However. she feeds her son (20) When(ever) an animal is a mammal. due to the prohibition against vacuous quantification (Chomsky 1982). . If the IL and SL distinction is. as presented. Kratzer is forced to readjust such a 5 In section 2. always. This is explained by the possession or lack of the spatiotemporal argument. I argue that they are just an apparent test for SL-hood. rooted in the argument structure.

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

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

we would expect that an overt generic operator could appear. 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. 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. Chierchia argues that the Gen operator is a strong quantifier. Besides. Clauses involving an IL predicate are about a prominent argument—a “category” in Kuroda’s (1972) terms. which I have already argued against in the two previous subsections. therefore.4 The IL/SL Contrast as a Categorical/Thetic Distinction. SL predicates are. Raposo and Uriagereka (1995). (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. As the following sentences show. Regarding Chierchia’s central proposal that IL predicates are inherent generics. IL clauses can be. This strict correlation leads this author to propose that when a predicate (like sick).12 Individuals in Time 1974). in some pragmatic sense. those referring to the event they introduce. that is not the case—another reason why I will not follow Chierchia’s approach on IL predicates. In turn. 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. are about the individual designated by the subject.” Clauses involving SL predicates . if we are referring to a chronic illness or an occasional one). called “categorical judgments. 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. simply. I consider that this conclusion makes the classification IL/SL lose its appeal since it does not have any predictive force. can be understood as stable or transient (for example. they define IL predicates as those that. Specifically. if a generic operator were actually in the constitution of IL predicates. (33) *Típicamente/*característicamente/*generalmente Juan es ignorante typically/ characteristically/generally Juan ser-PRES-3SG ignorant “Usually. Mary is cultivated” 2.1.

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

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

I have introduced the idea that equating permanency of a property and ILhood is inaccurate. see also Rapoport 1987 and Rothstein 1995. When the predicate is an NP → copula “don” c. ser and estar. . 9 For further discussion on the overtness of Hebrew copula. Spanish and Portuguese have two lexically different copular verbs. the postcopular NP (the predicate) alternates in case (nominative or instrumental) depending on whether the property is permanent (40) or nonpermanent or noninherent (41).Individual-Level Predicates 15 are instances of SL predications. When the predicate is a PP → copula “bè” (38) Ha-kli ha-ze *(i) patis he tool the this (COP)MASC-3SG hammer “This tool is a hammer” me’od ‘ayef (39) Dani (*hu) Dani (COP)MASC-3SG very tired “Dani is very tired today” byl durak (40) Oleg Oleg-NOM was fool-NOM “Oleg was a fool” ha-yom today 8 One of the closely related languages collectively called Manding. 1996.8 distinguish three types of copular verbs depending on the category the predicate belongs to (37). The Hebrew copula alternates between overt or null. the copula is null in the present tense but overt in the past tense (Kondrashova 1995. Bambara is one of the officially declared languages of Mali. Matushansky 2000). Other languages. In turn. I expand the discussion regarding what seems to be an adequate description for the IL/SL dichotomy. where the semantics of the two Spanish copulas are introduced. such as Bambara (Koopman 1992). In the past-tense cases. 2.2 When There Is More than One Copular Verb: Spanish Ser and Estar A good number of languages show copular alternations. and as secondary language it is employed to communicate nationwide. In the next section. (37) a. those clauses where the topic is an individual are instances of IL predications. hinging on whether the predicate expresses an inherent or definitional property of the subject (38) or not (39) (Greenberg 1994).9 In Russian. found across most of western Africa. I take up this issue also in chapter 7. When the predicate is an AP → copula “ka” b. A large part of the population uses Bambara as its mother tongue.

if. 11 See also comments by Milsark (1974) and Carlson (1977) about the same predicate. despite being ungrammatical with ser. or ser. I would like to note Demonte’s (1979) observation that this distinction poses an asymmetry in the account of copula distribution. if the subject is a physical entity (43). whereas estar is used to refer to those properties considered accidental.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. Roldán 1974) describe the distribution of the two copulas ser and estar on the basis of the Aristotelian dichotomy “essence” versus “accident. In the case of Spanish.” as cited previously (see section 2.12 10 Note that these descriptions fall in line with the ones given by Milsark (1974) for “properties” and “states.2.1 Distribution of Copular Verbs in Spanish Some traditional grammarians (Bello 1847. 2. the copula used can be estar. if not a mistake. discussed in section 2.1. Ese artefacto es/*está that artifact ser/estar-PRES-3SG a computer from 1960 “That artifact is a computer from 1960” una cabra b. The classic counterexample in this respect is estar muerto ‘be dead’.10 Essential properties are understood as permanent.11 which. Este animal es/*está that animal ser/estar-PRES-3SG a goat “That animal is a goat” professor c. there is no general agreement in the literature on how the copula alternation should be characterized. when the predicate is an NP. and accidental properties as temporary.1.1. if the subject denotes an event (44). un ordenador de 1960 (42) a. as . Pedro es/*está Pedro ser/estar-PRES-3SG teacher “Pedro is a teacher” When the predicate is a PP.1). which led to the claim that ser corresponds to permanent properties and estar to temporary ones. insufficient to characterize the copular alternation. other authors observed that the distinction between permanent and temporary was. How can the distribution of the two copular verbs be described? As traditionally observed. 12 Although I will not work out this difference here. obviously designates a nontemporary property. the copular verb must be ser (42). However.” The copula ser is used to refer to properties that are considered essential of the individual.

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

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

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

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

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

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

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

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

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

Since estar refers to a result state.2. its underspecification in aspect. female. happy). ser manifests a wider flexibility. PPs (when the subject is a mobile and immobile object) Estar + Table 2. since there is no reason to consider in the garden as a delimited predicate. That is.24 In a nutshell. whereas ser is underspecified with respect to aspect. nor is it an event or a process.) 2.1. APs (human. since the distribution of Portuguese copulas nearly parallels those of Spanish [see Table 2. scenarios incompatible with a result state are excluded for estar. PPs (when the subject is an event (party) or an immobile object) APs (tired. Accomplishments are those predicates that refer to a process with a inherent endpoint (write a letter or build a house).1.26 Individuals in Time (76) Imperfective predicate A(x) at times tj… tj+k However. old). However. APs (human. Portuguese and Spanish copular verbs The concrete claim of Schmitt (1992) is that estar involves aspectual properties.2 Schmitt (1992). round.) Copula Ser + Portuguese NPs. 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). its wider flexibility in combining with adjectives indicates that it does not involve internal temporal characterization. her reasoning is as follows. but. (Schmitt works on Portuguese. Predicates like be in the garden are not covered by Luján’s generalization.1]. In contrast. female. Schmitt provides several scenarios where estar and ser contrast in their possibility of appearance and argues that the explanation rests on the aspectual properties of estar. PPs (when the subject is an event (party)) APs (tired.2. Schmitt (1992) argues that the distinction between ser and estar is aspectual in nature. Estar corresponds to a result state of an accomplishment verb. States are eventualities with no duration or endpoint (be tall). which suggests. old). ser has no inherent temporal structure. happy). In the spirit of Luján.2). . swim). It is not a state. and processes or activities are those events with duration but without an inherent culminating point (walk. round. I take it that her conclusions about the topic can be applied to Spanish. for this author. (More on this in chapter 7. the description of SL predicates as perfective (in the sense of “delimited”) does not capture the intuition about the whole class of SL predicates. PPs (when the subject is a mobile object) Spanish NPs.

by virtue of the aspectual underspecification of ser. then. nice) can appear in the progressive. However.Individual-Level Predicates 27 (a) Possibility of appearing in pseudo-clefts and equative constructions Only ser can appear in pseudo-clefts and equative constructions. I will make two brief remarks. there is no result reading available. In this respect. estar predicates cannot. a house has been built but in John was building a house. since it would be the aspectual head that licensed the aspectual form. it is not surprising that just predicates of the kind like cruel are the ones that are grammatical in progressive contexts. which looks like a circular explanation. In sentences like John built a house. mean. where the copula does not play any role in terms of selection and does not add any semantic aspectual content to the construction. it seems that the incorporation of ser into the aspectual head (progressive) is what licenses the progressive. First. kind. With the progressive. if the reason for the grammaticality of (80) is the incorporation of ser into the . el chico del que te hablé (77) Juan es/*está Juan ser/estar-PRES-3SG the guy about whom you I-told “Juan is the guy I told you about” es culto (78) Lo que Juan es/*está what Juan ser/estar-PRES-3SG is cultivated “What Juan is is cultivated” (b) Possibility of appearing in the progressive Whereas ser predicates can appear in the progressive. Second. ser + some APs (such as cruel. Schmitt continues. Since the progressive favors subjects that have some control over the predicate. (79) María está *estando/siendo simpática Maria is estar/ser-PROG nice “Maria is being nice” Schmitt argues that this contrast is explained by the fact that the progressive cannot range over result states. the house has not been built yet. (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). Schmitt (1992) argues that ser incorporates into an aspectual head (the progressive in (80)) and becomes a predicate with internal temporal structure. according to Schmitt.

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

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

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

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

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

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

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

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

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

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

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

Because permanency and stativity are temporal concepts in nature. To accomplish this task. 3. culmination. are relegated to the next chapter. Differing from most previous literature. I will examine the temporal properties of these predicates. also called “aktionsart” or “situation aspect” (Smith 1991).1 Inner Aspect “Inner aspect” (Verkuyl 1989.1. I give a few introductory notes about inner aspect and its relevance in the description of the properties of sentences. the extent up to which the dichotomy between states and activities is real and grammatically relevant. I described individual-level (IL) predicates as predicates that are said of their subject independently from a given situation. Once we have become familiar with the different event types. I then present a survey of the distinct types of predicates traditionally distinguished in the literature.1 Inner Aspect and Event Types In this section I deal with two points..4—namely. I present one of them in section 3. I defend a finer grained distinction among IL predicates based on the different inner-aspect properties they show. First.Chapter 3 Event Classes and Individual-Level Predicates In chapter 2.e. One group tests out as states. 1993). Such properties tell us whether the event inherently tends . or delimitation. according to a set of tests. Inner aspect has to do with properties such as duration. I will demonstrate that copular adjectival constructions can be classified in two groups. I will show that not all cases in which the copula ser participates share the same inner-aspect properties. their internal temporal properties (inner-aspect properties). such as where the stem of the contrast resides (in the copular verb or in the adjective itself). Such a contrasting behavior raises a number of questions and invites reflection on some general issues. as has been widely assumed across the literature. in this chapter. According to classical and well-established aspectual tests. 3. I will bring the proofs onto IL predicates. the criteria to distinguish them. I also emphasized that this should not be taken to mean that IL predicates are stative and permanent predicates. and the tests to diagnose them. refers to the internal temporal properties of events in sentences. I start by investigating. whereas another can be justified as dynamic predicates (i. or whether there is more than one (IL) copular verb. as activities). I first introduce the notion of “inner aspect” and the different types of events that can be distinguished according to it. Other more specific issues. Specifically.

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

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

write novels. and it is the habit that is not understood as delimited in time. it lacks duration. Achievements Instantaneous events. in itself. hate cough.42 Individuals in Time like (4) works as an accomplishment. fall sleep. but. we also see that such templates cannot be a pure lexical matter. Events with duration but no endpoint die. 1967) (in addition to the class proposed by Smith). Semelfactives are not of great relevance to this work. be sick. collapse. be tall.” In (8).1.1 briefly characterizes each event type and provides some more examples. I return to the meaning of habituality in chapter 5. In (6). However. Verbs involving such a property are called “achievements. that is why. Both are conceived as habitual happenings. Event types Sentences (1)–(8) illustrate the event classification as proposed by Vendler (1957. That is. explode walk to the beach. This predicate does not take time or involve any kind of “process”. the latter two refer to a habit that does not have to end. read a chapter Accomplishments Activities States Semelfactives No-actions that hold in time but do not take time. the subject is involved in an action that is instantaneous and lacks an intrinsic endpoint. 1989. has neither duration nor endpoint. build houses and read novels are conceived as different building-houses and reading-novels events distributed over time. 1994 and Pustejovsky 1988). recognize. since they can be . sneeze. love. it is not an action in any sense. As noted earlier. there is a slight difference between activity-typed events like walk or push a cart and others like build houses or read novels. push a cart. fall asleep denotes an endpoint. each building-a-house or reading-a-novel event is conceivable to reach an endpoint. However.” The predicate in (7). belong. knock Table 3. know. Whereas the former two clearly refer to processes that need not ever end. arrive. Table 3. the same with a bare plural as an internal argument instead (build houses) behaves as an activity (see especially Tenny 1987. Eventualities like this are named “states. They can be considered as complex events with internal steps toward an end. awaken. with an endpoint but no duration Actions with a culminating point that take duration to be completed. build a house. importantly. Instantaneous and nonculminating events swim. They lack any kind of internal structure. be born. so I will not discuss them further. walk around the park be green. such events are called “semelfactives” (Smith 1991).

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

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

Juan traza un círculo Usually. on the other. whereas (SL) statives can be said to be ambiguous.2). and those that do not. 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 5.2 Activities and States versus Achievements and Accomplishments. and achievements and accomplishments.” as their reaction against combination with habitual adverbs proves. nonstatives are interpreted as habitual. Normalmente. Juan walks d. However. with stative though SL predicates. Juan draws a circle Nevertheless. In this subsection I go through a number of tests that separate activities and states from achievements and accomplishments. are completely grammatical. Juan se da cuenta de que su madre tiene razón Usually. Cinque 1999).Event Classes and Individual-Level Predicates 45 habitual (Comrie 1976. involving an IL stative predicate. whereas Juan walks does. according to the temporal adverbials that refer to the time an event lasts they admit. 4 I thank Tim Stowell for interesting discussion on this topic. I argue that the crucial condition for a predicate to appear in habitual form is its possibility of restarting.4 Whereas (11a).2. Juan realizes that his mother is right c. it is worthwhile to notice that when no (habitual) adverb is present. I show that this is a possibility for every type of predicate (including IL predicates) if a suitable context is created. (11) Interpretation as habitual in present tense a. . Juan knows mathematics b. is excluded in combination with a habitual adverb (which proves it cannot be interpreted as habitual). on the one hand. It seems clear that Juan is sick or Juan is at his best friend’s place does not have the habitual reading as the only (and preferred) reading. Normalmente. *Normalmente. States are argued to be the only type that cannot be interpreted as “habitual. The eventualities whose interpretation in the present is not habitual are states. Juan pasea Usually. I start out with the tests in (13) and (14). Bertinetto 1986. Juan sabe matemáticas Usually. sentences like Juan is usually sick or Juan is usually at his best friend’s place. Such compatibility cuts the pie of eventualities in two: those that have a habitual interpretation. A durative adverbial. it might seem that this test is a diagnostic for IL-hood rather than for stativity.1. which divide activities and states. Normalmente.3. 3.

*Pablo viajó en tres semanas Pablo traveled in three weeks c. Pablo viajó durante tres semanas Pablo traveled for three weeks c. Pablo se dio cuenta de que su madre tenía razón en tres semanas Pablo realized that his mother was right in three weeks The following contexts also group achievements and accomplishments together. *Pablo estuvo enfermo en tres semanas Pablo was ill in three weeks b. in-adverbials refer to the period of time it takes to complete a certain event. (13) In + x time a. the entailment of a delimit 5 Note that this test does not distinguish between IL and SL predicates. Pablo estuvo enfermo durante tres semanas Pablo was ill for three weeks b.46 Individuals in Time ending point.5 (12) For + x time a. as (13) shows. For a recent account about the different readings available in such sentences. 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. which is the reason why they are compatible just with eventualities entailing a delimit point. see Piñón 1999. . Pablo construyó una casa en tres semanas Pablo built a house in three weeks d. As the ungrammaticality of (14) shows. 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 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. Pablo está construyendo una casa Pablo is building a house d. *Juan arregló la lámpara.1. constitutes another diagnostic to differentiate accomplishments from activities. Pablo está paseando Pablo is walking b. (16) Perfect entailment from the progressive form a. this test and the for-adverbial one distinguish between permanent and nonpermanent properties. Pablo ha paseado Pablo has walked c. . and he is still walking b.Event Classes and Individual-Level Predicates 47 point in accomplishments and achievements makes them be incompatible with an assertion of noncompletion.2. unlike activities and states. Pablo ha construido una casa Pablo has built a house 3. accomplishments can be asserted to have taken place just when the process toward a culminating endpoint is over. and he is still arriving (15) a. rather than between IL and SL predicates. Whereas it is licit to establish (16b) as an entailment from (16a). *Juan llegó. The test in (17) reveals one property that activities and accomplishments share. and he is still fixing it b. in (16). y todavía sigue llegando Juan arrived. Juan estaba enfermo y todavía lo está Juan was sick and he still is The next test (Kenny 1963). but not when it is ongoing.7 as (15) proves. (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. However. since there is no endpoint privileged. With activities. y todavía sigue arreglándola John fixed the lamp. it is not licit to draw (16d) from (16c). it is legitimate to assert that the event has taken place at any point of its process. Juan salió a pasear por el parque y todavía sigue paseando Juan went for a walk.3 Activities and Accomplishments versus States and Achievements. (14) a.

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

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

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

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

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

Actually. To the extent that just some types of events can be undertaken and commanded. in principle.” Also. while volitionality and control are properties that just animate causers can embrace.” The cleft with “do” and the test of command imperative detect eventualities that can be “done. cannot hold with any willfulness on the subject’s behalf. First. although volition usually entails control. (20). activities and accomplishments may be the most persistent co-patterning of the table.2 A Brief Stop at “Agentivity” The set of tests in (22)–(25). 3. agency tests separate states and achievements from activities and accomplishments. 3. which gives bad results in contexts with adverbials such a deliberately. it is important to distinguish the semantic features it contains and to establish whether there is some feature more basic than the other. causation is the most basic notion involved in agency contexts. In the following section. The results of (17). agency tests detect eventualities that denote stuff that can be “brought about. three notions involved: causation. and (22)–(25) also suggest that states and achievements have characteristics in common. Likewise. not all controllers involve volition.1 A Cluster of Notions Although. I present a brief discussion about the agency tests above and about the notion of “agent” that I will observe in this work. but just holds. which diagnose agency. when I discuss the aspectual behavior of copular IL predicates. volitionality. In particular. Further discussion about the tests and about some concrete event types will come in section 3. To begin this investigation.” Then. Vendler (1967) himself considered states and achievements a natural class. In fact. at least.Event Classes and Individual-Level Predicates 53 properties. I will deal with the different notions that get blended in the term “agent. there are. As mentioned before. I consider two aspects related to this. In agency. and control. agency tests work as event-type tests. the term “agent” clearly denotes ‘the argument that brings about an event’. consider the classical minimal pair in (26) and (27).” and therefore “commanded. I will briefly introduce the relation between agents and event structure. (26) (27) El portazo rompió el cristal The bang broke the glass Juan rompió el cristal Juan broke the glass . As will become clear. may be used to discern properties of eventualities as well. In this section. stuff that is not performed.2.3.

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

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

or the intentionality of the agent. Cinque (1999). the adverb clearly refers to the DP subject. where they assign a role to the underlying subject. such an ambiguity poses a puzzle: how is it that it can also refer to the DP subject. I turn next to some discussion about the property of “volition. Some involve an additional characteristic— namely. since my judgments are based on them. (i) The doctors examined Juan deliberately Los médicos examinaron a Juan deliberadamente (ii) Juan was examined by the doctors deliberately Juan fue examinado por los médicos deliberadamente The interpretation of the adverb in cases like (ii) has been considered ambiguous in English. (I give the Spanish counterparts. proving its agentive properties. From now on.) (iii) Joe intentionally has seduced Mary Joe intencionalmente ha seducido a Mary 13 . the kind of adverb. we should say that when the DP is [+animate] the interpretation as a volition causer is possible. Consider (ii).” An agent is the cause of the event. I will reserve the term “agent” to causers with volition. also gives (iii) and (iv) as cases showing that some adverbs retain their orientation on the subject. it has also been extensively discussed whether such adverbs refer to the “subject” rather than to the “agent” in a sentence. in the passive (ii). The evidence mostly studied regarding the subject/agent orientation of the cited adverbs comes from passive sentences. Adverbials of the type of deliberately can be adjuncts at the V′ level. if it (assumedly) bears the role of “patient” (given the fact the sentence is in passive voice)? McConnell-Ginet (1982) proposes that the passive auxiliary be is interpreted as ‘to act’. or at the I′ level. If deliberately is an agency marker. it seems to be able to refer to the subject as well as to the DP in the agentive by-phrase complement.” Adverbials such as deliberately refer to the volition. However. and the position of the adverb matter. I wanted to show that “cause” is the most basic feature involved in the concept of “agency. volition—that is linked to the properties of the noun heading the DP subject (animacy).13 Actually. This is a very complex issue where the kind of verb. With this brief discussion. Zubizarreta (1987) argues that adjuncts such as adverbials assign “adjunct theta roles” to arguments. where they assign an adjunct role to the surface subject. but not all causers are on a par. 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.56 Individuals in Time (31) vP 3 *deliberately vP → nonvolition causer 2 DP[–animate] v More accurately. Whereas in the active counterpart (i). reporting an example from Jackendoff 1972. it is not an obligatory interpretation (see below for discussion of sentences like John hit him with no intention). the willfulness. which explains the agentive property of its subject. the passive form of (i).

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

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

which I take up later in the work. As is traditional. . (17) As a complement of parar de ‘stop’ a. the syntactic contexts reviewed in section 3. we can conclude that only dynamic events can have agentive subjects. Given the special status of the subject of meteorological verbs (cf.” yet nevertheless is grammatical.Event Classes and Individual-Level Predicates 59 3.) points out the interesting case of Paró de llover ‘It stopped raining’. Pablo paró de construir la casa Pablo stopped building the house c. repeated here).2 Agents in Event Structure In this subsection I offer a few introductory words about the possible role of agents in event structure. When the subject is [–animate] and. dynamic events yield grammatical sentences when they appear as complements of parar de ‘stop’ (see (17). In the second place. show a certain correspondence between events that are agentive and those that can appear in the progressive. We can prove that an event involves “internal movement” if it can express its progression in time. where the verb parar de does not have a subject that can be said to involve “control. At this point I just want to notice the consonance of the results of the tests diagnosing dynamicity and of those diagnosing agency. the verb parar de implies dynamic control by its subject. which suggests that only with dynamic events the subject has control over its ending. I will assume that the ability of an eventive verb to combine with the progressive is evidence that the event involves dynamicity. *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. the sentences are ungrammatical. Assuming that progression in time defines dynamic events. where their subject was a [+animate] DP such as Jupiter). In the first place. the infinitive is grammatical after parar de only when the subject can be agentive. I will not consider this kind of case as a severe counterexample to my suggestion. Latin. *Pablo paró de amar a María Pablo stopped loving María b. used as tests to identify the different event types.c. these events are typically activities and accomplishments. As the following contrasts show.15 15 Tim Stowell (p. thus cannot be understood as agentive. that is. Pablo paró de pasear Pablo stopped walking d.1.2.

the beginning is interpreted as alluding to the moment at which the state starts holding of the subject. 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 enfermo con 50 años Juan started being sick in his fifties b. with states. Juan empezó a estar a disgusto con su jefe cuando vio que desconfiaba de él Juan started being uncomfortable with his boss. both of which give completely grammatical results. but they contrast with activities and accomplishments. States (40a) are not completely excluded. Some stative predicates yield sentences more acceptable than others: (41) a. Empezar can have both [+animate] as well as [–animate] subjects. In this respect. too. 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. when he saw he did not trust him A crucial difference between states and activities and accomplishments is that. *Juan empezó a llegar a la meta a las tres de la tarde Juan started arriving at the goal at three in the afternoon c. . Juan/*la rueda paró de moverse Juan/the wheel stopped moving b. 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. however. Juan empezó a escribir una carta a las tres de la tarde Juan started writing a letter at three in the afternoon d. the beginning is understood as the moment at which the subject starts undertaking the process. there is an interesting difference with parar de. With activities and achievements. which is expected since the weight of their semantic import is on the resultant state (Pustejovsky 1988). *Juan empezó a estar en Madrid cuando consiguió trabajo Juan started being in Madrid when he got a job b. whose subjects can only be [+animate] and agentive.60 Individuals in Time (39) a.

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

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

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

*Juan fue esquimal en dos horas Juan was an Eskimo in two hours b. In principle. *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).64 Individuals in Time (46) For + x time a. As shown in section 3. Juan ha sido cruel con Pedro Juan has been cruel to Pedro 3.2.3. Juan está siendo cruel con Pedro Juan is being cruel to Pedro b.1. only activities. Juan fue rubio durante tres años Juan was blond for three years c. Once again. (48) Perfect entailment from the progressive form a. tests like (49) divide activities and accomplishments from states and achievements. such a predicate shows a nonstate patterning. where there is no endpoint. since this is only possible with eventive predicates.1. yield a perfect entailment from the progressive form.3.1. However. 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.2. As mentioned in section 3. a predicate like ser + AP would be unexpected as a complement of parar de ‘stop’.2. . (48b) is a licit entailment from (48a). the combination of ser and an AP such as cruel ‘cruel’ looks good. *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. *Juan fue esquimal durante tres semanas Juan was an Eskimo for three weeks b. and not accomplishments.3 Activities and Accomplishments versus States and Achievements.

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

20 Recall that. 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. except for cruel.1.4 Accomplishments versus All the Rest. they clearly contrast with the other two. ?Juan casi fue cruel con su adversario Juan almost was cruel to his opponent b. None of the AP predicates with ser. Only ser cruel is possible under a command imperative form (54) or in combination with adverbs like deliberately (55). it contrasts with other APs with ser. like (53c).3.1. The next set of contrasts aim at testing whether the process that be cruel can be considered is agentive or not. Tests (52) and (53) show some more contrasts consistent with the results from (49).5 Agentivity Tests. which suggests that dynamic properties are somehow involved in these predicates. The former reading is available with accomplishments.2. . (52) As a complement of finish a. and (b) that the subject did not even start out the process.1. as mentioned in section 3. confirm be cruel as an activity.4.66 Individuals in Time 3.3. All tests give the same results. Juan no empezó a ser cruel con su adversario Juan did not start being cruel to his opponent c. ?Juan terminó de ser cruel con su adversario (en el minuto 10 del debate) Juan finished being cruel to his opponent (in the 10th minute of the debate) The following sentences with almost. Although be cruel predicates do not give perfect results after finish. fit in canonical agentive contexts. which also test accomplishments versus all the other eventualities. *Juan terminó de ser esquimal.20 Although (53a) does not sound extremely natural. the latter is with activities. *Juan terminó de ser rubio Juan finished being blond c. (53) Interpretative entailments from the adverb almost a. Juan finished being an Eskimo b. whose combination with almost is impossible. since the unique licit entailment from (53a) is (53b). *Juan casi fue esquimal Juan almost was an Eskimo 3.

As argued in section 3.1. or regret (60). I take this to show that the subject of this predicate unequivocally involves “volition. *Juan fue esquimal deliberadamente Juan was an Eskimo deliberately b.” besides being the instigator (controller) of the process. because of their inherent semantic reasons. still. be cruel is a grammatical complement for verbs like persuade. (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.Event Classes and Individual-Level Predicates 67 (54) Occurrence in the command imperative form a. *Juan fue rubio deliberadamente Juan was blond deliberately c. Sé cruel con tu adversario Be cruel to your opponent! (55) Combination with volition adverbials (deliberately) a.2. . *Sé rubio Be blond! c. force (59). Be cruel may not be absolutely natural in do-cleft constructions (61). 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. but. all of which need. a sharp contrast with the other APs is observed. *Sé esquimal Be Eskimo! b. an agentive subject for their infinitival complements.

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

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

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

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

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

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

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

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

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

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

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

Event Classes and Individual-Level Predicates

79

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

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

80

Individuals in Time

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

Event Classes and Individual-Level Predicates

81

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

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

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

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

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

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

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

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

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

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

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

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

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

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

in italics in (58). although it seems intuitively clear that the PP specifies the goal addressed by the subject’s action.3. 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. One can. 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”). I will show that its syntactic presence correlates with the aspectual patterning of the construction.3 Relational Mental Properties: The Relational PP Complement As just described. 4. In this section. even if we finally concluded that the PP is an “affected argument”. I investigate the nature of the PP complement. to name just a few. 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). for example. offend or regale. 4. (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. Likewise. whether or not the PP expresses an “affected argument” depends on the concrete type of action undertaken by the subject. and discuss all the consequences derivable from this. can be considered as an “affected goal.” This captures his intuition that such a DP refers to the individual that gets affected by the (underlying) action undertaken by the subject DP. Note that referring to some action without specifying what particular action is involved is quite common among verbs too: think of verbs such as humiliate. harass.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. or by acting in a certain way. 1988. and such an action is qualified as cruel. 1994] and . abuse. set on fire and bother. Incidentally. Thus.96 Individuals in Time correlation with their aspectual characteristics. offend someone else either by saying something unpleasant. focusing on two aspects: its interpretation and its optionality or obligatoriness.

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

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

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

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

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

From these cases. Some authors have claimed that consider selects for SCs containing IL predicates. SCs complement of consider can perfectly have SL predicates. be claiming that the person is not cruel. However. I will conclude that cruel-type APs are not inherently relational. without any apparent contradiction. and rejects SL predicates. (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. Stowell (1991) was right in considering the simple versions of relational MPs as IL and the ones accompanied by the PP as SL. This could in principle lead us to think that. 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” . such as Demonte and Masullo (1999). at the same time.102 Individuals in Time (80) Juan no es cruel. pero ha sido cruel con Pablo Juan no ser-PRES-3SG cruel but has ser-PART-3SG cruel with Pablo alguna vez some time “Juan is not cruel. The first scenario where cruel cannot appear followed by a complement PP is small clause (SC) complements of epistemic verbs such as consider. Sentences (83) and (84) are based on examples from Demonte and Masullo (1999). have pointed out. as other authors. in fact. (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. I will concentrate on some syntactic scenarios where an overt relational PP cannot appear. In the remainder of the chapter I will leave aside these (somewhat confusing) interpretive contrasts. but he has been cruel to Pablo some time” Example (80) shows that we can be claiming that one person has been cruel to someone else and.

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

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

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

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

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

1 Lexicalist and Logico-Semantic Approaches The observed correlations among the presence of the PP relational complement. The pair in (108) from Borer (2005) shows an alternation stative/dynamic which correlates with the [±animate] properties of the DP subject.5. syntactic presence of a PP) as correlations since. kind.5 Justifying the Approach 4. I argued that such a contrast is grammatically relevant. (107) El delantero avanzó hasta la portería The foremost advanced until the goal (108) a. 4.108 Individuals in Time tioned thus far. María tocó la verja (dos veces) María touched the fence (twice) The alternation here is of states versus activities. From a lexical point of view. La pared tocaba la verja The wall touched the fence b.2) the contrast between animate and inanimate DPs in pairs like (109) and (110). by definition. 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. two different entries (stative and dynamic) for each adjective (cruel.). 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.4). both atelic eventualities. This perspective cannot capture any of the other properties (interpretation of the subject. we would be forced to assume that each aspectual behavior of adjectival copular cases (stative and nonstative) corresponds to different (but homophonous) lexical items. the thematic interpretation of the subject and particular aspectual properties. In (107) it is a PP which is exerting a crucial aspectual role in making the event an accomplishment. In the remainder of the chapter I will be concerned with such a contrast in the realm of adjectival copular clauses. since there are syntactic and semantic facts suggesting its actual substantiation.” Recall in this regard from chapter 3 (section 3. mean. I begin by discussing which approach to event representation can make sense of the observed correlations among the syntactic presence of a constituent. However. it does not allude to the other elements present in the syntactic configuration. whose contrast has been debated in the literature (see section 3. That is. etc. according to which aspect properties are fully encoded in the lexical items. Along the lines of Borer (2005) (among others). .

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

1989. among others. showed the relevance of arguments in the inner aspect properties of a sentence.2. 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. both would contain the term “hold.5. 1989. As discussed earlier (see section 3.” Second. and the presence or absence of other elements in the sentence (the PP complement in our cases) that. Tenny (1987.” given the fact that both are atelic and do not “culminate. which opens the possibility of establishing a direct relationship between the thematic properties of the arguments present in a sentence. 1994). this approach does not establish any relationship between the aspectual term (Hold). 1994) and van Voorst (1988) elaborate both on the observation that arguments grammaticize the points that give the temporal contour of an event. correlate with the aspectual outfits of the construction. Tenny (1987. as captured in the Event Correspondence Rule (116). each entity corresponds to an aspectual notion. Dowty (1979). (113) •---------------------------------------object of origin/actualization (114) •----------------------------------------• object of origin object of termination (115) ----------------------------------------• object of termination Activities Accomplishments Achievements (116) Event Correspondence Rule (Van Voorst 1988) Subject NP Direct object NP •----------------------------------------• event-object of event-object of termination origin/actualization . First. arguably. 4. and van Voorst (1988). these authors emphasized the role of the internal object as an event delimiter. authors such as Verkuyl (1972). there is no obvious way to distinguish between the two constructions in point (stative and dynamic copular clauses).1. the dynamic or stative properties.1). 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. its syntactic position and its role in the determination of the aspectual properties of the sentence.2 Syntactic Approaches 4. since. As van Voorst puts it. in principle.110 Individuals in Time This approach leaves several questions unanswered. In particular.5.1 Event Roles.

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

. In other words. Travis (1994.5. In sum. these approaches would have to assume that each aspectual behavior corresponds to different adjectival projections. as such. 2003). and. 2000). The car is the theme in both cases. she does nothing that can be called buying in (122). 2000. although both Tenny and van Voorst explicitly acknowledge the compositionality of events. but the position of event-roles seems to be so. As Rosen (1999) points out. 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. However. Furthermore. Note further that the NPs play different event roles in each one of them. For our present concerns. this assumption makes them similar to lexical approaches. different lexical items have different ways to project. the work of authors such as Borer (1994.2. 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. (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). 1998. 2005). and. Benua and Borer (1996). correspondingly. therefore. since. he does nothing that can be called selling in (123). a natural account of such compositionality cannot be derived from their lexicalbased approaches. 1998. they cannot capture aspectual alternations such as the one we are interested here in a natural way.112 Individuals in Time In these two sentences the theta-roles are the same. and Sanz (2000) elaborates on the idea that argument structure is licensed by functional syntactic structure. which I have argued are in strict correspondence. The position of thematic roles is not restricted. appears in the subject position. Compare (120) and (121) with (122) and (123). 4. regarding the stative/dynamic contrast in IL copular clauses. occupies the subject position. but not so their syntactic positions. for them. The old lady instigates the selling in (123). Although these event-role approaches provide accurate insights in important respects (the fact that arguments can affect the event type. Ritter and Rosen (1996. Incorporating the insights from the lexical argument-oriented approaches just reviewed. the systematic correlation between object and delimitation and subject and origin). the event-role perspective predicts the mapping of different arguments to particular syntactic positions more accurately than the UTAH. In a sense.2 Syntactic Structure as Event Structure. However.

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

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

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

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

and I will also be concerned with the aspectual treatment of the arguments expressed via a PP. From the description in the previous sections. the PP makes them a dynamic predicate. That is. I argue. and Demirdache and UribeEtxebarría (1997. all of them very likely emerging from animacy itself: the DP has to be able to behave as a controller agent. Stowell (1993). the dynamic aspectual behavior) actually resides in the preposition itself. since their intervention in the event structure is not obvious (given that they are oblique arguments). I will discuss the syntactic treatment of the opposition state/activity. it has become clear that the account for the cruel-type constructions has to capture. As I intimated before. Point (a) is discussed in section 4. In support of this proposal. that prepositions can be conceived of as heads encoding aspect properties.3. the following two properties: (a) When there is a relational PP.) in the active cruel-type constructions is a projection (the PP) that combines with the “basic” structure of the AP. Let us begin by discussing point (b) concerning the aspectual contribution of the preposition. I propose that the source of dynamicity and its related properties (agency etc. . 2000. I will address two theoretical points. but it correlates with the existence of the PP relational complement. As I advanced. which gains dynamic properties from the preposition heading the PP denoting the DP goal. I will take such a correlation seriously and I will propose that the source of the aspectual properties (specifically.6.1 Prepositions as Aspect Encoders I have shown that cruel-type APs.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. To capture (a)—that is. behave as activities. I propose that the dynamic properties become a part of the construction through the preposition heading the PP.Aspectual Alternations in IL Predicates 117 4.6. based on Hale (1984). the DP subject has to involve a particular set of properties. As I develop the proposal. (b) When an animate DP subject and the PP are present. 4. 2004). I will argue that the stative version of the cruel-type APs corresponds to their most “basic” structure. among others. To capture (b). the aspectual alternation (state/activity) is not unpredictable at all. as soon as they combine with a (relational) PP. the construction has characteristics proper of processes. 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. at least.

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

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

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

. Demirdache and UribeEtxebarría (1997. the locative preposition at directly combines with the verb in infinitival form to construe the progressive. in Dutch. similar cases are found in Romance languages. 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. such as from.Aspectual Alternations in IL Predicates 121 ‘(with)in/on/at’. as prepositions do. For example. to (143). in contrast to stative locative verbs participating in progressive forms. ‘to’) (144) or the English about to (145) are constituent parts of prospective forms. In turn. Specifically. indicating movement toward (centripetal movement). 2004) conclude that Aspect can be analyzed as a head establishing relations between two terms. 2000. (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. 23 Note also the presence of a verb denoting movement. prepositions of noncentral coincidence are present in the auxiliaries expressing perfect and prospective aspect across languages. (142) (143) (144) (145) Voy a hacer la cama I am going to make the bed Sto per uscire I am about to leave According to Givón (1982). they notice that. For example. are present in perfect aspectual forms in some languages. appear in the form used to express close future (going to). in Nigerian Margi and Palaung (from the AustroAsiatic family of languages).23 Prepositions like the Italian per (‘for’. the prepositions a (142). prepositions denoting ‘movement from’ (centrifugal movement). in Spanish and English. where “come from” can be used with the aspectual meaning of the perfect. Also.

In (149b) the TT is located “after” the span of time the event of cradling takes place. When I entered the room. which establish a temporal ordering between time-denoting arguments. Maria was going to cradle the baby //////////[------------The trees below schematize Demirdache and Uribe-Etxebarría’s (1997. . Following Reichenbach (1947). 2004) argue that Aspect can be conceived as a head establishing a temporal ordering relationship between two times. the time the speaker is talking about is a concrete time. Prospective AspP 2 t2 Asp′ t1 (within) t2 Asp′ t1 (after) t2 Asp′ t1 (before) 2 2 2 Asp Asp Asp 24 Hale (1984) proposes to analyze aspect and tense forms as central/noncentral coincidence relations. and in (149c). (150) a. the kind of the relations being either of central coincidence (progressive aspect) or noncentral coincidence (perfect and prospective aspect). Maria was cradling the baby [----------////////------------b. Based on Klein’s (1994. Maria had cradled the baby [----------------]///////// c. Progressive AspP 2 b. When I entered the room. 2000. When I entered the room. In (149). 1995) idea that Aspect orders the event time with respect to the time the speaker refers to in the sentence (the “Topic Time” or “Assertion Time”).” The discontinuous line represents the event and the slashes the TT. when he entered the room. 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. he characterizes tense distinctions as relations of priority.122 Individuals in Time Demirdache and Uribe-Etxebarría take as a point of departure Stowell’s (1993. This time is the Topic Time (TT). namely. 2000) proposal inspired in Hale 198424 and Stowell 1993. It is conceived as “contained” in the span of time the event of cradling takes place in (149a). (149) a. Some clarifying examples appear below. Demirdache and Uribe-Etxebarría (1997. Perfect AspP 2 c. Demirdache and Uribe-Etxebarría argue that Aspect can be conceived in parallel terms. 1996) proposal that the semantics of tenses can be described in analogy to prepositions like before or after. the TT is captured “before.

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

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

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.

The structure I thus propose for cruel-cases goes along the lines of (180). . a semantic function such as SHAPE. I propose that this dynamic preposition is present in all relational MP cases. para-paths are homogeneous. As suggested above. a directional preposition with the meaning of ‘toward’ (cf. which maps paths onto events. since the concatenation or sum of two para-paths is also a para-path. As a noncentral preposition. which in the aspectual realm translates into an unbounded eventuality. In this sense. it semantically entails that the destination has not been reached. When it is phonetically null. we can account for the event properties from the properties of the path itself. Morimoto 2001)—be it overt or not. Also. (179) GoalPathP 3 GoalPath PlaceP para/(∅) 3 Place DP (con) I propose that it is due to the dynamic properties of this preposition with the meaning of ‘toward’ that cruel-type constructions in the presence of the relational PP complement pattern with activities (instead of with states) in all the tests previously shown. which explains the directional meaning it acquires. where the GoalPathPP involves a null directional preposition of the type of para. I have argued that the PP has the properties of the preposition para (‘for’)—that is. assuming with Zwarts (2005.130 Individuals in Time Based on the existence in old and modern Spanish of a combination of prepositions to introduce the MP goal. para + con. 2006). it can be assumed that preposition con incorporates into its position. Jackendoff 1976. I argue that it is due to the properties of the path described by this preposition that the event is unbounded.

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

The examples (ii)–(v) from Davis (in prep. which argues in favor of the hypothesis that the preposition heading the complement is an actual dynamicity inductor. they gain dynamic and agentive properties.). a concluded process) are interpreted as past tense sentences. although. sentences with t’iq or tsicw. in inner aspect terms. to some extent. admit a relational PP complement. consider the temporal contribution that motion verbs make in Lillooet Salish. indicating a reached destination (i.) shows motion verbs in the mentioned language.3. but merely as a ‘process’ or ‘activity’.. shrewd) that. That is. but not in past. by their lexical meaning.e. in prep. Due to this reason. Interestingly. when these adjectival predicates combine with the PP. both refer to an ongoing process. we can say that such prepositions convey also the meaning of ‘before’ in the inner aspect domain. then. as mentioned in section 4. in other words. are interpreted in present tense (or future). The following table (Davis. Because of the same reasoning. no process has been fulfilled33. sentences containing this verb and lacking an explicit temporal marker. an incomplete process. the PP is not contextually inferred if it is not overtly present.132 Individuals in Time reached (“the destination has not been reached”). cunning. it cannot be said it has reached its destination. Finally. there are other adjectives (stupid. (184) Juan fue estúpido/ingenioso (*a propósito) Juan was stupid/ cunning (on purpose) (185) ?Juan fue estúpido/ingenioso con su editor (a propósito) Juan was stupid/ cunning to his editor (on purpose) This contrast shows that it is by virtue of the conjunction of the adjective and the PP in the syntactic configuration that the inner aspect nature of the In a similar vein. Since the figure is captured “before” the ground. which translates as ‘nonaccomplishment’. illustrate this point. (i) Motion toward speaker Motion away from speaker Destination not reached ts7as Nas Destination reached t’iq Tsicw 33 Verbs ts7as and nas share the property of referring to “movement” whose destination has not been reached. (ii) Ts7áswit ets7á Sát’a lhláku7 Lh7úsa “They are coming here to Sat’ from over there at Lh7us” (iii) T’iqwit ets7á Sát’a lhláku7 Lh7úsa “They came to Sat’ from over there at Lh7us” (iv) Náswit áku7 Lh7úsa lhelts7á Sát’a “They are going over there to Lh7us from here at Sat” (v) Tsícwwit áku7 Lh7úsa lhelts7á Sát’a “They went over there to Lh7us from here at Sat’ .3. In principle.

6. such as the progressive or as a complement of parar de ‘stop’. I repeat the contrasts below. one important property we want the structure proposed for the cruel-type captures is the correlation between the presence of the PP complement and the properties of the DP subject. the relational PP complement cannot appear. 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.3. eventuality properties are obtained from the very syntactic structure. And second. In sum. it suggests that the properties of the noun in the DP subject and the PP stand in a codependent relationship. As shown in section 4. as a complement of verbs like force or in combination with volition adverbials: (190) Occurrence in command imperative Juan/*imagen. when adjectives such as cruel are predicated of an inanimate subject. be cruel!” (191) As a complement of force El director forzó a Juan/*la imagen a ser cruel “The director forced Juan/the image to be cruel” . First.3 The DP Subject of Relational MPs As I mentioned in the beginning of the section.2. when the DP subject is inanimate.Aspectual Alternations in IL Predicates 133 construction is decided. and we could not explain why it cannot appear in sentences like (186). such as the imperative form. the PP complement would be possible independently from the properties of the DP subject. 4. when the subject is inanimate. the construction cannot appear in contexts proper of dynamic eventualities. If that were the case. (188) Juan/*esa imagen está siendo cruel Juan/that image estar-PRES-INDIC-3SG ser-ing cruel (189) Juan/*esa imagen ha parado de ser cruel Juan/that image has stopped ser-ing cruel Likewise. This fact suggests two things. the construction cannot appear in agency scenarios either. ¡sé cruel! “Juan/image. it suggests that the PP is not a plain complement of the adjective.

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

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

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

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

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

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

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

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

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

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

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

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

.

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

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

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

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

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

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

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

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

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

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

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

what the relationship is between the number of occasions alluded to by each viewpoint and the quantity nature of predicates (that is.1. giving rise to another interpretation. Let me start by considering sentences (22)–(24). As I will repeat later in section 5. The number of occasions the progressive and the perfective form allude to is just |1|. However.158 Individuals in Time of Quantity. Q<occ> can also have an existential value.2. the habitual imperfect.2 and next in section 5. However. as shown in (25). inner aspect. the three forms differ in the number of occasions they allude to. here in section 5.5. and second. I address these points in turn. besides the ordering component. I will consider two points—first. can be either |1| or |>1|. (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. aspectual viewpoints.3. The values of such a quantifier. whose head is a quantifier over occasions (Q<occ>). 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. and the progressive. (25) ----------------------x---------------------(26) ------x------x--------x-------x---------x-In the spirit of Verkuyl (1999). In this latter regard. as in (26). besides ordering. as the reader may have conjectured already. how many occasions each viewpoint alludes to. It appears . In this section I am chiefly concerned with the perfective. The continuous imperfect does not count the number of occasions a determined predicate takes place. I propose that aspectual viewpoints involve.1 Quantifying over Occasions As just mentioned. who observed that the difference between the progressive and the habitual rests on the number of occasions they refer to. the nonhabitual or “continuous” imperfect. 5. the functional projection which makes the eventuality heterogeneous and whose absence leaves it homogeneous. also contain information about the number of occasions an eventuality occurs. a quantificational component.

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

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

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

16 Likewise. where. as a consequence. in and of itself. simply. 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. several factors seem to play a role. refers to a plural number of occasions. as she conceives habituals to be. and. as Tim Stowell (p. where the predicate is understood as iterated: (i) Hice ese pastel durante años (como postre de navidad) I-made that cake for years (as a Christmas dessert) De Swart (1998) accounts for these cases by arguing for the existence of a coercer operator. Although I will discuss neither the process of coercion itself. For example.15 the perfective and the progressive can never denote a plural number of times. it is interesting to consider accomplishment predicates in perfective followed by a “for + time” adverbial. that cake is not physically the same cake year after year.) points out.162 Individuals in Time Regarding the number of occasions that the different viewpoints can allude to. it is interpreted as a progressive (ii). there is no need to interpret that the paper has been read completely or the sonata has been played entirely. where this interpretation is much more difficult to get. whereas the imperfect habitual. Once accomplishments are states. 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). which turns the event (make the cake) into a state. the iterative interpretation seems possible when the DP object can be understood as different instances of the same object description. Consider (38). Compare (i) to (ii). (iii) Leyó el periódico durante dos horas he-read-PERF-PRET-3SG the paper for two hours . (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. One can be reading just sections of the paper and playing just parts of the sonata. the point I want to make with (27) is that. However. the sentence is. nor an account for these cases. I would like to note that in the iterative interpretation with the perfective and for-adverbials. this does not mean that the number of occasions that an eventuality in the perfective takes place cannot be modified by an adverbial.c. their compatibility with for-adverbials gets explained. 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).

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

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

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

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

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

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

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

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

4).176 Individuals in Time 5. draw a conclusion in the same direction. If we say John walked from 2 to 3.” as mentioned in the previous section. In other words. (95) *Pablo nadó en una hora Pablo swim-PRET-PERF-3SG in an hour “Pablo swam in an hour” Other authors. However. Bach (1986). at least in Spanish. activities are not 100 percent homogeneous at a micro-level. The meaning of these sentences is that Pablo was once involved in the task of swimming (91) and walking (92). (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. which leads to the inference of “completeness” and “boundedness. I argue that a perfective form does not trigger telicity effects on predicates. In the same vein as before. With the perfective. 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. unlike states. Piñón (1995). a perfective form does not interfere with the quantity nature of the predicate. and others. such as Bertinetto (2000). he has had that car at every single moment between 1974 and 1985. de Swart (1998). 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. it does not mean that he has been walking at every single moment of that interval. Contrary to Mourelatos (1978).3.23 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. if we say John had that car from 1974 to 1985. As discussed back in chapter 3 (section 3. the TT is located ‘after’ the ET. who conceive the perfective as a modifier that makes an eventuality telic.2 Perfective Homogeneous Predications I turn my attention now to homogeneous predicates appearing in perfective form. In contrast.24 as subinterval and additivity proofs show. I will argue that. 24 23 .

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

I argued that.4. which does not make the sentence habitual. and not only with SL ones.26 The quantifier over occasions binds the eventive variable (e) which. |>1|. in Spanish. (98) ZP 2 Zi VP 2 ei VP (99) AspP 2 TT Asp′ 2 Asp0 ZP (after/before/in) 2 VP Z0 (|1|. independently from the kind of predicate.” habitual imperfect is at stake. 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). if “one. where there are no occasions to enumerate (think of stative homogeneous predicates such as Peter is tall). as Kratzer (1988) originally proposed. In the last section. I assume to be present with every type of predicate (SL and IL). I used Spanish habitual and perfective data to show that aspectual viewpoints do not have any impact on the inner-aspect properties of predicates. presented in (3) above (repeated here as (98)) about the internal constituency of ZPs. the perfective does not provide “quantity” to homogeneous predicates and habitual imperfect is not in compleRecall that an action can take place more than once (He hit him five times). and makes the prediction that the quantifiers (|1|.” perfective or progressive. the ordering component takes the quantified eventuality and orders it with respect to the TT interval. As will be shown in section 5. ∃) 2 e VP Thus. This unifies the working of Aspect. this is borne out.25 The existential quantifier corresponds to the continuous imperfect. See section 5. 1996). but these predicates are also compatible with the habitual quantifier (She was sick usually). 25 . When the quantifier denotes a number roughly describable as “greater than one. 26 This aspect form is involved with stative SL predicates (She was sick). ∃) should be possible with any type of predicate. following Stowell (1993.1 for further discussion.178 Individuals in Time cation) is depicted in (99).5. Specifically. Elaborating on Stowell’s (1993) proposal. |>1|.

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

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

Since I argued that be Eskimo and be blond pattern with states. whereas (105) can have it: (106) *Normalmente Pablo era esquimal/rubio usually Pablo ser-IMPF-PRET-3SG Eskimo/blond “Pablo used to be Eskimo” cruel (107) Normalmente Juan era usually Juan ser-PRET-IMPF-3SG cruel “Juan used to be cruel” Habituality entails. successive initial points and successive (inferred) ending points. then. I propose. Thus. It seems. that a Q<occ> with a value of greater than one is not present in (103) and (104). at least on a first approximation. that the structures corresponding to (103) and (104) involve an existential quantifier. it seems. but the property is attributed to the person as a whole. habituality entails that a given eventuality can hold or take place more than once. I argued. I want to show that.Outer Aspect and Individual-Level Predicates 181 cruel (105) Pablo era Pablo ser-IMPF-PRET-3SG cruel “Pablo was cruel” The imperfect form in (103) and (104) does not have a paraphrase expressing habituality. the absence of habitual interpretation with IL predicates such as Eskimo or blond can be explained by their inability to restart. it is not the case that they are incompatible with a quantifier over occasions by . where we are not counting the number of times an eventuality holds. therefore. However. that states are incompatible with such quantification. (108) AspP 2 TT Asp′ 2 Asp ZP (within) 2 Z0 V (∃) 2 e VP The structure of (108) corresponds to the interpretation of a continuous imperfect. although the habitual paraphrase is not the most salient one when (IL) stative verbs (such as Eskimo or blond) are at stake. That is to say.

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

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

they acquire activity-like properties. although the progressive form is correct with cruel. as noted in section 5.29 (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. (122)). whereas activities in imperfect (117) can also have a progressive paraphrase (118). this is just marginally available with be cruel (cf. (119) and (120)).2. the progressive is not a paraphrase saliently available from the imperfect form.3. a habitual interpretation emerges.5. which. 29 .184 Individuals in Time It is worth noticing that.1. rather than the habitual (cf. in Spanish. However. 30 For the reasons why the progressive is excluded with Eskimo and blond. see section 5. when a relational complement is added (123). as noted above. the progressive paraphrase of the imperfect is more natural in these contexts. Likewise.30 Finally. estúpido/inteligente (121) Juan era Juan ser-IMPF-PRET-3SG stupid/intelligent “Juan was stupid/intelligent” (122) ??Juan used to be stupid/intelligent muy estúpido con su hermano (123) Juan era Juan ser-IMPF-PRET-3SG very stupid to his brother “Juan was stupid to his brother” (124) “Juan used to be stupid to his brother” The examples are based on temporal subordinate clauses because. accordingly. the preferred interpretation of adjectives denoting intellectual aptitudes (121) is the continuous reading. make the habitual reading available. when a complement that can act as a “distributee” is added (125).

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

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

it is also possible to have the stative version of these adjectives together with the perfective. whereas processes take time and progress in time (although this progression in time does not mean they advance toward a culminating point). As can be seen. SL (141) and IL (142) statives are excluded with the progressive. and therefore reject the expression of progression in time. consistent with the description of the perfective above. cruel/inteligente toda su vida (140) Juan fue Juan ser-PRET-PERF-3SG cruel/ intelligent all his life “Juan was cruel his entire life” 5.Outer Aspect and Individual-Level Predicates 187 muy cruel/inteligente (137) Juan fue Juan ser-PRET-PERF-3SG very cruel/intelligent “Juan was very cruel/intelligent” (138) ‘Juan was very cruel to Peter’ (139) ‘Juan was very intelligent in that business’ However. according to which it is compatible with basically any type of predicate. .3 The Progressive and Adjectival IL Predicates I now turn my attention to the progressive viewpoint in Spanish. the progressive is possible with other class of homogeneous predicates. 31 As mentioned in chapter 3. states hold in time but do not take time.31 The progressive viewpoint does not seem to work like the imperfect and the perfective. 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). (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. in the sense that the progressive is not compatible with every kind of predicate.5. the progressive in English works differently in some contexts.

followed by Bertinetto (2000). can be argued to lack any input of energy.1) for the introduction of this concept. since estar enfermo ‘be sick’ and estar preocupado ‘be worried’ are states and. that is. In a similar vein. Aside from the intuitive understanding of the correlation between dynamism and progression in time. unexplained under this view. The input of energy correlates with the onset of the predicate. a third point can be established. Kearns (1991) explains the incompatibility between the progressive and states by alluding to their property of essentially lacking onsets and culminations.188 Individuals in Time The property usually invoked to explain the differences regarding progression (advancement) in time is “dynamism. 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. Some authors. nevertheless. . However. which. and. However. nevertheless. nondynamic eventualities do not. as the ungrammaticality of cases like *Juan estaba estando enfermo ‘Juan was being sick’ shows. does not allow us to use the progressive. progression in time is not possible for those eventualities that are “dense”—that is. it is difficult to find an explanation in more technical terms. See also chapter 6. eventualities with such a temporal structure that between every two temporal points. examples such as the following. accordingly. 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. we can distinguish different stages in. 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.”34 alludes to “temporal stages” to describe the distribution of the progressive. for example. According to these authors. an event is a stage of another event if the second can be regarded as a more developed version of the first. An eventuality can appear in the progressive if it is divisible in stages. According to Landman. based on Carlson’s (1977) notion of “stage. where the verb empezar ‘start’ refers to the onset of the predicate remain. In fact. a sickness. as a result. footnote 4.” related to the concept of “movement.” Dynamic eventualities progress in time.32.33 Landman (1992). they are excluded in the progressive form. argue that progression in time is not possible for those eventualities lacking internal development. strictly speaking. In some sense. if we can point at it and say ‘it’s the same event in a further stage of development. such as Landman (1991).

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

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

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

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

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

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

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

In examples like (7). predicts that the temporal reading of lifetime effects always arises.4. In chapter 2. On the one hand. 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. the past tense does not seem to locate in the past either the situation of his being from California or his lifetime. That is. but in none of them is a lifetime reading available. Consider the following examples as an illustration. Consider (7) as an example. these interpretive effects can be easily neutralized. (7) That day. First. 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. However. Throughout this work I have suggested that not all IL predicates can be considered alike. Harry was from California. such as hers. 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. based on the argument structure. I advanced that equating permanency to IL-hood is not accurate.196 Individuals in Time I will point out two kinds of facts (regarding temporal interpretation) that remain unaccounted for under Kratzer’s proposal as it stands. In chapters 3 and 4 I focused on the inner aspectual differences among them. the second part—that is.3 and section 6. In the next sections I will argue that the reasons for this differ. Intuitively. so he did not have to go through the immigration process. perfective viewpoint does not trigger lifetime effects. despite being parallel to the one mentioned by Kratzer Henry was French. The reading disappears. In section 6. 1997) noticed. which is why they do not appear in (8). lifetime effects arise more easily when the predicate is just temporally limited . On the other hand. Harry and I arrived in the USA. Harry was from California— does not trigger a reading like ‘he is no longer alive’. I give an account for the way contextual factors can contribute to block the lifetime reading. The second point I want to mention is that the lifetime reading does not arise with any type of IL predicate. as Musan (1995.

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

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

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

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

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

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

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

Musan (1995. Gregory was from America and Eva-Lotta was from Switzerland. The second condition is that the tense form has to be the past. This author proposed to consider in contrast sentences like (52). Musan argues as follows. Precisely. Gregory was from America. roughly. in the next sections. as any other operator. I would like to emphasize that these are not sufficient conditions for the lifetime reading to arise. “out of the blue” cases. since it was a consequence of their argument structure. as Kratzer’s examples were.208 Individuals in Time other verbs appearing in lifetime predications. 1997) As noted earlier. The third condition is that the viewpoint form must be the imperfect. it is not in (53). Westerståhl (1984) and Lewis (1986) observed that domains of quantification are contextually restricted. In the framework she assumes. ‘the day I was introduced to Gregory and Eva-Lotta’. 6. Kratzer’s (1995) account predicts a lifetime reading to be available with IL predicates in all circumstances.2 Introducing the Determining Role of Contextual Factors. (52) Gregory was from America (53) On that day. such as have brown eyes or have long legs. Musan’s idea can be informally depicted as in (54).4. Both examples here are from Musan (1995). tense is taken as a sentence operator. despite the fact that the sentence. the temporal operator of Gregory was from America in the second case is restricted by the previous context. I have already shown that the equation this author establishes between permanency and IL-hood is not accurate and I have suggested that the lifetime reading is expected to be salient when a lifetime property is at stake. This was first noted by Musan (1995. Although I am abstracting away from the mechanics of the semantic formalization.3.8 Thus. it can be restricted by the context through a variable C. Then. The sentence Gregory was from America is given a specific temporal context in which it is interpreted temporally.1. . whereas a lifetime reading is available in (52). I will show a number of situations where such a reading does not arise. To account for this fact. the lifetime reading does not necessarily come out. I was introduced to Gregory and Eva-Lotta. 1997). and others with a previous context. like (53). In this section I show that even when the predicate denotes a lifetime property. is exactly the same. As Musan observes. 8 As mentioned in the previous chapter. I return to this point in section 6.

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

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

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

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

nevertheless. Based on Keenan-Ochs and Schieffelin 1976. Kratzer (1977. 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). 6. Consider the following sentences. I assume that sentences are uttered against this background. following von Fintel (1994). a set of propositions that form the shared background of the participants in the conversation. I offer a preliminary description of the contextual background that DP subjects themselves can bring onto the sentence and the influence it plays with respect to the TT. I define “discourse topic” as the subject of immediate concern in the conversation. Among the elements constituting the discursive common ground.Tense and Individual-Level Predicates 213 interpretation of these two types (with and without context) varies (i.e. the content of their TTs is different) because their TTs get their values from distinct “discourse topics”. lifetime effects do not arise.” Following widespread proposals in semantics and pragmatics.1 When the Subject Is a QDP I am going to examine a set of examples without any previous explicit context. and von Fintel (1994). they are anaphoric to distinct “discourse topics. I assume.. where. The common ground determines the options that are live at any point in the conversation. As defined by Stalnaker (1972. In the next two sections. 1979). I assume that every natural language discourse takes place in a context.3 I will show how my proposal (to account for all the cases uniformly) works in a systematic way. among many others. Grice (1975). Regarding the specific relationship between the discourse topic and the elements influenced by it. 1981). (62) Todos los chicos eran esquimales Every guy was Eskimo (63) La mayoría de los chicos eran esquimales Most of the guys were Eskimo (64) Varios chicos eran esquimales Several guys were Eskimo (65) Muy pocos chicos eran esquimales Very few guys were Eskimo (66) Muchos chicos eran esquimales Many guys were Eskimo (67) Algunos chicos eran esquimales Some guys were Eskimo .4. that it can be described as an anaphoric relationship.4. In section 6. I will show that the TT is influenced by the discourse topic. a “context” contains a “common ground”—that is.

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

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

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

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

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

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

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

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

) I will deal with complement clauses and relatives in turns. to understand the temporal interpretation of a compound sentence is to understand the relationship holding between the two (or more) TTs. the subordinate clause will be temporally dependent on the main one. namely. TT and ET can be modified by time adverbs. Peter’s leaving occurs at 3 P. the prototypical case of this is the modification of the past perfect. the content of the RT is.5 The Lifetime Reading in Compound Sentences For the sake of a more complete survey of cases. Thus. I argued that what Tense locates is not the ET as such. whose nature I argue to be also a ZP. subordinate clauses (complement and relative). as in (i). the ordering between the TTs should depend on the specification of the embedded RT. by default. so that it can be controlled by the ET of the main clause.1. My purpose in doing this is to explore the temporal interpretation of these predicates when dependence with respect to another tense complex is syntactically established.1) I described the way temporal interpretation is accounted for following Stowell’s (1993. However. the embedded TT would become ordered with respect to the main ET. have noted.M. (Thus far. starting by the former ones. and its eventuality will be located with respect to the eventuality of the main clause. giving rise to different temporal readings. Peter’s leaving occurs prior to 3 P. if the RT of the subordinate clause is in the c-command domain of the ET. this could lead us to a situation where the TTs are not ordered between each other strictly speaking. yielding the reading in (ii) or the ET. yielding the interpretation in (iii). In this work. As Hornstein (1990) and Demirdache and Uribe-Etxebarría (2004). as the subindexes of the ZPs in (i) suggest.M. I have just dealt with IL sentences which establish a link with a previous one paratactically. the UT.15 However. According to him. This way. (i) (ii) (iii) 15 14 Peter had left at 3 P.1 Complement Clauses In chapter 5 (section 5. in compound sentences. . this does not mean that ETs always remain in the dark for interpretation. As mentioned in chapter 5. temporal interpretation is derived from the ordering relationship Tense establishes between the RT and the ET. yielding different interpretations. The time adverb at 3 can modify the TT. Following the logic of Stowell’s proposal discussed above. 6.5. since the RT would be controlled by the main ET rather than by the main TT. among others. I focus now on the interpretation of lifetime IL predicates in another type of syntactic environment.222 Individuals in Time 6. given that the ET is the closest ZP. In simple sentences.14 but rather the TT.M. 1996) work. instead of with respect to the TT. Specifically. the subordinate RT can be affected by the ET of the main clause.

only marking the event by an <e>. First. that Peter had washed the car at 3 P. where the TT and the ET can be proved to be disjoint in reference.M. it could also be the case that the embedded RT could be controlled by either of those yielding different readings. independently from the syntactic position it occupies with respect to other TTs. and at 3 P.--------say at 5 P. it seems that. for the time being I will just ignore the ET in the syntactic representation of the data. in effect. temporal ordering is not affected by the choice of TT or ET as RT controllers. Actually. unfortunately. as I will show through the study of IL predicates in relative clauses. 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. Just as time adverbs can modify the ET or the TT (see footnote 14).M.Tense and Individual-Level Predicates 223 (i) TP 2 RTi T′ 2 T0 AspP 2 TT Asp′ 2 Asp0 VP 2 ETk VP 2 e CP 2 TP 2 RTk T′ 2 T0 AspP 2 TTm However. Since. Consider (ii).M. (iii) ------wash at 3 P. Even in cases such as (i). This would yield the following ordering in (iii). Second.M. Nevertheless. we get the following: . in what follows I assume that the time-denoting argument counting for temporal interpretation is the TT for two reasons. (i) (ii) Peter had left at 3 P.M.M. there do not seem to be differences in the interpretation. modify the ETs. 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. John told me that Mary had said at 5 P.-------tell------------UTT Assuming now that the time adverbs at 5 P. it is the TT the ZP counting for interpretation.M. this fact raises the question whether an analysis in terms of closeness and control is adequate to fully capture the interpretation of the data.

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

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

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

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

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

(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 . (96) María dijo que Juan estuvo enfermo Maria said that Juan estar-PERF-PRET-3SG sick The unique temporal interpretation is a past shifted reading. What precludes. the possibility that an interval becomes ordered with respect to itself does not actually arise. but it has no content. and (b) that the subordinate TT is not an interval different from the main clause (TTj). by virtue of the content of the aspectual head. This analysis makes. their temporal values coincide. it is null. in contrast. sick sick ↓ ↓ (95) ----------------------//////X/////////-----------------UT---------↑ saying However. That is. Is this a technical inconvenience? I would like to argue that it need not. then. the interval the subordinate clause is referring to is the interval at which Maria said something. the same sentence in perfective. since the content of T is null. there is no ordering predicate. The second possibility represented in (94b) is to consider (a) that Tense does not have the meaning of ‘present’. the RT binds the TT and. If we follow interpreting the tree.Tense and Individual-Level Predicates 229 as usual. which seems desirable since simultaneity is only possible with such an aspectual form. but the same one of the main clause (TTi). where the main clause interval gets located “after” the subordinate interval. which is not ‘after’. therefore. but ‘within’. the interpretation can be worded as follows: the TTi corresponding to the interval of “saying” is included in the interval of “being sick”. crucial use of the content of the aspectual head. In other words. then. which seems to accurately capture the intuition about the interpretation of (88). Consider. then. 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”). with no further independent evidence. This way. Since there is no content in T. a past shifted reading? The content of T. as a consequence. what we get is that the interval (TTi) is included (‘within’) the eventuality of being sick.

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

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

Abusch demonstrated that the different temporal readings of RCs correlate with the interpretation of the DP the relative depends on. I will divide the task in two steps.232 Individuals in Time the TT is located in the past. when the DP is indefinite and acts as a complement of an intensional verb. Firstly. Donnellan 1966. among others). and the [+specific] interpretation to a wide scope position. as I have been doing thus far. look for > a girl b. I will summarize general points concerning the temporal interpretation of RCs. I will make some considerations about the interpretation of lifetime predicates in another syntactic environment: Relative Clauses (RCs). the DP can be considered as unambiguously [+specific]. Adriana. Summarizing very much.5. a lifetime interpretation is not expected from a syntactic situation such as (101). Secondly. I will pose what this analysis predicts with respect to the lifetime predicates and show whether it is borne out or not. The reading roughly described as [–specific] would correspond to a narrow scope position with respect to the intensional verb. 6. or Juan was looking for a particular girl. which may be determined by that of the DP . (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. whoever she might be. a girl > look for When the main verb is extensional. can be paraphrased in two different ways: either Juan was looking for any girl.2 Relative Clauses In this section. (103) a. If we take an example like (102). 1996) work. 1996) builds on Ladusaw (1977) and Abusch (1988) to elaborate his proposal about the temporal interpretation of RCs. it can have two interpretations. which can be described as a [–specific] and a [+specific] (Russell 1905. Rivero 1975. Stowell (1993. (104) Juan besó a una niña Juan kissed a girl Stowell (1993) argues that the tense interpretation of a relative clause can be predicted from its scopal position. following Stowell’s (1993. namely. we see that the interpretation of the DP a girl.

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

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

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

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

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

.

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

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

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

outer aspect.). Thus. which lack all such characteristics). once “acquired. I have argued that to attribute permanency to a property amounts to assigning it a determined temporal interval that the property extends over. Temporal unit 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). 7. if we say that SL and IL predicates differ in certain temporal terms. and gave concrete definitions to each. nonstable predicates. One of the difficulties in these descriptions is the vagueness as to what exactly the authors mean by such terms.D. I have considered that each of them is syntactically represented (quantity.” it lasts for the remainder of the individual’s lifetime (Ph. Differentiation between homogeneous and heterogeneous predicates Outer aspect Inner aspect Table 7. if we . Table 7. SL predicates are conceived as episodic. To understand a property as permanent means either that the property holds for the entire lifetime span of the individual (intelligent) or that. Chierchia 1995.242 Individuals in Time of them describe SL predicates as “involving aspectual properties. tense). 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. 1995. I distinguished three temporal levels (inner aspect. Different number of occasions eventualities occur and different orderings between the Eventuality Time (ET) and the Topic Time (TT) intervals.3 The IL/SL Dichotomy Is Not a Permanent/Episodic Distinction IL predicates have been defined in the literature as permanent predicates (Kratzer 1988. However. 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.2 summarizes the temporal units. among others).” or “temporally anchored” (in contrast to IL predicates.2.” “temporally bound. In this work. aspect. and the distinctions that can be made according to each one. In contrast. their definitions. I understand that the difference consists of something describable along the lines of the rightmost column.

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

temporal interpretation depends on the content of the TT and its relation with respect to the reference time. before. does not affect the choice of copula. 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. which is supposed to keep stable. depending on the quantifier (|1|. as I have argued. and orders it with respect to another interval taken as a reference. This has the limitation of predicting Lifetime Effects for all appearances of IL predicates. Temporal interpretation does not hinge on argument structure but depends on the content of the TT. thus. as shown in chapter 6. |>1|. the (spatio)temporal variable is bound by the quantifier over occasions (Q<occ>) discussed in chapter 5. The TT is the interval the speaker refers to and for which a particular predicate is asserted to hold. Crucially. 1996) in arguing for the presence of this variable in all types of predicates. In Kratzer’s account. One advantage of acknowledging a similar argument structure for both IL and SL predicates is that Tense is argued to work uniformly.244 Individuals in Time As cited in chapters 2 and 6. which is not accurate. the TT in the copular clause is exactly the same (8). this view has enabled me to capture the temporal interpretations potentially available with IL predicates without yielding any overgeneralization. Harry was sick and asked for help to get the suitcase” . is ordered by the Aspect predicate (after. which is different in (6) and (7). whereby it takes an interval. Harry estaba Harry and Maria arrived in the USA. Instead. whereas the external argument of an SL predicate is a spatiotemporal variable. Harry estar-PAST-IMPF-3SG enfermo y pidió ayuda para recoger la maleta sick and asked for help to get the suitcase “Harry and Maria arrived in the USA. the differentiation between IL/SL predicates cannot be set on the basis of the presence or absence of the spatiotemporal variable. In contrast. which is discourse sensitive. and their external argument is the DP subject itself. in contrast. however. The TT can vary without triggering any variation on the copula choice. (6) Harry y María llegaron a los EEUU. Tense takes and locates (with respect to a reference interval) the spatiotemporal variable in the case of SL predicates. I followed Stowell (1993. if an appropriate context is built up). This. within) with respect to the TT interval. 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. In these examples. namely the TT. Since. IL predicates lack a spatiotemporal variable. temporal interpretation was derived from argument structure. As suggested there. In my account. In Kratzer’s proposal. I have argued for a uniform treatment of Tense. ∃). The number of occasions obtained.

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

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

you get tanned in two weeks” Finally. the ones of our minimal pairs in (1) and (2)). There are other adjectives (enfermo ‘sick’) related to a verb (enfermar. only if the adverbial in +x time is present. with predicates other than participles and cut-short adjectives (i. therefore.3 2 For a detailed discussion about participles. but rather adjectives yielding verbs. from which participles derive. due to their atelic nature. estar-clauses admit in-adverbials with certain predicates (participles and perfective adjectives that have been described as “cut-short”2).e. Regarding cut-short adjectives. Bosque proposes that they are not adjectives derived from participles. I would like to suggest the idea that in the cases of participles (14) and cut-short adjectives (15). participial adjectives. and cut-short adjectives.Conclusions and Final Remarks 247 guapo en una hora (13) *Juan es Juan ser-PRES-3SG handsome in an hour “Juan is handsome in an hour” Second. The projection of AspQMAX gets justified by the properties of the participles themselves (or the adjectives derived or related to them). Following the theoretical framework I have worked with here. ‘sick-INF’. which come from heterogeneous verbs. behaving. the meaning of the adjectives (of the minimal pairs in (1) and (2)) with the in +x time adverbial is different from the meaning that ensues from the contrast in such minimal pairs. the verb they are related to has to be of heterogeneous nature too. rather than llenar > llenado > lleno (‘fill > filled > full’). see Bosque 1990. .. AspQMAX is projected. ‘get sick’) which. 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. estás with this cream estar-PRES-3SG dark-skinned in two weeks “With this cream. 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.

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

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

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

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

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

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

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

15 There are other types of adjectives derived from participles that. in cases such as (33)–(36).Conclusions and Final Remarks 255 accounted for. have an active or stative reading. leaves unexplained other copular combinations. However. either.15 The idea suggested previously that any predicate is. a metaphorical reading. As many authors have pointed out. the interpretation of the adjective with ser gets.14 muy seco (33) Pedro es Pedro ser-PRES-3SG very dry muy madura (34) María es María ser-PRES-3SG very mature muy abierta (35) María es María ser-PRES-3SG very open muy limpio (36) Su periodo presidencial fue his presidential period ser-PST-PERF-3SG very clean muy frío (37) El océano Atlántico es the Atlantic Ocean ser-PRES-3SG very cold muy caliente (38) El mar Muerto es the Dead Sea ser-PRES-3SG very hot Although. as described in chapter 2. the necessary combination of (perfect) participles with estar looks consistent. respectively.) (i) a. Also. then. this does not affect the discussion about the combination with the copulas. (Note that they correspond to different participial forms in languages such as English. Rather. cansado Juan está/*es Juan estar/ser. the idea that estar contributes the meaning of linking to a circumstance with any type of predicate is not accurate. 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 . the impossibility of participial adjectives to combine with ser is complex and seems subject to exceptions. as Fernández Leborans (1995) notes. combinable with ser. depending on their combination with ser or estar. the metaphorical reading proves that the hearer tries to give an interpretation in consonance with the structure.PRES-3SG tired cansado b. Likewise. Nominal predicates do not combine with estar (40). I will not investigate this issue here. Sometimes this alternation appears in correlation with the (in-)animate properties of the subject (i). in principle. such as copula + locative PP. note that there are adjectives of the type of (31) that can appear with ser. which only combine with estar (39).

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

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

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

3. hinging upon the type of the predicate following the copula and the telic inducing adverbial in + x time. Differences between ser and estar . Can appear in any outer aspect form The property is referred to the interval specified in the Topic Time Table 7. 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.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.

.

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

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

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

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

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

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

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

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

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

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

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

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

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

.

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

E. 143 Russell. 55 Rosen. 207. C. 115. T. G. N. 213. 5. 250 Rapoport. P. 71. 114 Schieffelin. L. 6. 138. 190. 55. 143. 43. 96. 59. 94. 71 Ramchand. G. 220. 254 Lewis. 14. 50 Landman. 147. 232. 222. 245 Martin. 191 Picallo. 116. 56 Sanz. 213 Kennedy. 113. 189. 226 Pagliuca. 202. 251. 237. 117. 165. 218 Phillips. R. 47 Keyser. 163. 102 Matthewson. 124. 18 Schmitt. A. 109 McNally. 193. 49. 242–244. L. E. 208. B. 106. 140. 240. C. 196. 79. 119. 148. L. 191 Ogihara. V. 215 Stalnaker. 151. 70. M. 85. 164. Y. 120 Pesetsky. 8. 106. 225. 37. 25 Quine. 8. 216 Reichenbach. 135 Kiparsky. M. 143. 15. P. 94 Morera. 86. D. 192. 214 Luján. 205. 96. 40. 205. 131. C. 41. 70. B. S. 11. 114 Kuroda. 233 Rohrer. 3. 150 Reinhart. 55. M. 130 Mourelatos. 3. 251 Krifka. 232 Lakoff. 74. S. 109. 3. 107. 37. 193–196. R. S. 33. M. R. 28. de 48 Milsark. 172. 136. 235. 14. 145. 218. 61. 194 Kratzer. de 30. 127 Swart. 244 Stump. 155. 12–14. 71. 61. 38 Mithun. 16.276 Individuals in Time Kearns. 37 Pustejovsky. 18 Piñón. 127. B. 43. T. 81. 16 Rosen. 154. J. 213 Stowell. T. 34. 116. W. 25–32. 43. 40. 15.-Y. W. 178. 60. 93. C. A. 73. 127. 118 Matushansky. 12. 11. 154. T. D. 61. 178.-Y. 188. 207. S. 176. 43. 3. 94 Musan. 112 Tungseth. J. 148. 78. 252. 120– 123. 80. E. 214 Longobardi. 12–14. 138. 3. 112. E. A. 26. 41. C. R. 240 Rivero. 119. 84. A. 3. 148–150. 40. J. 240. 240 . 233 Querido. 84. J. H. G. 145. 32. 144. 214 Perkins. H. 191 Partee. 126 Morimoto. P. 150. R. 29. 112. 38. P. 102. 15 McConnell-Ginet. 191 Larson. 70. 153. 191 Pollock. 127 Uriagereka. 189. 251 Quer. 113. 122. 122. 22 Kenny. 152. 116. 10. 250 Ladusaw. 40–44. R. 110– 112 Torii. 188 Keenan-Ochs. F. C. 152. S. H. 156. M. 33. 192. 56. 240 Rothstein. 1. 97. 113 Kitagawa. 251 Sportiche. E. 245. 243 Travis. 8. 75. 97. 152. 58. 57 Masullo. ter 148. 216 Uribe-Etxebarría. 36–38. C. 77 Roldán. J. M. 88 Lemmon. G. P. M. 8–12. M. 120 Parsons. 162. O. 38. 144. 86–90. D. 31. 194 Klein. 240 Kondrashova. 176 Munro. 208–212. M. 145. W. 32. L. J. 38. 162. 43. 236 Naumann. 113. 5. 144. 226. 126. B. 215. M. 167. 147. 122. 222. 136. A. 232 Ryle. 195 Leonetti. G. 144. C. S. 176 Szabolcsi. 214. 55. 109. 117. 211. T. 225. J. 208.-L. 89. 119. 80 Ritter. 15 Koopman. 213 Schmidt. 211. G. 112 Schein. Y. 12–14. 77. 46. 50. T. 84. 232–235. 29. 15 Raposo. 29 Talmy. M. J. 44. S. 194. 22 Meulen. 25. 156 Tenny. 40–43. 217–220. 225. 98. 38. 164 Miguel. 214 Svenonius. 61. A. W. 45.

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

.

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

189. 230. see Temporal interpretation Past-shifted reading 76–78. 240. 148. 224. 233–235. 135. 175. 246–248. 258. 240. 173. 176. 134 Instigator 55. 165. 159 Quantification restriction 11 Quantity 113–116. 148. 150. 79. 211. 230 of Narration 77–80 of Relative clauses 223. 227–229. 249 Existential coda 12 Existential quantification 8. 35. 148–152. 58. 136. 233 Perfect entailment 47. 142. 252 Realization function 6. 189. 100. 14. 183. 196. 226. 246. 148 Sequence of Tense 226. 129. 227. 181. 221–223. 187–192. 186. 162. 144. 244. 246. 130. 222. 159. 191 Event(uality) Time (ET) 3. 152. 147. 171– 175. 116. 189. 131. 122. 194. 186. 165. 245. 160. 192 Prospective 120–122. 112. 184. 241. 123. 134 There-sentences 5. 153–160. 136. 172. 187. 123. 242. 115. 106. 117. 257. 175–177. 186. 208. 212 Mereological properties 3. 248 Swahili 120 Temporal interpretation of Complement clauses 76. 162. 137 Subinterval property 70. 148. 180. see also Utterance Time Stativizer 115. 242. 12. 128–131. 187. 229. 156 Focus 13. 244. 192. 147. 248. 194 Reference Time (RT) 3. 168 French 121 Generic operator 11. 158. 148–152. 174–178. 234 Situation aspect 39 Small v 55 Spatiotemporal argument/variable 8–10. 186. 148. 113. 83. 258 Hindi 120 Homogeneous (predicates) 70. 123– 126. 176–178. 189– 192. 232. 132. 88. 11. 178. 229. 177–180. 249 Relative clauses temporal interpretation of. 167. 189. 236 Theme 105. 141. 125. 14. 215 Goal(Path) 96–98. 158–160. see also Subinterval property and Additivity property Narration temporal interpretation of. 158. 248. initiator 61. 116. 226. 183. 131. see Temporal interpretation Russian 15 Salish 118. 246. 178. 111. 131. 105. 192. 226. 11. see also Event(ive) argument Speech Time 122. 192 (Im)perfective contrast 25. 144. 76. 26. 86. 57. 172. 12. 233. 147. 181 Existential reading 6. 112 Juba Arabic 120 Kinesis 40 Krio 120 Lakhota 55. 134–136. 35. 209. 230. 141. 156 Hawaii-Creole 120 Hebrew 15 Heterogeneous (predicates) 72. 224. 257 Perform(er) 53. 173–180. 129. 250 . 94 Margi 121 Maxim of Quantity 85. 72. 180 Perfective (preterit) 31. 72. 184. 231. 179. 158. 212. 227. 147. 155. 201. 134–136. 118. 129 Progressive 153–163. 249. 186. 94. 202 Ground 119. 212. 258 Progressive paradox 191. 122. 222–225. 191. 14 Generic reading 6. 242. 72. 211.280 Individuals in Time Event-object 110. 178. 227 Simultaneous reading 76–79. 31 Imperfective paradox see Progressive paradox Initiation. 35 Figure 119. 207. 80. 55. 161. 64. 258 Imperfect Continuous 76. 144. 154. 37 Thetic (judgments) 12–14. 132. 158–160. 132. 79. 207. 6. 157. 158. 155–157. 184. 176. 172–174. 195. 154–157. 80. 221 Habitual 77. 67. 189. 94 Pomo 94 Prepositions Directional prepositions 127. 128. 148. 136 Place/situational prepositions 120.

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

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

Sign up to vote on this title
UsefulNot useful