Individuals in Time

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

Series Editors Werner Abraham
University of Vienna

Elly van Gelderen
Arizona State University

Advisory Editorial Board Cedric Boeckx
Harvard University

Ian Roberts
Cambridge University

Guglielmo Cinque
University of Venice

Ken Safir
Rutgers University, New Brunswick NJ

Günther Grewendorf
J.W. Goethe-University, Frankfurt

Lisa deMena Travis
McGill University

Liliane Haegeman
University of Lille, France

Sten Vikner
University of Aarhus

Hubert Haider
University of Salzburg

C. Jan-Wouter Zwart
University of Groningen

Christer Platzack
University of Lund

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

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

María J. Arche

John Benjamins Publishing Company
Amsterdam / Philadelphia



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

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


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

..........2.... A Brief Stop at “Agentivity”.......................2 Ser and Estar as an Aspectual Differentiation ........................ by Tim Stowell ............. Introduction to the Tests Discerning the IL/SL Status...............................................5 Summary of Section 2......................................4 The IL/SL Contrast as a Categorical/Thetic Distinction..................1 Inner Aspect................................................................61 3...................................................................1 The IL/SL Distinction as a Semantic Distinction.............. 1995) .............................................................53 3..........................39 3................................................ix Foreword....14 2...........................................2 .....1......3 IL Predicates as Inherently Generic Predicates................ When There Is More Than One Copular Verb: Spanish Ser and Estar.......................................1 Chapter 2 Individual-Level Predicates ..........2 ..........1......59 3................................1......16 2....................1 A Cluster of Notions ..........................1 Proposals about the Individual-Level/Stage-Level Dichotomy ..........................................................39 3.......2..... Inner Aspect and Event Types .......... Chierchia (1995) ....................................53 3...5 2...........3 Summary of Section 2.....................1 ...................................2..................12 2.............2..................................................38 Chapter 3 Event Classes and Individual-Level Predicates ........1.....39 3.............................................3 Aspectual Differences among Individual-Level Predicates ...............2.....51 3............5 2.........................................11 2...........61 ...2 Event Types and Event Structure..................32 2......Table of Contents Acknowledgments....................2......................................4 Summary of the Chapter .......... Raposo and Uriagereka (1995)........1...........8 2............................1.............. Kratzer (1988.............5 2...............1 ......1 Distribution of Copular Verbs in Spanish ..............................3............xi Chapter 1 Presentation of the Study ...40 3...3 Summary of Section 3...........3 Summary of Section 3.1.................2 Agents in Event Structure .......................................................................25 2.....2...1....15 2........ Higginbotham and Ramchand (1996) ..............2 The IL/SL Dichotomy as a Syntactic Distinction................. Milsark (1974) and Carlson (1977).......................1....................................2................33 2................. The Structure of Copular Constructions ....

vi Individuals in Time 3.........4 Summary of Section 3.....................................................................2 Be as a Copula Shifting a State into an Activity ........143 4.............3........ The Relational PP Complement.........137 4.......96 4...1 Lexical Semantic Classes of Adjectives ...............117 4.......................................98 4....................... 69 3.......................................81 Chapter 4 Aspectual Alternations in Individual-Level Predicates ...........1..............................1 Similarities between Activities and States .105 4.3...90 4.................2 Differences between Activities and States ........3..3 ........................143 ....3 Some Confusing Arguments in Defense of a State/Activity Distinction.............4 Summary of Section 4.................1 The Two-Copulas Hypothesis .........................62 3.....2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor123 4..................108 4....................6 An Account Based on the Relational PP Complement .......................2 Summary of Section 4.................................................2 Syntactic Approaches ....................1 Aspectual Tests on IL Predicate ......................6.............85 4.................................3 The Relational PP with Other APs .............116 4.3.............68 3...............................2......................2 ......4............5 Justifying the Approach ...........................................2 Summary of Section 3......................4..................... Summary of the Chapter .135 4................................1 Cruel-type Small Clauses Taken by Verbs Other Than the Copula ..............7........106 4..108 4......1 Previous Explanations of Nonstative Copular Clauses ......................3 Relational Mental Properties..................... Summary of the Chapter .........................3.............................3 The DP Subject of Relational MPs ..................................84 4..................... States and Activities: A Grammatically Relevant Difference?......................7..............5 .86 4...................2 Summary of Section 4.....................1...4 .......8.......2...........7...................................................133 4.....6.....4 The Issue: Aspectual Alternation in IL Copular Clauses ........................................7 .....4.......2 On the Optionality of the (Affected) Goal PP ...............................5.83 4...........................84 4.................76 3....5..........................5.............................................3 Summary of Section 4...............................................................4 Summary of Section 4........96 4........4....................................5..80 3................................110 4...........117 4.91 4........................1....... The Cruel-type as a Small Clause Integrant....................6..................................6........................................69 3..........................................3.....95 4.......138 4..........................................3 Adjectives Predicated of an Implicit Event Argument .......1 Prepositions as Aspect Encoders .........6 ..................1 Lexicalist and Logico-Semantic Approaches ....2 Adjectival Predicates Showing an Activity-like Behavior ..................................3 ..4..................1 On the Interpretation of the Relational PP ......................................72 3.107 4......

..193 6.................................. 179 5.......2 Differentiating Temporal Extensions of IL Predicates ..............199 6..2...3 On the Relation between the TT and the Habitual Q<occ> .....3.......................173 5...............171 5..........3............220 6............................5 The Lifetime Reading in Compound Sentences .....................................3...........197 6..........1 Tense and Aspect as Ordering Predicates .................... 186 5...3 Brief Notes about the Complement of Mental Properties APs ........................................................157 5................................................4....1 Necessary Conditions for Lifetime Effects ...................................................................... Summary of the Chapter ...............................................................2 Nonpermanent IL Predicates................................ 177 5...205 6.......212 6..........5.. A Brief Summary of Aspect Notions ............ 1996) ........... 187 5......................................1 Tense.......................1 The Imperfect and Adjectival IL Predicates ....................................................194 6...............148 5..152 5..................2.......... Kratzer (1988..... Musan (1995..2.. 1997) ............. 174 5.................................. 1995) ...........................1 Permanent IL Predicates ... 180 5...................205 6.218 6.........2...205 6............................................4...............................................................................................................................................................3........222 ......... Stowell (1993....3 The Progressive and Adjectival IL Predicates ...........................Table of Contents vii Chapter 5 Outer Aspect and Individual-Level Predicates .....5 Adjectival Individual-Level Predicates and Viewpoint Aspect .......................................2 Perfective Homogeneous Predications .....................................................2...4 Summary of Section 6...1................2 Aspect as a Quantifier over Occasions ..148 5....................................................................147 5.4 Summary of Section 6...3..................................3 The Arising of Lifetime Effects .............2 The Internal Argument of Tense: The External Argument of Aspect............. 192 Chapter 6 Tense and Individual-Level Predicates .....4.....202 6.........1 Temporal Interpretation as a Consequence of Argument Structure...........6...................................5......................................................1...................2 The Perfective and Adjectival IL Predicates ..........2 .........4...........215 6......2 The Habitual Interpretation: Iteration........ 176 5.....................................2......1 When the Subject Is a QDP ...3 Inner and Outer Aspect .................2 Introducing the Determining Role of Contextual Factors.... Proportion........................4 .1 Habitual Heterogeneous Predications............158 5...........................5.............2...........163 5....3 Articulating the Account...................1 Quantifying over Occasions .....4 The Determination of the TT Content and Lifetime Effects ........ and Systematicity .................................4...208 6........................................................... Aspect as an Ordering Predicate .........213 6........................3 The Content of the TT in the Arising of Lifetime Readings .....................................2 Context Associated to Individuals ..................................................210 6......199 6....................

...........3..................6....5.275 Subject Index..............241 7.......................................................................... Recasting the IL/SL Distinction .........................257 References .............................................249 7...................................232 6.245 7......................279 ...................................... Outer Aspect Does Not Affect the IL/SL Distinction .......................................................2..235 Chapter 7 Conclusions and Final Remarks ................... Some Remaining Questions .........................................242 7....................249 7.................................................2 Relative Clauses ..........1..............................................................................................8.......... The IL/SL Distinction Is Not a Matter of Inner Aspect…Completely ...................................................6.. Summary of the Chapter ................................. The IL/SL Dichotomy and the Ser/Estar Contrast .....................................viii Individuals in Time 6..254 7...................1 Complement Clauses .....5....................................................................................................5.........222 6....................4.....7.....................................261 Name Index ........239 7...........................................................................239 7........... The IL/SL Dichotomy Is Not a Permanent/Episodic Distinction ........................... Summary of the Conclusions ........................... Summary .........

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

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

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

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

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


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

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

Outer aspect and tense are conceived as dyadic ordering predicates that take two time-denoting arguments (Stowell 1993. kind. Specifically. 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 will propose a formal constitution of Aspect as a complex functional projection including an ordering predicate (which locates the so-called Topic Time [Klein 1994. I then apply such tests to different kinds of IL copular predicates and show that they can be divided in two groups: states and activities. 2004) and a quantifier over occasions. along the lines of Demirdache and Uribe-Etxebarría 2000. dynamicity is proposed to correlate with the presence of the relational complement and argued to be rooted in the prepositional phrase. I start the discussion of the temporal domain of inner aspect by first presenting the different event types acknowledged in the literature and the tests to diagnose them.Presentation of the Study 3 tense. Thus. I will then discuss the theoretical and empirical consequences for the IL/ SL dichotomy. In chapter 3. the dynamic properties of clauses where adjectives referring to mental properties appear (such as cruel. Chapter 2 introduces the notion of IL predicate and presents the distribution of the two copular verbs in Spanish. I will show that prepositions can contribute an inner aspect shift from states to activities. based on hypotheses that attribute aspectual content to prepositions (Hale 1984). the ideas defended in this monograph are sympathetic to the conception of syntactic structure as event structure. mean) are proposed to correlate with the presence of the relational complement of these adjectives (cruel to Mary). 1995] with respect to the totality of the Eventuality Time. this being understood in inner aspect terms (Borer 2005. Ramchand 2003. I present some reflections about habituality. In this respect. my proposal will concern the role of prepositions in event structure. I study the set of IL predicates patterning with activities and argue that this behavior is rooted in the nature of the syntactic frame intro- . In chapter 4. Ritter & Rosen 2000). 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). I argue that the IL/SL contrast should be rooted in the association to a particular circumstance (SL) or its lack thereof (IL). The study of tense focuses on the interpretive phenomenon known as ‘Lifetime Effects’ (whereby the individual referred to by an IL predicate in past can be understood as ‘no longer alive’). This book is organized as follows. which is described as a viewpoint referring to a plural number of occasions. In this vein. Focusing on copular clauses. In particular. 1996). The discussion of inner aspect is framed in the theoretical debate regarding the effects of syntactic structure on the semantics of predicates. giving the number of times a certain eventuality holds (along the lines of Verkuyl 1999).

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

and introduce some of the concrete aspects for which this work makes alternative proposals. lexically distinguish the two types of predication.Chapter 2 Individual-Level Predicates In this chapter I introduce the concept of the individual-level predicate. arguing for an aspectual distinction. who discuss the constituency of the SCs that copular verbs take (in particular. I also introduce proposals by Heycock (1994) and Becker (2000). (1) (2) a.1. This chapter is organized as follows. Section 2.1 The IL/SL Distinction as a Semantic Distinction. There are several policemen in the corner a. according to which the copular verb selects for a small clause (SC) containing the nominal argument and the predicate (AP. Section 2. There are several policemen available b. IL and SL. Likewise. ser and estar. For example. I concentrate on the most influential proposals about the contrast between individual-level (IL) and stagelevel (SL) predicates. NP). PP. 1981). I will first present the main descriptions of this notion found in the literature. according to most authors. I will introduce the syntactic analysis that I will assume for copular sentences. I will critically revise those analyses. I present the syntactic analysis I assume for copular cases—namely.3. specifically discussing arguments for semantic. I present other two aspects basic for my purposes in this work. but those in (2) are excluded. In section 2. the predicates of (1) are allowed. which. a part of this chapter presents the differences between the two Spanish copulas (ser and estar).2 is devoted to the differences between the two Spanish copulas. and pragmatic distinctions.1 Proposals about the Individual-Level/Stage-Level Dichotomy 2. Since the data in the work are from Spanish. the one offered by Stowell (1978.4 summarizes the chapter. 2. syntactic. the presence of functional projections) and its relationship with the IL/SL distinction.1. indicating some of the points on which my proposal differs. 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. *There are several policemen insane . In section 2. After that. Milsark (1974) and Carlson (1977). since I am chiefly concerned with copular cases. *There are several policemen intelligent b.

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

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

e) & (in the bathroom.1. such as Jäger (1999) and Becker (2000). 3 4 I will add a few more words about the eventive variable in chapter 4. As will be shown in section 2. 1995) claims that IL and SL predicates differ with respect to what they are predicated of. the event variable is existentially quantified). 1995) and Diesing (1992) but disputed by others. The trees in (13) just aim to capture Kratzer’s idea on the location of the external argument in the VP predicate. denoting nonactions but performing actions and proposed that events could be treated as individuals.3. When it is not projected. IL predicates do not.3 In Kratzer’s proposal. it is the DP referring to the individual itself which functions as external argument. whereas SL predicates project this extra argument (which can be understood as a spatiotemporal argument). the toast.2 The eventive variable is considered a primitive element in the logical semantics of events. . This gives rise to a difference in their argument structure which is the stem of the difference between the two types of predicates. as Diesing 1992 proposes. e) & (at midnight. 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. can be modified and quantified over by adjuncts. Kratzer (1988. (∃e) (buttered (Jones. SL predicates are predicated of something else: the eventive variable proposed by Davidson (1967). NP). such as Kratzer (1988. Whereas IL predicates are predicated of the individual directly. 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. with a knife.8 Individuals in Time This view has been adopted by some authors. He proposed that action sentences include an event variable as a primitive element in their logical semantics. e) & (with a knife. at midnight. and nouns. The Davidsonian argument is the external argument of the predicate. (i) Jones buttered the toast slowly. as in (i). (In the default case.4 2 Davidson (1967) took the classical differentiation from Plato between verbs. to mention just a few. Schematically. in the bathroom. from Davidson 1967. Kratzer’s proposal looks like (13). like nominal variables. PP. which. based on Williams’s (1981) Argument Linking description. denoting actions. e)). 1995) In the spirit of Carlson (1977). 2. Kratzer (1988.2 The IL/SL Dichotomy as a Syntactic Distinction. e) & (slowly.

Other interesting contrasts are accounted for by this spatiotemporal variable: (18) When(ever) Mary is in Paris. 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. potential variables to be bound are those that indefinites introduce (following Kamp 1981. rooted in the argument structure. . Stage-level predicates IP 1 I′ 1 I VP 1 <e> VP b. If the IL and SL distinction is. it is predicted to be a stable distinction. when(ever)-clauses of the sort of (18)–(20) involve an implicit quantificational adverb. When there is no variable to be bound. I discuss the semantics of when(ever)-clauses and. This is precisely the case of the IL predicate in (19). However. (14) Mary is at your disposal in the office (15) *Mary is a mammal in the office (16) Mary will be at your disposal next week (17) *Mary will be a mammal next week Whereas IL predicates (a mammal) cannot be temporally or spatially modified. she looks happy (19) *When(ever) Mary is a mammal. where the absence of indefinites and the absence of a spatiotemporal argument preclude the availability of a variable susceptible to binding. it feeds its son According to Kratzer. the sentence results in ungrammaticality. Kratzer is forced to readjust such a 5 In section 2. This is explained by the possession or lack of the spatiotemporal argument.3. always. In (18)–(20). Such an adverb unselectively binds all free variables it has in its scope. and the spatiotemporal variable (18). as the ungrammaticality of (15) and (17) shows. Heim 1982) (an animal in (20)). SL predicates (at your disposal) can. I argue that they are just an apparent test for SL-hood. as presented. differing from Kratzer’s suggestion.5 Let me now introduce another important aspect of Kratzer’s proposal. she feeds her son (20) When(ever) an animal is a mammal. due to the prohibition against vacuous quantification (Chomsky 1982).Individual-Level Predicates 9 (13) a.

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

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

in some pragmatic sense. (30) *There is [strong Q every computer] (31) There are [weak Q few computers] (32) *There are [students [strong Gen Q contestants] Chierchia assumes that the permanency or stability of the property is what makes IL predicates a natural class. it involves the same limitations affecting Kratzer’s proposal that the IL/SL nature of a predicate could be decided on the basis of its interpretation. called “categorical judgments. Specifically. which I have already argued against in the two previous subsections. Regarding Chierchia’s central proposal that IL predicates are inherent generics. if we are referring to a chronic illness or an occasional one). In turn. Juan is ignorant” (34) *Típicamente/*característicamente/*generalmente María es María ser-PRES-3SG typically/ characteristically/ generally muy culta very cultivated “Usually. SL predicates are.12 Individuals in Time 1974). they define IL predicates as those that. it can be classified as belonging to both classes of predicates. 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. those referring to the event they introduce.” Clauses involving SL predicates . Higginbotham and Ramchand (1996) Raposo and Uriagereka (1995) argue that the IL/SL distinction is not rooted in the lexicon but is a matter of differences in information structure. Raposo and Uriagereka (1995). are about the individual designated by the subject. Chierchia argues that the Gen operator is a strong quantifier.1. IL clauses can be. that is not the case—another reason why I will not follow Chierchia’s approach on IL predicates. can be understood as stable or transient (for example. As the following sentences show. therefore. (33) *Típicamente/*característicamente/*generalmente Juan es ignorante typically/ characteristically/generally Juan ser-PRES-3SG ignorant “Usually. if a generic operator were actually in the constitution of IL predicates. Clauses involving an IL predicate are about a prominent argument—a “category” in Kuroda’s (1972) terms. I consider that this conclusion makes the classification IL/SL lose its appeal since it does not have any predictive force.4 The IL/SL Contrast as a Categorical/Thetic Distinction. This strict correlation leads this author to propose that when a predicate (like sick). simply. Mary is cultivated” 2. Besides. we would expect that an overt generic operator could appear.

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

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

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. In turn. Spanish and Portuguese have two lexically different copular verbs. When the predicate is an NP → copula “don” c. hinging on whether the predicate expresses an inherent or definitional property of the subject (38) or not (39) (Greenberg 1994). such as Bambara (Koopman 1992). 2. and as secondary language it is employed to communicate nationwide. the copula is null in the present tense but overt in the past tense (Kondrashova 1995. (37) a. 9 For further discussion on the overtness of Hebrew copula. the postcopular NP (the predicate) alternates in case (nominative or instrumental) depending on whether the property is permanent (40) or nonpermanent or noninherent (41). I have introduced the idea that equating permanency of a property and ILhood is inaccurate. 1996. In the next section. I expand the discussion regarding what seems to be an adequate description for the IL/SL dichotomy. A large part of the population uses Bambara as its mother tongue.9 In Russian. I take up this issue also in chapter 7. . found across most of western Africa. Bambara is one of the officially declared languages of Mali. Matushansky 2000).8 distinguish three types of copular verbs depending on the category the predicate belongs to (37). When the predicate is an AP → copula “ka” b. Other languages. ser and estar. see also Rapoport 1987 and Rothstein 1995.2 When There Is More than One Copular Verb: Spanish Ser and Estar A good number of languages show copular alternations. those clauses where the topic is an individual are instances of IL predications. In the past-tense cases. where the semantics of the two Spanish copulas are introduced.Individual-Level Predicates 15 are instances of SL predications. The Hebrew copula alternates between overt or null.

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

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

6). one of the properties that Schmidt (1972) and Bache (1978) provide to distinguish classifying adjectives from qualifying ones is. Are origin adjectives qualifying or classifying? I will treat them as classifying because they behave as canonical classifying adjectives in the other three tests that Schmidt and Bache motivate. (b) impossible modification by a degree adverb (48). that their acceptability as predicates in copular constructions seems to be confined to those cases where the subject is a physical entity. precisely. see Bosque 1993. For more details about classifying adjectives. When the adjective is referred to the place where one is born. it is not gradable. These tests are (a) impossible participation in comparatives (47). If the subject is a resultative nominal.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. and Demonte 1999. that classifying adjectives are excluded as copulative predicates (The presidential trip vs. However. 14 Note.13. and those in (51) combine with both copular verbs. Further discussion is presented in chapter 7 (see section 7.14 (47) *Este chico es más africano que el otro15 “This guy is more African than the other one” (48) *Este chico es muy africano “This guy is very African” (49) *Africano/inafricano African/un-African The adjectives in (50) combine just with estar. and (c) impossible participation in binary correlations (49). 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. also. *The trip was presidential). 15 13 . Bosque and Picallo 1996.

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

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

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

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

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

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

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

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

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

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.2. However. (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. (See section 2. 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.28 Individuals in Time aspectual head. Schmitt also notes that estar. In other words. like all verbs in the perfect. 1995). as I already suggested. Chapter 4 develops a proposal about the strict correlation between the type of APs and this syntactic behavior.) (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). they may just show that estar predicates behave as states. 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. the reason why estar and the perfect are excluded as complements of perception verbs may be because both behave as states. (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 fact (also noted by Schmitt) that not all adjectival predicates combine with the progressive (cf. and (infinitive) states are not accepted as perception verb complements. is not allowed as a perception verb complement. such as when(ever)-clauses: .1. (81)) remains unaccounted for.

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

For this reason. kind. according to Schmitt (1992). I argue that if an appropriate context is built up (think. 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. Durative aspect would be opposed to “result state” in her proposal. which would lead to the conclusion that estar is excluded in such contexts.” Instead. mean. note that. This is what licenses. se arrepiente después whenever Juan ser-PRES-3SG cruel. for example. 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). I have formulated the distribution of whenever (which I consider a temporal distributor quantifier) as a subcase of distributivity in general. Chapter 5.30 Individuals in Time (87) Siempre que Juan es cruel. this may seem to be the case. whenever Juan is Eskimo. a permanent predicate all through someone’s life. he gets angry” In sum. under whenever. they select for some sort of durative aspect. he regrets right-after “Whenever Juan is cruel. which is dedicated to Aspect. If it finds a variable susceptible of referential variability. ser can be ruled out neither in when(ever)-scenarios nor in progressive ones. the presence of Eskimo. nice) are possible in these contexts. Schmitt argues that ser is allowed in when(ever)-clauses by virtue of its underspecification with respect to aspect. even adjectival predicates of the type of Eskimo fit. (89) (En todas sus reencarnaciones).26 (For the moment this is sufficient to introduce my point. it is licensed. of a tale). in this case. he bears a life full of difficulties and shortages” Here is a context where the existences of Juan can be iterated. expands on this account about iteration. se enfada whenever Juan estar-PRES-3SG dissatisfied with his work gets-angry “Whenever Juan is dissatisfied with his work. . and Jäger (1999) argue that whenever is correct with predicates that are “naturally iterable. which Schmitt treats in 26 De Hoop and de Swart (1989). Consider (89). Chierchia (1992). Although. at first sight. contrary to fact: (90) Siempre que Juan está descontento con su trabajo. siempre que Juan es esquimal.) As a last remark regarding when(ever)-clauses.

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

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

How this is brought about is a complex issue that I discuss only partially in chapter 7. the NP acting as their argument is generated in the specifier of such APs. which be takes. from where it later raises until reaching the specifier of Inflection Phrase (Tense). I will introduce here the syntactic analysis I assume for copular sentences. In this way. Stowell proposed to analyze the subject and the predicate of copular sentences as integrants of an SC. 1981) proposal that the copula takes (as its complement) a small clause (SC) containing the NP subject and the predicate. I will just debate whether SL-hood is a matter reducible to properties of (outer or inner) aspect or tense. the SCs are APs. leaving a trace. 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. In the aforementioned examples. I adopt Stowell’s (1978.3 The Structure of Copular Constructions Because this work concerns IL predicates in copular constructions. 2. (95) IP 2 Johni I′ 2 is AP 2 ti A′ g blond . Stowell proposed that the SCs were in fact projections of the predicate.Individual-Level Predicates 33 Ramchand 1996). In essence.

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

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

The evidence adduced to by Felser for the presence of an eventive argument is that perception verbs complements have to be “eventive. in the presence of a functional projection (Aspect) and an Event Phrase (in the spirit of Kratzer 1988.36 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. can host an argument (a subject) in its specifier. .28 (107) a. besides hosting the event argument. 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. Felser argues for the existence of an eventive argument and for the existence of a functional projection that. Felser presents evidence of a filled specifier position. I saw John draw a circle b. *I saw John know the answer As for the existence of an extra projection. 1995) or its lack thereof.1 for a discussion in which opposition (eventive/stative or IL/SL) is at stake. as in (107).” rather than stative. 28 See section 2.

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

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

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

I leave the arguments of this last issue for later. and I reserve the term “event” to allude to nonstate types. 1993) in future chapters. I will focus on those that offered insights that are generally adopted in some way or other. over. although in and of itself does not explain event structure. He distinguished between “states. 2003) have shown that adverbial modification.” The term “inner aspect” contrasts with what I refer to as “outer aspect” (also following Verkuyl 1989. Egg (1995). semantics.1. (To refer to all event types including states. 1 Other authors. culmination. Taking into account the presence or absence of the cited features (duration. As will become clear.” where it is. whose main insights’ spirit has been shared until today. those that involve culmination. or before its beginning. As Rosen (1999) points out. and “events. Grimshaw (1990). .” following Bach [1986].” where dynamicity is not involved at all (be red). making no further distinction. Investigations by Tenny (1987. named “kinesis-verbs” (e. whether it does not. 2005). This will be an important point when discussing the event types of the predicates I am interested in here. Pustejovsky (1988. 1989. In his work Metaphysics. 1998.40 Individuals in Time toward an endpoint. reserve the term “event” to telic predicates. as well as thematic relations and the syntactic projection of arguments. many researchers have attempted to identify a number of types into which events can be classified. 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.2 Event Types and Event Structure The first known study of event classification goes back to Aristotle. Dowty (1991). or delimitation) and their combination in a predicate. or De Swart (1998). can be derived from inner aspectual properties. build the house). This taxonomic line of study does not address questions as to how and where events are represented in the grammar (within the lexicon. 1991).g. such as Herweg (1991). 2000. or whether it involves no duration at all. event classification proves helpful in describing the basic characteristics of events that need to be explained.1) Among events. Aristotle proposed a classification based on the notions of dynamicity and terminus. I will use the term “eventuality. and Ritter and Rosen (1996.. van Voorst (1988). Borer (1994. 3. 1994). The events that lack an inherent endpoint are named “atelic”. My intention in doing so is to present the basic features describing the nature of events and the properties that more than one type share. Aristotle distinguished between those with an inherent terminal or culminating point. A large number of event classifications are offered in the literature. at this point I will briefly summarize the classifying work. “telic. or syntax) or which grammatical phenomena can be directly derived from it.

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

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

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

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

Juan walks d. *Normalmente. . Normalmente. 4 I thank Tim Stowell for interesting discussion on this topic. and achievements and accomplishments. whereas (SL) statives can be said to be ambiguous. according to the temporal adverbials that refer to the time an event lasts they admit. However. Juan pasea Usually. it might seem that this test is a diagnostic for IL-hood rather than for stativity. whereas Juan walks does. Normalmente. I start out with the tests in (13) and (14).2). on the other. (11) Interpretation as habitual in present tense a. In this subsection I go through a number of tests that separate activities and states from achievements and accomplishments.3. Juan realizes that his mother is right c. Juan draws a circle Nevertheless. Juan sabe matemáticas Usually. Such compatibility cuts the pie of eventualities in two: those that have a habitual interpretation.4 Whereas (11a). Juan knows mathematics b. Cinque 1999). are completely grammatical. I argue that the crucial condition for a predicate to appear in habitual form is its possibility of restarting. involving an IL stative predicate. sentences like Juan is usually sick or Juan is usually at his best friend’s place. 3. and those that do not.2 Activities and States versus Achievements and Accomplishments.Event Classes and Individual-Level Predicates 45 habitual (Comrie 1976. Bertinetto 1986. is excluded in combination with a habitual adverb (which proves it cannot be interpreted as habitual). on the one hand. nonstatives are interpreted as habitual. States are argued to be the only type that cannot be interpreted as “habitual. such as “for + x time” goes well with those predicates that refer to eventualities that extend over time with no involvement of an 3 I make a few more remarks on this in chapter 5 (section 5. which divide activities and states.2. The eventualities whose interpretation in the present is not habitual are states. A durative adverbial. I show that this is a possibility for every type of predicate (including IL predicates) if a suitable context is created. Juan traza un círculo Usually. It seems clear that Juan is sick or Juan is at his best friend’s place does not have the habitual reading as the only (and preferred) reading. Juan se da cuenta de que su madre tiene razón Usually.1.” as their reaction against combination with habitual adverbs proves. it is worthwhile to notice that when no (habitual) adverb is present. with stative though SL predicates. Normalmente.

see Piñón 1999. *Pablo viajó en tres semanas Pablo traveled in three weeks c. They express for how long someone has been engaged in a particular activity or for how long a particular state has held of an individual. Pablo viajó durante tres semanas Pablo traveled for three weeks c. *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. 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. (13) In + x time a. which is the reason why they are compatible just with eventualities entailing a delimit point. Pablo se dio cuenta de que su madre tenía razón en tres semanas Pablo realized that his mother was right in three weeks The following contexts also group achievements and accomplishments together. Pablo construyó una casa en tres semanas Pablo built a house in three weeks d. . as (13) shows.5 (12) For + x time a. Pablo estuvo enfermo durante tres semanas Pablo was ill for three weeks b. *Pablo se dio cuenta de que su madre tenía razón durante tres semanas Pablo realized that his mother was right for three weeks In turn. *Pablo construyó una casa durante tres semanas 6 Pablo built a house for three weeks d. As the ungrammaticality of (14) shows. For a recent account about the different readings available in such sentences.46 Individuals in Time ending point. the entailment of a delimit 5 Note that this test does not distinguish between IL and SL predicates.

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

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

what gets interpreted as interrupted is the habit of walking. *Pablo ha terminado de estar enfermo Pablo has finished being sick b. Pablo paró de pasear por unos minutos/*por una temporada Pablo stopped walking for a few minutes/for a period of time b. In a similar vein. it is grammatical if we think of a habitual interpretation like ‘(Usually) Juan used to realize about when his mother was right. in principle expected to not combine with dejar de. 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. rather than a concrete instance of it. (20) As a complement of finish a. *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. However. a durative signaling an extended span of time does not. which makes them close to accomplishments (‘John finished his [daily] walk for that evening’). among others. 3. the verb finish does not accept anything but accomplishments as complements. consider the judgment of (18d). There is a contrast with (17c) above.4 Accomplishments versus All the Rest. (19) a. This difference can be observed in the pair of (19). but he does not anymore’. for a period of time. The fact that both states and habits can appear after dejar de suggests that states and habits have some properties in common.Event Classes and Individual-Level Predicates 49 event of going walking as a habit. where an adverbial signaling an extended period of time fits perfectly. Pablo ha terminado de construir la casa Pablo has finished building the house c. . As Pustejovsky (1988). In this case what is interpreted that stops happening are different occurrences of ‘realizing’. suggests. ‘Juan finished walking for that evening’). an achievement. the habit of undertaking such an activity. The contrary is observed with dejar de. 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.2.1. ??Pablo terminó de pasear9 Pablo finished walking d. That is. which is understood as the interruption of a particular instance of the event.

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

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

the results of the tests (9)–(11). 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. States – – – + + / – + – / Activities + + + + + + + % + – Accomplishments + + + – + – – + – + + Achievements % + + – + – / – % – – (9) Progressive form (10) Complement of ‘happen’ (11)Interpretation as habitual in present tense (12) For + x time (13) In x time (14) and (15) Assertion of noncompletion (16) Perfect entailment from the progressive (17) Complement of parar de (18) Complement of dejar de (20) Complement of ‘finish’ (21) Ambiguous interpretation with the adverb ‘almost’ (22) Command imperative form (23) Cleft with ‘do’ (24) Combination with ‘deliberately’-like adverbs (25) Complement of ‘force’ – – – + + + + + + – – – – + + – Table 3.2. and a “%” if the event type shows a positive behavior in the test but just under a certain interpretation. as pointed out above in the description of each test.52 Individuals in Time indicates when the event type can have the interpretation described or can appear in the context expressed on the left of the table. (17). a “–” when it cannot. Tests for event types Among other things. The bracketed numbers in the leftmost column refers to the number of the test as they appear in the text above. (20).

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

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

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

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

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

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

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

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

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

1. The predicates I will test are a classificatory AP (Eskimo) and two qualifying APs (blond and cruel). as has been widely agreed on in the literature. *Juan estaba siendo esquimal Juan was being an Eskimo b.1 Aspectual Tests on IL Predicates The aim of this subsection is.1 Events versus States. Juan es cruel con Pedro Usually Juan is cruel with Pedro Despite all the observations introduced in section 3.” do not “take place. to identify the aspectual properties of IL predicates. taken to belong to the group of states. (43) Occurrence in the progressive form a. *Lo que sucedió fue que Juan fue esquimal What happened was that Juan was an Eskimo b.1 regarding the present-tense interpretation as habitual (see also chapter 5 for more important . Juan es esquimal Usually Juan is Eskimo b.2. Juan estaba siendo muy cruel con Pedro Juan was being very cruel with Pedro (44) Pseudocleft with happen (‘take place’) a. IL predicates are. *Normalmente.” They are considered to be eventualities that lack any bounding and are inherently durative. states are eventualities that do not “happen. The eventualities yielding ungrammatical results are taken as stative. To do that.” but just “hold. Juan es rubio Usually Juan is blond c. Normalmente. *Lo que sucedió fue que Juan era rubio What happened was that Juan was blond c. 3. in general.1. *Normalmente. I will concentrate on copular clauses and test out whether all IL predicates are appropriately grouped in the aspectual set of states.3.62 Individuals in Time 3. In (43)–(45) I reproduce the tests marking for stativity versus eventivity. simply. I will make use of the tests introduced above in the same order. As described above. *Juan estaba siendo rubio Juan was being blond c.3. ?Lo que sucedió fue que Juan fue muy cruel con su adversario What happened was that Juan was cruel with Pablo (45) Interpretation as habitual in present tense a.

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

and not accomplishments.64 Individuals in Time (46) For + x time a. the combination of ser and an AP such as cruel ‘cruel’ looks good. tests like (49) divide activities and accomplishments from states and achievements. *Juan fue cruel con su adversario en dos horas Juan was cruel with his opponent in two hours The test in (48) corroborates the results of (46) and (47). Juan ha sido cruel con Pedro Juan has been cruel to Pedro 3.1. since this is only possible with eventive predicates. (48) Perfect entailment from the progressive form a. yield a perfect entailment from the progressive form. In principle. where there is no endpoint. a predicate like ser + AP would be unexpected as a complement of parar de ‘stop’. *Juan fue esquimal en dos horas Juan was an Eskimo in two hours b.2.3. As shown in section 3. *Juan fue esquimal durante tres semanas Juan was an Eskimo for three weeks b. As mentioned in section 3. Once again. Juan fue rubio durante tres años Juan was blond for three years c.2. (48b) is a licit entailment from (48a).1.3 Activities and Accomplishments versus States and Achievements. which argues in favor of considering the predicate (ser cruel) as an activity. such a predicate shows a nonstate patterning.2.3. 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. *Juan fue rubio en dos horas Juan was blond in two hours c.1. only activities. Juan está siendo cruel con Pedro Juan is being cruel to Pedro b.

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

*Juan terminó de ser rubio Juan finished being blond c. The next set of contrasts aim at testing whether the process that be cruel can be considered is agentive or not. it contrasts with other APs with ser. Juan finished being an Eskimo b. confirm be cruel as an activity.1.3.5 Agentivity Tests. the latter is with activities. All tests give the same results. whose combination with almost is impossible.2. which also test accomplishments versus all the other eventualities. *Juan casi fue esquimal Juan almost was an Eskimo 3. (52) As a complement of finish a. since the unique licit entailment from (53a) is (53b).1.20 Although (53a) does not sound extremely natural.4 Accomplishments versus All the Rest. which suggests that dynamic properties are somehow involved in these predicates. The former reading is available with accomplishments. 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. None of the AP predicates with ser. and (b) that the subject did not even start out the process. ?Juan terminó de ser cruel con su adversario (en el minuto 10 del debate) Juan finished being cruel to his opponent (in the 10th minute of the debate) The following sentences with almost. Juan no empezó a ser cruel con su adversario Juan did not start being cruel to his opponent c. *Juan terminó de ser esquimal. Tests (52) and (53) show some more contrasts consistent with the results from (49).3. Although be cruel predicates do not give perfect results after finish. ?Juan casi fue cruel con su adversario Juan almost was cruel to his opponent b. (53) Interpretative entailments from the adverb almost a. . 20 Recall that. as mentioned in section 3.1. except for cruel. like (53c). fit in canonical agentive contexts. they clearly contrast with the other two. Only ser cruel is possible under a command imperative form (54) or in combination with adverbs like deliberately (55).4.66 Individuals in Time 3.

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

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

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

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

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

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

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

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

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

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

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

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

Event Classes and Individual-Level Predicates


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

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


Individuals in Time

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

Event Classes and Individual-Level Predicates


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

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

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

Rothstein (1999) argues (quite unspecifically) that “be + AP” constructions move quite freely among event types. the subject of sleep. Rothstein argues that. it is left unexplained why the stative reading is also available in such cases. if the small clause (SC) of (7) be polite is eventive by virtue of the presence of be. if it is the copula that shifts a state into an activity. As shown in chapter 3 (section 3. a temporary reading. since subjects of activities and accomplishments are typically understood as agents. associated to a particular eventuality.2 Be as a Copula Shifting a State into an Activity Rothstein (1999) proposes an alternative account to that based on two copular entries. leaving the eventuality in (7). The proposal by Rothstein leaves the following points unanswered. given the role of be to shift an eventuality from a state into an activity. (8) Dafna is polite Also. In a nutshell. whereas when the copula is overt (7). That is. To cover this question. (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). an activity. Rothstein (1999) resorts to the conversational principles (Grice 1975) of Quantity (“do not say more than required”) and Manner (“avoid obscurity”). Rothstein claims that when the adjective comes naked (6). Rothstein argues that. 1 .2 an agentive interpretation is the expected one given the event type of the outcome (an activity). the copula shifts a state into an eventuality with activity-like properties. with be present. 4.1 Regarding the interpretation of the subject.Aspectual Alternations in IL Predicates 85 And finally. is more salient. 2 Rothstein (1999) also notes this is not the case for all activities. it is understood as a property predicated of the whole person. it is not clear how the account can be extended to simpler inflected cases like (8). She offers a semantic investigation about the behavior of the predicates showing an agentive behavior as a complement of causative verbs like make. cannot have an agentive reading. locatable in time and space. would violate these maxims of Actually. depending on the context. First. Eskimo) do not appear in activity-agentive contexts. the analysis of the two copulas does not capture the fact that such “agentive properties” are activated with a specific set of adjectives. where Rothstein explicitly acknowledges that activity and agentive properties are not obvious. as a state.3).1. for example. other adjectives (blond.

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

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

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

Following Kratzer (1988. estar. in chapter 6 (section 6. that is not case. Ser is completely grammatical. In Spanish. and between the dyadic one and stage-level (SL) predication. However. necessarily.2.3). the event argument represented by the infinitive is implicit in simpler copular cases such as the ones under investigation here. 1995) account.1. I take up the discussion related to the interpretation of the “infinitival argument” in Spanish and its status (as argument or adjunct). and the contrast between the examples with ser and those with estar (27) can be captured by paraphrases such as the one in (28). 4 For details of Kratzer’s (1988.4 he takes it that the presence of such an argument is proper of SL predicates. Specifically. I will make two observations. 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. First.2. According to Stowell. Note that the contrast between ser and estar in this case falls in line with the contrasts studied in chapter 2 (section 2. 1995). the copular verb should be. (28)). (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. to admit there is an eventive argument that is implicit amounts to admitting that (25) is a SL predication. in his account. but he is not cruel to him usually” Later. see chapter 2 (section 2. (25) John was cruel to Peter Regarding Stowell’s analysis. this proposal would entail that in all instances of APs followed by a relational PP.2). I repeat one of the examples here in (29). he argues that the action denoting argument is an overt version of the (Davidsonian) eventive argument.Aspectual Alternations in IL Predicates 89 Stowell (1991) establishes a correspondence between the monadic version of these adjectives and individual-level (IL) predication. which is the lexical marking for SL-hood. . as (26) illustrates.1) regarding the APs that can appear with either copula (cf.

pero está muy (29) Pablo no es Pablo not ser-PRES-3SG handsome but estar-PRES-3SG very guapo con ese traje handsome in that suit “Pablo is not handsome. to account for nonstative IL copular clauses. I have argued in the previous section that the activity-like behavior should not be attributed to properties of the copular verb.90 Individuals in Time guapo. (8) above). Thus. These two facts lead me to discard this hypothesis based on the existence of implicit event arguments. it cannot be said it is the copula that induces the active features. which is proved by their acceptation of during-adverbials and the rejection of in-adverbials. I will consider these sentences as instances of IL predication. My purpose in the remainder of the chapter is twofold. More specific- .2 Adjectival Predicates Showing an Activity-like Behavior In chapter 3 (section 3. since they are not the same. As shown in the previous chapter. 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. as shown above. 4. 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. we observed that there were certain copular combinations that behave like states and others that behave like activities. these nonstative copular clauses pattern with activities and not with accomplishments. Since the activity-like behavior depends on the adjectival predicate combining with the copula. since it is not the copula that varies in the minimal pairs that can be construed (be Eskimo/be cruel) and. I will first elaborate an explanation as to what specific kind of adjectives are the ones that show an activity-agentive patterning and why. I will then offer an explanation regarding the stative reading also available with such predicates (cf. the event argument Stowell proposes corresponds to an accomplishment (21) or activity (22). the properties of copular cases such as (25) (the active-agentive characteristics) cannot be attributed to or derived from those of the implicit event argument. as a consequence.3). but he looks very handsome in that suit” Given the grammaticality of ser. Second. my proposal will not be based on properties of the copular verb.

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

*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. *Juan estaba siendo joven/viejo Juan was being young/old b. *La noticia era nueva/reciente a propósito The piece of news was new/recent on purpose (43) *La mesa era preciosa/horrible a propósito The table was beautiful/horrible on purpose (44) Juan era rápido/lento a propósito Juan was quick/slow on purpose (45) Juanera*apto/*capaz/*inteligente/*ingenioso/cruel/amable a propósito Juan was apt/capable/intelligent/cunning/cruel/kind on purpose . *La 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.92 Individuals in Time (35) a. *Juan era alto/bajo a propósito Juan was tall/short on purpose b.

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

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

Since their analysis is complex. Juan es inteligente en los negocios/ingenioso en sus bromas Usually. Thus. Consider the following contrasts: (56) a. 4.2 Summary of Section 4. Although slow or fast also refer to the way an individual does something. I will be chiefly devoted to analyze the properties of the agentive MPs (i..e. namely. As we can see. In the remainder of the work I will leave aside these adjectives and I will concentrate on the properties of MPs. I do not have a specific proposal to offer at this moment. They are predicates referring to ways of accomplishing a trajectory (physically or figuratively). 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. 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. relational ones). I will propose that such a property is in strict . Juan is intelligent at his business/cunning at his jokes When these adjectives appear in the progressive form. as is the case of MPs.Aspectual Alternations in IL Predicates 95 (54) Juan es inteligente/ingenioso Juan is intelligent/cunning (55) Normalmente. *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. we cannot say (57b) from (57a). Regarding the agentive-activity patterning of APs referring to velocity. the fact that they are predicates that can be interpreted in relation to another individual. although I will make some considerations about all the sets behaving as activities. since velocity APs work differently in many other respects. a complement of the type of those in (55) is understood.2 In this section I have described the semantic class that adjectives showing an activity patterning belong to: namely. I will deal with different aspects in turns. those referring to velocity and a subset of the APs referring to MPs. which is an (agentive) activity. giving pieces of the picture bit by bit. whereas out of (56a) we can assert (56b). they do not qualify the individual herself. I will start in the next section by treating the property that makes this group of adjectives a natural class.2.

can be considered as an “affected goal. and discuss all the consequences derivable from this. 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”). in italics in (58). whether or not the PP expresses an “affected argument” depends on the concrete type of action undertaken by the subject. One can. even if we finally concluded that the PP is an “affected argument”. set on fire and bother. for example. focusing on two aspects: its interpretation and its optionality or obligatoriness. (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. 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. although it seems intuitively clear that the PP specifies the goal addressed by the subject’s action. 1988.3. 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). In this section.” This captures his intuition that such a DP refers to the individual that gets affected by the (underlying) action undertaken by the subject DP. Thus. it is understood that Juan was the agent of an action that had Pedro as its affected argument. I investigate the nature of the PP complement.96 Individuals in Time correlation with their aspectual characteristics. Incidentally. harass. 1994] and .1 On the Interpretation of the Relational PP Stowell (1991) suggests that the DP inside the relational PP. offend someone else either by saying something unpleasant. or by acting in a certain way. I will show that its syntactic presence correlates with the aspectual patterning of the construction. offend or regale. Likewise. to name just a few. abuse. 4.3 Relational Mental Properties: The Relational PP Complement As just described. 4. and such an action is qualified as cruel. 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.

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

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

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

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

(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. which invites us to take the parallelism between habituals and statives with a bit of caution. states are excluded from such contexts (v) and (vi). note that. Examples (i) and (ii) illustrate this. To begin. habituals are stative. so that there is no crucial difference between (78) and (79). While we can have habitual predicates (go walking every day) combined with volition adverbials (iii) and as complements of verbs like force (iv). I will keep the idea that the habitual interpretation should not be considered stative in inner aspect terms. which suggest that habituals and statives should not be considered to be exactly the same thing.Aspectual Alternations in IL Predicates 101 (78) Juan es cruel (habitualmente) Juan is cruel (habitually) (79) Juan es sabio (*habitualmente) Juan is wise (habitually) In what follows I will show that the presence of the PP is optional and. These contrasts suggest that dynamic predicates keep as such even in the contexts they share with states. it is true that states and habituals behave the same way in certain contexts. In fact. as has been claimed in the literature. habituals of non state verbs keep their dynamic and agency properties in the habitual forms. (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. . One context is after dejar de ‘≈give up’. as already shown. as a consequence. the interpretation as habitual in present tense is not the unique reading available. whereas (74) above was considered a contradiction. the opposite assertion does not sound that strange: One could argue that.

102 Individuals in Time (80) Juan no es cruel. From these cases. 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. However. (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. in fact. without any apparent contradiction. I will concentrate on some syntactic scenarios where an overt relational PP cannot appear. Sentences (83) and (84) are based on examples from Demonte and Masullo (1999). The first scenario where cruel cannot appear followed by a complement PP is small clause (SC) complements of epistemic verbs such as consider. be claiming that the person is not cruel. (81) Juan considera a Pedro cruel Juan considers Pedro cruel (82) ??Juan considera a Pedro cruel con Maria Juan considers Pedro cruel to Maria To elucidate the kind of predicates licensed in the SCs of consider is quite delicate. I will conclude that cruel-type APs are not inherently relational. as their obligatory combination in Spanish with the copula estar proves: de vacaciones (85) Estoy/*soy estar/ser-PRES-1SG on vacation “I am on vacation” . Stowell (1991) was right in considering the simple versions of relational MPs as IL and the ones accompanied by the PP as SL. such as Demonte and Masullo (1999). This could in principle lead us to think that. Some authors have claimed that consider selects for SCs containing IL predicates. SCs complement of consider can perfectly have SL predicates. and rejects SL predicates. at the same time. To show that the PP is optional. 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. have pointed out. as other authors.

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

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

as mentioned before. also. while the subject of cruel is not an agent. Consider (101) and (102). which enables agency. it is not inferred from the context or interpreted as generic. not included in the group of relational MPs in the table above which. (104) is not the interpretation of (103). can also take a relational PP complement. its impossibility to appear with certain subjects would remain unaccounted for.3. as it is possible with the other adjectives (cruel. otherwise. conclude that the structure of cruel-type APs does not always involve a PP complement and aspectually behave as states in principle. I therefore. In the first place. I account for the relation between the two inner aspect forms. (101) ?Juan fue estúpido con el entrevistador (y no le contestó a ninguna pregunta) “Juan was very stupid to the interviewer (and did not answer any question)” (102) ?Juan fue muy inteligente con su jefe (y consiguió lo que quería) “Juan was very intelligent to his boss (and got what he wanted)” I will make two brief considerations in this respect. as Violeta Demonte makes me notice. This is because they describe a property that can be understood in relation to another animate entity.3 The Relational PP with Other APs There are other MPs. In the next sections. although these adjectives referring to human (or animate) aptitudes can appear in combination with a relational PP. since. dynamicity. they do not have the same relationship with the PP. which are not totally excluded.).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. (103) Juan es inteligente Juan is intelligent . since it needs to bear the required properties for that (and animacy is the most basic one). etc. I consider these cases with inanimate subjects as evidence that the PP complement cannot be an obligatory complement of cruel. with particular characteristics of the construction.” if we use traditional vocabulary. 4. kind. but just a “theme. If the PP complement is not overt. Summarizing. I will consider that the presence of the PP correlates with a subject involving particular properties and. namely.

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

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

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

which is located before now.Aspectual Alternations in IL Predicates 109 (109) (110) *El portazo rompió deliberadamente el cristal The bang deliberately broke the glass Juan rompió deliberadamente el cristal Juan deliberately broke the glass In essence. the theme of the event being a circle. The former corresponds to those eventualities that culminate—that is. The event holds now (at a time simultaneous with now). framed in the tradition of Davidson (1967). which has an experiencer (Martha). (112) says: there is an event. Such terms are two: Cul (for ‘culminate’) and Hold. Dowty 1979). Observe the following: (111) Martha drew a circle ((e) (drawing (e) & agent (e. 13 . distinct from the event variable itself. which applies to the event taking place at time t. The latter to those that do not culminate. argue for the explicit presence of the eventuality (events and states) as a variable in the logical representation of the sentence. mathematics) & Hold (e. The “active” meaning of be as a verb assigning an agentive role to its subject was also mentioned in chapter 3 (section 3. which is an event of drawing. and culmination. which has an agent. circle) & ((t) (t<now & Cul (e.t)) (112) Martha loves mathematics ((e) (loving (e) & exper (e. corresponding to the event type of the predicate in the logical representation. now) The logical formula in (111) reads: there is an event. The aspectual class of verbs gets registered in the logical representation. Parsons (1990) further adds extra terms.13 Logical-semantic approaches. which is an event of loving. according to which the dynamic properties found in these copular constructions were attributed to two lexical entries of be. Martha) & (theme (e. achievements and accomplishments. and a theme (mathematics). one stative and another one active. and it has a theme. Martha) & (theme (e. and there is a time (t). Predicates are aspectually different because they have different aspectual terms. to assume different entries (stative and dynamic) for each adjective would not be different from the proposals cited above (Partee 1977.2) 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). the agent of the event is Martha. but hold: activities and states.

the dynamic or stative properties. there is no obvious way to distinguish between the two constructions in point (stative and dynamic copular clauses). First. arguably. as captured in the Event Correspondence Rule (116). As van Voorst puts it. in principle.5. both would contain the term “hold. 1989.2.1 Event Roles. 1989.110 Individuals in Time This approach leaves several questions unanswered. and van Voorst (1988). Tenny (1987. 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. authors such as Verkuyl (1972). since. (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 . which opens the possibility of establishing a direct relationship between the thematic properties of the arguments present in a sentence. correlate with the aspectual outfits of the construction.1. these authors emphasized the role of the internal object as an event delimiter.” given the fact that both are atelic and do not “culminate.5.1). this approach does not establish any relationship between the aspectual term (Hold). and the presence or absence of other elements in the sentence (the PP complement in our cases) that. its syntactic position and its role in the determination of the aspectual properties of the sentence. As discussed earlier (see section 3. 1994) and van Voorst (1988) elaborate both on the observation that arguments grammaticize the points that give the temporal contour of an event.” Second. each entity corresponds to an aspectual notion. among others. Tenny (1987. showed the relevance of arguments in the inner aspect properties of a sentence. In particular.2 Syntactic Approaches 4. 4. Dowty (1979). 1994). 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.

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

. they cannot capture aspectual alternations such as the one we are interested here in a natural way. In sum. However. 2000). occupies the subject position. appears in the subject position. the event-role perspective predicts the mapping of different arguments to particular syntactic positions more accurately than the UTAH. Travis (1994. The car is the theme in both cases. correspondingly. this assumption makes them similar to lexical approaches. for them. these approaches would have to assume that each aspectual behavior corresponds to different adjectival projections. Furthermore. which I have argued are in strict correspondence. although both Tenny and van Voorst explicitly acknowledge the compositionality of events. Benua and Borer (1996). Compare (120) and (121) with (122) and (123). 2000. In other words. 2003). As Rosen (1999) points out. The old lady instigates the selling in (123). different lexical items have different ways to project. since. 2005). Incorporating the insights from the lexical argument-oriented approaches just reviewed. Note further that the NPs play different event roles in each one of them. the work of authors such as Borer (1994. Although these event-role approaches provide accurate insights in important respects (the fact that arguments can affect the event type. but not so their syntactic positions. he does nothing that can be called selling in (123). However. therefore. and Sanz (2000) elaborates on the idea that argument structure is licensed by functional syntactic structure. 4.112 Individuals in Time In these two sentences the theta-roles are the same.2.2 Syntactic Structure as Event Structure. and. Ritter and Rosen (1996. (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). regarding the stative/dynamic contrast in IL copular clauses. the systematic correlation between object and delimitation and subject and origin). 1998. a natural account of such compositionality cannot be derived from their lexicalbased approaches. In a sense. and. she does nothing that can be called buying in (122).5. the parameters controlling the event type are assumed to be encoded in the lexicon itself: the event type is determined by the way an event maps from the lexicon. but the position of event-roles seems to be so. as such. 1998. van Voorst’s and Tenny’s proposals do not provide a direct way to capture the correlation between the presence of a PP and the aspectual behavior as an activity. The position of thematic roles is not restricted. For our present concerns.

that is. see Kratzer 2004 and Kiparsky 1998. 16 The work by Borer has appeared progressively in a number of articles. When considering that functional projections are in charge of event-roles too. In other words. whereas a DP headed by a numeral. 1989. which I take as a frame of reference for the remaining discussion. Thus. Borer (2005) elaborates on the intimate relationship between quantification and telicity or lack thereof. Mostly along the lines of Verkuyl (1972. That is.4). Borer proposes that the core functional projection of the clause is Aspect. in turn. all the relevant grammatical processes get a unified account. 1999. she argues that aspectual properties of a determined event All the authors cited make proposals in a similar spirit in this regard. with telicity.14.17 Borer proposes that the core feature of aspectual distinctions is [±quantity]. events are not states or accomplishments from the lexicon. 2005) defends the idea that Case-checking processes are satisfied through aspectual projections. aspectual properties can be recast into quantity properties. that gives the type of aspectual interpretation. More accurately. 2001. These works are framed in the Minimalist Program (Chomsky 1995. a verb projects an unordered number of arguments whose interpretation depends just on the functional projection they move to. This head has the property of “quantity” (AspQMAX or Quantity).Aspectual Alternations in IL Predicates 113 and functional structure is interpreted as event structure. 14 . 17 This was introduced in chapter 3 (section 3. Compare: (124) John drank beer (125) John drank two beers A bare mass noun gives an atelic interpretation. events are not lexically specified for any type of aktionsart.16 For Borer. 1998. where functional projections are in charge of Case assignment and agreement processes. It is the syntactic structure. As mentioned in chapter 3. In turn. the quantification properties of the DP direct object make a direct difference regarding telicity. 2001a. Assuming the parallel distinctions between mass/count nouns and atelic/telic verb interpretation. I will refer to her book (Borer 2005). 1998. 2001b). I introduce some points of the work by Borer (2005). Likewise. argument structure is computed on the basis of the syntactic structure. Borer (1994. 2000). whose core is constituted by aspectual projections. an object that can be marked as [–quantity] correlates with atelicity. 2000) argue that functional projections responsible for Case processes (Agreement projections) assign the event-roles of delimitation and initiation. Event-roles are not assigned by a particular head but are simply derived as an entailment from the aktionsart of the whole event. a telic one. 15 For further discussion about the relationship between case and event structrure. A [+quantity] object. since it presents the latest versions of her ideas and proposals. 2000. Ritter and Rosen (1996.15 In what follows.

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

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

My discussion of inner aspect properties of IL copular sentences in Spanish takes Borer’s (2005) work as a frame of reference. I have suggested. 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. 4.5 In this section I introduced the approach I will take for the analysis of the aspectual properties of the IL copular constructions. I gave reasons to discard a lexical and a logical-semantic approach by arguing that they cannot make any prediction regarding the relationship among the syntactic presence of arguments. Such a perspective is in itself an account of compositionality.3. Since. in some sense. I also assume that heterogeneous and homogeneous eventualities are distinguished by a concrete syntactic projection (Quantity). and. More in the spirit of the seminal work by Hale and Keyser (1993). and the aspectual properties of the construction. I assume that it is syntax itself which. but it is the syntactic structure which gives the type of aspectual interpretation. provides the skeleton where lexical items merge and their interpretation is obtained as a by product. Regarding the main aspectual issue I am concerned with (the state/activity contrast observed in IL copular clauses). by the intervention of the PP complement that some adjectives can have: John was cruel to Pedro. I assume that the interpretation of arguments emerges as an entailment from the aspectual structure. . I will propose that activity properties and the related thematic characteristics of the subject (recall it is understood as an agent) become a part of the construction by the syntactic intervention of one particular projection. in the spirit of Borer (2005) and Ritter and Rosen (1996. accomplishments.5. their interpretation. is that the stative status is. by creating the structure based on functional nodes. I assume that eventualities are not specified from the lexicon as states. this complement is not obligatory. differing from Borer’s idea. I will discuss the claim that adjectival predicates can only be stative. Focusing on adjectival copular cases. which has been proved to be essential in explaining aspect properties. Likewise. 2000). as argued in section 4.116 Individuals in Time Regarding this specialized stative structure. that states are the type by default. in particular.” and dynamic properties come induced from separate projections. achievements or activities. The dynamic properties exhibited by the adjectival constructions studied here (those of the cruel-type) directly challenge this hypothesis. and its absence correlates with the stativity of the construction. the conclusion I draw.3 Summary of Section 4. differing from Borer. “default. Specifically.

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

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

(137) Central coincidence PP 2 figure P 2 in ground (138) Noncentral coincidence PP 2 figure P 2 to ground ◙ ○■ According to Hale. onto or into.Aspectual Alternations in IL Predicates 119 Hale argues that the oppositions underlying the spatial system are reducible to the more basic and abstract opposition of central coincidence versus noncentral coincidence of a figure with respect to a place or ground (Talmy 1978). kel7ßqsten-s-a ti-United States-a] sÚcsec [ni chief-3SG. into).POSS-DET DET-USA -DET fool absent-DET “The chief of the USA was a fool” (ii) a. in. Examples of prepositions of central coincidence are on. or through. Perkins. from. to. along. 21 . (ii)). such as the ones in the domain of time. out of. at. Bybee. Examples of noncentral coincidence are from. The drawings below illustrate these two relations: the white circle stands for the figure and the black square for the ground.21 Along the same lines. form the basis of the aspectual system of many languages. Invisibility or absence in the utterance place (i) correlates with a past interpretation (cf. in. not visible) president IS a fool kel7ßqsten-s-a ti-United States-a (iii) sÚcsec [ti PRES-DET chief-3SG. not visible) president WAS a fool b. *The (past. over. to. such as those expressing temporal relations (when and while clauses) and those expressing circumstance or condition (conditionals with if or comparisons with whereas). whereas proximity or presence in place (iii) correlates with the present tense.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. and Pagliuca (1994) argue that prepositions such as those just mentioned (on. In their invisibility. The (past. out of. at. over. along. the central coincidence versus noncentral coincidence opposition is a fundamental semantic distinction that is universally present in grammatical relations.

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

they notice that. In turn.23 Prepositions like the Italian per (‘for’. For example. (142) (143) (144) (145) Voy a hacer la cama I am going to make the bed Sto per uscire I am about to leave According to Givón (1982). . prepositions denoting ‘movement from’ (centrifugal movement). Demirdache and UribeEtxebarría (1997. the locative preposition at directly combines with the verb in infinitival form to construe the progressive. where “come from” can be used with the aspectual meaning of the perfect.Aspectual Alternations in IL Predicates 121 ‘(with)in/on/at’. ‘to’) (144) or the English about to (145) are constituent parts of prospective forms. the combination of a motion verb with an elative component and the preposition from yields a perfect interpretation: (146) Max COME FROM home ‘Max has come from home’ As Demirdache and Uribe-Etxebarría (1999) notice. in Dutch. in Spanish and English. the prepositions a (142). appear in the form used to express close future (going to). For example. similar cases are found in Romance languages. in Nigerian Margi and Palaung (from the AustroAsiatic family of languages). in contrast to stative locative verbs participating in progressive forms. to (143). Specifically. prepositions of noncentral coincidence are present in the auxiliaries expressing perfect and prospective aspect across languages. 23 Note also the presence of a verb denoting movement. (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. indicating movement toward (centripetal movement). Also. are present in perfect aspectual forms in some languages. such as from. 2004) conclude that Aspect can be analyzed as a head establishing relations between two terms. as prepositions do. 2000.

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

as seen in the examples above. conversely.1 Con: A Locative and a Directional Preposition. in these cases the preposition con is in complementary distribution with the directional preposition a ‘to’ when the DP complement is [– animate] and is interpreted as a location: .6.2. 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. it is conceivable that prepositions can also convey inner aspect properties.Aspectual Alternations in IL Predicates 123 Assuming that aspectual meaning is reducible to prepositions (and.2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor In this section I am going to describe the preposition heading the PP that refers to the (affected) goal in cruel-type construals in Spanish. denoting the goal. The preposition introducing the PP in Spanish is. 4. This preposition is associated with stative locational meanings. Preposition con establishes a central coincidence relation between the figure and the ground (in the terms introduced before). I am going to argue that the preposition introducing the (affected) goal carries inner aspectual properties affecting the aspectual nature of the whole construction. the preposition con. con ‘with’. and assuming with Demirdache and Uribe-Etxebarría that all temporal relations can be reduced to the same set of primitives. 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. that prepositions have aspectual meaning). I propose that this is the case. ¡corre con tu papa! Pablito. con-PPs can also appear with motion verbs. paying special attention to the APs I am taking care of in this chapter.6. 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. and. like in examples such as the following: (153) Pablito. 4.

with a goal. as a path with a goal (a Goal Path.25 As reported in the corpus. All this suggests that con is not (only) associated with stative locational meanings but can be analyzed as a predicate associated with directed motion— in particular. . XIII] (159) (…)& guardandosse segunt su aluedrio fue muy cruel contra los que lo non merecían & keeping himself according to his will. S. as (157) summarizes. (157) Con + [+animate] DPs A + [–animate] DPs Therefore. con (‘with’) and a (‘to’) can be said to share properties and can be both analyzed as prepositions denoting a trajectory with a destination. or hurt them… [Alfonso X : Siete partidas. Svenonius 2004). XIII] 25 I would like to thank Isabel Pérez for bringing this source into my knowledge. he was very cruel against whom did not deserve it [Alfonso X: General Estoria V. the prepositions introducing the relational complement in cruel-type cases were of same characteristics as con.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. S. and sobre ‘over’: (158) (…) Otrosi dizimos que si algund onbre fuese tan cruel a sus sieruos que los matase de fanbre: o les firiese mal (…) And we also say that if some man were so cruel to his servants that he killed them hungry. contra (‘against’). in previous periods of Spanish. S. the only difference between con ‘with’ and a ‘to’ is the nature of the DP acting as their complements. Consider first the following examples taken from the corpus of Spanish from the Illinois State University (Davies 1999). XIII] (160) Paulo orosio (…) que era muy cruel sobre los ciudadanos Paulo orosio (…) who was very cruel over the citizens [Alfonso X: General Estoria IV. It is interesting to note that. in previous centuries. the complement of adjectives of the type of cruel appears introduced by prepositions such as a (‘to’). Most were locative predicates that can also have a directional meaning.

Aspectual Alternations in IL Predicates


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

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



Individuals in Time

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

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

Aspectual Alternations in IL Predicates


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

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



Individuals in Time

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

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

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

Aspectual Alternations in IL Predicates


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

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

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

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

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

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

4. the property (cruel) and the trajectory described by the PP. the DP in its specifier is understood as the figure moving toward a goal. but it correlates with the possibility of the adjective of being able to have a relational PP complement. but emerge as an entailment from the aktionsart of the event. In particular. 194) makes the DP in the specifier of the PP be interpreted as an agent.6 In this section I analyzed the state/activity alternation in IL copular clauses.4 Summary of Section 4. the DP moves to the specifier of Tense. Thus. the idea that the DP is. I have argued that this aspectual alternation is not unpredictable at all.Aspectual Alternations in IL Predicates 135 Keyser 1993.6. because the preposition conveys dynamicity content and the PP can be considered as a predicational structure with the properties of activities. I have proposed that the source of the aspectual . simply. the interpretation that can be legitimately attributed to the DP inside the PP is. it moves to the specifier of cruel. in a sense. provided that the precise status (either affected or not) of the DP inside the relational PP is left in the dark. Borer 2005) that “thematic roles” are not “assigned”. Finally. From there. the subject of two predicates. (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. In a nutshell. (195) Juani T TP 2 T 2 ser 2 ser cruel (= SC) 2 ti cruel 2 cruel ti PP 2 P 2 P Pedro The DP subject is generated in the specifier of the PP. This plus the fact that the DPs subject of these constructions must be animate (cf. a ‘goal’. This proposal captures. in very simple terms.

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

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

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

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

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

triggering. rather than as an event. as a quantity head. or. or cannot be so.36 36 Before proceeding further. since it encodes the causative meaning. (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. According to the examples above. would make telic something which is not. for instance). whose markers (the explicit causative verbs) can be analyzed as aspect markers (in concrete. the sentences improve. sounds worse. whereas to make someone cruel. The tree wants to suggest that hacer. an “aspectual clash” with the consequent oddity. [+quantity]). therefore. the active SC seems excluded. it is expected it cannot be a part of a construction which is telic. to make someone cruel to someone else is not. when volver is a synonym of hacer ‘make’. as a property. Compare (i) and (ii) with (iii) and (iv). which seems the . I would like to remark on the nature of the DP inside the relational PP. it can be taken as the deglutinated version of a de-adjectival inchoative verb. It would be like having a causative with inchoative properties. as causatives are. I argue that is due to an aspectual mismatch too. The SC is understood. If we take it that cruel to someone involves and gives dynamic atelic properties. is grammatical (actually. In aspectual terms. at least. with a stative SC. something like encruelizar ‘cruelize’. a telic constellation against an atelic one: (209) AspQ 2 hacer cruel (–Quantity) 2 cruel con 2 Juan con con 2 Pedro We can assume that Quantity is associated to the lexical entry hacer ‘make’. which does not exist but could have existed. Causative constructions can be considered telic constructions.Aspectual Alternations in IL Predicates 141 However.

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

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

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

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


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

These predicates translate into the temporal domain as ‘before’. ter Meulen.2 One of the facts I alluded to was that. as Tense and Aspect are considered ordering predicates. Stowell (1993.1 Tense and Aspect as Ordering Predicates In this section I deal in greater detail with the conception of Tense and Aspect as ordering predicates. 5. posteriority.1). I argued that dynamicity properties (that activities involve) can be described through prepositions. based on Zagona 1990. atelic). In particular. in Spanish at least.148 Individuals in Time In section 5. and Wall 1993. that establish centripetal noncentral coincidence relations between a figure and a ground.5. 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.. Consistent with the conclusion that viewpoints do not interfere with quantity properties. I tried to show that all aspectual properties could be reduced to the same set of primitives. but in dynamicity versus lack of dynamicity. 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. this being why the eventuality is computed as “dynamic and unbounded” (i. Therefore.g. However.3 Authors such as Zagona (1990) and Stowell (1993. Lillooet Salish). 5. I show that states are compatible even with the perfective. I turn my attention to adjectival IL predicates and show the different interpretations available with each type of viewpoint mentioned. for example) and. inner aspect can be as well. specifically. by the same token. The idea is depicted in (1). therefore. which I mentioned briefly in chapter 4 (section 4.. Speech Time. inner-aspect characterization may work as an aspect indicator (an accomplishment can be taken as a perfect action. At that point.6. posteriority. in some languages (e. and the possible relationships that can be established among them (anteriority. and Eventuality Time).1 Tense. like to. For an introduction of such notions. the progressive seems to impose restrictions on the type of eventuality. Since the line of this work is not formal semantics I do not discuss them here.e.1. and inclusion are semantically defined as partial order relations fulfilling a number of properties. the interested reader is referred to Partee. Reichenbach gave a formalization of tenses in natural languages based on three intervals (Reference Time. I show that. I argued that inner aspectual properties could also be conceived in ordering terms. It turns out that this cannot be described in terms of presence or absence of quantity. aspect properties can serve as indicators of temporal properties (a perfect action is computed as finished and. as something that took place in the past). 3 Anteriority. 2 . and simultaneity).

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

In the absence of a controller (which is the case for matrix clauses). it gets the Speech Time as a default value. future means ‘before’. Past tense locates the RT. past. but the UT is one of the possible denotations the RT can have. future. tenses resemble prepositions like ‘after’ or ‘before’. In contrast to Zagona (1990). This way. in this sense. the location in time of the eventuality) is derived. the future does it before (5). (4) Mary took the book ---------------///////////-------UT (5) Mary will take the book --------------UT-------////////// (6) Mary knows mathematics -------------. Z binds the external eventive argument (e) of the VP (Kratzer 1988).5 Time-denoting phrases are viewed as referential or quantificational categories. they locate an interval with respect to another.. The tree in (7) represents (4). 1996) notices. respectively. consists of ordering the internal ZP with respect to the external ZP. The representations of (5) and (6) would be basically identical. the temporal interpretation (i.150 Individuals in Time Just as a determiner or a quantifier binds the variable acting as the external argument of the NP (Higginbotham 1985. the eventive variable is present only in SL predicates.e. the UT. Stowell (1993. Stowell 1989). What are the values of Tense? These are what we usually call “tenses”: present. and the present locates the RT coinciding with the ET (6). since they do the same job—that is. and present can be said to mean ‘in’ or ‘within’. It is worthwhile to note that in this framework the Utterance Time (UT) and the RT are not different primitive notions (as they are in Reichenbach 1947 or Hornstein 1990). The external ZP is subject to control.////////---------↑ UT As Stowell (1993. in sum. The job of Tense. Abney 1987. Past means ‘after’. depending on the value of the predicate (Tense) and the value of the RT. from the nominal realm. The external ZP is understood as an element comparable to PRO. after the ET (4). 5 Recall that. 1996) conceives the external ZP as an argument whose concrete features depend on syntactic conditions independently established. for Kratzer (1988). except that the content of T would change from ‘after’ to ‘before’ and ‘within’. which Stowell considers present in every type of predicate. . and its controller is the closest c-commanding ET ZP. who conceives the external argument as specifically denoting the Speech Time.

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

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

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

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

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

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

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

However. here in section 5. It appears . the imperfect in (24) refers to a plural number of occasions that can be described as greater than one (|>1|). Let me start by considering sentences (22)–(24). the habitual imperfect. the functional projection which makes the eventuality heterogeneous and whose absence leaves it homogeneous.2. giving rise to another interpretation. The values of such a quantifier. The continuous imperfect does not count the number of occasions a determined predicate takes place. However. I address these points in turn. inner aspect. the three forms differ in the number of occasions they allude to. (22) Pablo estaba paseando por el parque Pablo was walk-ing around the park (23) Pablo paseó por el parque Pablo walk-PRET-PERF-3SG around the park (24) Pablo paseaba por el parque Pablo walk-PRET-IMPF-3SG around the park These sentences all refer to a (homogeneous) eventuality that took place in the past. I propose that aspectual viewpoints involve. (25) ----------------------x---------------------(26) ------x------x--------x-------x---------x-In the spirit of Verkuyl (1999). what the relationship is between the number of occasions alluded to by each viewpoint and the quantity nature of predicates (that is.1. The number of occasions the progressive and the perfective form allude to is just |1|. Q<occ> can also have an existential value. and second. and the progressive. as shown in (25).1 Quantifying over Occasions As just mentioned. can be either |1| or |>1|. In this latter regard. how many occasions each viewpoint alludes to.158 Individuals in Time of Quantity. whose head is a quantifier over occasions (Q<occ>). a quantificational component. In this section I am chiefly concerned with the perfective. as in (26). besides ordering. also contain information about the number of occasions an eventuality occurs. As I will repeat later in section 5.2 and next in section 5. as the reader may have conjectured already. which I analyze as complex phrases containing a quantifier over occasions and ordering information. aspectual viewpoints. I will consider two points—first.5. 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.3. 5. besides the ordering component.

Outer Aspect and Individual-Level Predicates 159 with predicates that “hold. Perfective AspP 2 TT Asp′ 2 Asp Q<occ> (after) 2 1 walk Although I will not work on the prospective aspect. García (1999) mentions (i) as an example of continuous imperfect. therefore. Whereas for sentences like (i).” and can. 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. as in (iiia). (i) Voy a pasear I am going to go for a walk (ii) Sospecho que va a ser rubio cuando sea mayor I think he is going to be blond when he gets older (iii) a.” rather than “take place. it can be defended a cardinality of |1|. be conceived independently from a particular number of instances. Progressive AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 1 walk b. Prospective |1| AspP 2 TT Asp′ b.13 In (27) the differing ordering and quantifying components of each aspect form are represented: (27) a.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. where we can distinguish the habitual and continuous forms. the quantifier can be argued to be ∃ (iiib). I would like to point out that its articulation regarding the quantification over occasions is similar to the imperfect. it can be also present with eventive verbs.

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

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

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

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

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

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

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

Outer Aspect and Individual-Level Predicates


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

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

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

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


Individuals in Time


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

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



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

Outer Aspect and Individual-Level Predicates


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


Individuals in Time

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

Outer Aspect and Individual-Level Predicates


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

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

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

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

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

I will discuss the following points: (a) the compatibility of telic (quantity) predicates and habituality. atelic) predicates. are odd in the presence of such a modifier. and (b) the co-occurrence of perfective viewpoint and homogeneous (nonquantity. 5. To begin. either in the perfective or imperfect form.3. prepare the meal and write the report. consider now the following examples: la comida en media hora (86) Habitualmente Pablo preparaba habitually Pablo prepare-PRET-IMPF-3SG the meal in half an hour “Pablo used to prepare the meal in half an hour” . Their telic nature is tested by the appropriateness of the modifier in x time. which points to the conclusion that the viewpoint does not affect in any sense the nature of the predicate.1 Habitual Heterogeneous Predications The Spanish data show that the imperfect-habitual viewpoint is compatible with telic predicates. Bearing in mind this telicity proof.174 Individuals in Time relation between inner and outer aspect. 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. which tests the predicate swim as atelic. and offer a general overview of the relationships in the temporal system. (84) and (85). In contrast.

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

176 Individuals in Time 5. (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. if we say John had that car from 1974 to 1985. which leads to the inference of “completeness” and “boundedness. 24 23 .3. and others. In the same vein as before. 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. Contrary to Mourelatos (1978).2 Perfective Homogeneous Predications I turn my attention now to homogeneous predicates appearing in perfective form. de Swart (1998). I argue that a perfective form does not trigger telicity effects on predicates. I will argue that. draw a conclusion in the same direction. a perfective form does not make a homogeneous predicate heterogeneous. unlike states. As discussed back in chapter 3 (section 3. activities are not 100 percent homogeneous at a micro-level. Piñón (1995). In contrast. a perfective form does not interfere with the quantity nature of the predicate. If we say John walked from 2 to 3.4). such as Bertinetto (2000). at least in Spanish. it does not mean that he has been walking at every single moment of that interval. However. the TT is located ‘after’ the ET. The meaning of these sentences is that Pablo was once involved in the task of swimming (91) and walking (92). With the perfective. (95) *Pablo nadó en una hora Pablo swim-PRET-PERF-3SG in an hour “Pablo swam in an hour” Other authors. he has had that car at every single moment between 1974 and 1985. In other words.24 as subinterval and additivity proofs show. sentences such as these show that the presence of the perfective viewpoint does not contribute telicity. who conceive the perfective as a modifier that makes an eventuality telic.23 Observe these examples: (91) Pablo nadó Pablo swim-PRET-PERF-3SG “Pablo swam” (92) Pablo caminó Pablo walk-PRET-PERF-3SG “Pablo walked” Verbs such as swim and walk test out as homogeneous predicates.” as mentioned in the previous section. Bach (1986).

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

the ordering component takes the quantified eventuality and orders it with respect to the TT interval. (98) ZP 2 Zi VP 2 ei VP (99) AspP 2 TT Asp′ 2 Asp0 ZP (after/before/in) 2 VP Z0 (|1|. ∃) 2 e VP Thus. 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).26 The quantifier over occasions binds the eventive variable (e) which. I assume to be present with every type of predicate (SL and IL). in Spanish. where there are no occasions to enumerate (think of stative homogeneous predicates such as Peter is tall). which does not make the sentence habitual. |>1|. presented in (3) above (repeated here as (98)) about the internal constituency of ZPs. following Stowell (1993. if “one. This unifies the working of Aspect. When the quantifier denotes a number roughly describable as “greater than one.4. I used Spanish habitual and perfective data to show that aspectual viewpoints do not have any impact on the inner-aspect properties of predicates. and not only with SL ones. I argued that. this is borne out. 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).” habitual imperfect is at stake. 1996).1 for further discussion. In the last section. Specifically. independently from the kind of predicate. See section 5. ∃) should be possible with any type of predicate. and makes the prediction that the quantifiers (|1|.5.” perfective or progressive. |>1|. but these predicates are also compatible with the habitual quantifier (She was sick usually). as Kratzer (1988) originally proposed. 26 This aspect form is involved with stative SL predicates (She was sick). Elaborating on Stowell’s (1993) proposal.25 The existential quantifier corresponds to the continuous imperfect. As will be shown in section 5.178 Individuals in Time cation) is depicted in (99). 25 .

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

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

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

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

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

which. although the progressive form is correct with cruel.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. when a relational complement is added (123). they acquire activity-like properties.184 Individuals in Time It is worth noticing that. a habitual interpretation emerges.30 Finally. rather than the habitual (cf. 30 For the reasons why the progressive is excluded with Eskimo and blond. 29 .1. accordingly. when a complement that can act as a “distributee” is added (125). Likewise. (122)). as noted in section 5.2. as noted above. in Spanish. make the habitual reading available. see section 5. (119) and (120)). the progressive paraphrase of the imperfect is more natural in these contexts. 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. whereas activities in imperfect (117) can also have a progressive paraphrase (118). this is just marginally available with be cruel (cf.3. the progressive is not a paraphrase saliently available from the imperfect form.5. However. the preferred interpretation of adjectives denoting intellectual aptitudes (121) is the continuous reading.

the habitual viewpoint can be said to be compatible with any kind of predicate—stative IL (127).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. chino habitualmente (127) En sus reencarnaciones. the typo) are subject to referential variability for habitual interpretation to be possible. stative SL (129). 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. activities (130). nonstative IL (128). accomplishments (131). and any eventive predicate. . the objects (the house. Juan era in his reincarnations Juan ser-IMPF-PRET-3SG Chinese habitually “In his reincarnations. or achievements (132).

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

31 As mentioned in chapter 3. in the sense that the progressive is not compatible with every kind of predicate. . according to which it is compatible with basically any type of predicate. SL (141) and IL (142) statives are excluded with the progressive. it is also possible to have the stative version of these adjectives together with the perfective. the progressive in English works differently in some contexts. 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).31 The progressive viewpoint does not seem to work like the imperfect and the perfective. (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.3 The Progressive and Adjectival IL Predicates I now turn my attention to the progressive viewpoint in Spanish. As can be seen.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. the progressive is possible with other class of homogeneous predicates. states hold in time but do not take time. consistent with the description of the perfective above. whereas processes take time and progress in time (although this progression in time does not mean they advance toward a culminating point). and therefore reject the expression of progression in time. cruel/inteligente toda su vida (140) Juan fue Juan ser-PRET-PERF-3SG cruel/ intelligent all his life “Juan was cruel his entire life” 5.5.

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

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

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

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

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

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

That is. the “realization function.2). 1995) Kratzer (1988.1.”2 Specifically. Kratzer proposes that the differentiation between SL and IL predicates resides in their argument structure.” See chapter 2 (section 2.3 (1) a. temporariness and permanency is encoded in the argument structure of predicates: temporary predicates involve a spatiotemporal argument.194 Individuals in Time 6. The configurational difference between IL and SL predicates can be depicted roughly as in (1). she argues that the eventive argument is the external argument of (SL) predicates.1 Temporal Interpretation as a Consequence of Argument Structure: Kratzer (1988. 1995) is the first in establishing an explicit correlation between the temporal interpretation later dubbed “lifetime effects” and IL predicates. IL predicates IP 1 I′ 1 I VP 1 θ-subj Regarding temporal interpretation. Furthermore. First.1 I will briefly summarize her line of reasoning and show that there are cases it cannot account for. putting in direct correlation the presence of a spatiotemporal 1 2 For an earlier brief discussion on this phenomenon. This author assumes that the factor that draws the line between the two types of predicates is “being temporary” versus “being permanent. SL predicates IP 1 I′ 1 I VP 1 <e> b. a spatiotemporal or eventive (<e>) argument (Davidson 1967). I will mention two points of this proposal. In particular. As said before in chapter 2 (section 2. Kratzer argues that SL predicates involve an extra argument in their argument structure—namely. . she proposed that IL predicates applied directly to the subject. whereas SL predicates needed something “extra. 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. see Anderson 1973. in the same vein as Carlson (1977). Kratzer proposes that the spatiotemporal argument is syntactically represented.” Likewise.1. For Carlson (1977). 3 As I also said in chapter 2. as Diesing 1992 proposes.1). SL predicates involve an extra semantic operation. just aims to capture Kratzer’s idea on the location of the external argument with respect to the maximal projection of the predicate. whereas permanent ones lack such an argument. the representation in (1).

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

based on the argument structure. I give an account for the way contextual factors can contribute to block the lifetime reading. On the one hand. which is why they do not appear in (8). In chapter 2. so he did not have to go through the immigration process. In examples like (7). (7) That day. The reading disappears. Intuitively. In chapters 3 and 4 I focused on the inner aspectual differences among them. Consider (7) as an example. the second part—that is. That is. Consider the following examples as an illustration. predicts that the temporal reading of lifetime effects always arises. 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 of Harry was from California refers back to the moment at which the speaker and Harry arrived in the USA. these interpretive effects can be easily neutralized. I advanced that equating permanency to IL-hood is not accurate. In the next sections I will argue that the reasons for this differ. despite being parallel to the one mentioned by Kratzer Henry was French. Throughout this work I have suggested that not all IL predicates can be considered alike.3 and section 6. Harry was from California.4. perfective viewpoint does not trigger lifetime effects. First. lifetime effects arise more easily when the predicate is just temporally limited . The second point I want to mention is that the lifetime reading does not arise with any type of IL predicate. as Musan (1995. 1997) noticed. Harry was from California— does not trigger a reading like ‘he is no longer alive’. such as hers. a purely syntactic approach. However. but he is not anymore” rubio (de pequeño) (10) Juan era Juan ser-IMPF-PRET-3SG blond (when he was little) “Juan was blond when he was little” All of these sentences have the IL copula ser in the past tense. In section 6. Harry and I arrived in the USA. the past tense does not seem to locate in the past either the situation of his being from California or his lifetime. On the other hand. but in none of them is a lifetime reading available.196 Individuals in Time I will point out two kinds of facts (regarding temporal interpretation) that remain unaccounted for under Kratzer’s proposal as it stands.

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


Individuals in Time

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

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

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

Tense and Individual-Level Predicates


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


Individuals in Time

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

Tense and Individual-Level Predicates


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

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


Individuals in Time

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

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

the al + infinitive clauses can behave as true temporal adjuncts. It also has an interpretation very close to the one exhibited with ser. Pedro estuvo muy amable In walking-me home. 6 . 7 The meaning of the al + infinitive clause in these cases is not easy to capture. With estar. it is typical of estar. I did not want to ring you up (ii) No quise llamar porque era muy tarde I did not want to ring you up because it was very late However. As I intimated in chapter 2. Pedro estuvo muy amable al acompañarme a casa Pedro estar-PERF-PRET-3SG very kind in walking-me home b. al + infinitive clauses are interpreted as temporal as discussed above or causal clauses.7 However. According to Hernanz (1999). Al acompañarme a casa. no quise llamarte In being so late. whose paraphrase appears in (ii). which would be along the lines of a causal adjunct.204 Individuals in Time To the contrary.6 Although I cannot investigate further the status and meaning of al + infinitive clauses with ser here. this discussion at least shows the nontemporal nature of these clauses. like that in (i). 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. (i) Al ser tan tarde. it is necessary to point out that this is not the only meaning the al + infinitive clause has with estar. the linking to the particular situation is emphasized. 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. (36) Pedro estuvo muy amable al acompañarme a casa Pedro estar-PERF-PRET-3SG very kind in walking me home (37) ¿Cuándo estuvo amable Pedro? Al acompañarme a casa When estar-PERF-PRET-3SG Pedro kind? In walking-me home (38) a. when the copular verb is the SL estar.

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

but can give what can be called a “forward-lifetime effect” (46). esquimal (44) Pedro es Pedro ser-PRES-3SG Eskimo “Pedro is Eskimo” californiano (45) Pedro será Pedro ser-FUT-3SG Californian “Pedro will be Californian” (46) Pedro is not born yet . the predicate is understood as referring to school time. A sentence like (44). does not activate the reading in (43). The future tense does not yield the interpretation of (43). However. such as be Eskimo. either. (43) Pedro is no longer alive The second necessary condition is that the predicate must be tensed in the past.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). In the remainder of the discussion I will not make any other allusion to future tense cases. an interpretation like (43) is available when the predicate is a lifetime IL predicate. This property will hold of its subject all through his life since it refers to his origin itself and cannot be modified or lost. with an IL predicate that does not denote a necessarily lifetime property. In (40). in the present.

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

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

” An utterance such as (56) is appropriate if the situation is the one in (58) but not acceptable if it is as in (57). NP subjects provide the time of existence of the individual they denote as a value for C. she alludes to the Maxima of Quantity of information proposed by Grice (1975).Tense and Individual-Level Predicates 209 (54) Temp Operator [C] (Gregory was from America) ↓ [the day I was introduced to Gregory and Eva-Lotta] The sentence Gregory was from America is going to be temporally evaluated with respect to the contextual restriction “the day I was introduced to Gregory and Eva-Lotta.” Musan takes this as a proof .e. when sentences are uttered out of the blue and the variable C restricting the temporal operator lacks an interval explicitly supplied from the context. Musan does not seem to take this as a definitive explanation and argues that pragmatic considerations play a role in determining the arising of the reading in “out of the blue” cases. Specifically. (55) Temp Operator [C] (Gregory was from America) ↓ [Gregory’s time of existence] Since the interval of Gregory’s existence acts as a restrictor. Musan says. On my view. given the presence of such a contextual restriction. Regarding what makes the lifetime reading arise. it is the value of C that accounts for the arising or neutralization of the lifetime reading. (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). this could have been enough. Musan assumes that lifetime readings arise when the sentence is in a situation of temporal un-specification. That is. Musan assumes that past and present tense differ with regard to their “informativeness. the NP subject itself is able to supply a value for C. In my interpretation of Musan’s proposal. However. “out of the blue” sentences).. Musan (1995:56) argues that in temporally unspecific contexts (i. it is the different value of the contextual restrictor C that gives the different possibilities for lifetime readings. That is. the hearer would react by adding: “…and he still is from America. the lifetime reading gets neutralized.” Musan argues that. Specifically. the sentence is evaluated with respect to that interval.

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

Thus. it can be observed that the value of the TT is decisive in the arising of such readings. 1997) talks about “the neutralization of lifetime effects. From this.1.9 What needs to be defined is which content gives rise to the lifetime interpretation. section 5.” However.1.Tense and Individual-Level Predicates 211 sentence—namely. In other words. this interval is called “Topic Time” (TT) (Klein 1994). I thank Tim Stowell for this remark.. here assumed to be the Utterance Time (UT) (i. I assume that the nature of the TT is not different from other intervals like Reference Time (RT) and Eventuality Time (ET). Likewise. Tense is locating in the past not the whole interval the eventuality (be from California) holds. since we have proved that it is the content of the TT that matters.e. As I introduced in chapter 5 (section 5. in the past). I assume that the TT is a time-denoting phrase (ZP). In this case.2). 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.1). the interval at which Maria and Harry arrived in the USA. but the interval the speaker is referring to. it seems more accurate to talk about the arising or not of the interpretation. . In the spirit of Musan (1995). lifetime effects can be reasonably considered a phenomenon of temporal interpretation. since from the perspective I propose here it is not the case that the reading arises and then it becomes neutralized. The occurrence of the lifetime effect is the opposite side of the coin of their absence. the interval corresponding to the arrival in the USA. no lifetime reading is available. as Stowell (1993) proposes for the other two (see chapter 5. I propose that the lifetime reading arises when the TT is defined as the interval of the individual’s existence: 9 Musan (1995. is located within the span of time at which being from California holds and after the RT.

For sentences inserted in a context with a previous past interval mentioned. 6. focusing on sentences with lifetime predicates. Put in these terms. based on Musan (1995). Thus. also in the line of Musan. the reading that arises is a lifetime reading. it gets its content from salient elements in the conversation. In sum. In this sense.3). in (61). The next sections discuss the sources for the TT. a lifetime reading arises (or not) depending on the content of the TT. I argue that these two cases can be explained uniformly. For “out of the blue” examples. in this way differing from Musan.4 The Determination of the TT Content and Lifetime Effects Thus far. The leading idea in this regard will be that the TT is topic sensitive—that is. I conceive the lifetime reading as an interpretive outcome derived from the location in the past of a TT with a specific content—namely. I argue that the temporal TP 2 . In what follows. I proposed. we have seen two types of cases: “out of the blue” cases and sentences inserted in a context. I claimed. I will discuss the role of the contextual information conveyed by the DP subject itself in particular. and Tense restriction for contextual ones). the lifetime issue is reduced to the working of temporal interpretation in general. accounts for them by appealing to different resources (Gricean Maxima of Quantity for “out of the blue” cases. that referring to the lifetime span of the individual referred to by the DP subject.4. it is the whole span of time of Harry’s existence that gets located in the past. as I said before.212 Individuals in Time (61) RT (UT) T′ 2 T AspP (after) 3 TT Asp′ (Harry’s existence interval) 2 Asp VP (within) 2 ET VP (being from California) Since. that the TT takes its content from the lifetime interval of the individual referred to by the DP subject. that the TT takes its content from the previous interval mentioned (I will refine this account in section 6. who. 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).

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

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

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

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

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

Conversely. 5. The context relevant to find the “antecedent” for the TT does not have to be another past interval necessarily. so that all the surveyed cases can be accounted for in a uniform way. 4. 2. The TT refers to a specific interval. more systematically. 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.218 Individuals in Time situation. how this is articulated. That is. I have argued that this is brought about through the context variable restricting quantifiers (free variable C). I want to show now. Enç 1991a. as a consequence. it becomes the source for antecedents.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. 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. (77) Salient context " TT . Beghelli & Stowell 1996). This seems to suffice to convert the utterance situation as the background that counts as relevant for the antecedent of the TT. The relevant context can be brought to the scenario by the DP subject itself. I claimed that lifetime effects arise when the TT refers to an interval that can be described as the lifetime span of the individual referred to by the DP subject. In other words. where the individual referred to by João is present. along the same lines that specific DPs are generally understood (Pesetsky 1987. following von Fintel (1994). 3.4. The leading ideas introduced thus far are the following: 1. Elaborating on Musan 1995. as Musan (1995) suggests. and. The relationship between the contextual background and the TT or the free variable C has been taken as anaphoric. A specific interval is an interval that is linked to a previous context. The occurrence of lifetime effects depends on the content of TT of the sentence. 6. which contains contextual information associated to the individual referred to by the DP. the context acting as background constitutes the set where anaphoric elements and free variables find their antecedents. I also argued that proper name DPs involve a similar context variable. the utterance situation is the most salient context.

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

a lifetime effect does not arise. Result: a past form is allowed in the copular clause and. I propose. Since the DP subjects of copular sentences can be considered “topical elements. no lifetime effect arises. Finally.4.220 Individuals in Time crucial importance in the licensing and interpretation of a past form. 1995) and Musan (1995) assume. then. the value it takes by default is the individual himself and the interval associated to him.” they establish the relevant context. Once a context is fixed. I assume that. Result: lifetime arises With Musan (1995. I have argued that the establishment of the relevant context is brought about through the free variable DPs (more specifically. (80) Harry was from California.4 Summary of Section 6. the ZP . 6. Bearing in mind that the TT is a ZP context sensitive. when there is no specific context the free variable is to be linked to. The interpretive procedure is always the same and works the following way. let us consider a sentence where a lifetime reading can arise. (…) Topic of the sentence that may establish the connections to a previous background = Harry. As Kratzer (1988. since they are the same in all relevant respects—all of them are copular sentences in context. there is no reason to consider (59) different from (72) and (73). it finds its antecedents in the most salient context. Since the contextual variable directs us to a previous context containing a past interval. that in (59) it is the contextual information associated to the DP Harry that establishes the context that counts. the determiners) have. since it refers to the time of the arrival. these typically are out of the blue sentences. it is the information associated to the DP subject that counts. (82) Gregory was from America (83) Topic of the sentence that may establish the connections to a previous background = Gregory (84) Gregory c= undefined. Gregory himself TT → the whole interval overlapping Gregory’s lifetime span.4 In determining the context relevant for the specification of the TT content (which is what finally decides the temporal interpretation of a sentence). (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). by default. Actually.

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

yielding different interpretations. as the subindexes of the ZPs in (i) suggest. in compound sentences. have noted.M. and its eventuality will be located with respect to the eventuality of the main clause.15 However. the ordering between the TTs should depend on the specification of the embedded RT. giving rise to different temporal readings. since the RT would be controlled by the main ET rather than by the main TT. In this work.1. to understand the temporal interpretation of a compound sentence is to understand the relationship holding between the two (or more) TTs. this does not mean that ETs always remain in the dark for interpretation. (Thus far. the embedded TT would become ordered with respect to the main ET. temporal interpretation is derived from the ordering relationship Tense establishes between the RT and the ET. the UT. As mentioned in chapter 5. Specifically. whose nature I argue to be also a ZP. the prototypical case of this is the modification of the past perfect.M. instead of with respect to the TT. Thus. as in (i). yielding the interpretation in (iii). (i) (ii) (iii) 15 14 Peter had left at 3 P. so that it can be controlled by the ET of the main clause. starting by the former ones. among others. This way. yielding the reading in (ii) or the ET. My purpose in doing this is to explore the temporal interpretation of these predicates when dependence with respect to another tense complex is syntactically established. the content of the RT is. The time adverb at 3 can modify the TT.5 The Lifetime Reading in Compound Sentences For the sake of a more complete survey of cases. I focus now on the interpretation of lifetime IL predicates in another type of syntactic environment. Following the logic of Stowell’s proposal discussed above. this could lead us to a situation where the TTs are not ordered between each other strictly speaking. given that the ET is the closest ZP. 1996) work.222 Individuals in Time 6. In simple sentences. According to him. Peter’s leaving occurs at 3 P. TT and ET can be modified by time adverbs. subordinate clauses (complement and relative). As Hornstein (1990) and Demirdache and Uribe-Etxebarría (2004).1 Complement Clauses In chapter 5 (section 5. 6.1) I described the way temporal interpretation is accounted for following Stowell’s (1993. namely. . the subordinate RT can be affected by the ET of the main clause. if the RT of the subordinate clause is in the c-command domain of the ET. Peter’s leaving occurs prior to 3 P.M. the subordinate clause will be temporally dependent on the main one. by default. However. I have just dealt with IL sentences which establish a link with a previous one paratactically. I argued that what Tense locates is not the ET as such.) I will deal with complement clauses and relatives in turns.14 but rather the TT.5.

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

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

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

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

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

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

Since there is no content in T. and (b) that the subordinate TT is not an interval different from the main clause (TTj). but ‘within’. therefore. the interval the subordinate clause is referring to is the interval at which Maria said something. 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”). (96) María dijo que Juan estuvo enfermo Maria said that Juan estar-PERF-PRET-3SG sick The unique temporal interpretation is a past shifted reading. which seems to accurately capture the intuition about the interpretation of (88). which seems desirable since simultaneity is only possible with such an aspectual form. their temporal values coincide. a past shifted reading? The content of T. crucial use of the content of the aspectual head. in contrast. Is this a technical inconvenience? I would like to argue that it need not. the possibility that an interval becomes ordered with respect to itself does not actually arise. That is. there is no ordering predicate. This way. where the main clause interval gets located “after” the subordinate interval. which is not ‘after’. If we follow interpreting the tree. but it has no content. then. as a consequence. sick sick ↓ ↓ (95) ----------------------//////X/////////-----------------UT---------↑ saying However. What precludes.Tense and Individual-Level Predicates 229 as usual. the RT binds the TT and. it is null. since the content of T is null. then. by virtue of the content of the aspectual head. with no further independent evidence. This analysis makes. the interpretation can be worded as follows: the TTi corresponding to the interval of “saying” is included in the interval of “being sick”. The second possibility represented in (94b) is to consider (a) that Tense does not have the meaning of ‘present’. then. but the same one of the main clause (TTi). Consider. what we get is that the interval (TTi) is included (‘within’) the eventuality of being sick. (97) -----------be sick-------------said---------------UT----------I will thus assume that an analysis along the lines sketched in (94) (T as null and the subordinate TT being the same as the main clause’s one) is appropriate . the same sentence in perfective. In other words.

Consider in contrast (100). will examine whether a lifetime reading is available. as can be tested by the unavailability of paraphrases such as (99). the simultaneous one. I will follow what I said above. namely. with a stative SL predicate. . ‘Eva serfrom Albania in a previous life’. since it emphasizes the relevance of the aspectual head content. (98) Cristina dijo que Eva era de Albania Cristina said that Eva ser-IMPF-PRET-3SG from Albania The first characteristic to be observed is that there seems to be just one temporal reading. 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. this could be a paraphrase for Eva fue de Albania en una vida anterior. specifically.230 Individuals in Time to account for the simultaneous reading. I will examine the temporal interpretation of (98) and. The past shifted reading is absent. (99) *Cristina dijo que Eva había sido de Albania18 Cristina said that Eva had ser-PAST part from Albania (100) Cristina dijo que Eva había estado enferma Cristina said that Eva had estar-PAST part sick For analyzing the simultaneous interpretation.

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

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

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

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

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

I defended the lifetime reading as an interpretive outcome due to a specific content of the TT—namely. Consider (113) in relation to (111). I concluded. Among the issues I have not discussed in this work is the relationship between the TT content and time adverbials. Others (kind. in section 6. that the content of the TT is not sensitive to syntactic distance but to discourse prominence. In both cases the lifetime reading does not arise. (111) Harry and Maria arrived in the USA. Actually. 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. complement and relative clauses. Finally. blond. I examined the arising of the lifetime reading in syntactic composition—specifically. who defends the position that the lifetime characteristic comes specified in the lexical entry of the predicate. most of the properties are like the latter ones. For example. although it has become clear that the TT of an IL lifetime predicate can refer to an interval given by the discourse. I showed that a lifetime reading is not available in all of the cases where a lifetime predicate appears in the past tense. a lifetime reading will be available. when it refers to the whole interval of an individual’s lifetime. I also offered a number of examples of lifetime predicates in the past tense that. that interval cannot be “spelled out” by an adverbial.236 Individuals in Time individual’s lifetime that is their subject. although the syntactic dependence between the matrix and the subordinate tense complexes was different. which prevents a lifetime reading from arising. are not necessarily lifetime properties. In sections 6. even without the overt presence of another past tense.5.4. In this respect. and I argued that the arising of the lifetime reading depends on the content of the TT. their good combination with restrictive temporal adjuncts proves. if the TT refers to the interval corresponding to the span of time an individual’s lifetime extends over. so he did not have to go through immigration (112) TT refers to the time of Harry’s (and Maria’s) arrival (113) *When Harry arrived in the USA. I differ from Musan (1995. as. I argued that the contextual restriction of DPs can intervene in building up a previous background. After introducing Musan’s (1995.) can hold of an individual just for a determined period of time shorter than the individual’s lifetime. A large number of questions remain to which I have not even suggested a possible answer. etc.3 and 6. I assimilated her observation to the theoretical terms assumed here. I have shown that a good number of predicates that can be considered IL. therefore. Harry was from California. Thus. 1997). for example. do not have a lifetime interpretation. he was from California . based on independent grounds. 1997) insight that lifetime effects can be neutralized by the presence of another past interval in the context.

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


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

Specifically. dynamicity is proposed to correlate with the presence of the relational complement and argued to be rooted in the Prepositional Phrase. Third. ser and estar. Others. The restrictions between a ser-predicate and aspectual forms like the perfective or progressive are derived from the inneraspect properties of the predicate. but it is independent from that: cruel mean/kind to someone can appear with both copulas. kind. Topic Time has been argued to be a time interval taking its value from the salient discursive context. the temporal interpretation of IL predicates is derived. this study is a contribution to a better understanding of the role of prepositions in the properties of event structure. however. The book proposes a uniform treatment of tense. in the case of IL predicates. the aspectual characteristics of the copular constructions are derived from their syntax. among others). Ramchand 2003. this book gives further empirical support to recent hypotheses according to which aspectual properties are syntactically encoded (Borer 2005.240 Individuals in Time individual. based on hypotheses that attribute aspectual content to prepositions (Hale 1984. Ritter & Rosen 2000. such as Eskimo. differing from previous influential proposals like Kratzer 1988 where. are properties that can be true just of a limited period of time. or color-blind. I have proposed that the dynamic and active properties classically observed in certain copular constructions are due to the presence of a complement. 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. IL predicates are not all stative. I have accounted for a variety of contextual situations illustrating the sources from . I have shown that ser-predicates can appear in any aspectual form. Second. there are two types of IL predicates regarding inner aspect: stative and dynamic predicates.. As a result. Finally. Demirdache & Uribe-Etxebarría 2000). in Spanish. as is the case with any other type of predicate. I have argued that the dynamic nature of copular constructions does not correlate with an SL classification. In this respect. In particular. the dynamic properties of clauses where adjectives referring to mental properties appear (such as cruel. from the content of the so-called Topic Time (Klein 1994). such as blond or young. The investigation of Lifetime Effects (the interpretive phenomenon whereby the DP subject can be understood as ‘no longer alive’) in chapter 6 deepens our understanding of the concept of Topic Time. unlike what is commonly assumed. Rather. as is the case with any other type of predicate in Spanish (e. which clearly pattern with activities in an inner-aspect classification. gypsy. Crucially also. Thus.g. tense takes and orders the nominal argument (the NP subject). no stative verb can appear in the progressive). contrary to general belief. mean) have been proposed to correlate with the presence of the relational complement of these adjectives (cruel to Mary).

As was surveyed. the notion involved. where the definition of IL predicates has been mainly associated to the notions of permanent property and stativity. As noted in the beginning of the book. a number of authors have proposed that the IL/SL distinction should be cast in “aspectual” terms. most . these notions. tense orders the individual referred to by the DP subject Perfective and progressive viewpoints make an IL predicate SL IL predicates are stative predicates Notion involved Length of the interval the property holds Temporal ordering Semantic domain Tense Tense Delimitation and progress in time Internal temporal characteristic of the property Outer aspect Inner aspect Table 7. classically alluded to in the description of IL predicates. Table 7.1. Previous claims regarding IL predicates Throughout this work I have analyzed IL predicates in their relation with the temporal realms cited. By answering this question. Claim (in previous literature) IL predicates express permanent properties When an IL predicate is at stake. between IL/SL predicates) can be cast in temporal terms. we will be proposing something about the terms in which the IL/SL dichotomy is represented in the grammar.1 summarizes the claims previously made. One of the important questions here is whether the difference between ser/estar clauses (or. I will introduce some discussion in the following sections. One of the crucial questions is how the association to a situation is brought about. As mentioned in chapter 2. In examining this question we address the key difference between IL/SL predicates. and I have given an account for the different temporal interpretations that IL sentences can have without yielding inaccurate generalizations. most authors dealing with the ser/estar contrast concur with the idea that estar-predicates put the property in relation to a situation.Conclusions and Final Remarks 241 which the Topic Time can take its value.” leaving the property without any association to a determined situation. These conclusions differ from most previous literature. while ser is more “innocuous. more in general. 7. Ser simply classifies the subject into the category denoted by the predicate that appears in combination with it. Although I will not be able to offer a definitive answer to this. are temporal concepts in nature. and the semantic domain they belong to.2 The IL/SL Dichotomy and the Ser/Estar Contrast With major or minor differences.

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

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

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

it makes no difference between an SL and an IL predicate. In the first place.” I would like to make two remarks regarding these proposals. In this sense. it is not a state. the predicate may pattern with either states or activities. the inner-aspect type of one predicate does not depend only on the verb alone. Luján (1981) establishes a parallelism between estar and predicates like write a letter. Specifically. both undelimited predicates. which is why he did not have to go through immigration” TT = the time of the arrival (8) This constitutes another piece of evidence that the IL/SL dichotomy is not rooted in temporal terms of the realm of Tense. In turn. a delimited process. the notions they allude to. for Fernández Leborans (1999) the IL/SL dichotomy is founded on aspectual properties. in this respect. In a similar vein.4 The IL/SL Distinction Is Not a Matter of Inner Aspect…Completely In chapter 2 I surveyed the accounts by Luján (1981). the copular verb ser would not be very different from regular predicates. That is. ser lacks any inherent temporality and is described as “inert with respect to aspect. Harry was from California. an event or a process. Harry era Harry and Maria arrived in the USA Harry ser-PAST-IMPF-3SG from California. On my view. but the copular verb plus its complement can be asserted to behave either as a state or as an activity. I defined inner aspect in a . Schmitt (1992). I argued that depending on the adjective. and Fernández Leborans (1999). Whereas estar possesses internal temporal structure. As mentioned in chapter 3. when they establish the IL/SL distinction. or a process. Schmitt (1992). Second. Although these authors are not entirely explicit about the temporal realm they are referring to (or about the definition of some terms). seem to belong to the domain of inner aspect. 7. whereas ser is underspecified with respect to aspect. Likewise. it is not accurate to say that ser is not a state. Tense has been defined as a predicate ordering the TT with respect to a reference time. and Fernández Leborans (1999) what the term “aspectual” refers to or how inner aspectual properties are tested is not made explicit. The copular verb itself may be none of those. ser is conceived as a predicate parallel to write or admire. por lo que no tuvo que pasar la aduana California for it that not have to go through immigration “Harry and Maria arrived in the USA. an event. Schmitt (1992) argues that estar involves aspectual properties. In contrast. I have largely shown in chapter 4 that ser-clauses differ with respect to their inner-aspect properties. in the works by Luján (1981).Conclusions and Final Remarks 245 (7) de Harry y María llegaron a los EEUU. but on the verb plus its complement(s). ser has no inherent temporal structure. then.

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

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

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

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

Crucially.6 Demonte shows that adjectives have an IL interpretation when used as modifiers inside a DP. and the other is due to the intervention of one of the copular verbs. According to Higginbotham and Ramchand. but it is the entire predicate together with the copula that is IL or SL. Along the lines of Higginbotham and Ramchand (1996). . Categorical judgments are predications about an individual (x). or to propose that adjectives are of one kind by default. 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. Following Demonte (1999). precisely. among many others. I think that the properties of SL-hood consist of the ability to link a property to an external situation.250 Individuals in Time they can go with either of the copulas. describable in terms of categorical versus thetic judgment (Kuroda 1972). The next natural question is what such properties are. As I showed. the copular verb that appears in the paraphrases of the adjectives inside the DPs is ser and not estar. I argue. outer aspect. the external situational variable (s) exists in addition to the eventive variable 6 I leave aside here those adjectives that combine only with estar. temporal heads are not capable of contributing this external situation. whereas thetic ones are predications about a situation (s). whereas the semantics of IL-hood would consist in the absence of such an association. both options are quite close to each other). or tense. I want to consider that adjectives are IL by default. copular verb estar. makes the predicate SL. As introduced in chapter 2. Therefore. there are two possibilities: to propose that adjectives are not IL or SL. the semantics of SL-hood would consist. there are no differences between ser and estar predicates regarding inner aspect. (In a sense. in the association to a particular situation. From this perspective. Consider the following examples with the adjectives mentioned above: (20) Los niños guapos/morenos/graciosos ganaron el concurso the handsome/dark-skinned/funny won the contest “The handsome/dark-skinned/funny guys won the contest” (21) las niños que eran/*estaban guapos/ morenos/graciosos the guys who ser/estar-PRET-IMPF-3PL handsome/dark-skinned/funny “the guys who were handsome/dark-skinned/funny” As shown in (21). It is estar that introduces the properties typically associated to SL-hood. This way. That is. Higginbotham and Ramchand (1996) conceive the IL/SL contrast as a distinction pertaining to the Logical Form. and restrict my attention to those that can appear with either of the copular verbs. however. the interpretive status of those adjectives without the presence of any copula is the same as when they appear with ser.

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

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

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

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

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

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

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

258 Individuals in Time 7.8 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. outer aspect (by alluding to the completion or perfection of the property). regarding inner aspect. Restricting myself to the contrasts argued to correspond to the IL/SL dichotomy of (1) and (2).3 summarizes all of these points. I have argued so by showing. would locate the individual subject at the circumstance expressed by the predicate following the copula. Second. whereas estar is the only copula capable to constitute heterogeneous predicates with certain adjectives (participles and cut-short adjectives). so that no distinction can be made between them in this sense either. I have shown that. inner-aspect properties do not seem to be able to differentiate between the two copulas. Finally.16 Table 7. 16 . then. I have shown that both estar and ser clauses can appear with the same outer aspect forms. in the first place. The copula estar. links the property to a situation. or inner aspect (mereological properties). since we can construct both types of sentences with adverbials that restrict the duration of the property. (1) and (2). One difference that exists between ser and estar regarding inner-aspect properties is that estar cannot appear in the progressive under any circumstance. regarding the minimal pairs where the IL/SL opposition is detected. I have argued that the lexical entry of estar includes as a part of its meaning ‘circumstance at which an individual is found’. both predicates ser and estar are homogeneous predicates. by virtue of its lexical characteristics. 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. that the length of the interval a property holds does not differentiate estar from ser clauses. Therefore. I have contributed the idea that the association to a particular situation cannot be recast in terms of tense (the length of the interval).

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


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

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

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

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

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

MIT Press. E. Berlin: Zentrum für Allgemeine Sprachwissenschaft. Hornstein. C. 1976. dissertation. M. 1985. As Time Goes by: Tense and Universal Grammar. On semantics. Stage levels. The logic of perceptual reports: An extensional alternative to situation semantics. 1996. 1988. Cambridge. Ohio State University. Higginbotham. Ramchand. R. Ph. Higginbotham. A Compositional Semantics for Aktionsarten and NP Reference in English. 1993. J. 29–48. 1996. Cambridge. Heycock. dissertation. Linguistic Inquiry 16:547–593. Hale and S. Higginbotham. Jackendoff. Linguistics Inquiry 7(1):89–150. 1995. de Swart. Herweg. 1981. . and the semantics of the copula. Demonte (eds. 1972. Papers in Linguistics 14:65–94. Stokhof (eds. 1983. M. Kempson (ed. MA. J. Columbus. Kamp. telicity. J. Janssen. models. El infinitivo. In Formal Methods in the Study of Language. Proceedings of GLOW in Asia 2:72–82. In Gramática Descriptiva de la Lengua Española. North East Linguistic Society 25:223–238. Cambridge: Cambridge University Press. University of Massachusetts. J. and perhaps even quantification in English. Heycock. Hernanz.D. In Mental Representations. The Journal of Philosophy 80:100–127.266 Individuals in Time Hale. The Semantics of Definite and Indefinite Noun Phrases. 1999. 2000. The proper treatment of measuring out. Amsterdam: Mathematical Centre Tracts. 1994. A theory of truth and semantic representation. MA: MIT Press. New York: Garland. MA: MIT Press. I. I. What Counts. 1991.).D. The internal structure of small clauses: New evidence for inversion. Jackendoff. 277–322.). University of Oxford. and M. Jäger. Over indefiniete objecten en de relatie tussen syntaxis en semantiek. In The view from Building 20. Amherst. MA: MIT Press.. Higginbotham. R. 1989. Jackendoff. de and H. The stage-level/individual-level distinction and the Mapping Hypothesis. J. J. Ph. J. Perfective and imperfective aspect and the theory of events and states. and G. Cambridge. R. Layers of Predication: The Non-Lexical Syntax of Clauses. Bosque and V.). E. T. J. Madrid: Espasa Calpe. On argument structure and the lexical expression of syntactic relations. Glot 12:19–35. Toward an explanatory semantic representation. 2000. states. Cambridge.). 1990. H. C. Higginbotham.-L. 2197–2356. Semantic Interpretation in Generative Grammar. Herburger. Ms. Natural Language & Linguistic Theory 14:305–354. Keyser (eds. 1986. Linguistics 29:969–1010. and S. Hoop. K. Accomplishments. 53–109. and meanings. K. Groenendijk. G. N. R. Contexts. H. Hinrichs. 1982. Keyser. 1999. Heim.

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

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

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

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

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

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

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


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

81. 205. P. G. 126 Morimoto. 3. 8–12. 106. 214 Luján. 11. R. 37. 33. R. 156. 38. 31. 70. 12–14. 152. C. 145. P. 59. G. V. 56 Sanz. R. J. 120 Parsons. 32. B. 240 Rivero. 131. 40. M. 61. P. 215 Stalnaker. W. 1. 144. 214 Svenonius. de 48 Milsark. 192. M. 144. P. J. 57 Masullo. 215. S. 143. 213 Stowell. ter 148. 40. 232. 94 Morera. 122. 116. 191 Larson. 245 Martin. D. T. 33. 213 Schmidt. 73. 154. 152. 113. 18 Piñón. 165. 114 Schieffelin. 38. 12–14. 193–196. 109. 55. 37 Pustejovsky. 178. J. M. 155. 213 Kennedy. 150. 29. Y. 122. L. 178. J. 3. 34. 12–14. 22 Meulen. 251 Krifka. 10. 15 Koopman. R. 225. S. 96. 236 Naumann. S. 232–235. 14. 217–220. 28. 153. 194 Kratzer.-Y. 55. 226 Pagliuca. 3. 163. 127. M. 138. 176. S. 119. N. 240. B. 150 Reinhart. 106. 40–43. 43. 208. 119. 135 Kiparsky. 176 Munro. 70. 41. A. 8. T. B. 216 Uribe-Etxebarría. 40–44. A. 154. 56. W. 32. E. 14. 50 Landman. 25 Quine. D. 113. 71. 148. 120– 123. E. A. C. 43. 8. de 30. 41. G. 245. T. 3. 220. 232 Lakoff. 5. 78. 235. 191 Picallo. 216 Reichenbach. A. 164 Miguel. 74. 109. 16. C. C. 61. 71. 226. 191 Partee. 75. 172. 115. L. 218. 188 Keenan-Ochs. 237. M. 136. 112 Schein. 77 Roldán. H. 251 Quer. 213. 16 Rosen. M. 145. 140. 214 Perkins. G. 116.-Y. 80 Ritter. 71 Ramchand. 3. 26. 250 Rapoport. 233 Rohrer. 61. 188. 110– 112 Torii. 38 Mithun. 119. L. 109 McNally. 143 Russell. 194 Klein. 147. 46. 43. P. E. 233 Querido. 15. 45. 22 Kenny. 127.276 Individuals in Time Kearns. 120 Pesetsky. 211. 138. 40. M. T. 116. 44. 122. T. 251 Sportiche. 96. 38. 11. 113. 77. 208. 225. C. 70. 189. 80. 189. 8. 113 Kitagawa. W. 136. 147. D. S. 190. 112 Tungseth. 15 McConnell-Ginet. 18 Schmitt. 143. 60. 240 Rothstein. 222. R. 94 Musan. 47 Keyser. 84. 167. 58. 43. 243 Travis. 102. 205. 151. C. 36–38. H. M. 107. 85. 251. 176 Szabolcsi. 49. 118 Matushansky. 162. J. 5. A. 12. 192. 145. 97. 152. 191 Pollock. R. 218 Phillips. M. 250 Ladusaw. E. O. 196. 97. 164. 252. 38. M. 114 Kuroda. 144. G. 84. 6. 240. 232 Ryle. 55. 214. S. 195 Leonetti. C. 126. C.-L. 208–212. 37. 29 Talmy. 79. 148–150. 15 Raposo. J. 86–90. 112. 207. 117. 102 Matthewson. 25–32. 225. F. 222. 194. J. 3. 156 Tenny. 50. J. L. M. 25. 244 Stump. T. G. S. 162. 117. 211. 61. 130 Mourelatos. 112. B. 124. 242–244. E. 240 Kondrashova. Y. W. 29. 93. 55 Rosen. 88 Lemmon. 86. 214 Longobardi. 94. 193. 148. 98. 240 . 84. 127 Uriagereka. 202. 127 Swart. A. H. 43. 89. 144. 15. 191 Ogihara. 254 Lewis. 207.

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


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

115. 128–131. 184. 186. 168 French 121 Generic operator 11. 129. 123– 126. 246. 236 Theme 105. 148. 131. 233. 155. 234 Situation aspect 39 Small v 55 Spatiotemporal argument/variable 8–10. 249. 252 Realization function 6. see also Event(ive) argument Speech Time 122. 159 Quantification restriction 11 Quantity 113–116. 244. 196. 180. 192 (Im)perfective contrast 25. 207. see Temporal interpretation Past-shifted reading 76–78. 37 Thetic (judgments) 12–14. 122. 14. 241. 144. 118. 221–223. 35. 201. 187. 192 Prospective 120–122. 175–177. 134 Instigator 55. 192. 153–160. 129 Progressive 153–163. 12. 94 Margi 121 Maxim of Quantity 85. 64. 147. 240. 35. see also Subinterval property and Additivity property Narration temporal interpretation of. 176. 227. 181. 148–152. 134–136. 248. 180 Perfective (preterit) 31. 152. 192. 189– 192. 76. 189. 227 Simultaneous reading 76–79. 148. 212. 134–136. 72. 183. 215 Goal(Path) 96–98. 156 Focus 13. 137 Subinterval property 70. 202 Ground 119. 186. 135. 79. 249 Existential coda 12 Existential quantification 8. 187–192. 258 Progressive paradox 191. 189. 157. 194. 156 Hawaii-Creole 120 Hebrew 15 Heterogeneous (predicates) 72. 158. 154–157. 72. 11. 249 Relative clauses temporal interpretation of. 11. 224. 184. 246. 229. 172. 112 Juba Arabic 120 Kinesis 40 Krio 120 Lakhota 55. 136 Place/situational prepositions 120. 148. 181 Existential reading 6. 134 There-sentences 5. 94. 160. 258 Imperfect Continuous 76. 186. 35 Figure 119. 123. 177–180. 125. 246–248. 183. 189. 258 Hindi 120 Homogeneous (predicates) 70. 207. 128. 172–174. 233 Perfect entailment 47. 222. 132. 212. 178. 173–180. 258. 148. 155–157. 6. 231. 129. 174–178. 178. 176. 162. 105. 240. 242. 224. 147. 144. 244. 165. 227–229. 144. initiator 61. 175. 159. 248. 94 Pomo 94 Prepositions Directional prepositions 127. 148–152. 194 Reference Time (RT) 3. 230 of Narration 77–80 of Relative clauses 223. 116. 88. 14.280 Individuals in Time Event-object 110. 257 Perform(er) 53. 130. 67. 165. 131. 158–160. 161. 230. 246. 242. 31 Imperfective paradox see Progressive paradox Initiation. 173. see Temporal interpretation Russian 15 Salish 118. 57. 195. 158. 122. 248 Swahili 120 Temporal interpretation of Complement clauses 76. 100. 142. 211. see also Utterance Time Stativizer 115. 208. 178. 147. 148 Sequence of Tense 226. 136. 132. 226. 112. 179. 186. 86. 227. 232. 158. 123. 230. 141. 167. 229. 250 . 233–235. 209. 131. 187. 191 Event(uality) Time (ET) 3. 184. 14 Generic reading 6. 111. 172. 83. 191. 226. 116. 150. 176–178. 26. 245. 158. 80. 186. 58. 154. 222–225. 162. 106. 148. 72. 117. 79. 158–160. 171– 175. 113. 132. 141. 242. 211. 212 Mereological properties 3. 55. 226. 80. 189. 189. 257. 147. 12. 136. 221 Habitual 77.

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

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

Sign up to vote on this title
UsefulNot useful