Studies in Linguistics and Philosophy – serie

In the past few decades, the development of theoretical linguistics has proved to be successful in shedding light on the intricate nature of language and knowledge of grammar, which contributes to a deeper understanding of the human mind. This book discusses various issues in syntax and logical structure of natural language from theoretical perspectives. The primary data on which theoretical claims are made is drawn from Japanese and Japanese-type languages, but it also contains discussion of related phenomena in English which have never been discussed from the same viewpoint in the current literature. Although the book is written in the format of a version of the Extended Standard Theory tradition, informally referred to as the Principles and Parameters Approach or 'Government and Binding (OB) Theory', it should be of interest to a much wider audience. The reader interested in other theoretical frameworks will find the discussion in this book easily translatable in the framework of his or her choice - in fact, I would like to claim that the problems posed by this book are inevitable in any theory of syntax and semantics of natural language.

In the past few decades, the development of theoretical linguistics has proved to be successful in shedding light on the intricate nature of language and knowledge of grammar, which contributes to a deeper understanding of the human mind. This book discusses various issues in syntax and logical structure of natural language from theoretical perspectives. The primary data on which theoretical claims are made is drawn from Japanese and Japanese-type languages, but it also contains discussion of related phenomena in English which have never been discussed from the same viewpoint in the current literature. Although the book is written in the format of a version of the Extended Standard Theory tradition, informally referred to as the Principles and Parameters Approach or 'Government and Binding (OB) Theory', it should be of interest to a much wider audience. The reader interested in other theoretical frameworks will find the discussion in this book easily translatable in the framework of his or her choice - in fact, I would like to claim that the problems posed by this book are inevitable in any theory of syntax and semantics of natural language.

Bound and Referential Pronouns.- Logical Form and Barriers in Navajo.- Towards a Modular Theory of Coreference.- Head Government in LF-Representations.- Logical Structure in Syntactic Structure: The Case of Hungarian.- In Defense of the Correspondence Hypothesis: Island Effects and Parasitic Constructions in Logical Form.- Construing WH.- Two Properties of Clitics in Clitic-Doubled Constructions.- LF Movement in Iraqi Arabic.- List of Contributors.

In ordinary discourse we appear to ta1k about many things that have seemed mysterious to philosophers. We say that there has been a hitch in our arrangements or that the solution to the problem required us to examine all the probable outcomes of our action. So it would seem that we speak as if in addition to eloeks, mountains, queens and grains of sand there are hitches, arrangements, solutions, probiems, and probable outcomes. It is not immediately obvious when we must take such ta1k as really assuming that there are such to develop tests for things, and one of the tasks in this book is discerning what has eome to be called ontological commitment, in naturallanguage. Among the entities that natural language appears to make reference to are those connected with temporal and modal discourse, times, possibilities, and so on. Such entities play a crueial role in the kind of semantieal theories that I and others have defended over many years. These theories are based on the idea that an essential part of the meaning of a sentence is constituted by the conditions under whieh that sentenee is true. To know what a sentence says is to know what the world would have to be !ike for that sentence to be true.

In ordinary discourse we appear to ta1k about many things that have seemed mysterious to philosophers. We say that there has been a hitch in our arrangements or that the solution to the problem required us to examine all the probable outcomes of our action. So it would seem that we speak as if in addition to eloeks, mountains, queens and grains of sand there are hitches, arrangements, solutions, probiems, and probable outcomes. It is not immediately obvious when we must take such ta1k as really assuming that there are such to develop tests for things, and one of the tasks in this book is discerning what has eome to be called ontological commitment, in naturallanguage. Among the entities that natural language appears to make reference to are those connected with temporal and modal discourse, times, possibilities, and so on. Such entities play a crueial role in the kind of semantieal theories that I and others have defended over many years. These theories are based on the idea that an essential part of the meaning of a sentence is constituted by the conditions under whieh that sentenee is true. To know what a sentence says is to know what the world would have to be !ike for that sentence to be true.

Preface This book is about semantics and logic. More specifically, it is about the semantics and logic of natural language; and, even more specifically than that, it is about a particular way of dealing with those subjects, known as Discourse Representation Theory, or DRT. DRT is an approach towards natural language semantics which, some thirteen years ago, arose out of attempts to deal with two distinct problems. The first of those was the semantic puzzle that had been brought to contempo­ rary attention by Geach's notorious "donkey sentences" - sentences like If Pedro owns some donkey, he beats it, in which the anaphoric connection we perceive between the indefinite noun phrase some donkey and the pronoun it may seem to conflict with the existential meaning of the word some. The second problem had to do with tense and aspect. Some languages, for instance French and the other Romance languages, have two morphologically distinct past tenses, a simple past (the French Passe Simple) and a continuous past (the French Imparfait). To articulate precisely what the difference between these tenses is has turned out to be surprisingly difficult.

During the period I have been working on this project I have received institutional support of several kinds, for which I am most grateful. I thank the Institute for Advanced Study at Stanford University, and the Spencer Foundation, for a stimulating environment in which the basic idea of this book was developed. The Max Planck Institute for Psycholinguistics at Nijmegen enabled me to spend several months working on the the manuscript. ANational Science Foundation grant to develop Discourse Representation theory, and a grant from The University Research Institute of the University of Texas, allowed me time to pursue this project. I also thank the Center for Cognitive Science at the University of Texas for research support. I thank Helen Aristar-Dry for reading early drafts of the manuscript, Östen Dahl for penetrating remarks on a preliminary version, and my collaborator Gilbert Rappaport for relentIess comments and questions throughout. The individuals with whom I have worked on particular languages are mentioned in the relevant chapters. I owe a particular debt of gratitude to the members of my graduate seminar on aspect in the spring of 1990: they raised many questions of importance which made a real difference to the working out of the theory. I have benefitted from presenting parts of this material publicly, including cOlloquia at the University of California at Berkeley, the University of California at San Diego, the University of Pennsylvania, Rice University, the University of Texas, and the University of Tel Aviv.

Formalization plays an important role in semantics. Doing semantics and following the literature requires considerable technical sophistica­ tion and acquaintance with quite advanced mathematical techniques and structures. But semantics isn't mathematics. These techniques and structures are tools that help us build semantic theories. Our real aim is to understand semantic phenomena and we need the technique to make our understanding of these phenomena precise. The problems in semantics are most often too hard and slippery, to completely trust our informal understanding of them. This should not be taken as an attack on informal reasoning in semantics. On the contrary, in my view, very often the essential insight in a diagnosis of what is going on in a certain semantic phenomenon takes place at the informal level. It is very easy, however, to be misled into thinking that a certain informal insight provides a satisfying analysis of a certain problem; it will often turn out that there is a fundamental unclarity about what the informal insight actually is. Formalization helps to sharpen those insights and put them to the test.

Formalization plays an important role in semantics. Doing semantics and following the literature requires considerable technical sophistica­ tion and acquaintance with quite advanced mathematical techniques and structures. But semantics isn't mathematics. These techniques and structures are tools that help us build semantic theories. Our real aim is to understand semantic phenomena and we need the technique to make our understanding of these phenomena precise. The problems in semantics are most often too hard and slippery, to completely trust our informal understanding of them. This should not be taken as an attack on informal reasoning in semantics. On the contrary, in my view, very often the essential insight in a diagnosis of what is going on in a certain semantic phenomenon takes place at the informal level. It is very easy, however, to be misled into thinking that a certain informal insight provides a satisfying analysis of a certain problem; it will often turn out that there is a fundamental unclarity about what the informal insight actually is. Formalization helps to sharpen those insights and put them to the test.

What can the philosophy of language learn from the classical Indian philosophical tradition? As recently as twenty or thirty years ago this question simply would not have arisen. If a practitioner of analytic philosophy of language of that time had any view of Indian philosophy at all, it was most likely to be the stereotyped picture of a gaggle of navel­ gazing mystics making vaguely Bradley-esque pronouncements on the oneness of the one that was one once. Much work has been done in the intervening years to overthrow that stereotype. Thanks to the efforts of such scholars as J. N. Mohanty, B. K. Matilal, and Karl Potter, philoso­ phers working in the analytic tradition have begun to discover something of the range and the rigor of classical Indian work in epistemolgy and metaphysics. Thus for instance, at least some recent discussions of personal identity reflect an awareness that the Indian Buddhist tradition might prove an important source of insights into the ramifications of a reductionist approach to personal identity. In philosophy of language, though, things have not improved all that much. While the old stereotype may no longer prevail among its practitioners, I suspect that they would not view classical Indian philoso­ phy as an important source of insights into issues in their field. Nor are they to be faulted for this.

The purpose clause is a common fonn of adverbial modification in English. The bracketed phrases below are purpose clauses, and they look and sound unremarkable. We hear and see these things all the time. John came [to play with the children] [to play with] I brought John along Insofar as purpose clauses appear to be adverbial, they frequently occupy a relatively low place on the scale of important things for syntactic theory to address itself to. In this book I assume the theoretical framework that has come to be known as 'Government-Binding' theory (GB), initiated in Chomsky (1981). The general fonn of the analyis of purpose clauses in GB dates roughly from Chomsky (1977). where several kinds of constructions akin to purpose clauses are identified. Within GB. this analysis is so widely accepted that it deserves to be considered the standard theory. This book, then. is about a few syntactically peripheral ell~ments that have enjoyed a relatively long-lived. virtually universally accepted. theoretical treatment What is perhaps an obvious question arises in this context. Why write a GB book about purpose clauses? This book. I hope, will supply an interesting answer. Simply put. purpose clause:;, and related constructions, have various properties that are not accounted for in the standard theory. In this book I propose an alternative analysis of purpose clauses, an analysis from which. I think. more of their properties follow more naturally.

In comparative syntax a general approach has been pursued over the past decade predicated on the notion that Universal Grammar allows of open parameters, and that part of the job of linguistic theory is to specify what values these parameters may have, and how they may be set, given primary linguistic data, to determine the grammars of particu­ lar languages. The papers presented in this volume are also concerned with language variation understood in this way. Their goals, however, do not strictly fall under the rubric of comparative syntax, but form part of what is more properly thought of as a comparative semantics. Semantics, in its broadest sense, is concerned with how linguistic structures are associated with their truth-conditions. A comparative semantics, therefore, is concerned with whether this association can vary from language to language, and if so, what is the cause of this variation. Taking comparative semantics in this way places certain inherent limitations on the search for the sources of variability. This is because the semantic notion of truth is universal, and does not vary from language to language: Sentences either do or do not accurately characterize what they purport to describe. ! The source of semantic variability, therefore, must be somehow located in the way a language is structured.

The articles in this volume grew from papers presented at the workshop on control held at the Massachusetts Institute of Technology, March 1989. The work of the various authors comes at a moment in linguistic theory that is notable for two developments. First, there has been increasing involvement of syntactic theory in semantics and of semantic theory in syntax, with the result that the sorting of facts into syntactic and semantic has become a more complex and theory-laden affair. Second, there has been an enormous growth both in the breadth and depth of studies in languages other than English. Both of these develop­ ments have left their mark on the authors, directly and indirectly. They have also been responsible for the shifts that have given the key terminology its present range of application. In this introduction we discuss the background to the issues that were particularly prominent both at the workshop and in the authors' final drafts. We also com­ ment on the spirit of inquiry that they represent. Our goal is to provide some orientation to the specific contents of the essays and to supply material for reflection on a set of problems that will doubtless develop and deepen as rapidly in the foreseeable future as they have in the recent past.

This book integrates the research being carried out in the field of lexical semantics in linguistics with the work on knowledge representation and lexicon design in computational linguistics. Rarely do these two camps meet and discuss the demands and concerns of each other's fields. Therefore, the book is interesting in that it provides a stimulating and unique discussion between the computational perspective of lexical meaning and the concerns of the linguist for the semantic description of lexical items in the context of syntactic descriptions. The volume is divided into three sections. Part one deals with the fundamentals of lexical structure and what constitutes a lexical semantic representation for a grammatical theory. Part two deals with the specific mappings that are necessary from a lexical semantic representation to the syntax. Finally, Part three addresses the possibility of modeling lexical knowledge with computational uses in mind. The unifying assumption behind all these works is that methodologies in computational linguistics and theoretical linguistics have converged, and that the goals of both fields are the same: to investigate the phenomenon of language in a rigorous and analytic way, for the purpose of establishing how it fits into the larger domain of intelligent human behavior.

Reference to Abstract Objects in Discourse presents a novel framework and analysis of the ways we refer to abstract objects in natural language discourse. The book begins with a typology of abstract objects and related entities like eventualities. After an introduction to `bottom up, compositional' discourse representation theory (DRT) and to previous work on abstract objects in DRT (notably work on the semantics of the attitudes), the book turns to a semantic analysis of eventuality and abstract object denoting nominals in English. The book then substantially revises and extends the dynamic semantic framework of DRT to develop an analysis of anaphoric reference to abstract objects and eventualities that exploits discourse structure and the discourse relations that obtain between elements of the structure. A dynamic, semantically based theory of discourse structure (SDRT) is proposed, along with many illustrative examples. Two further chapters then provide the analysis of anaphoric reference to propositions VP ellipsis. The abstract entity anaphoric antecedents are elements of the discourse structures that SDRT develops. The final chapter discusses some logical and philosophical difficulties for a semantic analysis of reference to abstract objects.For semanticists, philosophers of language, computer scientists interested in natural language applications and discourse, philosophical logicians, graduate students in linguistics, philosophy, cognitive science and artificial intelligence.

The goal of this book is to integrate the research being carried out in the field of lexical semantics in linguistics with the work on knowledge representation and lexicon design in computational linguistics. Rarely do these two camps meet and discuss the demands and concerns of each other's fields. Therefore, this book is interesting in that it provides a stimulating and unique discussion between the computational perspective of lexical meaning and the concerns of the linguist for the semantic description of lexical items in the context of syntactic descriptions. This book grew out of the papers presented at a workshop held at Brandeis University in April, 1988, funded by the American Association for Artificial Intelligence. The entire workshop as well as the discussion periods accom­ panying each talk were recorded. Once complete copies of each paper were available, they were distributed to participants, who were asked to provide written comments on the texts for review purposes. VII JAMES PUSTEJOVSKY 1. INTRODUCTION There is currently a growing interest in the content of lexical entries from a theoretical perspective as well as a growing need to understand the organization of the lexicon from a computational view. This volume attempts to define the directions that need to be taken in order to achieve the goal of a coherent theory of lexical organization.

Preface This book is about semantics and logic. More specifically, it is about the semantics and logic of natural language; and, even more specifically than that, it is about a particular way of dealing with those subjects, known as Discourse Representation Theory, or DRT. DRT is an approach towards natural language semantics which, some thirteen years ago, arose out of attempts to deal with two distinct problems. The first of those was the semantic puzzle that had been brought to contempo­ rary attention by Geach's notorious "donkey sentences" - sentences like If Pedro owns some donkey, he beats it, in which the anaphoric connection we perceive between the indefinite noun phrase some donkey and the pronoun it may seem to conflict with the existential meaning of the word some. The second problem had to do with tense and aspect. Some languages, for instance French and the other Romance languages, have two morphologically distinct past tenses, a simple past (the French Passe Simple) and a continuous past (the French Imparfait). To articulate precisely what the difference between these tenses is has turned out to be surprisingly difficult.

At a superficial examination, English has different types of nominals with similar meaning and distribution: (1)a. John's performance ofthe song b. J ohn' s performing of the song c. John's performing the song d. the fact that John performs the song These nominals are also perceived by English speakers to be related to the same sentential construction: (2) John performs the song A more accurate inspection reveals, however, that the nominals in (1) differ both in their distribution and in the range of interpretations they allow. An adequate theory of nominalization should explicate rigorously how nominals of the types in (1) are related to sentential construction (2), and should also account for their distributional differences and meaning differences. The task of this book is to develop such a theory. I defend two main theses. The first is that, in order to provide an adequate semantics for the nominals in (1), one needs to distinguish among three types of entities in the domain of discourse (in addition to the type of ordinary individuals): events, propositions, and states xiii XIV PREFACE of affairs. I argue that the nominals in (1) differ in their ability to denote entities of these types and that predicates differ in their ability to select for them.

All work is work in progress. The ideas developed in this work could be (and probably will be) developed further, revised, and expanded. But it was time to write them down and send them out. Some of these ideas about linking had their origins in my 1987 dissertation. However, this work has grown beyond the dissertation in a number of important ways. The most important of these advances lie in, first, articulating aspectual roles as linguistic objects over which lexical semantic phenomena can be stated, and over which linking generalizations are stated; second, recognizing that syntactic phenomena may be classified as to whether or not they are sensitive to the core event of event structure; and third, recognizing the modularity of aspectual and thematic/conceptual structure, and associating that modularity with a difference between language-specific and universal language generalizations. The three chapters of the book are organized around these ideas. I have tried to state these ideas as strong theses. Where they make strong predictions I have meant them to do so, as a probe for future research. I hope that other researchers will take up the challenge to investigate, test and develop these ideas across a wider realm of languages than I --as one person --can do.

All work is work in progress. The ideas developed in this work could be (and probably will be) developed further, revised, and expanded. But it was time to write them down and send them out. Some of these ideas about linking had their origins in my 1987 dissertation. However, this work has grown beyond the dissertation in a number of important ways. The most important of these advances lie in, first, articulating aspectual roles as linguistic objects over which lexical semantic phenomena can be stated, and over which linking generalizations are stated; second, recognizing that syntactic phenomena may be classified as to whether or not they are sensitive to the core event of event structure; and third, recognizing the modularity of aspectual and thematic/conceptual structure, and associating that modularity with a difference between language-specific and universal language generalizations. The three chapters of the book are organized around these ideas. I have tried to state these ideas as strong theses. Where they make strong predictions I have meant them to do so, as a probe for future research. I hope that other researchers will take up the challenge to investigate, test and develop these ideas across a wider realm of languages than I --as one person --can do.

Part I of this book presents a theory of modal metaphysics in the possible-worlds tradition. "Worlds" themselves are understood as structured sets of properties; this "Ersatzist" view is defended against its most vigorous competitors, Meinongianism and David Lewis' theory of existent concrete worlds. Related issues of essentialism and linguistic reference are explored. Part II takes up the question of lexical meaning in the context of possible-world semantics. There are sceptical analyses of analyticity and the notion of a logical constant; and an "infinite polysemy" thesis is defended. The book will be of particular interest to metaphysicians, possible-world semanticists, philosophers of language and linguists concerned with lexical semantics.

Part I of this book presents a theory of modal metaphysics in the possible-worlds tradition. `Worlds' themselves are understood as structured sets of properties; this `Ersatzist' view is defended against its most vigorous competitors, Meinongianism and David Lewis' theory of existent concrete worlds. Related issues of essentialism and linguistic reference are explored. Part II takes up the question of lexical meaning in the context of possible-world semantics. There are skeptical analyses of analyticity and the notion of a logical constant; and an `infinite polysemy' thesis is defended. The book will be of particular interest to metaphysicians, possible-world semanticists, philosophers of language, and linguists concerned with lexical semantics.

This volume of papers grew out of a research project on "Cross-Linguistic Quantification" originated by Emmon Bach, Angelika Kratzer and Barbara Partee in 1987 at the University of Massachusetts at Amherst, and supported by National Science Foundation Grant BNS 871999. The publication also reflects directly or indirectly several other related activ­ ities. Bach, Kratzer, and Partee organized a two-evening symposium on cross-linguistic quantification at the 1988 Annual Meeting of the Linguistic Society of America in New Orleans (held without financial support) in order to bring the project to the attention of the linguistic community and solicit ideas and feedback from colleagues who might share our concern for developing a broader typological basis for research in semantics and a better integration of descriptive and theoretical work in the area of quantification in particular. The same trio organized a six-week workshop and open lecture series and related one-day confer­ ence on the same topic at the 1989 LSA Linguistic Institute at the University of Arizona in Tucson, supported by a supplementary grant, NSF grant BNS-8811250, and Partee offered a seminar on the same topic as part of the Institute course offerings. Eloise Jelinek, who served as a consultant on the principal grant and was a participant in the LSA symposium and the Arizona workshops, joined the group of editors for this volume in 1989.

Plurality, Conjunction and Events presents a novel theory of plural and conjoined phrases, in an event-based semantic framework. It begins by reviewing options for treating the alternation between `collective' and `distributive' readings of sentences containing plural or conjoined noun phrases, including analyses from both the modern and the premodern literature. It is argued that plural and conjoined noun phrases are unambiguously group-denoting, and that the collective/distributive distinction therefore must be located in the predicates with which these noun phrases combine. More specifically, predicates must have a hidden argument place for events; the collective/distributive distinction may then be represented in the part/whole structure of these events. This allows a natural treatment of `collectivizing' adverbial expressions, and of `pluractional' affixes; it also allows a unified semantics for conjunction, in which conjoined sentences and predicates denote groups of events, much like conjoined noun phrases denote groups of individuals.

This extended collection of papers is the result of putting recent ideas on quantification to work on a wide variety of languages. A central perspective of many of the papers follows the recognition of two broad types of quantificational strategies, one associated with nominal structures and determiners, the other with adverbial and other non-nominal expression ('D-quantifiers' and 'A-quantifiers'). The papers demonstrate both the unity and the variety of natural language quantificational forms and meanings. Many of the papers also shed new light on questions of language typology and syntactic and morphological variation. The languages discussed include English, Dutch, Italian, American Sign Language, Hindi, and a number of languages of Australia, Greenland, and the Americas. These comparative studies provide initial data for a typology of quantificational structures in natural languages, with important implications for the study of universal grammar. The book consists of research papers aimed at linguists, philosophers, and psychologists interested in semantics and linguistic form.An introduction presents a sketch of the background of this research and some of the central issues discussed, with pointers toward the included papers.

Semantic underspecification is an essential and pervasive property of natural language. This monograph provides a comprehensive survey of the various phenomena in the field of ambiguity and vagueness. The book discusses the major theories of semantic indefiniteness, which have been proposed in linguistics, philosophy and computer science. It argues for a view of indefiniteness as the potential for further contextual specification, and proposes a unified logical treatment of indefiniteness on this basis. The inherent inconsistency of natural language induced by irreducible imprecision is investigated, and treated in terms of a dynamic extension of the proposed logic. The book is an extended edition of a German monograph and is addressed to advanced students and researchers in theoretical and computational linguistics, logic, philosophy of language, and NL- oriented AI. Although it makes extensive use of logical formalisms, it requires only some basic familiarity with standard predicate logic concepts since all technical terms are carefully explained.

Temporal Logic: From Ancient Ideas to Artificial Intelligence deals with the history of temporal logic as well as the crucial systematic questions within the field. The book studies the rich contributions from ancient and medieval philosophy up to the downfall of temporal logic in the Renaissance. The modern rediscovery of the subject, which is especially due to the work of A. N. Prior, is described, leading into a thorough discussion of the use of temporal logic in computer science and the understanding of natural language. Temporal Logic: From Ancient Ideas to Artificial Intelligence thus interweaves linguistic, philosophical and computational aspects into an informative and inspiring whole.

Tense, Attitudes, and Scope is a model-theoretic inquiry into the semantics of tense in natural language. The book presents the view that the semantic contribution of tense is made in relation to structurally higher expressions (the `relative tense theory') and argues against the view that tenses are all indexicals. This idea is formally encoded as a de se analysis of attitudes, originally proposed by Lewis, coupled with a sequence-of-tense rule posited for English. An auxiliary proposal is made to account for some exceptional cases (e.g. so-called double-access sentences), which invokes de re attitudes about temporal entities (states or intervals). Since the proposed account assumes that the interpretation of tense is structure-dependent, it also correctly predicts scope interactions between tenses and NPs. Tense, Attitudes, and Scope is intended for scholars and graduate students in formal semantics, syntax-semantics interface, philosophy of language and Japanese linguistics.