Date: Mon, 12 Dec 2005 14:24:08 +1100 From: Andrea Schalley Subject: Semantics: Critical Concepts in Linguistics, Vol. I-III
EDITOR: Gutiérrez-Rexach, Javier TITLE: Semantics SUBTITLE: Critical Concepts in Linguistics, 6 volumes SERIES: Critical Concepts in Linguistics PUBLISHER: Routledge (Taylor and Francis) YEAR: 2003
Andrea C. Schalley, School of Languages, Cultures and Linguistics & Language and Cognition Research Centre, University of New England (Australia)
[This is the first part of a two part review, covering Volumes I - III. The evaluation may be found in the second part. -- Eds.]
This set of six volumes is a collection of 101 reprinted extracts, chapters and articles from various monographs, dissertations, book collections, journals, and conference proceedings. The contributions, which are all from the field of formal semantics, first appeared between 1892 and 1999, with one contribution from 2002 that is yet to appear. They are grouped according to topics (not chronologically), an arrangement that has proven successful and done justice to the depth and breadth of formal semantics. The volumes each have an overall theme - Volume I: Foundational Issues, Volume II: Generalized Quantifiers and Scope, Volume III: Noun Phrase Classes, Volume IV: The Semantics of Predicates and Inflection, Volume V: Operators and Sentence Types, and Volume VI: Discourse and Dynamics - and apart from Volume II all of them are further divided into thematic parts (which are listed in the synopsis of each volume).
The collection starts with a table of contents of all six volumes, a chronological table of all of the reprinted articles and chapters, and a list of all contributors to the volumes. The preface to the collection gives a brief general historical overview of (formal) semantics, touches on current developments in the field, gives an overview of the anthology and indicates the rationale for the selection of the contributions. The selection is based on two main criteria: on the one hand cornerstones or classics are included, and on the other hand contributions published in conference proceedings of narrow diffusion are reprinted. The editor explicitly does not want this collection to be understood as an introduction to semantics. The level of technical difficulty of some papers is indeed high and the majority of the papers requires some acquaintance with the tools and methods of formal semantics.
In addition, each volume comprises acknowledgements (indicating the reprint permissions) and a short introduction introducing the papers of the respective volume. Volume VI closes with an index of the overall collection. In the following synopsis of the volumes, the author, title, and source for each contribution will be listed, along with the chapter number and page range it has in this collection. (Please note that the source information has been taken from the volumes and only two obvious errors have been corrected.)
VOLUME I - FOUNDATIONAL ISSUES [xxxix+352pp.]
The first volume of the collection contains 14 papers, divided into two parts. Part A, 'Truth and denotation' (4 contributions), comprises seminal works that can be seen as truly foundational papers of semantics in general and are concerned with denotation and referring. Part B 'Semantics and grammar' (10 contributions) deals with the question of how to integrate semantics as a component into grammar. The introduction to the volume (pp. 1-3) gives a brief overview of the papers and puts the contributions into context.
Part A - 'Truth and denotation'
The first paper (Chapter 1) by Gottlob Frege - ''On sense and reference'' (pp. 7-25; Source: The Philosophical Review 57, 1948, 207- 230, originally published in German in 1892) - is rightly seen as contribution founding the semantics discipline. Frege was the first to address meaning systematically, bringing in both compositionality as well as the distinction between sense and reference. According to Frege, the reference of a sentence is its truth value, an idea which is discussed at some length in the paper and which has proved highly influential in formal semantics.
Chapter 2 - Bertrand Russell's ''On denoting'' (pp. 26-38; Source: Mind 14, 1905, 479-493) - deals with what Russell calls ''denoting phrases'' such as 'the man', 'a man', 'some man', 'any man', 'every man', 'all men' etc. Thereby he is one of the first dealing with quantification, formulating the first theory of logical form and its relation to surface structure. Also, he proposes solutions to well- known puzzles of meaning (as, e.g., the famous 'The present King of France is bald'), utilising his philosophical theory of denoting.
In the well-written Chapter 3, ''On referring'' (pp. 39-60; Source: Mind 59, 1950, 320-344), Peter F. Strawson's aim is to show that Russell's account contains some fundamental mistakes. Strawson argues that Russell confuses sentences/expressions with the uses of sentences/expressions. He sees meaning as a function of the sentence or expression, whereas referring and truth or falsity are seen as functions of the use of the sentence or expression. In addition, the ''question of whether they [sentences, A.S.] are being used to make true or false assertions does not arise except when the existential condition is fulfilled for the subject term'' (p. 59). As a consequence, sentences such as 'The present King of France is bald' are not false (as Russell claims) but indeterminate, because we are dealing with a failure of reference for the definite description in the subject ('the present King of France').
Chapter 4, ''Extensions and intensions'' (pp. 61-69; Source: Meaning and Necessity; Chicago: University of Chicago Press, 1947, 23-52) by Rudolph Carnap, appears to be an extract from a monograph, which is in this case unfortunate, as the reader is not provided with some crucial definitions and important information regarding notation (although in the course of the extract one can infer the meanings for most of the notational terms). The extract reprinted here is the monograph's section after the one introducing the notions of 'intension' and 'extension'. These notions are extended in the present extract and applied to sentences. The extension of a sentence is its truth value, whereas the intension of a sentence is the proposition expressed by it.
Part B - 'Semantics and grammar'
Jerrold Katz and Paul Postal propose in Chapter 5, ''The semantic component'' (pp. 73-89; Source: An Integrated Theory of Linguistic Description; Cambridge, MA: MIT Press 1964, 12-29), a semantic component of a linguistic description that is taken to be a projective device in the sense of Katz and Fodor (1963). This projective device is to consist of two parts: the dictionary (providing meanings for lexical items) and a finite set of projection rules (assigning a semantic interpretation to each string generated by the syntactic component). The two parts of the semantic component are discussed in more detail in the present monograph extract, outlining both the components of dictionary entries and different types of projection rules. The paper rightly advocates that a semantic component necessitates rich lexical entries, readings of which are then, on the basis of syntactic structure and projection rules, combined to form the overall semantic interpretation.
Richard Montague starts Chapter 6, ''Universal grammar'' (pp. 90-109; Source: Theoria 36, 1970, 373-398), with the important claim: ''There is in my opinion no important theoretical difference between natural languages and the artificial languages of logicians; indeed, I consider it possible to comprehend the syntax and semantics of both kinds of languages within a single natural and mathematically precise theory.'' (p. 90) He then sets out to show this, providing a grammar for a fragment of English, which is semantically interpreted via translation into a system of intensional logic. This idea has been truly influential for formal semantics. As the first approach towards such a mathematically or logically universal grammar, it still has rather strong limitations: the fragment of English is limited and constructed, and as Montague states in Note 2 (p. 107), the basic aim of semantics for him is to characterise ''the notions of a true sentence (under a given interpretation) and of entailment''.
David Lewis' paper in Chapter 7, ''General semantics'' (pp. 110-153; Source: Synthese 22, 1970, 18-67), is not intended as a contribution to an empirical linguistic theory but to a theory of philosophy of language. He aims to provide a convenient format for semantics that is general enough to work for a great variety of logically possible languages. In attempting this, Lewis criticises Katz and Postal's semantics as 'Markerese Semantics' and suggests instead a categorially-based transformational grammar with a possible-worlds semantics. An editorial comment seems appropriate here: the references list of this paper is incomplete, stopping approximately half- way with letter L and John Lyons.
Chapter 8, Noam Chomsky's ''Deep structure, surface structure, and semantic interpretation'' (pp. 154-196; Source: Danny Steinberg and Leon Jakobovits (eds.), Semantics. An Interdisciplinary Reader in Philosophy, Linguistics, and Psychology; Cambridge: Cambridge University Press 1971, 183-216), is an very good contribution which - convincingly with respect to its examples - criticises semantically- based models of grammar. It is concerned with the relation of syntactic structure to semantic representation in generative grammar, stressing the contribution of the surface structure for delimiting meaning. In consequence, it suggests a reconstruction of the standard theory of grammar as earlier proposed, which apart from the deep structure takes other levels such as transformations and the structure determined by the phonological interpretation of the surface structure into account in semantic interpretation.
In Chapter 9, ''On generative semantics'' (pp. 197-224; Source: Danny Steinberg and Leon Jakobovits (eds.), Semantics. An Interdisciplinary Reader in Philosophy, Linguistics, and Psychology; Cambridge: Cambridge University Press 1971, 232-252), George Lakoff presents the alternative conception criticised by Chomsky in Chapter 8: Semantics is taken as foundational from a generative point of view - semantics plays a central role in syntax, and the role of transformations and derivational constraints in general is to relate semantic representations and surface structures. On this theoretical basis, Lakoff studies how quantified sentences are analysed in generative semantics.
Chapter 10, ''The proper treatment of quantification in ordinary English'' (pp. 225-244; Source: Jaako Hintikka, J. Moravcsik and Patrick Suppes (eds.), Approaches to Natural Languages. Proceedings of the 1970 Stanford Workshop on Grammar and Semantics; Dordrecht: Reidel 1973, 221-247), is the last paper written by Richard Montague before his death. Known by the acronym PTQ, it is well-known and has proved extremely influential. The paper presents a much more detailed treatment of another still somewhat limited fragment of English, which in the present paper also includes intensional predicates, intensional locutions with anaphoric pronouns, scope and quantification. The contribution follows an approach similar to Chapter 6: the syntax for the fragment is explicitly introduced, the semantics indirectly in that an artificial language (a tensed intensional logic) is set up and the English fragment is interpreted through a rigorous translation into this artificial language.
Barbara Partee suggests in Chapter 11, ''Some transformational extensions of Montague grammar'' (pp. 245-267; Source: Journal of Philosophical Logic 2, 1973, 509-534), to enrich Montague's rules in such a way that the theory can deal with reflexives, passives, passive agent deletions, tough movement, and subject and object raising. Apart from these valuable extensions, this easily accessible article was an important step in presenting Montague's views to linguists. It sketches Montague's PTQ paper, is equipped with an excellent introduction that outlines the broader picture, and it explains the assumptions and basic elements of the theory clearly.
Chapter 12, ''Logical form as a level of linguistic representation'' (pp. 268-299; Source: Logical Form: Its Structure and Derivation; Cambridge, MA: MIT Press 1985, 1-30), an extract from Robert May's 1985 monograph (which is based on his 1977 dissertation), discusses what the relation of a sentence's syntactic form to its logical form is. May proposes to include a level of representation, Logical Form (LF) derived from other linguistic levels. LF represents in this view the properties of syntactic form that are relevant to semantic interpretation or, put differently, those aspects of semantic structure which are expressed syntactically. This is seen to centre around three basic concerns: (i) the formal nature of the representation, (ii), how representations are derived, and (iii) constraints on well-formedness. Discussing these issues, he concludes that LF arises from and is restricted by very general mechanisms of Universal Grammar in the Chomskyan sense and supposedly is part of our innate endowment. It is worth noting that this is the first contribution which - at least to some extend - discusses a language other than English as object language, namely Chinese.
Richard Larson and Gabriel Segal's ''Knowledge of meaning and theories of truth'' (pp. 300-318; Source: Knowledge and Meaning; Cambridge, MA: MIT Press 1995, 25-42) in Chapter 13 is a well and clearly written book extract pursuing semantics as a theory of speaker knowledge. They follow Davidson's work who proposes the existence of a truth theory (for short, T theory), ''a deductive system that has the resources to prove something about the truth value of every sentence'' (p. 300). Central assumptions are that humans are designed to acquire a T theory, to treat any T theory they acquire as interpretive, and to learn a T theory that is in fact interpretive. After briefly outlining T theories (including a sample derivation and a discussion of the nontriviality of T theories), T theories are then introduced as theories of meaning, that is: the knowledge of meaning amounts to the knowledge of a T theory.
In the last contribution to the first volume (Chapter 14), ''Language, meaning, and interpretation: Chomsky against the philosophers'' (pp. 319-352; Source: [an adapted excerpt of a paper that is to appear elsewhere]), Carlos P. Otero explores some of the traits of Chomsky's current theory of meaning and its consequences for the philosophy of mind. It is well written and gives an short but interesting historical overview of the development of logic and the philosophy of mind and their relation to Chomsky's work. Some of the papers of this volume are briefly put into context (e.g., Strawson's paper in Chapter 3). Nevertheless, it is not fully clear why this biased contribution with its exposed study of Chomsky's ideas has been included in this semantics collection and, more specifically, in the present volume on foundational issues of formal semantics.
VOLUME II - GENERALIZED QUANTIFIERS AND SCOPE [viii+337pp.]
Volume II is the first of two volumes dealing with the semantics of noun phrases. The 13 papers of this volume specifically address the topic of generalised quantifiers and scope. In the Frege-Russellian works proper names were treated differently from quantificational elements (the former as constants, the latter with quantifiers). However, since they belong to the same syntactic category, following Montague's principle (cf. Chapter 10) they should have objects of the same type as denotations.
Such a systematic and uniform treatment of quantification in natural language is what the first contribution to Volume II, Chapter 15 by Jon Barwise and Robin Cooper entitled ''Generalized quantifiers and natural language'' (pp. 4-61; Source: Linguistics and Philosophy 4, 1981, 159-219), aims at. Barwise and Cooper propose to treat noun phrases uniformly as generalised quantifiers, referring to the traditional notion of generalised quantifiers in mathematical logic. This excellent paper introduces first the nature of generalised quantifiers and their relationship to English syntax in general, then develops a logic containing generalised quantifiers, shows how this logic may be formally related to a syntax fragment of English, and discusses some general implications of the notion of generalised quantifier for linguistics.
In Chapter 16, ''Semantic constraints on the English partitive construction'' (pp. 62-74; Source: Proceedings of the West Coast Conference on Formal Linguistics (WCCFL) 1; Stanford, CA: CSLI Publications 1982, 231-2442), William Ladusaw improves and builds on Barwise and Cooper's contribution (Chapter 15), extending their analysis to characterise the semantic difference between 'both' and 'the two'. He also discusses the phrase-internal semantics of English partitive noun phrases, closing with the remark that ''the semantics of the partitive construction cannot be completely reduced to quantification over contextually specified sets.'' (p. 73)
Johan van Benthem's technical paper in Chapter 17, ''Determiners and logic'' (pp. 75-105; Source: Linguistics and Philosophy 6, 1983, 447-478), is concerned with a semantic characterisation of all possible determiners (both in a 'global' and an 'inductive' sense), additional constraints for focusing upon the 'logical' determiners, and extensions of this study to arbitrary types in categorial grammar. It thereby achieves a generalisation of results obtained for logical determiners to arbitrary ones. The contribution presents a logical perspective on generalised quantifiers.
The purpose of Franciska de Jong and Henk Verkuyl's contribution in Chapter 18, ''Generalized quantifiers: the properness of their strength'' (pp. 106-126; Source: Johan van Benthem and Alice ter Meulen (eds.), Generalized Quantifiers in Natural Language; Dordrecht: Foris 1984, 21-43), is to provide a positive answer to the question whether semantic properties of noun phrases can be related to their syntactic properties. The authors suggest that this answer requires that partial interpretation of noun phrases are allowed and argue in favour of it. In the course of the paper, they propose to reduce the property of properness to the property of strength, revising Zwarts' (1981) classification of Dutch quantifiers. Their resulting quantification presented in this excellent and clearly written paper leads to a descriptively more adequate account of the data and also to a better understanding of the relation between the syntactic and semantic properties of determiners.
Chapter 19, ''Determiners and context sets'' (pp. 127-151; Source: Johan van Benthem and Alice ter Meulen (eds.), Generalized Quantifiers in Natural Language; Dordrecht: Foris 1984, 45-71) by Dag Westerståhl, addresses the issue of context dependence of determiners. Westerståhl distinguishes between the discourse universe, which is constant over pieces of discourse, and context sets, which are variable in discourse. Following this, determiners are seen as restricted to context-set variables, a relativisation that allows to treat definites and partitives uniformly. Barbara Partee presents excellent ideas in Chapter 20. Entitled ''Noun phrase interpretation and typeshifting principles'' (pp. 152-180; Source: Jeroen Groenendijk, Dick de Jongh and Martin Stokhof (eds.), Studies in Discourse Representation Theory and the Theory of Generalized Quantifiers; Dordrecht: Foris 1987, 115-143), this well-structured paper attempts a resolution between two approaches to noun phrase interpretation - one being Montague's uniform treatment, the other being authors distinguishing between referring, predicative and quantificational noun phrases (or uses of noun phrases). Partee argues that the insights of both sides are basically correct and mutually compatible. This is shown in the present paper, implemented through the idea of a type-driven interpretation and the use of several type-shifting rules involving lifting and lowering of types. In doing so, Partee keeps from Montague's approach the requirement of a systematic category-to-type correspondence. However, instead of requiring each syntactic category to correspond to a single type, she allows each category to correspond to a family of types.
Chapter 21, ''Polyadic quantifiers'' (pp. 181-209; Source: Linguistics and Philosophy 12, 1989, 437-464) by Johan van Benthem, is the first contribution dealing with multiple quantification, that is, dealing with constructions in which more than one quantifier occur. These are of particular difficulty, because questions of scopal dependence and independence come into play. In the present paper, van Benthem argues for the existence of non-reducible polyadic quantifiers and shows this by way of discussing several linguistic phenomena. He also suggests some conditions that characterise polyadic quantification in order to be able to distinguish these cases from unary Fregean iterations. The paper is highly technical, with not all abbreviations and notations explained, which makes it together with its many typos difficult for the reader to access.
Edward Keenan introduces in Chapter 22 ''Semantic case theory'' (pp. 210-232; Source: Renate Bartsch, Johan van Benthem and Peter van Emde Boas (eds.), Semantics and Contextual Expression; Dordrecht: Foris 1989, 33-56) the notion of 'semantic case' as a way of extending NP interpretations, an idea developed from the scope treatment in van Benthem (1986). A semantic case is an extension of a basic generalised quantifier function. The resulting 'semantic case theory' (which is introduced in the paper, together with its foundational axioms) essentially allows to account for the occurrence of quantifiers in different argument positions, and also allows to derive the narrow- scope and wide-scope interpretations of noun phrases. After introducing the theory, Keenan investigates its consequences and compares it with other approaches. The present paper is one of those taking language data from non-English languages into account, thereby adding to the data pool that needs to be explained - and it is often in the discussion of such data that semantic case theory proves superior to other approaches that were designed more or less exclusively to account for English.
Gila Sher in Chapter 23 discusses specifically those entities mentioned in her title ''Ways of branching quantifiers'' (pp. 233-260; Source: Linguistics and Philosophy 14, 1990, 393-422). Branching quantifiers, as polyadic quantifiers (cf. Chapter 21), cannot be reduced to linearly-ordered iteration. The aim of the paper is to develop a notion of branching quantification which is applicable not only to linguistics but also to philosophy and 'pure' logic (Sher prefers a wide notion of quantifier, which satisfies van Benthem's (1986) 'logicality' but not any other property attributed to natural language quantifiers). The paper was essentially part of her PhD thesis on 'Generalised Logic: A Philosophical Perspective with Linguistic Applications', which explains that linguistics is not in the focus (although this makes it questionable whether the article is enough 'linguistic' to appear in volumes on 'Critical Concepts in Linguistics').
Donka Farkas' Chapter 24 ''Quantifier scope and syntactic islands'' (pp. 261-267; Source: Proceedings of the Chicago Linguistics Society (CLS) 7, 1981, 59-66) is the first of the last four papers in the present volume to focus on scope, leading to the observation that not all natural language quantifiers are scopally equal. She claims that 'every' is not only island-bound but also clause-bound, whereas indefinite noun phrases are, on the other hand, island-free and clause- free.
In Chapter 25 by Fengh-Hsi Liu, ''Scope dependency'' (pp. 268-274; Source: Scope and Specificity; Amsterdam: John Benjamins 1998, 9- 15), the notion of scope dependency is introduced, describing how a noun phrase occurrence in an expression may be within the semantic scope of other expressions, in which case the noun phrase is scope dependent on the other expressions. Listing different observations in this monograph extract, Liu shows how noun phrases interact with other operators such as negation, predicates of various sorts, modals, and other quantified noun phrases. The availability of scope- dependent readings is sensitive to the type of noun phrases: ''Individual-denoting NPS don't induce dependency or depend on other NPs. Universally quantified NPs also don't depend on other NPs, although they induce dependency. Only when NPs other than these two classes are concerned do we get the possibilities of two dependent readings, the case that has been discussed most often in the literature.'' (p. 272)
In Chapter 26, ''Object wide scope and semantic trees'' (pp. 275-294; Source: Proceedings of Semantics and Linguistic Theory (SALT) 3; Ithaca, NY: CLC Publications, Cornell University 1993, 19-37), Dorit Ben-Shalom shows that object wide scope readings - in which object noun phrases have scope over subject noun phrases - are severely restricted, in contrast to what has often been assumed. Departing from Liu's (1990) work, Ben-Shalom claims that object wide scope readings are only possible when the object noun phrase is interpreted as a principal filter (the definition of which is given on p. 281). In addition (cf. p. 293), object wide readings are derived by a binary quantifier in which the interpretation of the object noun phrase defines a crucial domain, and this claim can be motivated within an algebraic approach to generalised quantifiers, which is called 'semantic trees'.
Anna Szabolcsi's ''Strategies for scope taking'' (pp. 295-337; Source: Anna Szabolcsi (ed.), Ways of Scope Taking; Dordrecht: Kluwer 1997, 109-154), the final chapter (Chapter 27) of Volume II, deals with the issue of scope dependence and its consequences at the syntax- semantics interface. Taking data from Hungarian as empirical basis, she proposes to fine-tune the structure of the Logical Form and to view it as the input for a component of semantic representation modelled on the discourse representation structures of Discourse Representation Theory (Kamp and Reyle 1993). The main concrete modification she suggests pertains to widening the class of discourse referents.
VOLUME III - NOUN PHRASE CLASSES [ix+446pp.]
The third volume of the collection comprises 19 papers, grouped around the topic of noun phrase classes. The volume is divided into two parts: Part A, 'Indefiniteness and definiteness' (11 contributions), and Part B, 'Plurals and mass nouns' (8 contributions), challenging subareas that have been debated a lot. The interpretation of indefinites cannot easily be achieved in a unified way because it is often depending on other elements in a clause; in particular, indefinites are not uniformly existential. Plural noun phrases are also not easily interpreted in a uniform way; they appear to have a number of readings which may depend on the presence of other elements such as certain modifiers or predicate types, for instance.
Part A - 'Indefiniteness and definiteness'
In Chapter 28, ''Adverbs of quantification'' (pp. 7-19; Source: Edward Keenan (ed.), Formal Semantics of Natural Language; Cambridge: Cambridge University Press 1975, 3-15), David Lewis focuses on the interpretation of adverbs of quantification such as 'always', 'usually', 'never', 'almost never' etc. He discusses as possibilities interpretations as (i) quantifiers over times, or (ii) quantifiers over events, or (iii) as quantifiers over cases, concluding that (iii) is correct. A case is regarded as the tuple of its participants, that is, cases are admissible assignments of values to the variables occurring free in the open sentence modified by the adverb. Adverbs of quantification unselectively bind the variables they have scope over. This has consequences for the interpretation of indefinite phrases, which hence get their quantificational force from the adverbs of quantification.
Lauri Karttunen's excellent Chapter 29 ''Discourse referents'' (pp. 20- 39; Source: James McCawley (ed.), Syntax and Semantics 7; New York: Academic Press 1976, 363-385) takes as its departure point the question 'When is there supposed to be an individual associated with an indefinite noun phrase?' Or put differently, when is there a discourse referent established by an indefinite noun phrase? (By 'establishing a discourse referent' it is meant that there may be a coreferential pronoun or definite noun phrase late in the discourse.) Karttunen is thus concerned with a particular feature a text interpreter must have according to him: it must be able to recognise when a novel individual is introduced by the input text and it must be able to store it along with its characterisation for future reference (cf. p. 20). There are several aspects of sentences that are important in determining whether an indefinite noun phrase establishes a discourse referent, which are discussed in more detail in the present chapter.
Chapter 30, Gary Milsark's ''Toward an explanation of certain peculiarities of the existential construction in English'' (pp. 40-65; Source: Linguistic Analysis 3, 1977, 1-29), is also a very fine contribution. Milsark shows that existential constructions ('there is/are' noun phrase) are restricted in that not all noun phrases can appear in the postcopular position. He distinguishes two types of determiners of noun phrases: 'weak' and 'strong' ones, with the latter characterised semantically as expressions of quantification, whereas the former are nonquantificational. Thereby, he explains the above mentioned restriction: existential sentences are themselves already understood as quantificational, which blocks noun phrases with strong and hence quantificational determiners in postcopular position, since this would result in double quantification. Furthermore, he notes that only strong noun phrases may appear as subjects of property-naming predicates.
Following Janet Fodor and Ivan Sag in Chapter 31, ''Referential and quantificational indefinites'' (pp. 66-107; Source: Linguistics and Philosophy 5, 1982, 355-398), indefinites are - semantically - inherently ambiguous. They propose two readings, a referential and a quantificational interpretation, and analyse a wealth of empirical data to support this proposal. Finally, they develop a formal semantics for referential indefinites, adapting Kaplan's formal treatment of demonstratives, due to the observation that a referential indefinite is more like a demonstrative than like a definite description.
Irene Heim sees neither definites nor indefinites as quantifiers in Chapter 32, ''File change semantics and the familiarity theory of definiteness'' (pp. 108-135; Source: Rainer Bäuerle, Christoph Schwarze and Arnim von Stechow (eds.), Meaning, Use and Interpretation of Language; Berlin: Walter de Gruyter 1983, 164-189). Reviving the familiarity theory of definiteness (a definite is used to refer to something that is already familiar at the current stage of conversation, whereas an indefinite is used to introduce a new referent), she deploys Karttunen's discourse referents (cf. Chapter 29) to develop a theory of conversation as file keeping, with meanings being understood as file-change potentials.
The primary purpose of the too broadly entitled Chapter 33, Edward Keenan's ''A semantic definition of 'indefinite NP''' (pp. 136-164; Source: Eric Reuland and Alice ter Meulen (eds.), The Representation of (In)definiteness; Cambridge, MA: MIT Press 1987, 286-317), is to define the class of English noun phrases that naturally occur in 'existential there' contexts with the existential reading. Keenan calls this class of noun phrases 'existential'. An existential noun phrase is headed by an existential determiner, which is always interpreted by an existential determiner function f (defined in the following way: a function f from properties to sets of properties is existential if and only if for all properties p,q holds: p is element of f(q) if and only if 1 is element of f(p&q), with 1 being the domain of evaluation). This proposal results from an analysis of determiners, and is then investigated with respect to its explanatory adequacy and compared to both Milsark's work and Barwise and Cooper's generalised quantifier proposal (cf. Chapter 15).
In Chapter 34, ''Existential sentences and predication'' (pp. 165-183; Source: Paul Dekker and Martin Stokhof (eds.), Proceedings of the Eight Amsterdam Colloquium; Amsterdam: ILLC, University of Amsterdam 1991, 601-621), Alessandro Zucchi revises several approaches to the definite restriction, including those reprinted in this collection: Barwise and Cooper (Chapter 15), Milsark (Chapter 30), and Keenan (Chapter 33). He proposes a presuppositional account, suggesting that 'there are/is' sentences are only felicitous in contexts that entail neither that the set denoted by the noun of the postverbal noun phrase is empty nor that it is not empty.
In Chapter 35, ''The semantics of specificity'' (pp. 184-211; Source: Linguistic Inquiry 22, 1991, 1-25), Mürvet Enç asks the question what it means for a noun phrase to be specific. She pursues an analysis of specificity that is independent of scope relation (and thus in particular gives up the hypothesis that specific noun phrases need to have wide scope) and independent of truth conditions. For her, specificity is a notion akin to partitivity, which can be covert or overt. Based on Turkish data (where indefinites in the object position are always unambiguously specific or nonspecific), she characterises the semantics of specific indefinites and explores partitive indefinites, the relationship between definiteness and specificity and between indefiniteness and specificity, and the semantics of 'certain'. The analysis she proposes predicts that there will be no non-specific definite noun phrases, and the analysis leaves the specificity of indefinites open.
Chapter 36, ''Deriving logical representations: a proposal'' (pp. 212- 222; Source: Indefinites; Cambridge, MA: MIT Press 1992, 1-11), appears to be the introduction to a monograph by Molly Diesing. She concentrates mainly on two questions: (i) What are the possible semantic interpretations of indefinite and quantificational noun phrases? and (ii) What role does the syntactic representation (in the framework of generative grammar) play in the derivation of the semantic representation of noun phrases? Her proposal is a mapping from the syntactic logical form into the logical representation as such. The resulting 'mapping hypothesis' suggests that material from the verb phrase is mapped into the nuclear scope of a quantificational operator, and material from the higher inflectional-phrase constituent is mapped into its restrictive clause. Since this extract is merely the introduction and starting point to Diesing's monograph, the made claims are not supported by empirical data or explained otherwise in detail. It is a bit unfortunate to have a contribution in the collection which - due to its provenance - only outlines and lists claims (and it boasts - in contrast to all other contributions of the whole collection - footnotes instead of endnotes).
In Chapter 37, ''Semantic incorporation: a uniform semantics for West Greenlandic noun incorporation and West Germanic bare plural configurations'' (pp. 223-242; Source: Proceedings of the Chicago Linguistic Society (CLS) 31, 1995, 171-186), Veerle van Geenhoven presents data which is not English focused (however, additional translations of the German data would have improved the readability): she primarily presents evidence from West Greenlandic showing that West Greenlandic noun incorporation and West Germanic bare plural configurations have important discourse semantic properties in common. The presented analysis regards this common denominator as 'semantic incorporation', which constitutes a mechanism or operation that derives the narrow scope of indefinite descriptions.
The last chapter in Part A, Chapter 38 ''Semantic universals and choice function theory'' (pp. 243-254; Source: Francis Corblin, Carmen Dobrovie-Sorin and Jean-Marie Marandin (eds.), Empirical Issues in Formal Syntax and Semantics; The Hague: Holland Academic Graphics 1999, 59-73), proposes a new implementation of the theory of choice functions in the analysis of indefinite noun phrases. Yoad Winter claims that in this reformulation of the theory, the restriction of existential closure to choice functions follows from the universals of generalised quantifier theory (and needs not be stipulated), which in his eyes motivates the choice-function theory linguistically.
Part B - 'Plurals and mass nouns'
Greg Carlson argues in Chapter 39, ''A unified analysis of the English bare plural'' (pp. 257-300; Source: Linguistics and Philosophy 1, 1977, 413-456), that the apparently distinct uses of the bare plural in English - the generic and the existential (or 'indefinite plural') use - are merely facets of a syntactically and semantically unified phenomenon. As he shows by way of his excellent data and argumentation, these two uses are complementarily distributed, with the distribution being predictable from the context. In other words, the bare plural in itself is never ambiguous in a given context, and a unified analysis is not only desirable but also necessary. Such a unified analysis is then proposed; it treats bare plurals as proper names of kinds (with kinds being thought of as abstract individuals), and ontologically distinguishes between individuals and stages of individuals. Another point convincingly discussed is that the 'null determiner' of a bare plural is not to be regarded as the plural of the indefinite article 'a'.
Chapter 40, Remko Scha's ''Distributive, collective and cumulative quantification'' (pp. 301-326; Source: Jeroen Groenendijk, Theo Jansen and Martin Stokhof (eds.), Truth, Interpretation and Information; Dordrecht: Foris 1984, 131-158), describes a treatment of quantification which accounts for a variety of readings of plural noun phrases, specifically for the distributive, collective, and cumulative readings. He abolishes the traditional dichotomy between distributive and collective verbs. This is achieved by deriving distributive readings from collective ones by means of meaning postulates. As a rather underexplored phenomenon, cumulative quantification is discussed. In order to generate cumulative readings, the also proposed grammar for a fragment of English can translate a sequence of noun phrases into one single quantifier, ranging over the Cartesian product of the extension of the nouns.
Godehard Link presents ''The logical analysis of plurals and mass terms: a lattice-theoretical approach'' (pp. 327-349; Source: Rainer Bäuerle, Christoph Schwarze and Arnim von Stechow (eds.), Meaning, Use and Interpretation of Language; Berlin: Walter de Gruyter 1983, 302-323) in the technical Chapter 41. He claims that the logic of plurals and the logic of mass terms share a lattice structure. As the only difference between the respective lattice structures he perceives that the former leads to an atomic structure while the latter does not. In addition, he introduces the 'star operator' (pluralisation operator), which allows for a compositional treatment of plural constructions. Further claims are that plural terms and collective terms are equivalent in that they are interchangeable in invariant context (which, however, does not make them co-referential), and that collective predication becomes possible in a unified way (due to the fact that many predicates are not marked with respect to distributivity).
In Chapter 42, ''The readings of plural noun phrases in English'' (pp. 350-369; Source: Linguistics and Philosophy 10, 1987, 199-219), Brendan Gillon addresses two questions: (i) What is the nature of the variation in construal to which plural noun phrases are liable? (Cf. the collective vs. distributive reading of, e.g., 'The men wrote operas.') (ii) What is the range within which these construals vary? With respect to the first question, he claims that the discrepancy between collective and distributive construals of plural noun phrases is a matter of ambiguity (not vagueness). With respect to the second question, he provides what he calls ''a more descriptively adequate characterisation of the range of variation'' (p. 350), a specification of what readings a plural noun phrase is susceptible to. For instance, he argues that the readings of English plural noun phrases in subject position bijectively correspond to the 'minimal covers' of the set denoted by the noun phrase.
Peter Lasersohn criticises Gillon's theory in Chapter 43, ''On the readings of plural noun phrases'' (pp. 370-374; Source: Linguistic Inquiry 20, 1989, 130-134), defending that verb phrases and not noun phrases are ambiguous and that verb phrases are two ways ambiguous (with a collective and a distributive reading).
In Chapter 44, ''Against groups'' (pp. 375-390; Source: Martin Stokhof and Leen Torenvliet (eds.), Proceedings of the Seventh Amsterdam Colloquium; Amsterdam: ILLC, University of Amsterdam 1989, 475- 493), Roger Schwarzschild discusses two contrasting approaches to plurals, the sums approach and the group approach. The former presupposes a domain of discourse having individual entities as well as plural entities (or sums), which correspond to sets of individuals. The latter presupposes a richer domain of discourse with entities corresponding to individuals and sets of individuals, but also to higher order sets. After outlining and investigating the approaches, he concludes in favour of the sums approach, which is ontologically simpler, and he draws on pragmatics to explain some of the examples that motivate the groups approach.
Chapter 45, ''On conceptional neuterality'' (pp. 391-414; Source: Linguistic Individuals; Stanford, CA: CSLI Publications 1991, 161-183), a book extract by Almerindo Ojeda (therefore unfortunately not providing enough of the overall context of the monograph), studies the linguistic category Otto Jespersen (1924) called 'the conceptional neuter'. Based on evidence from Spanish, he proposes that the conceptional neuter is [+Pronominal, +Individual], the class of pronouns whose mereological component is the entire universe of discourse. Ojeda 'expects' that his characterisation of the conceptional neuter of Spanish will apply with full crosslinguistic generality, so that it might be regarded as a category of universal grammar constituted by pronouns which can denote without regard to mereology.
Gennaro Chierchia's objective in Chapter 46, ''Partitives, reference to kinds and semantic variation'' (pp. 415-446; Source: Proceedings of Semantics and Linguistic Theory (SALT) 7 [not 4 as listed in the volume]; Ithaca, NY: CLC Publications, Cornell University 1997, 73- 98), is to provide a contrastive analysis of partitives in English and Italian. He sketches the Neo-Carlsonian view of bare plurals in English, which provides him with the starting point for his analysis. Full partitives are then analysed in a rather canonical way, whereas the analysis of bare particles is the novel aspect in this contribution. They are derived from the same structure as full partitives, involving the shifting of the meaning of the restriction of full partitives into an argument, with the result that a new determiner with an existential meaning is created. The absence of bare partitives in some of the Romance languages follows from blocking. This excellent paper is unfortunately not only full of typos that make it difficult to follow, but it disappointingly demonstrates that no final proof-reading has been carried out. On the pages 442-444 the text is broken three times (content-wise, not with respect to visual appearance) and chunks of text are displaced. It takes a while to figure out that ''The question, in'' (in the first paragraph on p. 442) is followed by ''more technical terms...'' (p. 443), that (45) a. and b. (p. 444) are actually followed by c. on p. 442, and that ''syntactic categories (and move'' (p. 443, in the text followed by ''more technical terms'', cf. above) should be followed by ''material to them)...'' (p. 444).
[This review continues in the next issue. -- Eds.]
REFERENCES (for both parts)
Benthem, Johan van 1986. Essays in Logical Semantics. Dordrecht: Reidel.
Groenendijk, Jeroen and Stokhof, Martin 1990. 'Dynamic Montague Grammar.' In: László Kálmán and László Polós (eds.), Papers from the Second Symposium on Logic and Language, Akadémiai Kiadó, Budapest, 3-48.
Jespersen, Otto 1924. The Philosophy of Grammar. London: George Allen and Unwin.
Kamp, Hans and Reyle, Uwe 1993. From Discourse to Logic. Dordrecht: Kluwer.
Katz, Jerrold J. and Fodor, Jerry A. 1963. 'The structure of a semantic theory.' In: Language 39, 170-210.
Lewis, David 1979. 'Attitudes de dicto and de re.' In: Philosophical Review 88: 513-543.
Liu, Fengh-Hsi 1990. Scope Dependency in English and Chinese. PhD diss., UCLA.
May, Robert 1977. The Grammar of Quantification. PhD diss., Cambridge, MA: MIT.
Rooth, Mats 1985. Association with Focus. PhD diss., University of Massachusetts, Amherst.
Zaefferer, Dietmar 2002. 'Polysemy, polyvalence, and linking mismatches. The concept of RAIN and its codings in English, German, Italian, and Spanish.' In: DELTA - Documentação de Estudos em Lingüística Téorica e Aplicada 18 (spe.), 27-56. Special Issue: Polysemy.
Andrea C. Schalley is Australian Research Council Postdoctoral Fellow at the University of New England, Australia. Trained in theoretical linguistics (PhD University of Munich, Germany, 2003), she is primarily interested in semantic representation frameworks, questions of modelling adequacy in theories about language, the relationship between language and cognition, and constraints on linguistic variability.