Читать книгу A Companion to Chomsky - Группа авторов - Страница 18

Chomsky's work on grammar sees language as a bridge between sound and meaning, and has always aimed to explain certain facts about linguistic meaning. “Meaning,” of course, is a controversial and polysemous term, and the different entries in this section address some of the very different concerns to which it is attached.

One issue that has always been central to generative grammar is why certain readings of sentences are and others are not possible for certain sentences. For example, why can The man called the woman from Montana mean “The man called the woman, who was from Montana,” and “The man called, from Montana, the woman,” but not “The man, who was from Montana, called the woman”? And why are superficially very similar sentences understood so differently (John is easy to please entails It is easy to please John, but John is eager to please does not entail *It is eager to please John)? Equally, why do certain distinct strings have related or identical meanings (as with an active sentence and the related passive, for example)? Chomsky and other generativists provide answers in terms of underlying hierarchical sentence structures and constraints which explain both which structures can be generated and transformed, and how their constituents relate to each other: that is, by providing a syntactic theory. It is in this respect that generative syntax can be said to be essentially concerned with meaning.

Chomsky has, however, always opposed the widely held functionalist view of language according to which the purpose of language is communication, and the associated methodology that seeks to explain syntactic facts –which kinds of configurations of linguistic items are possible– in terms of semantic function, i.e., what they are used for. (Such views are discussed in Newmeyer's chapter in section 3.) Chomsky's view is that if language has any purpose at all, it is the expression of thought. He notes that many, indeed the vast majority of grammatical sentences, are hardly usable, because they are too long – e.g. example (2), or hard to parse (3), or combine words that are syntactically but not semantically compatible (4), for example:

1 (2) John and Mary and […+ two million conjuncts…] went to the party.

2 (3) Mice cats dogs chase bite squeak.

3 (4) Colorless green ideas sleep furiously.

Chomsky is, however, also skeptical about much work in linguistic semantics, in particular formal truth‐conditional semantics – at least if it is taken in full metaphysical seriousness, involving commitments to extralinguistic objects, as it often is in contemporary philosophy of language – since it treats word‐meanings as denotational (e.g. water denotes H₂O, and London denotes the city of that name) and sentence meanings as involving truth conditions. Influenced on this point by the later Wittgenstein, JL Austin and Peter Strawson among others, he argues that it is language users who refer, not the words themselves considered in abstraction from use. Sentences should not be seen as contributing to meaningful propositions that are true or false by themselves independently of usage, but rather constraints on the thoughts that they can be used to express (when used literally).

In addition to these influences from ordinary language philosophers, and Chomsky's rejection of formal semantics, his conception of grammar as a system of computations over representations has been influenced by the formal, “analytic” tradition in the philosophy of language exemplified by the work of Frege, Russell, and the early Wittgenstein, as John Collins discusses in Chapter 24. Of course, for Chomsky the formal system is intended as a description of a particular aspect of cognition, the language faculty. Collins argues that it is a significant achievement of Chomsky's to propose that “language is its own thing, an object of interest regardless of its poor design relative to the ends to which formal systems are developed,” and to show how an explicit theory of that object can be provided.

Paul Pietroski's Chapter 25 explains in some detail the centrality to generative grammar of accounting for different readings in terms of underlying structures (discussed above). He also sets out some of the currents in modern philosophy of language that Chomsky opposes, in particular the truth‐conditional, referential conception of linguistic meaning found in the work of Donald Davidson and David Lewis, and Hilary Putnam's semantic externalism. He explains Chomsky's challenges to these views and shows how they point the way to an alternative, internalist conception of meaning.

In Chapter 26, Michael Glanzberg explores the influence that Chomsky has had on natural language semantics. He explains the extent and the specific targets of Chomsky's skepticism about semantics and the related issue of whether syntax is in some important sense ‘autonomous’ of semantics. He notes that much of the leading research on truth‐conditional semantics is generativist and indebted to Chomsky's methodology in syntax. But he goes on to explain Chomsky's reasons for skepticism about the foundations of such work, which seems to him to be committed to the existence of referents of words such as foible and average family, as when we talk about “the foibles of the average family.” Chomsky's point is not that it's impossible to devise workarounds for such problems, but that they suggest that, as we've already noted, the mechanisms of language in abstraction from its use don't involve reference, so linguistic semantics should be recast. Glanzberg also discusses what Chomsky sees as a more productive line of enquiry: the study of features within the I‐language, which both have syntactic effects and encode aspects of meaning. Chomsky's suggestions here have helped to foster the now thriving field of lexical semantics.¹⁰

As mentioned above, Chomsky's view is that truth and reference are not properties of linguistic expressions but of speech acts performed by language users. He has also stressed what he calls “the creative aspect of language use”: the ability to produce and understand sentences in ways that are appropriate to circumstances but not determined by them, an ability that is in part explained by the generative, recursive nature of human language which enables us to produce and parse an indefinite number of novel sentences. So he might be thought sympathetic to the study of communication and utterance meaning in context: linguistic pragmatics. Indeed, in the context of his skepticism about truth‐conditional linguistic semantics, he says that “[i]t is possible that natural language has only syntax and pragmatics” (Chomsky 1995a, 26). But he is also skeptical about the prospects for pragmatic theory. In Chapter 27, Nicholas Allott and Deirdre Wilson examine his reasons and argue that they do not rule out systematic work. On the contrary, they claim, Chomsky's work provides a blueprint for the study of mental systems in terms of an explicit theory of their proprietary principles, and they argue that work in theoretical and experimental pragmatics implements these recommendations in investigating the system that generates interpretations for overt communicative acts.