The aim of the present paper is to understand what the notions of explanation and prediction in contemporary linguistics mean, and to compare various aspects that the notion of explanation encompasses in that domain. The paper is structured around an opposition between three main styles of explanation in linguistics, which I propose to call ‘grammatical’, ‘functional’, and ‘historical’. Most of this paper is a comparison between these different styles of explanations and their relations (Sections 3, 4, 7, and 8). A second, more methodological aspect this paper seeks to clarify concerns the extent to which linguistic explanations can be viewed as predictive, rather than merely descriptive (Sections 2, 5, and 6), and the problem of whether linguistic explanations ought to be causal, rather than noncausal (Section 6). I argue that the notion of prediction is equally applicable in linguistics as in other empirical sciences. The extent to which the computational model of generative syntax can be viewed as providing a causal or psychologically realist model of language is more controversial (Sections 5–9).

Explanation and prediction in grammar (and semantics)

Explanation in Linguistics

https://www.onlinelibrary.wiley.com/doi/pdf/10.1111/mila.12286

(Implicit) Knowledge, reasons, and semantic understanding

OF DISSERTATION INFLATIONARY TRUTH-THEORETIC SEMANTICS I argue that satisfaction and reference—and therefore, truth—are multiply realizable properties. I advocate a novel approach motivated by a commitment to the robustness and fruitfulness of truththeoretic approaches to natural language semantics. DEFLATIONISM: Philosophers keen on deflating the metaphysical pretensions of truth theories claim that we need not appeal to a substantive truth-property. Recently, however, some philosophers have sought to combine deflationism about truth with the view that our concept of truth or the truth-predicate can play an important role in natural language semantics. TRUTH-THEORETIC SEMANTICS: The goal of a formal semantic theory of a natural language is to provide both the semantic values of that language’s lexically primitive items as well as the semantically significant modes of combining those basic elements into meaningful and more complex expressions. Most approaches have in common a commitment to finite stateability and compositionality as well as a commitment to something like Davidson’s “Convention T.” PLURALISM: Pluralists about truth argue that different areas of discourse have different truthproperties. Can pluralism successfully be combined with a commitment to truth-theoretic semantics? OPEN SEMANTIC FUNCTIONALISM: The pluralist approaches to truth are unsatisfactory for a variety of reasons. The only option, I argue, is to regard truth as multiply-realizable. Specifically, we should view the set of truth’s realizers as possessing non-actual members—as being “open.” Truth is defined in the usual way in terms of reference and satisfaction, but these latter two relations are to be understood as multiply realizable but open. The property of truth can be specified using the Ramsey/Lewis method. My final view—Open Semantic Functionalism— respects compositionality and finite stateability, avoids triviality, handles plurality, and fits with robust, explanatorily significant natural language semantic theories.

/pdf/inflationary-truth-theoretic-semantics-aexswgrp20.pdf

Inflationary truth-theoretic semantics

Davidson on truth and reference

There are three competing frameworks for answering the foundational questions of linguistic theory. Platonism holds that linguistics is about abstract entities, whose essential properties grammarians discover, by using nonempirical reasoning, as in mathematics. Nominalism takes linguistics to be about concrete physical tokens that comprise conventional systems of communication; grammars explain how inscriptions and the like can be, e.g., grammatical, co-referential, or contradictory. Cognitivism takes linguistics to be a branch of psychology, seeing grammars as hypotheses about the tacit knowledge that every competent speaker possesses. I argue that the epistemological side of the platonist position is undermined by W. V. Quine’s attack on the notion of nonempirical modes of inquiry. Jerry Katz contends that Quine’s epistemology is inconsistent, because it entails that principles of reasoning are simultaneously revisable and unrevisable. I show that Katz’ “revisability paradox” overlooks the distinction between our principles of reasoning and our theory of those principles. Drawing this distinction eliminates the threat of inconsistency. Further, I argue that linguists’ claims concerning the infinitude of language need not signal an ontological commitment to abstract entities. Rather, they reflect the lawlike, counterfactual-supporting character of linguistic generalizations, as well as a principled idealization away from mortality, memory constraints, and motivational factors.

Explanation and prediction in grammar (and semantics)

Citations

Explanation in Linguistics

(Implicit) Knowledge, reasons, and semantic understanding

Inflationary truth-theoretic semantics

Davidson on truth and reference

Cognitivism and Nominalism in the Philosophy of Linguistics

Related Papers (5)

Entailment rules in semantic theory.

Universal grammar : 15 essays

A grammar systems approach to natural language grammar

Semantics in generative grammar: A critical survey

Semantic constraints on utterance interpretation: complementary perspectives from Construction Grammar and Relevance Theory