Robert Freidin

Exquisite connections : some remarks on the evolution of linguistic theory*

Abstract This article discusses the motivation for the shift from GB theory to the Minimalist Program. It explores notions of economy and conceptual naturalness in linguistics and the physical sciences — in particular, symmetries across levels of analysis. As an example of such symmetries, it proposes a new analysis for some Principle C effects. It concludes with some observations on the conduct of science.

Core Grammar, Case Theory, and Markedness

Core grammar (CG) is universal grammar’s substantive contribution to the grammar of a particular language. It consists of a set of general rules or rule schema (e.g. “move α”), conditions on rules (e.g. the recoverability condition for deletions), and filters (e.g. *[that [NP e]]) — all of which provide a severely limited set of possible grammars from which certain properties of language follow.1 The contents of the grammar are organized into various components, each with its specific properties. The various components interact modular fashion, in a way determined by the general organization of grammars. Following Chomsky and Lasnik (1977), we adopt the organization in (1).

A note on methodology in linguistics

Evans & Levinsons (E&Ls) critique of Universal Grammar fails because their methodology is flawed, as illustrated in their discussion of the Subjacency Condition. The lack of explicit analysis leads the authors to a false conclusion that is refuted by work published in this journal twenty years ago. They miss the point that unanalyzed data cannot disprove grammatical hypotheses.

Adult language acquisition and Universal Grammar

The current conception of the relation between UG and the grammar of a language rules out the no-access hypothesis, but does not distinguish between the full-access and partial-access hypotheses. The former raises the issue of why language acquisition in child and adult should be so different. The evidence presented in Epstein et al.s target article seems inconclusive regarding a choice between hypotheses.

Recursion in Generative Grammar

The term recursion has been used in generative grammar to describe two separate but related phenomena, one concerning the internal structure of linguistic expressions and the other, the computational mechanisms that generate this internal structure. From an internalist perspective, the latter is primary, while the former is apparently of secondary importance. This becomes clear when we consider the history of modern generative grammar, in particular the rationale behind the reduction of transformational grammar (i.e., phrase structure rules and transformations) to the single operation Merge. The recursive structure-building character of Merge accounts for the unbounded character of human language, involving a discrete infinity of interpretable expressions. To the extent that Merge is the sole structure-building operation in a grammar, it is a prime candidate for what is unique about language and languages. The criticisms of this proposal generally fail to make the appropriate distinction between the grammar in the mind of a speaker, a computational system and a lexicon (i.e., an I-language), and the set of linguistic expressions generated by the grammar. Given that former determines the properties of the latter, we also need to consider the role of the lexicon in this discussion of recursion.

Disjoint Reference and Wh-Trace 1981

This article explores the ramifications of an analysis of disjoint reference for the theory of core grammar.1 The analysis we will adopt accounts for coreference possibilities between pronouns and Wh-traces and, following May (1981), also provides an explanation for the COMP-to-COMP condition on Wh-Movement of Chomsky (1973). The central assumption of the analysis — that Wh-trace binding is exempt from the Propositional Island Condition (PIC) and the Specified Subject Condition (SSC) — allows for a simplification of the theories of binding and indexing, and also provides an argument that the Subjacency Condition is properly interpreted as a condition on representations rather than a condition on derivations.2