Derivations in Minimalism

Abstract

This pathbreaking study presents a new perspective on the role of derivation, the series of operations by which sentences are formed. Working within the Minimalist Program and focusing on English, the authors develop an original theory of generative syntax, providing illuminating new analyses of some central syntactic constructions. Two key questions are explored: first, can the Extended Projection Principle (EPP) be eliminated from Minimalist analysis without loss, and perhaps with a gain in empirical coverage; and second, is the construct 'A-Chain' similarly eliminable? The authors argue that neither EPP nor the A-chain is in fact a property of Universal Grammar, but rather their descriptive content can be deduced from independently motivated properties of lexical items, in accordance with overarching principles governing derivation. In investigating these questions, a range of new data is introduced, and existing data re-analyzed, presenting a pioneering challenge to fundamental assumptions in syntactic theory.

Some (not all) nonsententials are only a phase

“…Some types of nonsententials are generable without the derivation of a full sentential structure, while other (apparent) nonsententials require the derivation of a full sentential structure, the unpronounced portion of which is later elided. More precisely, I will argue that as a consequence of the architecture of the independently motivated phase-based model considered here (Epstein and Seely, 1999, 2002, 2006Epstein et al, 1998), certain subsentential objects converge on their own, without being embedded in a larger structure; however, others do not. When these types of subsentential objects are Spelled Out and evaluated by the interfaces, they contain no unvalued features (which are uninterpretable at the interfaces), and so do not crash.…”

“…The analysis of nonsententials proposed herein is couched within the Derivation by Phase (DBP) framework-specifically, within the variant of DBP proposed in Epstein and Seely (1999, 2002, 2006.…”

“…(Chomsky, 2001a:22) By hypothesis, vP and CP are the only phases because only they are relatively independent; they are propositional; and they are categories within which all theta roles borne by the head are discharged. Epstein and Seely (1999, 2002, 2006 and Epstein (forthcoming) argue that it is incorrect to single out vP and CP as phases, and that instead Spell Out must occur following every rule application; that is, following completion of each and every category in a derivation, not just vP and CP. These works, which contain arguments that the singling out of vP and CP as the only phases is both empirically and methodologically problematic, advance the hypothesis that all categories constitute phases.…”

“…For expository purposes, we refer to all these as ''phonological''.' For Epstein and Seely (2006), in contrast, Spell Out is not so much a stripping of a certain kind of feature from the narrow syntax, but the requirement that 'the derived representation is interpreted by both PF and LF' (p. 178). More precisely, 'LF and PF necessarily evaluate linguistic entities at every point in the derivation.…”

“…Thus, if X and Y are merged creating C, then C is necessarily input to both LF and PF, which interpret as much of C as possible, leaving a ''syntactic residue'' of material that may serve as input to subsequent derivational operations.' (Epstein and Seely, 2006) An important consequence of the proposed architecture is the possibility of 'nonfatal crashing'. While 'fatal crashing' (i.e., crashing under its usual conception) has the effect of halting the derivation once and for all, nonfatal crashing yields a much less severe result.…”