Minimality effects in syntax (review)

This book is a collection of papers focusing on the encoding of locality constraints in grammar. The editors offer an introductory section stating the basic questions that the data raises with respect to locality and the minimal link condition (MLC), and describe the phenomena examined in the different papers.

Elena Anagnostopoulou introduces the notions of minimal domain and equidistance in double-object constructions. These notions explain restrictions on the type of elements that the DP receiving nominative case can ‘skip’ over on its way to tense. Željko Bošcović discusses stylistic fronting (SF) in Icelandic and shows that related long-standing problems are accounted for if we assume a merger at the phonetic form (PF) interface level. The element targeted by SF moves to the specifier of a functional-affixal head above VP, which must merge under PF adjacency with a verb. Gisbert Fanselow considers the fact that the MLC is observed in all cases of head movement but faces numerous counterexamples in cases of wh-movement. He proposes that the MLC decides between syntactic objects that share the same partial logical form (LF). In cases of operator movement, information structure plays a role and thus the MLC cannot prevent the movement operation of a syntactic element over an intervener.

Susann Fischer investigates SF from a diachronic view introducing data from Old Catalan. SF is the result of movement of an item to the specifier of a projection above VP in order to check a strong V-feature. She identifies this projection as ∑P, the locus of sentential operators. This ties SF to the information structure of the sentence. Hubert Haider investigates restrictions on the crosslinguistic behavior of wh-in-situ and concludes that the MLC is inadequate to capture the variation observed. A set of four independent grammatical restrictions is proposed: the obligatory wh-element in situ is obligatorily an operator; moved and in-situ wh-elements cannot both range over higher-order semantic types; there is a domain requirement for semantic integration of an adverbial wh-element; and a biuniqueness restriction on the licensing of wh-in-situ. John Hale and Géraldine Legendre reconstruct the MLC as a set of violable constraints within optimality theory (OT). They argue that the inability of a remnant phrase to undergo the same kind of movement that the antecedent of the trace inside the remnant has undergone can be explained adequately in OT syntax if we assume constraints that measure the lengths of movement chain links by counting the maximal projections that they cross. Winfred Lechner reworks the MLC in order to simplify the theoretical assumptions that are related to the operations of Merge and Move. He proposes an adjustment of the MLC so that the effects of the economy principle of Merge over Move can be captured. Other properties of the MLC are reduced to aspects of the linear correspondence axiom. Hanjung Lee explores word-order freezing where movement of an object NP in front of a subject NP is blocked if the two NPs are of the same type, for example, have the same morphological case. Lee presents an account of these structures within an OT version of lexical functional grammar.

Gereon Müller points out that constraints like the MLC have no place in a purely derivational approach as they presuppose a large search space. He proposes that superiority effects with wh-movement in German and English can be derived from a strengthened version of the phase impenetrability condition, where the latter applies not to phases but to phrases. Geoffrey Poole and Noel Burton-Roberts return to stylistic fronting in Icelandic and related phenomena in Breton and propose a new understanding of the phonological component in which a set of ‘representational conventions’ that vary crosslinguistically play a role in mapping the...