In Locality in minimalist syntax, Stroik proposes that displacement in language results fromRemerge of SyntacticObjects (SO) that contain unchecked features. Stroik attempts to demonstrate how his theory accounts for a variety of recalcitrant data, primarily involving wh-constructions. The book has some shortcomings with respect to organization—for example, repetition of essentially the same analyses in different chapters, discussion of a single topic spread across different chapters, etc. There are also enough typos to occasionally create confusion. However, this book presents several interesting proposals regarding the structure of the human language faculty that are worthy of examination.
In Chapter 1, “Optimal design for human language”, Stroik discusses problems with Minimalist Program (MP) (Chomsky 1995) accounts of displacement (e.g., reliance on an EPP feature, reliance on economy conditions, etc.), presents his new proposal, survive, and shows how it overcomes these problems. survive is presented as a purely derivational theory (Brody 2002), since once elements areMerged into a derivation they can no longer be accessed by further operations. Stroik accomplishes this feat via the proposal that an SO can be copied multiple times from a Numeration. Another proposed advantage of survive is that it is crash-proof (Frampton and Gutmann 2002), since ill-formed constructions are underivable.
In Chapter 2, “The survive Principle”, Stroik elaborates on the nature of survive (1)while attempting to demonstrate the advantages of this proposal over featuredriven “Greed-based” and “Attract-based” (p. 37) movement analyses of the MP. [End Page 156]
1. The survive Principle:
If Y is an SO in an XP headed by X and Y has an unchecked feature incompatible with (i.e., cannot potentially be checked by) the features of X, Y must Remerge from the WorkBench with the next head Z that c-commands XP. (p. 45)
According to survive, Lexical Items originate in a Numeration and are merged together in a workspace. The WorkBench is “the union of the WS [worskpace] and the Numeration” (p. 44). An SO remains in the WorkBench, from where it can be copied and remerged multiple times into the derivational workspace until all of its features are checked. If an SO has any features that are not compatible with a head that it is Merged with, then the SO survives and must Remerge with the next head that is Merged into the derivation. For example, in (2), survive propels the subject DP they both to repeatedly Remerge from its base position with all subsequent heads until it arrives in its final position where all of its features are checked. The possibility of quantifier floating of both is given as evidence for this proposal.
2. They (both) were (both) expected (both) to (both) have (both) been (both) elected to the Senate. (p. 39)
An important component of survive is that if an SO is not incompatible with a local head, where feature compatibility is “a could-potentially-be-checked-by relation” (p. 37), the SO cannot undergo Remerge. In (3a), with the simplified derivation in (3b), both wh-phrases undergo Remerge with every higher head until they arrive in the embedded C. C contains feature-checking features that check the wh-features of where. These feature-checking features on C then become deactivated; however, even though these feature-checking features are deactivated, they are not incompatible with what, since they are the type of features that check wh-features. As a result, what is not repelled and the derivation stalls.
|(3)||a. *Where did Chris tell you what to put||(p. 41)|
|b. [CP what [where [C [TP to [VP put (what) (where)]]]]]||(p. 42)|
Stroik goes on to explain how survive accounts for Super-Raising and some multiple wh-constructions.
In Chapter 3, “Some Wh puzzles”, Stroik demonstrates how survive accounts for a variety of wh-construction data, primarily but not exclusively in English. Stroik relies heavily on features of a wh-phrase (4), which he claims are subfeatures of a [WH] feature.
a. [OP]: operator feature
b. [REF]: referential feature
c. [REF/WH]: “reference-dependent-on...