LINGUIST List 14.1519

Tue May 27 2003

Disc: Reply to Review: Dimensions of Movement

Editor for this issue: Karen Milligan <karenlinguistlist.org>

Directory

1. Sjef Barbiers, Review of Alexiadou et al. Dimensions of Movement

The discussion of my paper 'Remnant stranding and the theory of
Movement' in Jonathan White's review of Alexiadou et al. Dimensions of
Movement (Linguist 14.1360) contains some crucial mistakes and
misunderstandings that I would like to correct.

My paper primarily provides strong evidence for vP as an intermediate
landing site. Chomsky (1986, 2000, 2001) proposes on theoretical
grounds that vP is an intermediate landing site, but so far there was
hardly any evidence for this claim. The evidence provided in this
paper involves three different types of remnant stranding in
intermediate landing sites of long extraction. The example that White
gives involves short P-stranding which does not show anything about
stranding in intermediate landing sites.

The second major claim of my paper is that, at least in Dutch, the
left edge of an embedded clause is never an intermediate landing
site. Evidence for this claim is the fact that stranded material of
long extraction does not occur at the left edge of embedded clauses,
whereas it does occur at the edge of the matrix vP.

I give two reasons in the paper why stranding at the left edge of
embedded clauses is impossible:

(i) Propositional clauses are not phases, because they lack Force. As
a result, they do not trigger movement to their left edge, so neither
full constituents nor remnants can surface there. Contrary to what
White says, I do not argue that stranding is impossible with
propositional clauses; I claim that it is possible, but only at the
edge of matrix vP, not at the left edge of an embedded propositional
clause.

(ii) Factive clauses do have Force and are phases. Constituents may
therefore move to their left edge. However, they have to stay there
because extraction from factive clauses is impossible in Dutch. I
provide ample and partially novel evidence for this claim in this
paper (and in Barbiers (2000)) and conclude from this that factive
clauses are strong islands in Dutch because they are adjuncts. This
evidence is ignored by White who simply says that Melvold (1991) has
argued that factive clauses are not strong islands. As a consequence
of the general impossibility of extraction from factive clauses,
remnant stranding at the left edge of a factive clause and at the edge
of a matrix vP is impossible. According to White, the fact that no
material can be stranded at the left edge of factives weakens my claim
that factive clauses are phases. However, it seems to me that the fact
that constituents can move to the left edge of a factive clause but
have to stay there because of general conditions on movement is
sufficient evidence for the phase status of factive clauses.

Subject-Language: Dutch; Code: DUT