Freezing in it-clefts

1. The Issue

Freezing is the cover term for the restriction on extraction from constituents in a derived position. The traditional Freezing cases are illustrated here with topicalization in (1a), heavy-NP shift in (1b), and extraposition in (1c).

1. (1).

   1. a. * Who_i did Mary say that [to t_i]_j she sent her manuscript t_j.

   2. b. ??Who_i did he donate t_j to a museum [a vase of t_i]_j?

   3. c. ??Who_i did you see [a picture t_j] yesterday [of t_i]_j?

Originally, Freezing has been analyzed as a syntactic restriction, namely the Freezing Principle (= FP) (Wexler and Culicover 1980).¹ The FP has been revived and redefined in more recent work (for example, see Rizzi 2006 assuming Criterial Freezing, Müller 2010, 2011 proposing extraction from last merged specifiers as impossible; see also Kitahara 1994, Takahashi 1994, Stepanov 2007, and references therein). Here, I use the term Freezing as an empirical generalization that A′-moved constituents are frozen; that is, extraction from these is blocked, independent of the precise explanation.² Thus, I take Freezing to be a diagnostic for A′-movement and apply it to it-cleft sentences. I report an empirical study testing whether the clefted constituent (= pivot) in it-cleft sentences is frozen. If it is, this suggests that the pivot is A′-moved. If not, this shows that the pivot is not A′-moved. This is relevant for the syntactic analysis of it-clefts, a construction that has received a large range of quite different analyses.

The squib is structured as follows. Section 2 introduces the syntactic analyses of it-clefts and explains how Freezing is relevant. Section 3 reports the results of a rating study, showing that sub-extraction from the pivot is indeed possible. Section 4 discusses the results in relation to the syntax of it-clefts. Section 5 concludes the squib.

2. Syntactic Approaches to it-Clefts

Research on the syntax of it-clefts can be divided into two major camps (see Huber 2002 and Reeve 2012 for more detailed overviews).

2.1 Monoclausal movement analyses

Monoclausal movement analyses propose that the pivot is extracted from the cleft clause to move into a focus-related projection (see Chomsky 1977, Rochemont 1986, Meinunger 1998, É. Kiss 1998:245), illustrated in (2).

1. (2).
   

2.2 Copular analyses

Copular analyses propose that it-clefts belong to the general class of copular clauses. They can be further distinguished by the type of predication they assume, namely predication versus specification (see Higgins 1979 for seminal work on the classification of copular clauses).

In predicational analyses, the pivot and the cleft clause are in a predicative relationship, while it is inserted as an expletive into the subject position (see Williams 1980, Delahunty 1982, Heggie 1988).

Specificational analyses assume that it-clefts share the major syntactic and semantic features of specificational copula clauses (= SCC), such as My best friend is Peter. They can be further subclassified with respect to how the three major elements — it, pivot and cleft clause — are combined to form a SCC. In one type of approach, it in subject position is linked to the cleft clause in extraposed position, either by movement (e.g., Akmajian 1970, Pinkham and Hankamer 1975, Percus 1997, Huber 2002) or indirectly through some interpretive process (see Gundel 1977; Hedberg 1990, 2000). A second group assumes that the cleft-clause syntactically modifies the pivot, whereas it is the subject of predication (see Reeve 2011, 2012) or an inverted predicate (see den Dikken 2013).

Concerning Freezing, the movement analyses predict that extraction from the pivot in it-cleft sentences is impossible, because the pivot is A′-moved to a focus position. Under the copular analyses, the pivot is not A′-moved, thus, extraction from the pivot should be available. It has been repeatedly reported that both extraction of and sub-extraction from the pivot seems available, (3a), though the status of (3b) is not entirely clear.

1. (3).

   1. a. Which book_i do you think it is t_i that every linguistics student should read?

      (den Dikken 2013:7)

   2. b. ?Who_i was it...