how much do islands matter in sluicing? susumu kuno and soo-yeon kim since ross (1969) introduced the notions of ‘islands’ and ‘sluicing’
How much Do Islands Matter in Sluicing?
Susumu Kuno and Soo-Yeon Kim
Since Ross (1969) introduced the notions of ‘islands’ and ‘sluicing’
in the study of syntax, an extensive body of literature has been
produced on the visibility or invisibility of islands in sluicing
constructions. These studies, despite their divergence in resolution
of sluicing constructions with respect to their interpretations and
internal structures (if there is any), agree on the fact that the
presence of ‘islands’ somehow affects the acceptability status of
sluicing constructions, leaving debatable questions on why, how, and
how much it does. This paper focuses on data that have been claimed to
be island-sensitive in the literature to show that many of the
critical island-related data in sluicing are graded in nature and that
this gradient nature of the acceptability status makes it necessary to
incorporate in the analysis of sluicing non-structural factors that
control their acceptability status. One such factor relates to whether
the hearer is aware of the correlate (or antecedent) of the wh-expression
in a sluicing sentence. There are various ways for the antecedent to
be in the awareness of the hearer: by being explicit (as in canonical
examples of island repair in sluicing with overt antecedents); by
being in the dominant part of the sentence (in the sense of
Erteschik-Shir 2007 for implicit antecedents in non-islands); by
co-occurring with lexical items that are closely associated with it,
or by having a leading statement in the discourse that activates the
hearer’s awareness of the antecedent. We show that the claimed
structural constraint (e.g., an island constraint) can easily be
proven to be moot by manipulating the crucial sentences in such a way
as to satisfy “awareness” and some other nonstructural constraints
that are discussed in the paper. The data of this paper include (i)
English sluicing with an implicit correlate (Chung, Ladusaw, and
McClosky 1995, 2010, 2011); (ii) Japanese sluicing with an overtly
Case-marked remnant (Hiraiwa and Ishihara 2010, Fukaya 2012); (iii)
Contrast sluicing in English and Japanese (Merchant 2008, Fukaya 2012,
Barros 2013); and (iv) Korean pseudo-sluicing with an implicit
correlate (Ok and Kim 2012).