Long-Distance Bound Local Anaphors in Korean: An Empirical Study of the Korean Anaphor Caki-casin
Ji-Hye Kim
UIUC Linguistics
4PM, Thursday, April 20
Lucy Ellis Lounge, FLB
Since the important work of Pollard Sag (1992), we know that languages like English that lack genuine long-distance anaphors (LDAs) still allow local anaphors to be bound long-distance (LD) as exempt anaphors (logophors). For example, an anaphor lacking a locally accessible antecedent – either because there is none (cf.1a) or because the local antecedent cannot be the actual antecedent (cf. 1b) – behaves as an exempt anaphor in English.
(1) a. Bill remembered that the Times had printed
[a picture of himself] in its Sunday edition.
b. Bill was worried about [the administration’s
opinion of himself].
Since Korean possesses both local (caki-casin, pronoun-casin) and LD anaphors (caki, casin), the possibility arises that speakers of Korean would simply reject a local anaphor bound outside its GC, resorting to LDAs instead. This possibility is especially likely under approaches that hold all LDAs are logophors (Y. Huang 1994).
The present study investigated whether and how Korean monolingual speakers would accept exempt binding of Korean local anaphor caki-casin. Monolingual speakers of Korean were tested on different types of sentences where caki-casin was forced to bypass the local antecedent in favor of a LD-antecedent – in violation of both the Tensed-S Condition (TSC) and the Specified Subject Condition (SSC) -- as shown in (2). This is because core anaphors in languages like Korean and Chinese are not constrained by the TSC, as is well-known.
(2)Inphyo i-nun [kyenchalcheng j -i [caki-casin i -i I-top police-nom self-nom swumki-n cungkemwul]-ul chacanayssta]-ko malhay-ss-ta hide-rel exhibit–acc found-comp say-past-decl ‘Inphyo said that the Police Agency found out the exhibit self (he) had hidden.’
In addition, the question of whether speakers who accept exempt binding show a preference for the strict reading in VP-ellipsis was also investigated, since preference for strict readings has been argued to be another diagnostic of exempt binding (Huang and Liu 2001, Ying 2005).
Overall results with Grammaticality Judgment Task (designed to investigate whether Korean native speakers regard exempt binding of local anaphors as grammatical) coupled with a Preferential Sentence Interpretation Task (designed to test the preference for sloppy vs. strict reading in VP-ellipsis) showed that the local anaphor caki-casin can be LD-bound, and when it is, it behaves as an exempt anaphor.
The results indicate that the local vs. long-distance property of anaphors is orthogonal to the core-exempt distinction. Therefore, the existence of LDAs in a language does not rule out exempt local anaphors in the same language, as also shown in Oshima (2004) and Kim and Yoon (2006). Approaches that take LDAs to be exclusively exempt anaphors (Y. Huang 1994) do not fare well in light of these results.