Abstract: (PDF version)

Given a certain linguistic intuition, we do not know a priori what aspects of it are due to the language faculty proper and what others are due to factors outside it. The general goal of this talk is to provide an illustration of how one might proceed to separate the contributions of the language faculty from those of the factors outside it. A specific empirical goal of this talk is to argue that the nature of the sloppy identity reading in surface anaphora is distinct from that in deep anaphora. In particular, I maintain, following Hankamer & Sag 1976, that the ellipsis site in surface anaphora, such as empty VP in VP ellipsis (VPE) in English (and an empty IP in one type of comparative in Japanese) is fully represented at LF, in such a way that the necessary conditions for the availability of a sloppy identity reading in surface anaphora are satisfied. I further maintain that deep anaphora, more precisely, the categories or elements that are considered to exhibit properties as such, on the other hand, is not fully represented at LF in the way the surface anaphora is.
　
Descriptively, I discuss VPE, do the same thing, do that, stripping in English, and the so-called null object construction, soo su 'do so', stripping, and comparatives in Japanese. (I might also discuss sluicing in Japanese and English.) One of the main points of the paper is that what appear to yield the "same interpretation(s)" sometimes have radically different formal properties. This holds cross-linguistically as well as language-internally. Cross-linguistic instances of this can be seen when we examine VPE in English and the null object construction in Japanese, and language-internal instances of this can be seen when we look at (i) VPE and do the same thing in English and (ii) case-marked versions of stripping and comparatives in Japanese on the one hand and their non-case-marked versions on the other. We can observe the kind of contrast noted in (ii) in English as well, by considering the differences between NY, too and To NY, too.
　
The presentation will be based on (portions of) a (hopefully improved) version of my "Surface and Deep Anaphora and Experiments in Syntax" (to appear in Anaphora: A Reference Guide, eds. A. Barss, and T. Langendoen, Blackwell,) which in turn is based on:

1. Ch. 5 "Sloppy Identity in Japanese" of Hoji 1990 Theories of Anaphora and Japanese Syntax.
2. "Formal Dependency, Organization of Grammar and Japanese Demonstratives," (1998) Japanese/Korean Linguistics 7, N. Akatsuka, H. Hoji, S. Iwasaki, S.-O. Sohn, and S. Strauss, eds., Center for the Study of Language and Information, Stanford., pp. 649-677.
3. "Null Object and Sloppy Identity in Japanese," (1998) Linguistic Inquiry 29.1 pp. 127-152.
4. "Sloppy Identity and Formal Dependency," (1997) WCCFL 15, Center for the Study of Language and Information, Stanford. pp. 209-223.
5. "Sloppy Identity and Principle B," (1997) ATOMISM & BINDING, H. Bennis, P. Pica and J. Rooryck, eds., Foris Publications, Dordrecht, pp. 205-235.
6. "Stripping and Sluicing in Japanese and Some Implications," (1999) (With Teruhiko Fukaya) In Bird, Sonya, Andrew Carnie, Jason D. Haugen, and Peter Norquest, eds., Proceedings of the 18th West Coast Conference on Formal Linguistics, Cascadilla Press, Somerville, Massachusetts. pp. 145-158.

If you would like to read some of these, prior to his presentation at KACL, please contact him at hoji@mic.atr.co.jp.