Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.philol.msu.ru/~otipl/new/fdsl/abstracts/mitrenina.pdf
Дата изменения: Sun Nov 9 20:30:20 2008
Дата индексирования: Wed Jan 14 13:52:44 2009
Кодировка:

The Syntax of Correlatives in Russian
The purpose of this study is to provide an analysis of correlative constructions in Russian. Correlatives have been actively studied in recent 15 years (Keenan1985, Izvorski 1996, Vries 2002, Bhatt 2003, LiptАk 2005, Dikken 2005), but Russian constructions were usually not considered, although they provide useful material for better understanding of syntax of relativisation and topicalisation (Zalianyak and Paducheva 1975, Shvedova 1980, Lutikova 2008). Correlativisation is a non-local strategy of relativisation; in a simple correlative the subordinate clause contains relative XPrel and precedes the main clause that contains anaphoric XPana associated with XPrel. It can be basically schematized like this: (1) [CorC
P(subordinate clause)...

XPrel i...]i [IP

(main clause)...

XPana i...]

Following Strvastav (1991), Rajesh Bhatt (2003) describes differences between Headed Relative Clauses and Correlatives. The Russian correlatives reveal the same properties: (a) unlike headed relatives, the `head' can appear in either Srel or Smain, or both; (b) the demonstrative requirement; (c) multi-head relative clauses are only possible with correlatives; (d) while externally headed relative clauses allow for stacking, correlatives do not. Depending on the realivized element Russian correlative can be nominal or adverbial, there are also two types of comparative correlatives. One of the possible structures of correlatives that were described in the literature (Bhatt 2003, LiptАk 2005) assumes that the correlative is base-generated adjoined to Dem-XP and is optionally moved out of there together with the optional scrambling of Dem-XPs. Arguments in favor of the low-adjunction analysis come from reconstruction effects. But in the case of Russian correlatives reconstruction effects reveal no movement. (2) [ k i ]k, k i tk. Which yacht Mishai likes Dem hei Prt buys `Mishai buys whichever yacht hei likes.' If CorCP [ k i ]k (Which yacht Mishai likes) was moved from IP in (15), then this sentence should be ungrammatical, because i c-commands the "trace" of [ k i ] that includes its coindexed R-expression i. So, there is no movement from IP and low-adjunction does not work for Russian correlatives. Another structural account assumes that the correlative clause is base-generated adjoined to the matrix clause (IP). Rajesh Bhatt shows that is does not work for simple correlatives in Hindi, and AnikС LiptАk proves that it does not work for Hungarian either. This analysis was rejected because the relationship between the Correlative Clause and the Demonstrative Phrase is subject to islands. The same is true for Russian. No islands when there is no NP: (3) , , [CorCP ],[ , -> Who honestly worked, that-him you tries to-fire, but who thieved you want ]]. [IP that I Dem again take to myself to work `You want me to fire those who worked honestly, and as for those who stole, you want me to hire them again'. 1

Complex NP island: (4a) *[CorCP ]i, [NP, [IPi ]]]. *[CorCPWho badly worked ]i, I issued [NPdecree, in-order [IP Demi to-fire]]]. `As for those who worked badly, I issued a decree to fire them'. (4b) *[CorCP ]i, [NP , [IP ]] *[CorCPWho sees through walls]i, there-are [NP hospitals, where [IP Dem are-treated for-free]] `As for those who can see through walls, there are hospitals where such people are treated for free'. So, as Rajesh Bhatt says, something must be moving. But it is moving not from IP (according to the previous section). Aniko Litpak suggests her own structural account of correlatives by introducing Correlative Topic Phrase. This analysis works well for Hungarian where focus is always overly marked and everything that precedes it is in topic. In Russian the situation with topic/focus although Russian correlativisation is definitely connected with topicalisation. The possible structural account for Russian can be the following: (5) [TopP/FocP [CorCP]j [FocP (Dem-XPi) [IP ... ti... ]]]

This analysis explains also the ungrammaticality of (4a-b). References 1. Bhatt R. 2003. Locality in Correlatives. In: Natural Language and Linguistic Theory, Volume 21, Number 3, August, pp. 485-541. 2. Dikken M. den. 2005. Comparative Correlatives Comparatively. In: Linguistic Inquiry. Volume 36:4, pp. 497 532. 3. Izvorski R. 1996. The syntax and semantics of correlative proforms. In: K. Kusumoto (ed.), Proceedings of NELS 26. Amherst, GLSA Publications. 4. Keenan E. 1985. Relative Clauses. In: T. Shopen, ed., Language typology and syntactic description, Cambridge, England: Cambridge University Press. Volume 2, pp. 141 170. 5. LiptАk A. 2005. Correlative Topicalization. uLCL, Leiden University ms. 6. Lutikova E. 2008. Zagadki russkih otnositelnyh predlozhenij Report at the conference "Syntactic Structuures-2". Moscow. 7. Shvedova, N. (ed.). 1980. Russkaya grammatika. II. Syntax. Moscow. 8. Vries M de. 2002.The Syntax of Relativization. PhD thesis, University of Amsterdam. 9. Zalianyak, A, Paducheva, E. 1975. K tipologii otnositelnogo predlozhenija. In: Semiotika I informatika, Volume 6, Moscow.

2