Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.philol.msu.ru/~otipl/new/fdsl/abstracts/madariaga.pdf Дата изменения: Sun Nov 9 19:52:24 2008 Дата индексирования: Wed Jan 14 13:52:07 2009 Кодировка:

A MINIMALIST APPROACH TO DP-INTERNAL SCRAMBLING IN RUSSIAN GOAL: We study Russian DP-internal scrambling analyzing it to the edge of a DP-phase whereby the scrambled element cyclically up to the highest phase where it checks a pragmatic additional evidence in favour of the 'strict' version of the Phase as instances of movement of an element escapes the nominal domain and raisig ([+F], [+topic]) feature. Thus, we provide Impenetrability Condition (PIC).

THE PUZZLE: Russian DP-internal scrambled elements trigger specific pragmatic effects, just as they do at the sentence level (Bailyn 2003): (i) nuclear stress assignment is usually associated to the scrambled element (1b); (ii) they must bear narrow focus and can be perceived contrastively (2a) (1) a. Ja postiral [DP krasnye noski]. (Neutral word order) I w as hed re d socks `I washed the red socks.' b. Ja postiral [DP NOSKI krasnye]. (Scrambled-DP word order) `I washed the red socks.' (Narrow focus + nuclear stress on noski, cf. 2) (2) a. Postiraj [DP NOSKI krasnye], a ne [DP STANY krasnye]. (Narrow focus on noski) wash socks red and not trousers red `Wash the red socks, and not the red trousers!' b. # Postiraj [DP NOSKI krasnye], a ne [DP STANY]. (Broad focus on noski) With the pragmatic value: `It is the red socks you must wash, and not the trousers!' THE PROPOSAL: an analysis of DP-internal scrambling in common DPs We will argue that the syntactic, semantic, and phonological behaviour of DP-internal scrambled elements are due to their focal nature and we will propose an analysis based on the combination, in a phase-theoretic model, of the derivational analysis of focus in Irurtzun (2007) and earlier accounts proposing a left periphery in nominal domains (cf. i.a., Haegeman (2004), Szabolcsi (1994)). Following Irurtzun (2007), we argue that the [+F] feature is potentially assigned to various tokens of the numeration and `projected' derivationally by means of Merge. In other words, when an element and an element undergo Merge both of them bearing a [+F] feature, the new syntactic/set theoretic object will also be a set containing just [+F] featured lexical items)). Thus, when an element/set of [+F] featured items is merged with an element that does not bear the [+F] feature itself, the new object will not be a set containing only [+F] featured material, hence, it won't be interpreted as focal (4): (3) {F, {F, F}} (4) {F, {,{F {F, F}}}} 3 3 F F {F, {F, F}} 3 F F Hence, the adoption of this system makes it possible to construct the different F-structure possibilities depending directly on the elements selected by the numeration (thus, without any violation of the Inclusiveness Condition). At the same time, it will allow us to account for a number of problematic data for the Nuclear Stress Rule-based theories given that the focus structure will already be set in the syntactic component before Spell Out. We will show that the focal structure of examples like (1b) (the impossibility of focus projection and the nuclear stress assignment pattern) is derived in a natural way with the adoption of the derivational approach to focus structure proposed in Irurtzun (2007). Regarding their syntactic behaviour, we argue that the DP in structures like (1b) derives from the movement the [+F]-marked N/NP to the edge of the DP (5), where it obtains the discourse-related 'edge-semantics' typically associated to internally merged elements (cf. Chomsky (2004)): (5) [DP N/NP[+F] [DP D [XP N/NP[+F] ...]]]

1


Due to this displacement to a phase edge position, the moved element is able to escape the Spell Out domain of the DP phase and thus it becomes accessible to check its [+F] feature with the Comp (6).
PhaseHead3 PhaseHead2 PhaseHead1

(6) [CP C ...

[v

P

v... [DP N/NP[+F] [DP D [XP N/NP[+F] ...]]]]]

Therefore, we will show that Russian DP-internal scrambling provides additional evidence in favour of the 'strong' version of the PIC (also called PIC2) as stated in (7) (see, i.a., Chomsky (2001), Richards (2007), Boeckx (2008) and MЭller (2008)). (7) PIC2: Spell-Out the complement of Phase1 as soon as Phase2 is merged. In a nutshell, the focal element that stays in the edge of a DP-phase remains active until the merger of Comp, which can check its [+F] feature. We will provide additional evidence in favour of such a phasal analysis of the movement in (5), regarding the position of adverbs and extraction from islands. EXTENSIONS: an analysis of DP-internal scrambling in Russian Approximative Inversion (AI): We will relate this type of scrambling with the well-known phenomenon of AI in Russian (8b), which is analysed as an instance of DP-internal movement, with the expected pragmatic effects. (8) a. AlКna priglasila [QP desjat' podruzek]. (Neutral word order) Alena invited ten friends GN.PL. `Alena invited ten of her friends.' (Broad focus; answering to `Who did Alena invite?' or `How many of her friends did Alena invite?') b. AlКna priglasila [QP podruzek desjat']. (Approximative Inversion) `Alena invited more or less ten of her friends.' (Narrow focus on desjat'; answering only to `How many of her friends did Alena invite?') In the case of AI, an additional semantic effect is observed (besides the topicalization of the raised N and focus on the stranded numeral): the numeral functions as an approximate quantifier (8b). We follow previous hypotheses (cf. Yadroff (1999), Pereltsvaig (in press)) in that AI in (8b) is the movement of an N head to a higher position inside DP to check a strong [+approximative] feature. In AI, the only way to have the approximative semantics without topicalizing the scrambled N is by inserting a noun classifier (celovek `people' or stuk `things'), which we assume to be the overt realization of the Measure head proposed by Yadroff (1999). As it is lower than the edge of DP (the position that leaves open the way to cyclic movement up to CP), no [+topic] feature can be checked in this configuration; see structure (10), corresponding to (9): (9) Ja kupil stuk desjat' xorosix knig. (AI with no pragmatic effects) I bought units GN.PL. ten [good books] GN.PL. `I bought more or less ten good books.' (10) [QP [Q stuk desjat'] [MeasureP stuk [NP xorosix knig]]]]] In addition to the evidence for a phasal analysis of the movement of N/NP in common DPs, we will offer evidence in favour of such an analysis with pragmatic effects for AI, including the following: (i) strong topicalization of the whole AI-scrambled QP and the sentential li-focus position are impossible, given that the N element has been already raised to check a [+topic] feature.
REFERENCES: Bailyn, J., 2003, `A (purely) Derivational Approach to Russian Scrambling'. Barbara Partee et al (eds.) Formal Approaches to Slavic Linguistics 11. Boeckx, C., 2008. "Treelets not trees", ms: Harvard U. Chomsky, N. 2001. "Derivation by Phase", in M. Kenstowicz (ed.), Ken Hale: A Life in Language, Cambridge (Mass.): MIT Press, 1-52. Chomsky, N. 2004. "Beyond Explanatory Adequacy", in A. Belletti (ed.), Structures and Beyond. Oxford: OUP,104-131. Haegeman, L., 2004, `DP-periphery and clausal periphery: possessor doubling in West Flemish. Relating Nominal Periphery to Clausal Periphery'. David Adger et al. Peripheries. Irurtzun, A., 2007, The Grammar of Focus at the Interfaces, PhD.Thesis: UPV/EHU. MЭller, G., 2008, "On Deriving the CED Effects from the PIC", ms: U. Leipzig. Pereltsvaig, A. (in press) `Passing by Cardinals: in support of Head Movement in Nominals' Formal Approaches to Slavic Linguistics 14. Richards, M., 2007, "Deriving the Edge. What's in a Phase?", ms: U. Leipzig. Szabolcsi, A., 1994, `The noun phrase'. Ferenc Kiefer & Katalin E. Kiss (eds) The Syntactic Structure of Hungarian. Yadroff, M., 1999, Formal properties of functional categories: the minimalist syntax of Russian nominal and prepositional expressions, PhD.Thesis, Indiana University.

2