Movement versus Merge: Stylistic and semantic perturbations versus neutral style
Helen Trugman (Holon Academic Institute of Technology)
Since the advent of Kayne's Antisymmetry theory (1994) numerous attempts have been made to obtain all possible word-order permutations by the asymmetric merge of various satellites to the left of the head, followed by successive movement of relevant phrases. Specifically, this approach was thoroughly developed for the nominal domain in Cinque 1994, 2005a/b, Bošković 2005, among others, setting forth an ambitious goal to eliminate head movement as a possible kind of syntactic movement cross-linguistically.
In this talk, I will focus on Russian extended (in the sense of Grimshaw 1990) nominal projections and will challenge the above claim both on theoretical and empirical grounds. Most importantly, I will show that the attempt to reduce permissible movement types to XP-movement ends up in unjustified complication of the derivational mechanism. Moreover, the DP-structure gets greatly expanded by postulating multiple AgrPs, which serve as landing sites for XPs raised across filled specifiers of intermediate functional projections (see Cinque 2005a, (2)). In addition, the nature of the attraction for various DP-internal movements remains yet unclear. Though some nominal feature might be assumed, what underlies its weakness/strength across languages in triggering various instances of XP-raising or XP-rolling-up stays unanswered.
For instance, in contrast to previous accounts (Babby 1975), Cinque (2005a) proposes that relative clauses are merged prenominally either below Dem (Determiner) or below Num(eral) (see (1)), and end up phrase-final due to the successive movement of the NP or AP to some higher locus. Consequently, the order in (2a) is the result of merge below the numeral head, which in Russian is signaled by the genitive marking on the RC; and the one in (2b) will require the NP to move across a lower RC to the Spec, AgrP dominating the AP, which I ignore here.
If the reduced relative clause is merged above the NumP, it will be marked as nominative and the merged word order is claimed to be, as in (3a). For deriving the postnominal variant in (3b), one will need to move the whole NumP across the RC to the Spec, AgrP dominating the RC.
However, Russian abounds in structures with reduced relative, or adjectival clauses found before ordinal numbers or even determiners, as in (4). Does it mean that we have to assume two more possible merging sites for reduced RCs in Russian or do we have to allow for RC-extraposition in (4), as is clearly the case in the following example: Neobxodimye dlja pervoj zadaèi, vo vtoroj zadaèe sve šest' uravnenij okazyvajutsja nepremenimymi. '[RC Needed for the first task], in the second task [DP all six equations] turn out to be inapplicable.'? The examples in (4) seem to demonstrate that 'a fixed hierarchical structure resulting from Merge' in (1) coupled with only 'upward movements of phrases containing the NP', cannot derive all the variations of the universal structure, as attempted in Cinque (2005a: 327). In some cases, a RC extraposition must be assumed to supplement the analysis, hence the latter seems to be too weak on itself to accommodate all possible word-order permutations.
Russian also seems to undermine another claim put forth in Cinque (2005a/b), who follows Kayne (2002) in assuming that prepositions heading the PP-complements of N are merged DP-externally, and attract their 'complements', with the remnant raising to their left, thus making PP-complements DP-final. This assumption accounts for their being stranded at the end of the DP and not being dragged along by the NP and its modifiers (Cinque 2005a: 327). Russian seems to challenge this claim as well, by allowing PP-complements and genitive complements of N to intervene between the N-head and the postnominal relative clause, whose postnominal position is derived via NumP-raising across the RC, as in (5). Countered by similar data in other languages, Cinque has to assume that a RC can be merged either lower or higher than P (Cinque 2005b:42), thus further flexing the universal hierarchy of prenominal modifiers he proposes. Yet even such a modification will fail to account for cases like (6), when a PP or a reduced RC intervenes between a head noun and a classificatory adjective without the latter being focalized or topicalized. Moreover, with a PP-modifier merged in the initial position, the whole DP is ruled out.
Interestingly, Russian (and Ukrainian) noun phrases exhibit another type of DP-internal scrambling, which necessarily strands the PP and GenP complements of N, as in (7a). Moreover, this movement seems to derive a very specific semantic effect, namely, obligatory classificatory reading of the postnominal modifiers (see Trugman 2005). This might point to a different nature of the movement involved, with a well-defined trigger, such as a semantic feature on the Determiner or some other functional element of the projection, (8).
In my talk, I will argue that various permutations of the word-order in Russian DPs can be better explained by semantic (above) or pragmatic considerations, when stranding or fronting of reduced relative and adjectival clauses is governed by focus/topic features, as demonstrated in (9).
It will also be shown that allowing for some symmetry of modifier merge, one can account for various patterns of prenominal and postnominal modifiers in Russian with less complexity.
(1) [Quniv...[Dem...[Numord... [RC... [Numcard...[RC... [Classifier...[A... [NP]]]]]]]]]
(2) a. [Numcard pjat' [RC vvedennyx vyše [NP priznakov]]]
five introduced.gen above features.gen
b. [Numcard pjat' [NP priznakov] [RC vvedennyx vyše tNP]]
3) a. [Numord pervye, [RC vvedennye vyše, [Numcard pjat' [NP priznakov]]]]
first introduced.nom above five features.gen
b. [Numord pervye, [Numcard pjat' [NP priznakov]] [RC vvedennye vyše, tNumP]]
(4) a. [RC vvedennye vyše, [Numord pervye [Numcard pjat' [NP priznakov]]]]
b. [RC vvedennye vyše, [Dem èti [Numcard pjat' [NP priznakov]]]]
introduced.nom above these.nom five features.gen
(5) a. [[NumP Troe [PP iz nix]], [RC obessilevšie] tNumP], k veèeru dobreli do poselka.
Three from them.gen exhausted.nom by evening reached to settlement
b. [NumP Sto [NP èkzempljarov[GenP brošjury]]] [RC podgotovlennye k otpravke] tNumP naxodjatsja v sejfe.
Hundred copies brochure.gen prepared.nom for dispatching are found in the-safe
(6) a. Mineral'naja, [PP s privkusom tuxlyx jaic] voda ne ponravilas' Pete.
Mineral with aftertaste rotten eggs water not liked Peter.dat
Peter did not like mineral water with the aftertaste of rotten eggs.
b. *[PP S privkusom tuxlyx jaic] mineral'naja voda ne ponravilas' Pete.
(7) a. vina suxie krasnye [GenP urožaja 2001 goda]
wines dry red harvest.gen 2001 year.gen
b. *vina [GenP urožaja 2001 goda] suxie krasnye
'dry red wines of 2001'
(8) solod pyvovarnyj jaèminnyj svitlyj (Ukrainian)
malt brewing barley.adj light
'light barley beer-brewing malt'
(9) Petr byl [NP èelovekom] [molodym, neopytnym tNP]
Peter was man young inexperienced
Bošković, ®eljko. 2005. On the locality of left branch extraction and the structure of NP. Studia Linguistica 59(1), pp.1-45.
Cinque, Guglielmo. 1994. Evidence for partial N-movement in the Romance DP. In G. Cinque, et al. (eds.) Paths Towards Universal Grammar. Georgetown University Press, 85-110.
Cinque, Guglielmo. 2005a. Deriving Greenberg's Universal 20 and Its Exceptions. LI 36:315-332.
Cinque, Guglielmo. 2005b. The dual source of adjectives and phrasal movement in the Romance DP. Ms., University of Venice, December 2005.
Kayne, Richard S. 1994. The Antisymmetry of Syntax. Cambridge, MA: MIT Press.
Kayne, Richard S. 2002. On some prepositions that look DP-internal: English of and French de, Catalan Journal of Linguistics, 1.71-115.
Trugman, Helen. 2005. Rudiments of Romance N-to-D movement in Russian, Talk presented at FDSL-6, Potsdam, Germany, December 2005.