[lingtalks] TODAY: Andrew Carnie (Linguistics Colloquium)

[Klinton Bicknell]

TODAY at 2pm, Andrew Carnie (University of Arizona; http://dingo.sbs.arizona.edu/~carnie/ 
  ) will give the final colloquium of the quarter in the UCSD  
Linguistics Department, in AP&M 4301.

:: Abstract ::

Constituency as feature checking: dependency and morphophonological  
constraints on syntax

Andrew Carnie
University of Arizona

Assuming post-syntactic linearization to be an operation that applies  
in the PF component, it is surprising that most approaches to word  
ordering are primarily syntactic in nature. The fact that this  
component has access to both phonological and syntactic atomic  
structures predicts that under certain circumstance the phonological  
and/or morphological properties of component items in the constituent  
structure should affect linearization. This paper proposes that the  
linearization of constituent structures can be due to  
morphophonological considerations in addition to syntactic ones,  
broadening the mechanisms for linearization beyond Kayne’s (1996)  
Linear Correspondence Axiom. Evidence for this claim comes from two  
sources. First, I show that the position of realization of a word in a  
movement chain follows straightforwardly for the elsewhere condition  
on the insertion rule in Distributed Morphology. The position of  
realization of a word corresponds to the best fit of a vocabulary  
item’s features with the checked features at each stage in the chain.  
The effect here will be shown to be a late-insertion notational  
equivalent to multidomination.

Second, I present evidence from Scottish Gaelic that shows a situation  
where the linear order of object pronouns is determined exclusively by  
stress. In Scottish Gaelic, the object pronoun can appear after the  
verb or after any verbal adjunct provided the resultant prosodic  
structure is trochaic. I show that the stress pattern is not a  
consequence of the syntactic structure, but appears to be a primary  
motivation for the ordering. In this sense then, the linearization  
algorithm appears to be working as a repair strategy to insure a well- 
formed prosodic structure. The range of the possible positions for  
these object pronouns is seen to be a direct consequence of a set  
theoretic constituent structure, as follows from Chomskyan Bare Phrase  
Structure, and from the observation of Uriagereka (1997) that in a set- 
theoretic system constituency need not form a connected graph. Instead  
overlapping membership in sets (e.g. we allow A = {a, b}, C= {a, c},  
such that A and C overlap in membership but neither set contains the  
other) found in the case of adjuncts predicts precisely the range of  
constituents that the object pronouns can be linearized with, a notion  
that cannot be captured with traditional tree representations. Instead  
it is claimed that constituent structure is an epiphenomenal  
consequence of a dependency based theory of feature checking.

-- 

For further information about the Linguistics department colloquia
series, including the schedule of future events, please visit http://ling.ucsd.edu/events/colloquia.html 
  .