Jila Ghomeshi

Non-Projecting Nouns and the Ezafe: Construction in Persian

With the advent of functional as well as lexical structure within the noun phrase, there exists much syntactic structure that appears to be unused. While there has been discussion as to which inflectional categories project structure, the question of whether lexical categories project has not been raised. In this paper it is claimed that common nouns in Persian do not project to maximal categories, i.e. do not appear with a complement and a specifier position. Several facts follow as a consequence. First, we explain an intriguing property of the Ezafe construction in Modern Persian, namely that all modifiers of nouns must be heads themselves. We also have an explanation for the presence of the Ezafe vowel itself, which is analyzed as a linker affixed to X0s at PF to identify elements forming a nominal constituent. Finally, we explain why only one possessor is possible within Persian noun phrases. Since nouns cannot project complement and specifier positions, the only position available for a possessor is [SPEC, DP].

Plural marking, indefiniteness, and the noun phrase

This article proposes that plural marking on nouns in Persian is licensed only if those nouns are contained within D/QPs. This proposal accounts for why plural-marked nouns are construed as definite unless an overt marker of indefiniteness appears, and why plural marking does not cooccur with numerals unless the noun phrase is definite. It is also shown that the indefinite marker in Persian is quantitative rather than cardinal and is thus associated with higher functional structure within the noun phrase than in English. In English, on the other hand, number marking, the indefinite article, and the grammatical distinction between count and mass nouns, are all realized at the level of NumP. Differences in the interpretation of bare noun phrases in English and Persian are therefore explained by the claim that argument noun phrases must minimally be NumPs in English while Persian lacks this projection altogether.

Topics in Persian VPs

In this paper the Persian morpheme -râ is examined. The examination sheds light on a number of current issues in the theory of case such as whether multiple case marking is an instance of case assigned by a functional head to more than one specifier, and what kinds of semantic properties case can correlate with. The morpheme -râ is of great interest in that (a) it appears to be a case marker, yet can appear more than once in the same clause, and (b) it appears to be a definiteness marker yet can cooccur with indefinite marking. It is argued that -râ case-marks noun phrases that are adjoined to VP. Syntactically such adjunction is free; however, it is claimed that all such noun phrases must be thematically licensed. Thematic licensing occurs either when the noun phrase is construed as the direct object (via theta-role assignment) or when the noun phrase is construed as a topic (via coindexation with a thetamarked element). Semantically, it is argued that -râ marks noun phrases that satisfy a number of different properties all correlating with high transitivity, such as definiteness, animacy or topic-hood. Especially with respect to the latter, the syntactic and semantic functions of -râ converge.