This dissertation is concerned with a range of syntactic phenomena related to sentential complementation in Modern Greek and the consequences they have for the Theory of Grammar. Two types of complements are examined: factive complements and their pattern of wh-extraction, and subjunctive complements and their implications for the theory of control and the licensing of case. ;In particular, evidence from the syntax of Modern Greek--more specifically, from the distribution of an A$\sp\prime$-bound pronoun--is drawn against an operator approach to factive complements. (...) It is argued that the strong islandhood observed in Modern Greek factive complements is due to the fact that they are not real complements of the subcategorizing verb but independent clauses standing in a paratactic relation to an empty nominal complement of the matrix predicate. The crosslinguistic variation attested with respect to extraction is attributed to the ability of languages such as English to form certain types of A$\sp\prime$-chains which evidently are impossible in Modern Greek. ;Furthermore, upon investigation of the so called subjunctive complements in Modern Greek it is shown that the position of their understood subject may be occupied sometimes by a pronominal pro and sometimes by an anaphoric PRO. The licensing of the two empty categories is derived through the aspectual and tense properties of these constructions. It is claimed that the presence of PRO in these configurations and the set of anaphoric properties it displays provides evidence that the distribution of this category can be derived on a case theoretic account, along the lines of Bouchard or Chomsky and Lasnik , and does not depend on the notion of government as in Chomsky. (shrink)

This article examines the discourse competence of high-functioning children with autistic spectrum disorders to participate in narrative introduction sequences with family members. The analysis illuminates the children’s own efforts to launch narratives, as well as their ability to build upon the contributions of others. Ethnographic, discourse analytic methodology is integrated with the theory of discourse organization and the weak central coherence account of autism. Introductions of both personal experience narratives as well as fictional narratives are examined. The children were especially (...) competent in the use of stable introductory practices when launching fictional narratives, pre-organized by the media of expression. Their challenge was not in the introduction, but in the narrative co-telling, which often was not globally organized over an extended course of propositions. The heterogeneity of the ASD children’s discourse competence and its implications for discourse analysis are discussed. (shrink)

In this paper, I examine the syntax-semantics of subjunctive clauses in (Modern) Greek. These clauses are headed by the particle na and contain a dependent verbal form with no formal mood features: the perfective nonpast (PNP). I propose that the semantics of na is temporal: it introduces the variable now (n) into the syntax. This is necessary because the apparent present tense in the PNP cannot introduce n. The PNP, instead, contains a dependent time variable. This variable cannot be interpreted (...) as a free variable – hence it cannot be identified with the utterance time of the context. This analysis relies on two premises. One is the (quite influential) idea that pronouns and tenses are analogous creatures (Partee 1973, 1984, Heim 1998, Kratzer 1998, and others). The other premise is that at least some polarity items are expressions that contain variables that cannot be interpreted deictically (Giannakidou 1998, 2001). In the present work I suggest to enlarge the domain of phenomena that can receive a unified treatment across individuals, worlds, and tenses, and treat the subjunctive mood as a non-deictic time, thus an instance of a polarity dependency of the temporal kind. It is my hope that the analysis proposed here for the PNP can be used to analyze verbal subjunctives in Romance languages, and perhaps also infinitival forms in English, but investigation of this question will have be left for the future. (shrink)