Theta-Theory Research Papers - Academia.edu (original) (raw)

While the LGB literature (Chomsky 1981) contains extensive discussion of empty categories in languages typologically like English on the one hand (non-pro-drop) and Italian on the other (pro-drop), relatively little has been said about languages like Russian, Polish or Czech, which fit neither of these two types of languages. The main purpose of this paper is to correct this typological imbalance by attempting to demonstrate whether current approaches need to be substantially revised in the light of data from Slavic and some other languages. The paper proceeds as follows: after making explicit the basic assumptions underlying the several versions of the theory of empty categories (EC), I will argue on the basis of Russian, Polish and Czech null-subjects and null-objects and their distribution that a number of revisions in theory's rules and basic principles is indeed necessary. Whereas the Minimalist program seems to correctly predict - on the basis of the principle of FULL INTERPRETATION (FI) and some additional assumptions (economy, greed, last ressort) - the existence of EC's, there is a problem with the notion of impoverished AGR as the only licensing formal condition, since languages without morphology, being at the same time pro-drop (e.g. Chinese, Japanese), cannot be explained this way due to lack of inflection. Although the notion of Morphological Uniformity of inflectional paradigm (Jaeggli/Safir 1989) seems to correctly predict the existence of Null-Subjects, there is no way how to include Null-Objects into this kind of explanation (p. 34). In both cases, pro must meet the licensing conditions of locality, agreement, case assignment, and, in case of thematic (referential) null subjects and objects, assignment of a referential theta-role. For languages without rich morphology, we therefore propose that both, null-subjects and null-objects, are bound - similarly to variables, by an Operator. This hypothesis is explained by the following facts that are observable in both pro and variables: 1) both variables and null-subjects/-objects stand in A-position; 2) Variables must be bound from A-bar positions, and null pronouns can be bound (and so they must be bound) from A-bar-positions (long scrambling); 3) null-subjects/-objects must have case, this is also a property of variables; 4) null-subjects/-objects must receive a theta-role, so do variables; 5) both, pronouns and variables are /- anaphors/.