Department People Alumni Students Courses Facilities News Home


Mürvert Enç

Ph.D. Dissertation, 1981


In model-theoretic semantics, it has been customary to distinguish between two classes of expressions claimed to differ in the kind of context dependency they exhibit. One class contains expressions like I, now, traditionally called indexicals. The second class contains expressions assumed to be non-indexical (though still context dependent), including nouns and verbs. One motivation for this distinction has been the assumption that indexicals behave differently from non-indexicals in that their interpretation is independent of the scope of negation or intensional operators such as tense, modals and propositional attitude verbs, whereas the denotation of a non-indexical is taken to vary according to the scope of sentential operators. In this study, I challenge the assumption that nouns and verbs are basically different from indexicals. I show that certain readings are not representable using traditional scope analysis, and that the scope analysis can be saved only by trivializing scope and introducing arbitrary operators. An adequate analysis is possible, however, if we treat nouns and verbs as indexicals. This treatment has further advantages. It allows different denotations for different occurrences of the same noun in a sentence, and automatically gives us the domain of quantification for noun phrases. It also allows a unified treatment of definite descriptions by letting the difference in the individual concept, directly referential and generic readings fall out of differences in contexts. Another reason for distinguishing indexicals from non-indexicals has been the assumption that indexicals contribute their denotation to propositions, whereas non-indexicals do not. This was motivated by a pre-theoretic notion of propositions as objects of mental attitudes. Treating nouns and verbs as indexicals gives us the kind of propositions which fully characterize mental attitudes. The analysis I propose also has the following advantages. (1) The interpretation of verbs in different languages is essentially similar, whether or not languages mark tense obligatorily. (2) There is no need to posit distinct syntactic derivations to account for ambiguities in definite descriptions, adverbs of quantification or scope of operators. (3) There is a close correspondence between the semantic system and surface forms of natural languages. Thus the picture that emerges from this kind of analysis is one of greater unity and simplicity.

Department of Linguistics University of Wisconsin-Madison

Questions?Questions about the Department?
Questions/Comments about this website?