Abstract:
The issues discussed in this mini-course are based on language surveys in Dayal, V. (ed.). The Open Handbook of (In)definiteness: the Hitchhiker’s Guide to Interpreting Bare Arguments, Open Handbook of Linguistics, MIT Press and related work.
Day 1: Demonstrative to Definite: What Changes and What Stays the Same
I posit that a demonstrative has two parts to its meaning, an indexical part and a contrastive part. The first demarcates a possibly proper sub-domain in the context of evaluation within which a unique referent must be established. It must also be possible for the referent to be contrasted with another entity in the context of evaluation or outside it, the property of anti-uniqueness. A definite lacks anti-uniqueness for sure, it may also lack also the indexical part, simply requiring its referent to be unique in the context of evaluation.
I show how this distinction accounts for certain well-known facts. The definite article is compatible with proper names in many languages. A proper name, suitably adjusted for type, satisfies the uniqueness requirements of the definite. Languages make a parametric choice between projecting a D. There is no cross-linguistic variation with respect to demonstratives — proper names are uniformly unacceptable. Strikingly, their unacceptability can be ameliorated under exclamatives – a fact that I suggest can only be explained with reference to the property of anti-uniqueness.
Finally, we consider cases where demonstratives require the support of a full DP to piggy-back on and argue that such cases crucially require distinguishing between indexicality and anti-uniqueness.
Day 2: Demonstratives, Definites, Bare Nouns: What Competes with What
Since Schwarz (2009), cross-linguistic studies have embraced the distinction between a strong familiarity-based definite and a weak uniqueness-based definite. But so far only Fering and German have been shown to have two distinct lexical forms of definite articles. What has actually been attested is a division of labor between either one definite determiner and a demonstrative (English), or a bare noun and a demonstrative (Mandarin).
I introduce the neo-Carlsonian account of kind terms in order to ground the discussion of bare nouns in languages that have articles (English, Italian, Akan) as well as those that do not (Russian, Hindi, Xhosa). I show that various aspects of the semantics of bare nouns, definites and demonstratives can predict their distribution when combined with standard theories of competition namely, Blocking and Maximize Presupposition. The picture of nominal systems that emerges does not require us to recognize strong definite articles as a distinct category in the nominal system of universal grammar.
Based on the diagnostics of anaphora and the ability to introduce discourse referents, I also comment briefly on the so-called ambiguity between definite and indefinite readings typically ascribed to bare nouns in article-less languages.