LONGOBARDI 0. Introduction: variation in syntax and interpretation Principles-and-parameters theory is likely to be the main achievement of the linguistic sciences after and along with the historical-comparative method. It proposes a model of grammatical variation according to which differences in the morphosyn-tactic components of human languages are discrete, finite and more limited in number than they appear on the surface (i.e. several superficial differences turn out to cluster together at

more »

... the appropriate level of analysis). An important current question is whether a comparable state of affairs holds for semantic properties of natural languages, namely 1) whether differences really exist in the interpretive components, 2) whether they display the same parametric properties pointed out above, and 3) to which extent they can be connected to independently manifested syntactic differences. In simpler words, one may ask whether comparative semantics is possible, which form it may take, and how independent it can be of (comparative) syntax. Crucial evidence should in principle be provided by possible cases of crosslinguis-tic syntactic homonymy, i.e. phrasal expressions with roughly the same surface structure across different languages, but mapping to clearly distinct logical representations. In this article I will push forward a line of inquiry initially suggested in Longo-bardi (1994), arguing that Romance and English bare nouns differ in meaning, though not in shape, in a formal and grammatically predictable way, and improving descriptively on my own as well as on other recent accounts. The main proposal will be that Romance bare nouns are nothing but indefinites (quantificational variables, existentially or generically bound) in Heim"s sense, while English bare nouns are ambiguous between this quantificational interpretation of indefinites and a referen-tial (i.e. kind-referring) one.