Testing weak and strong definites experimentally across languages

Schwarz (2009) proposed a distinction between weak and strong definite articles, reflected in Standard German in the presence or absence of contraction of the article with certain prepositions (e.g., vom vs. von dem). Semantically, the analysis took the former to be a situationally restricted uniqueness article, and the latter an anaphoric article bearing an index, just like a pronoun. In subsequent work, numerous authors have applied this distinction to analyze contrasts between definite forms in a wide range of languages. While this cross-linguistic evidence supports the availability of the ingredients of the distinction in natural language in general (with some analytical variations and adjustments, and quite possibly further aspects in play in certain languages), the basic semantic contrast remains subtle and has generally not been captured systematically in experimental investigations. I present data from a simple picture selection task paradigm where the contrast in anaphoricity is pitched against another factor, that of typicality, which do provide quantitative support of the posited contrast. Next, I extend this paradigm to English to turn to a question vexingly left open in this literature, namely whether articles in languages that do not seem to make such a contrast and only use one form for definites throughout, like English, are ambiguous or map onto either weak or strong articles. The final part of the talk adds in data from various ongoing collaborations that investigate yet more languages to work towards a more systematic and quantitative comparison of different forms across languages. I close by considering challenges that lie ahead in integrating this data theoretically.

Location: Royce Hall, Room 156