Variable Degrees of Constituency: Frequency Effects in the Alternation of pa vs. para in Spoken Discourse

Abstract

This study examines how repetition in discourse determines variable degrees of constituency between the preposition para 'to, for' and surrounding words, thus predicting greater reduction of para in cases where the degree of constituency is tightest. Our study contributes to the overall understanding of the role of usage in explaining reductive phonetic processes in contemporary Spanish. The findings of this corpus-based study of Caracas Spanish reveal that the frequency of the collocation is a key factor in predicting reduction of the form para. Further, frequency of use is a better explanation for the reduction of para than more traditional variables such as phonetic context. This result provides evidence to support the hypothesis that frequent combinations are accessed as autonomous units, due to their tighter degree of constituency, as in the case of I don't know in American English (Bybee and Scheibman 1999).