La pluralización de haber presentacional en el español peninsular: datos de Twitter [The pluralization of presentational haber in Peninsular Spanish: Data from Twitter]

In this paper we explore whether Twitter data can be used in dialectological studies of morphosyntactic alternations in Peninsular Spanish. Specifically, drawing on a corpus of tweets we evaluate the hypothesis that in Peninsular Spanish, as in Caribbean Spanish, the pluralization of presentational haber (e.g., había(n) fiestas 'there were parties') is constrained by two domain-general cognitive constraints on language production: markedness of coding and statistical preemption. Then, we analyze the geographic distribution of haber pluralization in Peninsular Spanish. The results show that haber is pluralized in 11% (N= 611/5,500) of the cases. The mixed-effects logistic regression analysis supports that nouns which can easily be imagined as starting points of a series of events, the absence of negation, and verb tenses other than the preterit favor pluralization. To explore the geographic distribution of haber pluralization in Peninsular Spanish, we use a generalized additive mixed model, which generates a line that runs approximately from Lérida in the north of Spain to Almería in the south, crossing Teruel and Albacete. East of that line, the pluralized variant has a higher probability of occurrence, but it is not absent from other areas of the Peninsula. We conclude that (i) haber pluralization in Peninsular Spanish is constrained by domain-general cognitive constraints on language production and that (ii) tweets constitute a viable alternative to spoken language data in dialectological studies of morphosyntactic variation.

http://ift.tt/2fHo9Nc