engleski

Grammaticality judgments and language usage data.

In this paper we discuss the results of a corpus-based study of clitic placement and clitic clusters in Croatian, accompanied by results from experiments on gradual grammaticality judgments using acoustic sentence repetition, and text correction tasks. On the one hand, we suggest that the uncertainty about specific low-frequent constructions in corpora, which could be due to errors of writers and reviewers, or to real variation and soft grammatical constraints of native speakers, can be eliminated on the basis of results from psycholinguistic experiments. On the other hand, the rather plausible conclusion that the relative frequency of particular constructions in corpora correlates with the certainty of grammaticality judgments of native speakers, as well as their sensitivity to identify normative or prescriptive deviations, can find empirical support, using quantitative analysis of corpora and psycholinguistic experiments.

grammaticality judgements; Croatian; psycholinguistic experiments

nije evidentirano