Vývojová dynamika obouvidových sloves cizího původu na základě korpusových dat

This paper describes how the Czech National Corpus can be used in researching what is called the ‘loss of biaspectuality’ process. We have tried to establish that the process manifests itself as a gradual year-by-year decrease in frequency of foreign-origin biaspectual verbs in (sub)corpora with a balanced distribution of texts. We deem this decrease to be inversely proportional to the increase in frequency of newly formed perfective correlates. Though these hypotheses have not been unambiguously verified, we have nonetheless acquired a great quantity of valuable data. We have identified four general types of verbal correlation in our database: (1) the “akceptovat” type stands for biaspectual verbs with no identified perfective correlates in the Czech National Corpus; (2) the “ilustrovat” type includes biaspectual verbs with infrequent perfective correlates; (3) the “dokumentovat” type represents biaspectual verbs with moderately frequent perfective correlates (both components of the correlation can ordinarily express perfective meaning); and finally, (4) the “likvidovat” type stands for originally biaspectual verbs with very frequent perfective correlates with a tendency to become the exclusive vehicle of the perfective meaning.