Keywords: morphology, verbal inflection, paradigm structure, corpus linguistics, linguistic variation, text statistics, child language, historical linguistics, old Estonian
The article discusses the implicational patterns present in the Estonian verb paradigm: which paradigm slot acts as the base form, which slots act as other principal parts, and what their hierarchical dependency looks like. The argumentation relies on data from three different sources: childrens acquisition of Estonian as their first language; verbal inflectional classes as reconstructed from the 17th century Tallinn variety of Estonian; and statistics from different contemporary corpora. The article arrives at a different implicational schema than the one traditionally considered to hold. The article suggests that the base form is the bare stem (which is used as the 2nd person imperative and prohibitive, as well as for the negation of the present indicative), and that the other three principal parts are the infinitive; the word-form representing simultaneously the past participle impersonal and the past indicative negative impersonal; and the 3rd person singular of the past indicative. The supine, which is traditionally regarded as the base form, is relegated to being dependent on the 3rd person singular of the past indicative. The article acknowledges that the proposed schema causes difficulties with the algorithm of generating paradigm slots for words that now exhibit a strengthening gradation pattern, traditionally considered to be unproductive for Estonian words and even completely missing for verbs.
Heiki-Jaan Kaalep (b. 1962), PhD, University of Tartu, Institute of Computer Science, Associate Professor in Language Technology (Narva mnt 18, 51009 Tartu), firstname.lastname@example.org