CNL Lunch

Natalia Slioussar & Kira Gor

Russian Verb Morphology and the Structure of the Mental Lexicon

Thursday May 2nd, 12:30pm, 3416 Marie Mount Hall

There was a big debate in linguistics about formation of English Past tense. It was one of the "classic" things that connectionist networks were taught to do. These models performed considerably well, and this fact let their proponents make a number of hypotheses both about how Past tense forms are acquired and stored, and about mental lexicon in general. This theory, however, received strong criticism from such linguists as Steven Pinker, Gary Marcus and others, who proposed an alternative view on how mental lexicon functions and, in particular, what is going on with the Past tense. These linguists postulated separate mechanisms for the processing of regular and irregular forms, while in connectionist approach there was no such distinction.

As the issue became very exciting, speakers of languages with rich inflectional
morphology decided to participate in the debate. I take part in the research project studying Russian verb morphology from this perspective. Our experiments were designed to show how the notion of regular versus irregular processing becomes much more complex and controversial when a language such as Russian is involved. English has one regular verb class and a number of irregular verbs. Russian has 23 verb classes, a number of irregular verbs and several anomalous ones plus different endings for number and person. Regular verb classes differ in productivity, type frequency and morphological complexity. How do children acquire these riches? How do second language learners cope with them? How do native speakers produce forms from nonse verbs? Are different forms in the paradigms of productive regular classes, which are claimed to be assembled on line, equal?