Slavic evidence for the existence and organization of word-level paradigms

One of the first, and most urgent, tasks for cognitive linguists is to resurrect taxonomy, marginalized by generativists,
as a legitimate concern of linguistic analysis. For morphologists, this means attention to the psychological underpinnings of
word-level paradigms. Although it is true that such paradigms are, in simple terms, merely sets of words established on the basis of one or more shared attributes (e.g., tense, aspect), the activity responsible for them -- classification -- is one which characterizes our daily lives. There is, accordingly, no reason not to assume that it characterizes our interaction with language as well, and is therefore a component of linguistic competence as deserving of our attention as any other. The importance of paradigms to the morphologist, however, goes well beyond their mere existence because, although the fact is frequently ignored in contemporary theories of word-structure, paradigms provide the domains of opposition within which word-level forms are evaluated by speakers for the synchronic assignment of grammatical meaning to desinential phonemes, and within which such forms evolve diachronically. Thus, for example, the evaluation of -te as a minimal sign meaning 'second person plural' in certain Slavic languages is based on the occurrence of the relevant forms (e.g., Macedonian second person plural present nosite 'you carry') within six-member paradigmatic unities in which forms are opposed in person and number. In short, the analysis of word-level constituent structure, synchronically and diachronically, is impossible without prior knowledge of, or assumptions relating to, paradigmatic organization. The bases of this organization, therefore, are central to a fuller understanding of the cognitive relationship between form and meaning in word-level units.

In this paper, I will present Macedonian verbal data to support the existence, or psychological reality, of word-level paradigms, and, in so doing, I will attempt to demonstrate the role they play not only in the assignment of grammatical meaning to desinential phonemes, but also in the evolution of such units. I will argue that the shared attribute(s) constituting the basis of word-level paradigms may be grammatical or nongrammatical, and that paradigmatic structure within the conjugational unit may go well-beyond the six-member paradigms based on tense (i.e., the paradigms we traditionally label present, imperfect, aorist and imperfect) that we are accustomed to recognizing. That is, I will argue that subsets of such paradigms may themselves constitute superordinate paradigmatic units.