Why markedness is always local: the case of Russian aspect

Abstract

Markedness is the observation of an encoding asymmetry in which higher complexity (both in terms of form and meaning) tends to pattern with lower frequency. Given that markedness focuses on the relationships between form-meaning patterns and usage patterns, markedness is of inherent theoretical interest for cognitive linguistics. Traditionally it is assumed that the markedness values of Russian aspect are perfective as marked vs. imperfective as unmarked. However, this assumption is not without controversy and conflicts with patterns observed in other languages. Furthermore, neither complexity of form nor corpus frequency support the traditional markedness assumption at the level of the category. We drill down to the levels of both the lexeme (groups of verbs defined by the major patterns of aspectual morphology) and the subparadigm (tense) and find better support for observation of markedness at these levels. While one group of verbs supports the traditional assignment of perfective as marked, two other groups of verbs support the opposite, with imperfective as marked. The subparadigm level of tense presents a special challenge since, due to confounding factors of homonymy and non-contiguous forms, no previous study has presented an accurate measurement of the incidence of future tense forms in Russian. We overcome this obstacle by examining a stratified set of verbs, sampling and manually tagging forms, and then using the sample data to extrapolate a reasonable estimate of future tense forms. We find that perfective future tense forms are approximately fourteen times more frequent than imperfective future tense forms. Russian future tense forms give strong support to the recognition of imperfective aspect as marked due to higher morphological complexity and much lower frequency. We conclude that it makes more sense to evaluate markedness patterns at local levels rather than at the category level.