Effects of the relationships between forms within and across paradigms on lexical processing and representation

The frequency and distribution of forms within a lexeme’s paradigm affect how quickly forms are accessed (e.g., Kostić, 1991; Milin, Filipović Đurđević, & Moscoso del Prado Martín, 2009; Moscoso del Prado Martı́n, Kostić, & Baayen, 2004). The distribution of forms across paradigms, in contrast, has received little experimental attention. Theoretical studies investigate the distribution of forms across paradigms because forms vary in how predictive they are of other (unknown) forms. Such investigations have uncovered typological tendencies (e.g., Ackerman & Malouf, 2013; Stump & Finkel, 2013) and contribute to explanations of language-specific phenomena (e.g., Sims, 2015; Parker & Sims, To appear). The intersection of these research approaches raises questions about how the distribution of forms within and across paradigms affects lexical access and representation. Based on forms of Russian nouns representing two morphosyntactic property sets and lexemes from three inflection classes, it is shown that speakers are sensitive to differences in form and morphosyntactic property set in a visual lexical decision task. In a priming task, nominative forms prime locative forms better than vice versa regardless of suffix, despite differences between the same forms in the lexical decision task. These results suggest that speakers make generalizations about forms across classes, including at the level of word forms and morphosyntactic property sets.