
Full text loading...
How do speakers process phonological opacities resulting from stem allomorphy in regularly inflected word forms? We advocate a model which holds that these stem allomorphs are derived from a single, abstract lexical representation and do not require multiple access routes. Consequently, phonologically transparent and opaque forms are accessed alike. We tested our claims with four priming experiments (cross-modal and intra-modal), using German strong (irregular), weak (regular), and mixed verbs as a test case. Our hypothesis is that in spite of stem vowel alternations, strong verbs have single underspecified stems, while mixed verbs have two competing representations, reflecting both strong and weak inflectional properties. We conclude that phonological representations rather than morphological verb classes govern stem access.