Inflectional morphology plays an important role in language. But how do speakers attach the correct inflectional affixes to the words they produce? This question is addressed in my dissertation project, and I will present the main results here.
The implicit priming task was used throughout. Subjects first studied three associated prime-target pairs. Next, upon seeing the prime they had to produce the target. This task is sensitive to phonological and morphological overlap between the three targets. When there is overlap, subjects show a preparation effect: They are faster to produce the targets.
This preparation effect was obtained for Dutch compounds (werkdruk, werkplaats, werkboek). But when an inflected verb (werkte) was inserted, the preparation effect was strongly reduced. No such reduction was found when using deverbal adjectives or diminutives.
We conclude that the crucial difference lies in the `inflectional frame' of these words. This frame helps combining the stem with the suffixes, and contains one `slot' for every possible suffix. Dutch nouns have two slots (stem, number), verbs need three (stem, tense, number), and the preparation effect is reduced by this difference. Adjectives have two slots (stem, gender), similar to nouns. As a crucial test, we looked at strong past tense forms. These were predicted to have two slots (stem, number) because tense is expresssed by the stem. Indeed, no reduction of preparation was found for these words.
Referat in der Gruppe Sprache I, Montag, 29. März 1999, 11:00, HS 17
