| Summary: | Abstract: This study investigated the nature of the retrieval architecture of Semitic morphemic entities in word production in Hebrew, a language with a non-concatenated morphology. By taking advantage of the potential for dissociation of form and meaning in Hebrew, we explored the relative contribution of word-form and semantics to morphological processing. In two picture–word interference (PWI) experiments we investigated the lexical status of the two primary morphemes: the root and the word pattern (WP), with the effects of morphological, phonological and semantic relatedness between written distractors and picture names. Facilitation effects were observed for both the root and the WP. Moreover, neither semantic nor phonological similarity affected picture naming in the absence of a morphological relation. Although both the root and the WP induced equally robust morphological effects, we postulate a typical and distinct processing mechanism for each type of morpheme. We interpret our findings as evidence for a parallel processing mechanism of morphemic entities, which is implemented in part at the lemma level and in part at the lexeme level, where the output of both processes is integrated. Implications of these findings for the retrieval architecture of morphemic entities are discussed.
|
|---|