Root and Pattern Effects in the Processing of Spoken Non-Words in Arabic

Abstract

Words in Arabic are formed by mapping roots into patterns. Conducting a series of priming experiments, Boudelaa and Marslen-Wilson (2015, p. 955) have concluded that “root and word pattern morphemes function as abstract cognitive entities, operating independently of semantic factors and dissociable from possible phonological confounds” In the present study, plausibility of this conclusion is tested by investigating native Arabic speakers’ sensitivity to the presence of roots and patterns when processing spoken non-words in Arabic. 50 native Arabic speakers were given a 7-point word-likeness rating task. In this task, participants were asked to rate the word-likeness of 132 auditorily presented non-words in Arabic. 88 of these non-words were created by using real Arabic roots (e.g. /mlk/) that varied in their type and token frequencies and were mapped into two different types of pseudo patterns. Results have shown that native Arabic speakers are sensitive to the presence of roots in the non-words. Specifically, root type frequency had the strongest effect on subjects’ ratings of the non-words in both types of patterns. Implications of these findings to theories of the Arabic mental lexicon will be discussed.