Tracking the Stem and Root Morphemes in Arabic: Evidence from Visual Morphological Priming.

Abstract

The status of the consonantal root in Arabic (and Semitic, in general) has been subject to controversy in both theoretical and experimental research. While most priming experiments reported root effects in Arabic, a few reported either inconsistent effects or no such effects at all, questioning the morphemic status of the root and pointing to the possible role of the stem, instead. We carried out a visual lexical decision masked priming experiment to explore the extent to which the stem may prime lexical access in Arabic. The same target (ya-ʕtamid "he approves") was preceded by a prime that was either (i) an inflectional relative, member of the same lexeme, with which it shares the surface stem (na-ʕtamid "we approve"), (ii) a derivationally close relative, member of a separate lexeme, with which it shares the same abstract stem (mu-ʕtamid "approving"), or (iii) a derivationally distant relative with which it shares the root consonants only (ʕtimaad "approval"). A phono-orthographic condition (iv) was used as a baseline in which the target was preceded by an unrelated prime with which it shares all but one or two sounds of the surface form (yaʕtaqid "he believes"). Results show a gradient facilitation effect across all four conditions and a significant main effect of priming in the three related conditions. They further reveal that Condition (i) showed significantly more priming than Conditions (ii) and (iii). In Condition (iv), reaction times were much slower compared to Conditions (i) and (ii), but not as much when compared to Condition (iii). These results suggest that words sharing the inflectional stem prime each other more than words sharing the derivational stem or abstract consonantal root and argue in favor of gradient salience of the stem and root and their roles in Arabic lexical organization. We discuss the implications of these results to the theoretical debate surrounding the status of the root and stem in Arabic and argue for a hybrid model of the Arabic lexicon and morphology in which a word can be accessed both through its root and through its stem.