Degree of productivity differentially affects priming of suffixed words in English

Convergence Theory (CT) suggests that morphological effects are an emergent product of the convergence of orthographic, phonological, and semantic codes. At its core, CT attempts to capture the quasi-regularities ubiquitous in "morphological" phenomena. Consistent with the CT framework, recent research with adults has demonstrated that the magnitude of priming for pairs such as, teacherteach, is moderated by the degree of semantic and phonological overlap between the prime and the target. An additional factor that seems central to "morphological" processes is productivity. Productivity is central to theories about the development of different morphemes and may affect processing speed for complex words. The purpose of the current project was to look at certain productive and nonproductive (less productive) affixes and how they affect the speed of processing for complex English words. The role of productivity in priming suffixed English words was examined using a lexical decision paradigm. The first study focused on the agentive morphemes -mall, -cr, -iall, and -ist which vary in productivity; -er and -man are highly productive whereas -ist and -ian are not. The second study focused on the adjectival morphemes -y, -ish, and -ous which also rary ill productivity; -y is "cry productil'c whereas -ish alld -ous arc less productil·e. Although the results of both experiments did not support the specific prediction that more productive suffixes would prime more. the results were consistent with the general hypothesis that suffixes with different degrees of productivity would produce graded priming effects. Degree of Productivity Differentially Affects Priming of Suffixed Words in English Inflectional morphology has been at the center of a long-standing debate concerning how morphologically complex words are represented. Rules have been used to capture the productive aspect of our generative language capacities and it is the productivity of rules that make them so powerful and useful in a domain where there is considerable regularitl. At the no-rules end of the extreme are those that argue that complex words are stored as "gestalt" forms (whole words) (Butterworth, 1983). At the all-rules end are those who suggest that complex words are stored according to their stems with appropriate ~ffixes added according to rules (Taft & Forster, 1975). However, most models draw a line somewhere in between, suggesting both the use of rules and whole word memory (Marslen-Wilson, Tyler, Waksler, & Older, 1994). Interestingly, the very notion of a complex word presupposes that it derives from more primitive constituents (i.e. morphemes) and, consequently, morphemes are defined as the "minimal meaning bearing units" i.e. the foundational semantic building blocks. It is these building blocks that must be used (via rules) in order to form complex words. For a canonical example, the word cars is considered to be composed of the morphemes car (a vehicle for transportation) and 'sO (the pluralization of a word). Cars would constitute a complex word that is produced by the application of a rule: add's' to a singular noun to make it plural. In I If morphology is also quasi-regular as has been argued for by others (Bybee. 1985: Harm & Seidenberg. 1999: Plaut. ~tclelland. Seidenberg. & Patterson. 1996: Seidenberg & ~teClelland. 1989). then there ought to be degrees of productivity to capture that quasi-regularity. ~tore on this below.