ABSTRACT
INTRODUCTION Models for the processing of morphologically complex words range from those in which no word-internal structure is represented (Butterworth, 1986) to those which strip all words to their basic, morphological atoms in a mandatory stage prior to lexical access (Bergman, 1990; Taft, 1985). Among the analytic models,
there is sufficient evidence for processing of morphological stems based on stem frequency effects (Taft, 1979; Taft & Forster, 1975) and patterns of facilitation when stems are repeated (Stolz & Feldman, this volume). Experimental evidence for processing of affixes, however, is less compelling. Moreover, the effects of prefixed and suffixed affixes are not always comparable.
