Abstract
The Chinese lexicon is characterized by its typologically unique one-to-one-to-one mapping of morphemes, syllables, and orthographic characters. This architecture poses practical difficulties for the psycholinguist wanting to study lexical processing in Chinese. More seriously, seen as a natural experiment, Chinese challenges assumptions that processing models traditionally make about the roles of phonemes, morphemes, lemmas, and words in lexical access. It is argued that cross-linguistic variation in lexical processing cannot be accommodated by simply modifying lexical processing models, but instead what is needed is a universal learning model. Suggestions are given for how such a model could be tested empirically by extending methods already used for testing language-specific lexical processing.