Introduction

Theories of lexical access and orthographic processing must accommodate an obvious difference between alphabetic/syllabic systems and logographic systems: the componential units of the latter encode semantic information. Although linguistic history has known several logographic and hybrid-logographic systems (e.g. Ancient Egyptian, Sumerian cuneiform, Mayan, Sino-Vietnamese, and Sino-Japanese), written Chinese stands alone as the only system that is both primarily logographic and the chief written system of a living language. Despite a “phonographic tendency,” novel Chinese characters are not freely formed from existing components; Chinese is not “spelled.” Furthermore, the existing phonographic representation is incomplete; phonological information is encoded only loosely at the level of the component, only in some characters, and in phonological units that span a greater distance than the phoneme. The looseness of the phonological mapping arises in part because phonographic Chinese characters represent the multiple phonologies of the speakers who standardized them thousands of years ago. Nevertheless, present-day readers of Chinese manage to cope with the extreme fossilization of its ancient written forms across centuries of phonological change. Therefore theories of reading and word recognition must account for the apparent success of the Chinese reader. In the present chapter we discuss the foregoing issues and their implications for research on the reading of Chinese.

Specifically, we identify several issues that arise when one investigates the processing of Chinese characters, issues that are not necessarily obvious to the non-Chinese psycholinguist.