Hemispheric Asymmetries in Face Recognition and Naming: Effects of Prior Stimulus Exposure

Abstract

Detailed models of the cognitive operations subserving performance in a number of processing domains such as reading, spelling and object recognition have become increasingly prevalent in both cognitive psychology and neuropsychology. Such models attempt to distinguish between functionally separable components of processing, and to describe their organisation and operation. The study of face recognition has recently been the subject of the development of such models, and in the last three or four years at least four models have been offered (Bruce, 1983; Ellis, 1983; Hay & Young, 1982; Rhodes, 1985). These models are similar in that they distinguish between the following processing components (or stages) in face recognition: 1) visual-spatial processing; 2) a face representational system which permits recognition of known or familiar faces; 3) semantic information; and 4) naming. The precise details of the model differ, however, in many respects and in particular in the characterisation of how the system which permits the recognition of a face as a known or familiar face (without necessarily fully accessing semantic and name information) is assumed to operate. Ellis’ (1983) model posits a “familiarity check”, whereas Bruce’s (1983) model would suggest that the “structural code” would be accessed, and in Hay & Young’s model known faces may be classified as such when they gain access to “face recognition units” which may be seen as being face analogs of input logogens in Morton’s (1979) model of visual word recognition. Finally, Rhodes’ (1985) model would appear to involve the accessing of her posited processing stage labelled “view-independent topographical representation”.