Abstract
According to Farah's (1991) framework, recognition of different objects can be characterized by a continuum with faces at one extreme involving holistic processing, words at the other extreme involving part-based processing, and other objects somewhere in between. Face perception requires fine, subordinate-level individuation of similar objects, and it is thought that holistic processing is developed as an optimal strategy to meet this recognition demand. More recent research has challenged the unique status of faces by showing that holistic processing is also developed for other homogeneous object categories after one has acquired expertise in individuating these objects. These findings suggest that holistic processing might be a more general hallmark of expertise, and raises the possibility that holistic vs. part-based processing may not be the critical dimension on which to differentiate processes involved in face, object, and word processing. We tested whether holistic processing would also occur for word perception, which has been shown to be a different type of perceptual expertise emphasizing basic-level categorization instead of fine individuation. We applied the often-used face composite task to English words and asked participants to match the target halves (left or right) of two 4-letter words while ignoring the irrelevant half. Clear evidence of holistic processing was found for words similar to that typically found for faces. In a second experiment, we found that holistic processing was larger for more frequently encountered stimuli (i.e., words compared with pseudowords), and larger for experts (i.e., native compared with second-language readers), indicating that holistic processing of words is sensitive to the amount of experience. Ongoing experiments are showing holistic processing for Chinese characters in experts. Overall, results cast doubt on the simple distinction between holistic face processing and part-based word recognition, and call for a richer cognitive model for explaining perceptual phenomena of different object categories.
This research was supported by the Direct Grant (2020939) from the Chinese University of Hong Kong and the General Research Fund (452209) from the Research Grants Council of Hong Kong to A.W. and by a Faculty Research Grant from the University of Richmond to C.B., and Undergraduate Research Grants from the University of Richmond to S.L., and E.G.