ABSTRACT
Visual words and faces differ in their structural properties, but both are objects of high expertise. Holistic processing is said to characterize expert face recognition, but the extent to which whole-word processes contribute to word recognition is unclear, particularly as word recognition is thought to proceed by a component-based process. We review the evidence for experimental effects in word recognition that parallel those used to support holistic face processing, namely inversion effects, the part-whole task, and composite effects, as well as the status of whole-word processing in pure alexia and developmental dyslexia, contrasts between familiar and unfamiliar languages, and the differences between handwriting and typeset font. The observations support some parallels in whole-object influences between face and visual word recognition, but do not necessarily imply similar expert mechanisms. It remains to be determined whether and how the relative balance between part-based and whole-object processing differs for visual words and faces.
Acknowledgements
We thank Oded Rock and Mehar Singh for comments and review, C Barton for contributing a handwriting sample, and J Davies-Thompson and S Corrow for lending us their (morphed) facial images, with permission.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 There are exceptions in languages written in boustrophedon fashion (alternating the direction of writing between lines), such as the vertically reflected letters in some ancient Greek texts (Jeffery, Citation1948) and the inverted words of rongorongo on Easter lsland (Métraux, Citation1940).
2 Note, though, that the term ‘word form’ in this study referred not just to whole words, but also to other familiar units such as graphemes, syllables and morphemes.
3 One might argue that the contrast between words of high and low frequency in a familiar language is a better parallel. The parallel may also depend on whether the unfamiliar language uses similar components – e.g. English/French – or not – e.g. English/Chinese.
4 Mooney faces take greyscale images and depict them in binary black and white values, which is purported to emphasize configural processing.