There is growing evidence that face recognition is ''special'' but les
s certainty concerning the way in which it is special. The authors rev
iew and compare previous proposals and their own more recent hypothesi
s, that faces are recognized ''holistically'' (i.e., using relatively
less part decomposition than other types of objects). This hypothesis,
which can account for a variety of data from experiments on face memo
ry, was tested with 4 new experiments on face perception. A selective
attention paradigm and a masking paradigm were used to compare the per
ception of faces with the perception of inverted faces, words, and hou
ses. Evidence was found of relatively less part-based shape representa
tion for faces. The literatures on machine vision and single unit reco
rding in monkey temporal cortex are also reviewed for converging evide
nce on face representation. The neuropsychological literature is revie
wed for evidence on the question of whether face representation differ
s in degree or kind from the representation of other types of objects.