The "two-route model of face recognition" proposed by Bauer (1984) and adop
ted by Ellis and Young (1990), has become a widely accepted model in studie
s of face processing disorders, including both prosopagnosia and the delusi
onal misidentification syndromes. We review the origin and application of t
he two-route model of face recognition in examining both the neuroanatomica
l pathways and the cognitive pathways to face recognition. With respect to
the neuroanatomy, we conclude that face recognition is subserved by a singl
e pathway, the ventral visual pathway, as there is no evidence to suggest t
hat the dorsal visual pathway is capable of visual recognition or of provid
ing an affective response to familiar stimuli. We demonstrate how operation
of the ventral visual pathway and its connections to the amygdala can pars
imoniously account for the findings in the literature on prosopagnosia and
delusional misidentification syndromes. In addition, we propose a cognitive
model of face processing stemming from the work of Bruce and Young (1986).
Our model involves two pathways subsequent to the system responsible for f
ace recognition: one pathway to a system containing semantic and biographic
al information about the seen face, and a second pathway to a system respon
sible for the generation of an affective response to faces that are familia
r. We demonstrate how this cognitive model can explain the dissociations be
tween overt and covert recognition observed in prosopagnosia and the Capgra
s delusion.