An important question for research exploring the systems (or
perhaps system) that code facial expression and identity is where
in the face processing hierarchy their representations interact (Calder
& Young, 2005). Skinner and Benton (2012) addressed this
question by testing whether two properties that relate to early
structural processing; idiosyncratic variation in expression production
and invariant aspects of facial features, can account for identity
effects in expression adaptation. Their results showed that
identity effects emerge even when variations in expression production
are controlled (see also Ellamil, Susskind, & Anderson,
2008) and that measures of feature position do not correlate with
aftereffect magnitude. This indicates that the effect of identity congruence
in expression adaptation reflects the operation of higherlevel
facial representations (Skinner & Benton, 2012). In the present
study we tested whether the modulating effect of identity in
cross-emotion aftereffects is explained by differences the intensity
or salience of the adaptor expression (disgust) across individuals.
Results from Experiment 2 indicate that this is not the case since
larger aftereffects were found for same-identity adaptor–target
pairings when the same faces are used as same- and differentidentity
adaptors. Furthermore, in both experiments a size-transformation
between adaptor and target face was used to reduce
low-level adaptation effects. Thus, it would appear that aspects
of identity effects in cross-emotion adaptation also arise at
higher-level representation.