| |
Abstract:
The perception of upright faces may differ from that of
inverted faces in the degree to which they are processed and
represented holistically. To examine this possibility, participants
were instructed to selectively attend to a single feature of a
stimulus face and then to decide whether this feature was the same
or different as that seen in a second, sequentially presented face.
If faces are represented in visual memory as a collection of
explicitly represented facial features, participants should have
the ability to selectively attend to only the relevant feature
while blocking out irrelevant visual information. If, however,
faces are encoded in a holistic fashion with little or no part
decomposition, then the overall similarity of the two faces should
affect participants' ability to compare individual facial features.
Our results show that for upright faces, the alteration of
distractor features had significant effects on both the speed and
accuracy of same-different judgments. Yet when the same pairs of
faces were presented upside-down, distractor features had no effect
on response accuracy and rarely influenced reaction times. These
findings suggest that the human visual system may employ
qualitatively different processing strategies for dealing with
upright and inverted faces: upright faces may be processed by
specialized holistic mechanisms that mandate the direction of
attention to the global form, whereas inverted faces may be
processed by more generalized mechanisms that represent objects in
terms of their component parts.
|
