hide
Free keywords:
-
Abstract:
To what extent is visual attention driven by the semantics of individual objects, rather than
by their visual appearance? To investigate this
we continuously measured eye movements, while observers searched through displays of common objects for an aurally instructed
target. On crucial trials, the target was absent, but the display contained object
s that were either semantically or visually related to the
target. We hypothesized that timing is crucial in the occurrence and strength of semantic influences on visual orienting, and
therefore
presented the target instruction either before, during, or af
ter (memory
-
based search) picture onset. When the target instruction was
presented before picture onset we found a substantial, but delayed bias in orienting towards semantically related objects as
compared to
visually related objects. However, this delay
disappeared when the visual information was presented before the target instruction.
Furthermore, the temporal dynamics of the semantic bias did not change in the absence of visual competition. These results po
int to
cascadic but independent influences of
semantic and visual representations on attention. In addition. the results of the memory
-
based
search studies suggest that visual and semantic biases only arise when the visual stimuli are present. Although we consistent
ly found that
people fixate at locat
ions previously occupied by the target object (a replication of earlier findings), we did not find such biases for visually
or semantically related objects. Overall, our studies show that the question whether visual orienting is driven by semantic c
ontent
is better
rephrased as when visual orienting is driven by semantic content.
