Abstract
Attentive selection can be constrained by simple features, in addition to locations and objects. Using fMRI, Saenz et al [Nature Neuroscience, 5, 2002, 631–632] showed that task-irrelevant stimuli in the attended color receive facilitated processing even when presented at an ignored location. However, this result appears to conflict with previous ERP studies in which clear modulations of feedforward sensory processing (i.e., modulations of P1 amplitude) have been observed only for spatial attention and in which nonspatial attention effects have been observed only at attended locations. To resolve this discrepancy, we recorded ERPs in a paradigm modeled after that used by Saenz et al (2002). Random dot patterns consisting of a mixture of isoluminant green and red dots moved randomly within a circular aperture in one visual field. Participants were instructed to attend to either red or green in separate blocks and to make a button response whenever they detected a luminance change of the attended color. This was a demanding task that required subjects to attend selectively to one color and ignore the other color. In the opposite visual field, task-irrelevant random dot patterns (probes) with either green or red dots were flashed every 400–800 ms. Participants were instructed to fixate a central point and EOG recordings were used to ensure compliance. The task-irrelevant probe in the attended color elicited a larger contralateral P1 component (with an onset at 125 ms) than the probe in the ignored color, even though the probes were presented at an unattended location. The same effect was observed, but with an onset latency of approximately 105 ms, when the stimulus contrast was increased. Thus, color-based selective attention influences feedforward sensory processing as indexed by P1 component, even at unattended locations.
This research was made possible by grants R01 MH63001 and R01 MH65034 from the National Institute of Mental Health