Parietal cortex has been implicated in the updating, after eye movements, of a salience map that is required for coherent visual experience and for the control of visually guided behavior. The current experiment investigated whether TMS over anterior intraparietal cortex (AIPCx), just after a saccade, would affect the ability to update and maintain a salience map. In order to generate a salience map, we employed a paradigm in which an uninformative cue was presented at one object in a display to generate inhibition of return (IOR)—an inhibitory tag that renders the cued object less salient than others in the display, and that slows subsequent responses to visual transients at its location. Following the cue, participants made a saccade to either left or right, and we then probed for updating of the location of IOR by measuring manual reaction time to targets appearing at cued location of the cued compared to an uncued object. Between the time of saccade initiation and target appearance, dual-pulse TMS was targeted over right (Experiment 1) or left AIPCx (Experiment 2), and a vertex control side. Updating of the location of IOR was eliminated by TMS over right, but not the left, AIPCx, suggesting that right parietal cortex is involved in the remapping of IOR. Remapping was eliminated by right AIPCx, regardless of whether the saccade was made to the left (contralateral), or right (ipsilateral) visual field, and regardless of which field the target appeared in. We conclude that right AIPCx is the neural substrate for maintaining a salience map across saccades, and not simply for propagating an efference copy of saccade commands.

Parkinson patients were tested in two paradigms to test the hypothesis that the basal ganglia are involved in the shifting of attentional set. Set shifting means a respecification of the conditions that regulate responding, a process sometimes referred to as an executive process . In one paradigm, upon the appearance of each stimulus, subjects were instructed to respond either to its color or to its shape. In a second paradigm, subjects learned to produce short sequences of three keypresses in response to two arbitrary stimuli. Reaction times were compared for the cases where set either remained the same or changed for two successive stimuli. Parkinson patients were slow to change set compared to controls. Parkinson patients were also less able to filter the competing but irrelevant set than were control subjects. The switching deficit appears to be dopamine based; the magnitude of the shifting deficit was related to the degree to which l-dopa-based medication ameliorated patients' motor symptoms. Moreover, temporary withholding of medication, a so-called off manipulation, increased the time to switch. Using the framework of equilibrium point theory of movement, we discuss how a set switching deficit may also underlie clinical motor disturbances seen in Parkinson's disease.

The contributions of the superior prefrontal cortex (SPFC) and the superior parietal lobule (SPL) in generating voluntary endogenous and reflexive visually guided saccades were investigated using transcranial magnetic stimulation (TMS). Subjects made choice saccades to the left or right visual field in response to a central arrowhead (endogenous go signal) or a peripheral asterisk (exogenous go signal) that were presented along with a single TMS pulse at varying temporal intervals. TMS over the SPFC increased latencies for saccades made in response to an endogenous go signal toward the contralateral hemifield. No effects were observed when the go signal was exogenous and TMS was over the SPFC or when TMS was over the SPL for either saccade type. The delayed contralateral endogenous saccades observed in this study are likely a consequence of disruption in the normal operations of the human frontal eye field.

Five patients with visual extinction following unilateral brain injury were briefly presented with colored letters in either or both visual fields, and required to report and locate the colors or the shapes. On double simultaneous stimulation, they tended to miss the event contralateral to their lesion. This extinction was increased when the two stimuli were the same on the reported dimension, Similarity on the irrelevant dimension had no effect. These data suggest that extinguished colors and shapes may be correctly extracted by the visual system (when task-relaant) even though they are unavailable for verbal report. An analogy is made with the phenomena of “repetition blindness” in normal observers, and it is proposed that extinction may reflect failure in a token-individuation process for correctly extracted visual types.