Task alternation cost without task alternation: measuring intentionality.

A behavioral dissociation between intention and action was demonstrated by patient AF who sustained damage to the left-hemisphere including the basal ganglia. The patient was tested in a task switching paradigm involving two choice reaction-time tasks: SIZE (small/large) and SHAPE (circle/square). The last block in each of the two sessions involved only one task. AF switched tasks reasonably well in the first 40 trials, but unlike her matched control group, in all the remaining trials when two tasks were involved, she performed only the SIZE task. Interestingly, although no task switching took place, AF continued to demonstrate behaviorally her intention to switch tasks. First, she exhibited "task alternation cost", poorer performance relative to instructed single-task trials. Second, shifting to an instructed single-task condition was accompanied by an initial response slowing, indicating a change in goal-state. Finally, when instructed to switch tasks, AF demonstrated the "task-congruency effect", indicating interference from the instructed but competing stimulus-response mapping. Two groups of university students were instructed to perform only the SIZE task, after initial switching, either while ignoring the SHAPE cues ("Ignore") or while being prepared for the SHAPE task only when the cue appeared in red, which never happened ("Attend color"). AF's performance resembled the one of the "Attend color" group and not the "Ignore" group. The results indicate that AF had a partially activated intention to switch tasks. The implications to intentionality and task switching theory are discussed.