Parallel response selection in dual-task situations

Semantic priming and response priming were studied in a dual-task procedure. In two experiments, reaction times to the first and second stimuli were faster when the finger required for the Task 1 response was the same as the finger required for the Task 2 response. Such printing suggests that Task 2 response information was generated prior to the completion of Task 1 response selection. These data pose a potential challenge to the response-selection bottleneck (RSB) theory of dual-task performance, since they may indicate a violation of the discrete-stage processing assumption on which the underlying locus-of-slack logic depends. Accommodating these data while preserving the essential bottleneck character of RSB theory may be possible but may also alter the very nature of the bottleneck itself.