Psychological refractory period

A tone asks for one response; a color arrives soon afterward and asks for another. You may recognize the color promptly yet answer it late. This second-response delay is the psychological refractory period. What makes the second response wait: choosing what to do, timing the action, or deciding which task goes first?[1] [2]

Imagine hearing a tone that tells you where to look, followed quickly by a color that tells you which finger to press. Both signals can enter the brain while the second response waits. What event allows its processing to resume? One account places the limit at response selection, meaning choosing the appropriate action; an initiation-timing account places it at programming when that action begins.

In a 2025 human study with learned sound-to-eye and color-to-finger mappings, brain patterns linked to the two tasks queued serially in a multiple-demand network and relevant motor areas. Sensory processing was largely parallel. The multiple-demand network is a set of regions recruited by many demanding tasks; the observed ordering supports a central processing limit in this experiment.

A competing review argues that response-initiation timing explains important behavioral costs. The imaging study’s Granger analysis asks whether earlier activity helps predict later activity elsewhere; it does not experimentally establish which region opens a causal gate.[1] [2]

A 2025 imaging study found sequential brain-activity patterns in trained tasks linking sounds to eye movements and colors to finger presses. That supports a queue in this setting, but does not establish which event releases it across other tasks or whether motor regions must always be involved.[1] [2]

The imaging result narrows the candidate stages but does not show that motor cortex is necessary for every such delay. Its authors explicitly leave other task combinations to be tested. Consider keeping the same color choice but postponing the finger press: a queue tied to making the choice and one tied to starting the action need not predict the same change. That is a proposed distinction, not a result demonstrated by this study. Separating choice demands from action timing would help determine what the observed serial order means.[1] [2]

Multitasking can fail even when each task is easy on its own. Understanding the waiting period would help explain how recognizing two signals differs from responding to both. It would also help distinguish a change that reduces interference from one that merely gives a different task priority.[1] [2]

This is a question about how flexible cognition is coordinated. A system that can recognize several signals together may still need to arrange their decisions or actions in sequence, so parallel sensory processing does not guarantee parallel performance.

The distinction also changes how a multitasking improvement should be interpreted. If a hypothetical instruction makes the color response faster by delaying the tone response, the second task’s gain alone is not evidence that the total coordination cost fell. Explaining both responses would connect a neural processing account to the tradeoffs people actually display, without assuming one universal scheduling rule.[1] [2]