14 January 2016·Attention and Memory
Attention Is Not One Thing
Ordinary speech treats attention as a single quantity. People say they have run out of it, that a task demands more of it, that a child has too little. Experimental psychology has never been able to use the word that simply. When the topic was first studied under laboratory conditions, the question was narrow: given far more sensory information than a person can act on, what determines which part gets through?
The classic answer was a gate. Signals arrive in parallel, the argument ran, and something early in the chain selects one stream for fuller handling while the rest fade. Listening experiments, in which two different spoken messages are delivered at once and the listener is asked to repeat only one, gave the picture its early support. Listeners could follow one voice well and report almost nothing about the words in the other.
Where the simple gate broke down
The difficulty was that the unattended stream did not go entirely dark. Certain material, notably a listener's own name, tended to break through. That result forced a revision: if nothing about the ignored message were processed, nothing in it could be recognised as worth interrupting for. Later accounts moved the gate later in the chain, or replaced the gate with a graded attenuation, or dropped the single bottleneck altogether in favour of limited resources shared across tasks.
Permalink: /notes/attention-is-not-one-thing