Associative-activation errors are triggered by an internal train of thought firing a related but non-target action

What it is

The trigger for this class of error is internal association — something the person happens to be thinking about automatically pulls in another, related but not currently intended action. A classic associative-activation error looks like this: while meaning to call one person, another related person happens to come to mind at the same moment, and the hand ends up dialing the number of the person who was just recalled instead. What triggers it isn't any externally presented information — it's the person's own train of thought drifting between a few related notions, and cashing out one of the notions along that drift into an action.

Why it happens

Once a thought is activated, it spreads activation along semantic or situational associations to other related thoughts, bringing them close to their own activation threshold — this is how associative memory normally works, and it's usually harmless because only the thought the person actually intends to act on gets cashed out as an action. An associative-activation error happens when one of these "incidentally warmed" thoughts along the chain unexpectedly crosses its activation threshold and gets cashed out into an action ahead of the intended one, at a moment when the person's attention wasn't specifically checking whether the action their hand was carrying out actually matched the thought they meant to act on.

Studying it

This error type was also first collected through diary studies: participants recorded episodes of "thinking about one thing but doing another related thing," and retrospectively noted which related associations had been active in their mind at the time. This kind of self-report faces a specific difficulty — the association itself is rarely clearly conscious, so reconstructing the trigger chain often depends on recalling "what else was I thinking about at the time," and that after-the-fact reconstruction easily misses associative nodes that flashed by quickly and unnoticed, meaning the collected cases may only represent the more self-detectable slice of associative-activation errors.

Where it stops holding

Classifying an error as associative-activation requires confirming that the intruding thought existed only in the person's own mind at the time, not also in perceptual input — without ruling out the latter, there's no way to tell it apart from an error triggered by external data. This boundary depends entirely on whether the person can accurately recall, after the fact, "did I just think of that on my own, or did I also actually see or hear something" — which is just as easily contaminated by recollection, and is why this error type is hard to measure precisely.

Related

• Same group: A10.11.2 This error typically occurs between actions with a semantic or situational link to the current one · A10.11.3 Highly automatized users are more prone to being triggered by an associative chain · A10.11.4 Interrupting the associative chain prevents this error better than a reminder after the fact
• Nearby: A10.10 Data-driven errors · A10.05 Capture errors · A6.08 Retrieval cues and context dependency
• Search terms: associative-activation error · associative memory · action slip