Theory 22/Life, Medicine & Mind /Memory science
Reconsolidation Needs Surprise
Reconsolidation is triggered by recall. It may require recall that contradicts.
A memory reconsolidates only when the prediction error at retrieval exceeds a threshold, so retrieval that confirms expectations leaves the memory stable and cannot be modified.
01The evidence
Reconsolidation is observed in some retrieval paradigms and not others, with the difference tracking whether the outcome was surprising.
02The tension
Models treat retrieval as sufficient. That predicts modification whenever a memory is recalled, which is not observed.
03The mechanism
A memory is updated only when it fails to predict. The error signal opens a window in which the trace can be rewritten; without it, retrieval strengthens rather than destabilises.
That makes prediction error the gate, and predicts that confirming retrieval should prevent modification.
04What this is not
Those models use recall as the trigger. This uses surprise, which explains the failures.
05Predictions
- P1Modification should require prediction error.
- P2Confirming retrieval should strengthen without updating.
- P3Error magnitude should gate the window.
06Refutation
Modification without prediction error would remove the gate.
07What it changes
It would change how traumatic memories are treated.