What gaslighting takes is not the memories. It is the calibration.
Paige Sweet's survivor interviews describe the residue precisely: people who retain accurate recollections and have lost the ability to rely on them. Chronic second-guessing. Deferring to other people's accounts of events they witnessed. Decision paralysis, because deciding requires trusting your own read of the situation. A learned reflex of asking someone else whether that really happened.
The reason this cannot be repaired with reassurance is chapter 65's finding. Loftus established that repeated authoritative contradiction genuinely degrades confidence in accurate recall — the confidence is not merely emotionally suppressed, it has been reduced by an actual process operating on an actual memory system. Telling someone their memory is fine does not restore a calibration that has been trained out.
The repair therefore has to be evidential.
Rebuild a documented timeline. Not from memory alone, which is the thing in question, but from artifacts: messages, emails, calendar entries, photographs with dates, bank records, medical notes, receipts, journey histories. These have timestamps, and timestamps do not argue.
What survivors consistently report on doing this is that the artifacts corroborate them. The events happened as remembered. The confidence was wrong and the memory was right, and seeing that demonstrated in a form that cannot be contested is the mechanism of the repair.
Third-party corroboration does the second half, and it is worth pursuing even when it feels like an imposition. One person who was present, or who was told at the time, converts a private account into a checkable one — and Sweet's central sociological point is that gaslighting was enabled by other people's willingness to doubt you, which means restoring credibility is partly a social task rather than an internal one.
Calibration practice is the ongoing part, and it is straightforward. Make small predictions and record them. Write down what you expect to happen, and check. Note your confidence and see whether it tracks. This is the same discipline forecasters use, and it rebuilds the association between a feeling of certainty and being right, which is what was severed.
Two cautions. Do not do this in contact with the person who caused it, because each interaction re-runs the mechanism. And expect the process to be slower than it seems it should be, because a trained reflex extinguishes on its own schedule.
This counters Law 17.
Counters Law 17 — Repeat It Until It Feels True