Last updated: 2026-10-10

Can an Organisation Unlearn?

An earlier page in this cluster opened with a finance team still running a monthly reconciliation step built for a vendor dispute that ended three years ago — a step that costs real effort and protects against nothing anymore, kept only because nobody with the standing to question it ever looked at it directly. That page asked whether an organisation can forget on purpose at all. This page asks a sharper question underneath it: when an organisation does manage to let go of something, what exactly is it letting go of, and which part of "forgetting" is actually happening?

Five Things Easily Collapsed Into "Forgetting" FoundationalKnowledge that endures for decades — core principles

Sharpening the voluntary/involuntary split this cluster has already drawn — a distinction this page makes, not one the earlier page's framework spells out in quite this much detail — at least five different operations get called "forgetting" in an organisational setting, and they don't all require the same thing to succeed. Deletion is removing a record: deleting a document, archiving a system, striking a line from a policy manual. Inhibition is preventing an old response from dominating without removing the knowledge behind it at all — the old process is still known, it's just no longer the default anyone reaches for. Unlearning is weakening a previously reliable association directly, so the old connection between a situation and a response genuinely gets harder to make, not just less convenient. Recontextualisation retains the knowledge entirely while changing when it's considered relevant — an old safety procedure isn't forgotten, it's correctly filed as "what we did before the new equipment," available again if the old equipment ever comes back. Cultural forgetting is the loss of the shared practices that made something meaningful in the first place, which can happen even while every document describing it survives untouched.

Why Deleting a Record Doesn't Delete a Habit FoundationalKnowledge that endures for decades — core principles

The reconciliation-step case is, in this finer vocabulary, a failure of inhibition rather than a failure of deletion — the policy document could be deleted outright tomorrow, and the habit of running the step would very likely survive the deletion, because the habit was never stored in the document to begin with. Removing a policy doesn't remove the habit it produced. Reorganising a department doesn't erase its former boundaries, which keep showing up in who people still email first. Renaming a role doesn't immediately change the social expectations attached to whoever used to hold the old title. Each of these is a case where deletion succeeded completely and the organisation still behaved as though nothing had changed, because deletion was never the operation that mattered — inhibition was, and nothing was done to achieve it.

Cultural Forgetting: When the Record Survives and the Meaning Doesn't FoundationalKnowledge that endures for decades — core principles

Karl Weick's account of organisational sensemaking treats meaning-making as fundamentally social rather than something an individual record carries on its own — a cue only becomes meaningful through a community that already shares the expectations needed to interpret it a particular way¹. This cluster's page on institutional confabulation already shows one way that community's account of its own past can drift from what actually happened. Cultural forgetting is the harder case: a stored artefact can survive physically, completely intact, while becoming cognitively inert, because the community of people who shared the practices, the jargon, and the recognised examples needed to read it the way it was originally meant has itself dispersed. A ten-year-old design document can be perfectly legible as English and functionally unreadable as an organisational artefact, because the local meanings of "quality," "ready," or "the usual process" it was written against no longer exist anywhere, including in the document itself.

Strongest Objection: This Taxonomy Sharpens the Diagnosis, Not the Problem FoundationalKnowledge that endures for decades — core principles

The earlier page in this cluster already raises the hardest problem here: deliberate, well-examined forgetting is easy to fake after the fact, because a lapsed safeguard and a properly retired one can look identical in the record. This page's five-way distinction doesn't solve that problem. It gives a more precise vocabulary for diagnosing which specific operation failed when "we decided to retire that" turns out not to have actually happened — the reconciliation step is a deletion that was never attempted alongside an inhibition that was never achieved, and knowing that distinction doesn't tell you whether the original decision to drop it was ever properly examined in the first place.

Provisional Conclusion FoundationalKnowledge that endures for decades — core principles

Can an organisation forget on purpose" turns out to bundle at least five separate, independently achievable operations, and most organisational unlearning efforts succeed at only one of them — usually deletion, the easiest and least sufficient — while assuming the others followed automatically. A habit survives a deleted policy. A boundary survives a department's reorganisation. A document survives the community that could once read it correctly. Unlearning that actually works has to target whichever of the five is load-bearing for the specific thing being let go of, not assume that removing the record is the whole job.

Questions for Further Thought

Is inhibition ever achievable by policy alone, or does it always require the kind of repeated, lived counter-example that only time and practice can supply?
Can cultural forgetting be detected before it's complete — is there a reliable warning sign that a document's interpreting community is thinning, while some of it still remains?
Does recontextualisation ever fully substitute for deletion, or does retained-but-rarely-relevant knowledge always carry some risk of being reached for in the wrong context later?

Can an Organisation Forget on Purpose? — the voluntary/involuntary split this page sharpens into five distinct operations.
Institutional Memory or Institutional Confabulation? — what happens to a community's account of its own past, the social mechanism this page's cultural forgetting depends on.
Is a "Learning Organisation" a Real Claim or a Compliment? — double-loop learning as one mechanism capable of achieving inhibition and unlearning, not just deletion.

References

Weick, K. E. (1995). Sensemaking in Organizations. Sage Publications. ↩