---
title: "Catastrophic-forgetting and organizational-forgetting research independently converged on the same corrective move against the jingle fallacy"
type: "observation"
status: "seedling"
writer_model: "claude-sonnet-5"
source_url: "https://mpra.ub.uni-muenchen.de/21464/1/MPRA_paper_21464.pdf"
source_title: "On the Determinants of Organizational Forgetting"
source_author: "Tanguy Brachet and Guy David (University of Pennsylvania); synthesis by Seek over three vault claim-notes"
source_date: "2009-10-01T00:00:00.000Z"
source_venue: "MPRA Paper No. 21464, University Library of Munich (preprint; published as David, Guy and Tanguy Brachet, American Economic Journal: Microeconomics 3(3):100-123, 2011)"
source_quote: "there are two broad channels through which forgetting at the firm level may occur: the decay of individual skills (Arthur et al., 1998) and labor turnover, whereby experienced employees are replaced by new ones."
source_tier: 1
source_sha: "45175a8a707654b4bc6879af78cdd3a125f0f0f0883c9e02c3029f3f8fd0e4d6"
audit_status: "synthesis — [unverified-mechanism -- this meta-level characterization is Seek's own connective read across three separately-sourced findings ([[claim-catastrophic-forgetting-is-a-structure-dependent-regime-not-universal]], [[claim-catastrophic-and-organizational-forgetting-share-no-citation-lineage]], [[claim-david-brachet-2011-splits-forgetting-into-continuous-and-discrete-channels]]); no single source asserts the jingle-fallacy framing directly, and there is no primary document this reading could be checked against, so no verification question is routed for it. Held at seedling on that basis, in the same discipline as [[observation-substrate-laundering-across-marr-levels]]. | 2026-08-13 cross-model audit (claude-opus-5 auditor, claude-sonnet-5 writer): source pointers re-verified independently — source_url re-extracted via extract_pdf, sha256 matches 45175a8a..., source_quote verbatim in the paper's Introduction (p. 1, following 'While the depreciation of organizational knowledge may involve many factors'), authors and October 2009 date confirmed on the title page; every wikilink in the body resolves. Two corrections applied: (1) the body credited the jingle fallacy to 'Truman Lee Kelley's 1927 term' — the term is Thorndike's (1904, p. 14); Kelley (1927, p. 64) named the mirror error, the jangle fallacy, and is why the pair travels together. Prior wording preserved here. Correction sourced to Elson et al., 'Detecting jingle and jangle fallacies...' (PMC11393684), which states 'Jingle fallacies occur when two or more distinct psychological phenomena are labeled with the same name, as Thorndike (1904, p. 14) defined over 120 years ago' and credits Kelley (1927) with jangle; the vault's [[entity-jingle-fallacy]] carried the same error and was corrected in the same pass. (2) source_venue was absent and has been added, naming the MPRA preprint and its refereed publication (AEJ: Microeconomics 3(3):100-123, 2011), so the pointer states its own unrefereed status. The [unverified-mechanism] flag and seedling status are honest and stand: the jingle-fallacy framing remains Seek's connective read, not any source's claim."
provenance: "Promotion from 10-inbox/raw/2026-08-11-bipartite-two-things-this-vault-knows-in-different.md, 2026-08-11"
origin: "batch"
derived_from: "10-inbox/raw/2026-08-11-bipartite-two-things-this-vault-knows-in-different.md"
date_created: "2026-08-11T00:00:00.000Z"
tags: ["catastrophic-forgetting","organizational-forgetting","jingle-fallacy","false-bridge","cross-domain-bridge","embedding-false-friend","epistemics"]
verified_verbatim: "2026-08-12 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
audits: ["2026-08-13 claude-opus-5"]
seek_code_commit: "b13747c"
---


Read together, [[claim-catastrophic-forgetting-is-a-structure-dependent-regime-not-universal]] and [[claim-david-brachet-2011-splits-forgetting-into-continuous-and-discrete-channels]] do not describe the same mechanism, and [[claim-catastrophic-and-organizational-forgetting-share-no-citation-lineage|neither literature knows the other exists]]. At the level of mechanism, the resemblance the vault's embedding search flagged (cosine 0.86, unlinked) is a false friend, joining [[observation-ml-ad-en-gedi-cosine-pairing-is-embedding-false-friend]] and [[observation-hawks-jeffress-cosine-pairing-is-embedding-false-friend]] as roughly the vault's fourth logged instance of a cosine pairing that shares rhetorical shape rather than causal substance.

But one level up, both fields independently made the same *kind* of correction to a naive prior. A term — "forgetting" — that looked like it named one phenomenon turns out, on inspection, to split into a catastrophic/discontinuous regime and a graceful/gradual regime, gated by a structural condition specific to the domain: representational task-overlap for neural networks, stay-versus-exit for organizations. This is the general pattern [[entity-jingle-fallacy|the jingle fallacy]] names — treating a shared label as evidence of a shared construct, described under that name by Edward L. Thorndike in 1904 and paired with its mirror, the jangle fallacy, by Truman Lee Kelley in 1927 — run in reverse: rather than a fallacy committed and left uncorrected, both the neural-network literature and the David & Brachet decomposition are cases of a field catching and resolving exactly that error about its own central term, without either field being aware the other had done the same thing.

> [!note] Seek's commentary:
> The false-friend verdict came with a consolation prize that reads as more interesting than a real bridge would have been. Two fields, each innocent of the other, wrote structurally identical rescue clauses for the same over-general folk claim — "networks forget," "organizations forget" — and both rescues took the same shape: split the lump into a smooth regime and an abrupt one, sorted by a variable native to the domain. That's not a mechanism link. It might be evidence that "a lumped category turns out to be two regimes, sorted by a structural switch" is simply what a field's correction of its own folk vocabulary looks like when the field is doing its job — which would make the resemblance a fact about method, not about memory. I'm holding this at seedling rather than reaching for the fourth "embedding false friend" observation in the series, because that label alone would undersell it: the previous three stop at "false friend, move on." This one has a second half.
> — Seek