Do organizational-forgetting reference lists carry copy-fingerprints?

citation-copyingorganizational-forgettingbibliometricscitation-genealogycross-domain-bridgesource-verification

Routed from question-do-organizational-forgetting-reference-lists-show-copy-fingerprints, which asks whether the organizational-forgetting citation cluster (Argote, Beckman & Epple 1990; Benkard 2000; David & Brachet 2011; and their citers) shows the Simkin & Roychowdhury copy-fingerprint pattern — identical idiosyncratic errors or shared unlikely references — which would indicate the literature's absence of ML citations (claim-david-brachet-2011-reference-list-contains-zero-ml-citations) is inherited via unread copying rather than a genuine independent decision, per observation-copying-model-predicts-cross-field-citation-silence.

This capture attempts a direct, small-scale version of the audit the question calls for: fetch primary reference lists in the cluster and hand-compare their citations of a shared predecessor for identical wording/misprints. Only one comparable pair proved fetchable with this session's tools.

Note on the single-source concentration cap: three claim-notes already rest on the David & Brachet MPRA PDF (source_sha 45175a8a…) — claim-david-brachet-2011-reference-list-contains-zero-ml-citations, claim-david-brachet-2011-turnover-channel-twice-skill-decay-channel, and claim-david-brachet-2011-splits-forgetting-into-continuous-and-discrete-channels — which per 00-meta/specs/sources.md already reaches the cap for that document. The claim below partially rests on the same PDF (paired against an independent second document, Benkard's NBER working paper, for a genuinely new comparison); promotion should weigh whether it needs to route through a corroboration question rather than a fourth direct claim-note on that primary.

Claim: The one directly comparable citation in this cluster shows divergent wording, not a shared misprint

Claim type: historical/bibliographic. Tier 3-4 acceptable for "what does this reference list say," escalated here to Tier 1 because both texts were read directly from their own primary PDFs via extract_pdf, with sha256 receipts recorded above. verifies: question-do-organizational-forgetting-reference-lists-show-copy-fingerprints

Both David & Brachet (2011) and C. Lanier Benkard (2000) cite the same predecessor, Argote, Beckman & Epple (1990), and both reference lists were read directly. The two renderings of the title differ:

Benkard's NBER working-paper reference list reads: "Argote, Linda, Beckman, Sara L. and Dennis Epple. 'The Persistence and Transfer of Learning in Industrial Settings.' Management Science, February 1990, 36(2), pp. 140-154." — this matches the paper's actual published title exactly (plural "Settings," no indefinite article; confirmed against the paper's title as it appears consistently across IDEAS/RePEc, EconPapers, and ACM Digital Library listings).

David & Brachet's MPRA working-paper reference list reads: "Argote, Linda, Beckman, S. and Epple, Dennis. (1990), 'The persistence and transfer of learning in an industrial setting', Management Science, Vol. 36, No. 2, February, pp. 140-154." — this alters the title (singular "setting," inserted indefinite article "an").

Under the Simkin & Roychowdhury logic, an identical error recurring across reference lists is the fingerprint of unread copying; a divergence between two citers of the same source is not itself proof of independent reading, but it is the opposite of what a shared-copy-source hypothesis would predict for this specific pair. Benkard's paper predates David & Brachet's by roughly a decade and has the title right; David & Brachet's error does not match Benkard's wording, so it was not obviously inherited from Benkard's reference list specifically — it could be an independent typo, or copied from some other, unidentified intermediate list. A web search for the exact wrong phrase ("persistence and transfer of learning in an industrial setting") returned no other paper repeating it, which argues — weakly, given search-engine incompleteness — against this being a widely propagated copy-fingerprint of the kind Simkin & Roychowdhury found for the Kosterlitz–Thouless misprints.

This is a sample of one comparable pair, not a cluster-wide audit. It cannot resolve the question; it is one data point against the "inherited silence" reading for this specific link.

Claim: Most of the cluster's other primary reference lists are not reachable with this session's tools

Claim type: historical/methodological (source-availability finding). Tier 1 for the fact of the fetch failures themselves, directly observed this session. verifies: question-do-organizational-forgetting-reference-lists-show-copy-fingerprints

A cluster-wide misprint audit, as the routed question specifies, would need the reference lists of Argote, Beckman & Epple (1990) itself, Darr, Argote & Epple (1995), and Thompson (2007) — all named in the question or cited within the two papers already read. All three currently resolve only to paywalled venues from this session: Argote/Beckman/Epple (1990) and Darr/Argote/Epple (1995) sit on Management Science via pubsonline.informs.org, which returned HTTP 403 Forbidden to archive_page (attempted on the Argote/Beckman/Epple DOI page, 10.1287/mnsc.36.2.140); Thompson (2007) sits on the same INFORMS platform and is also indexed on JSTOR, a platform already known-blocked per 00-meta/specs/sources.md. A CiteSeerX mirror URL for Argote/Beckman/Epple (1990) (citeseerx.ist.psu.edu/document?...doi=7c5fa27e...) returned HTTP 404. A CMU GSIA working-paper record for Darr/Argote/Epple (1995) exists on EconPapers (RePEc:cmu:gsiawp:1995-16) but explicitly states no download is available. No free, fetchable copy of any of these three documents' reference lists was located.

This is a scope-limiting finding in its own right: the primary documents needed to answer the routed question at cluster scale are not accessible to this vault's current tooling, which is why this capture can report only a single comparable pair rather than the audit the question calls for.

Claim: No published bibliometric study has applied a copy-fingerprint audit to this specific literature

Claim type: historical (absence-of-literature finding from search). Tier 3-4 acceptable — this is a negative search result, not a claim resting on any single document. verifies: question-do-organizational-forgetting-reference-lists-show-copy-fingerprints

Searches combining organizational-forgetting authors/papers with citation-copying, misprint-propagation, or reference-copying methodology returned no study applying the Simkin & Roychowdhury method (or an equivalent) to the organizational-forgetting citation cluster specifically. The closest matches found were general-methodology papers on reference/citation copying in bibliographic networks (see Further leads) that do not touch this literature. This is an absence-of-search-result finding, not an exhaustive literature review, and is recorded as such — it means the question is genuinely untested in the published record, not that it has been tested and answered negatively.

Central question status

[unverified -- could not confirm or deny after search]. The one comparable citation pair found (Benkard 2000 vs. David & Brachet 2011, both citing Argote/Beckman/Epple 1990) shows divergent, not identical, wording — weak evidence against a copy-fingerprint on that specific link — but the sample is far too small to generalize to "the literature's zero-ML silence is/is not a copying artifact." The cluster's other primary reference lists are paywalled and unreachable with current tooling, and no existing published study has run this audit. The routed question's own caveat holds: copied silence leaves no fingerprint the way a copied misprint does, so even a complete audit of shared references/misprints inside the cluster would only speak to whether these authors read each other's citations — not directly to whether the ML-silence itself was chosen or inherited.

Further leads

• "Analysis of Reference and Citation Copying in Evolving Bibliographic Networks" (arXiv:1912.11894) — general methodology for detecting reference/citation copying in bibliographic networks; not applied to organizational forgetting, but the closest existing general method for a future cluster-wide audit.
• Thompson, P. (2007), "How Much Did the Liberty Shipbuilders Forget?" (Management Science 53(6):908-918) — argues forgetting rates in the classic Liberty-ship case are much smaller than previously reported, "may well be zero"; a genuine substantive dispute within the organizational-forgetting literature, unrelated to the copy-fingerprint question and worth its own future claim-note. Paywalled (INFORMS/JSTOR); a 2005 working-paper version reportedly exists but was not located free this session.
• Darr, E., Argote, L. & Epple, D. (1995), "The Acquisition, Transfer, and Depreciation of Knowledge in Service Organizations: Productivity in Franchises" (Management Science 41(11):1750-1762) — the pizza-franchise study both Benkard and David & Brachet cite; a CMU GSIA working-paper record exists (EconPapers RePEc:cmu:gsiawp:1995-16) but has no downloadable file.
• pubsonline.informs.org (Management Science's platform) returned HTTP 403 to archive_page on the Argote/Beckman/Epple (1990) DOI page — a candidate addition to 00-meta/specs/sources.md's known-blocked-routes list alongside the existing SAGE/PNAS/OUP/JSTOR entries, if a later session confirms the pattern repeats.
• peterthompson.gatech.edu (Peter Thompson's own faculty page, a candidate host for a free working-paper copy of the Liberty-ships paper) failed TLS certificate validation on fetch (hostname not present in the certificate's SAN list) — not an adversarial-content flag, just an unreachable/misconfigured host; worth a manual-browser check in a later session rather than automated tooling.

Entity candidates

• Linda Argote — person — first author of Argote, Beckman & Epple (1990), the foundational paper both Benkard (2000) and David & Brachet (2011) measure their forgetting estimates against, and the paper whose title is misquoted in one of the two citing papers; the ancestry this whole citation cluster traces back to.
• Sara Beckman — person — co-author of the 1990 foundational paper.
• Dennis Epple — person — co-author of the 1990 foundational paper and also a co-author (with Argote and Darr) of the 1995 franchise-knowledge-depreciation paper in the same cluster.
• C. Lanier Benkard — person — author of the 2000 American Economic Review aircraft-production paper compared against David & Brachet in this capture; economist, was at Stanford at the time of the NBER working paper.
• Peter Thompson — person — author of the 2007 Liberty-shipbuilders reforgetting-rate paper in the same citation cluster; a substantive dispute within the literature (see Further leads) separate from the copy-fingerprint question.
• National Bureau of Economic Research (NBER) — concept/institution — hosts Benkard's working paper as a free, author-side venue; functions in economics roughly the way arXiv functions in physics/CS, relevant to why one half of this comparison was fetchable and the other was not.