Capture: Does the Brode, Werbos & Dunn TSP manual's actual imprint read 1975 or 1977?

werbostsp-manualbibliographyverificationhistory-of-mlmyth-ledger

Follow-up on question-werbos-tsp-manual-year-1975-vs-1977, raised while promoting claim-werbos-1988-rhw-special-case-of-dynamic-feedback: Werbos's own 1988 Neural Networks bibliography cites "Brode, Werbos, & Dunn, 1975" while the CrossRef record behind claim-werbos-tsp-manual-first-publication carries the manual's date as 1977. This session reads both underlying documents directly (not just the CrossRef metadata layer) to see whether the discrepancy sharpens or resolves. It does not resolve — no title page, library catalog record, or digitized copy of the manual itself was located — but it sharpens considerably: the 1975 date is not a stray citation but Werbos's own, repeated three separate times in one paper, and the 1977 date traces to one specific, identifiable, independent secondary citation rather than to Werbos himself.

Claim: Werbos's own 1988 paper cites the TSP manual as 1975, consistently, in three separate places within the same document

Claim type: historical/bibliographic. Source is Werbos's own published paper, citing his own co-authored work — Tier 1 for the fact of what Werbos himself wrote and when.

verifies: question-werbos-tsp-manual-year-1975-vs-1977

Paul J. Werbos, "Generalization of Backpropagation with Application to a Recurrent Gas Market Model," Neural Networks 1 (1988): 339–356, cites the manual three times, all dated 1975, not once dated 1977:

"(See Brode, Werbos, & Dunn, 1975; Dennis & Schnabel, 1983; SAS Institute, 1986, Werbos, 1988a.)" (p. 340)

"The results were documented in Werbos (1974), embedded (and documented) in user-oriented software in an MIT version of the Time-Series Processor (Brode et al., 1975), and discussed in Werbos (1977) …" (p. 342)

"Brode, J., Werbos, P., & Dunn, E. (1975). TSP in the Datatran language. Cambridge, MA: MIT Cambridge Project. (Also distributed by the MIT Information Processing Services Department and DOD/AFDSC as a contract product.)" (References, p. 356)

The full reference-list entry also gives a different publishing/distributing body than the one attached to the 1977 date elsewhere (see next claim): "MIT Cambridge Project," with separate distribution credited to "the MIT Information Processing Services Department and DOD/AFDSC," rather than "MIT Laboratory of Architecture and Planning."

Provenance:

• source_url: https://gwern.net/doc/ai/nn/rnn/1988-werbos.pdf
• source_sha: 1c5be985e963f74a566ad05951c82aa5f9fe83ec55d920e865da0dbc6e4d0ad9
• source_author: Paul J. Werbos
• source_date: 1988
• source_title: "Generalization of Backpropagation with Application to a Recurrent Gas Market Model"
• source_venue: Neural Networks, Vol. 1, pp. 339–356 (Pergamon Press / Elsevier)
• source_tier: 1
• source_quote: "Brode, J., Werbos, P., & Dunn, E. (1975). TSP in the Datatran language. Cambridge, MA: MIT Cambridge Project. (Also distributed by the MIT Information Processing Services Department and DOD/AFDSC as a contract product.)"

Claim: The 1977 date in the vault's existing claim-note traces to one specific independent citation — Klensin & Romberg's 1989 Springer chapter — not to Werbos or a library catalog

Claim type: historical/bibliographic. Source is a named-author academic chapter citing a work it did not author — Tier 3, secondary citation of a document neither author co-wrote.

verifies: question-werbos-tsp-manual-year-1975-vs-1977

John C. Klensin and Roselyn M. Romberg, "Statistical data management requirements and the SQL standards — an evolving comparison," in Statistical and Scientific Database Management (SSDBM 1988 proceedings), Lecture Notes in Computer Science vol. 339 (Springer, 1989), pp. 19–38 — the chapter behind DOI 10.1007/BFb0027503 that CrossRef's structured-reference metadata drew from — gives its own reference list entry as:

"Brode, J., P. Werbos, and E. Dunn, TSP in the Datatran Language, Cambridge, MA, USA: MIT Laboratory of Architecture and Planning, 1977."

This is the entire textual basis for the 1977 date already recorded in claim-werbos-tsp-manual-first-publication (via CrossRef's API record for the same DOI). Klensin & Romberg give no further sourcing for the date or the publisher attribution — it appears as a bare reference-list line among ~40 others, with no indication of how they obtained it (their own copy, a library catalog, a citation chain). Notably, their own reference list elsewhere cites a different, related MIT Laboratory of Architecture and Planning manual dated 1975 ("The Consistent System, Janus Reference Manual … 1975"), showing the same lab was issuing dated manuals in that period — which cuts both ways and does not itself resolve which date is right for the Brode/Werbos/Dunn manual specifically.

Provenance:

• source_url: https://link.springer.com/chapter/10.1007/BFb0027503
• source_sha: e77b9259e565bd315204a9db3c863b39c1f3070634f4ce594be194a3fb7cf467
• source_author: John C. Klensin and Roselyn M. Romberg
• source_date: 1989 (chapter publication; SSDBM 1988 conference)
• source_title: "Statistical data management requirements and the SQL standards—an evolving comparison"
• source_venue: Statistical and Scientific Database Management (SSDBM 1988), Lecture Notes in Computer Science vol. 339, Springer-Verlag Berlin Heidelberg
• source_tier: 3
• source_quote: "Brode, J., P. Werbos, and E. Dunn, TSP in the Datatran Language, Cambridge, MA, USA: MIT Laboratory of Architecture and Planning, 1977."

Claim: No title page, library catalog record, or digitized copy of the manual was located this session, so the imprint question itself remains unconfirmed

Claim type: historical/negative-existence finding — directly answering the core question of what the physical document says, which neither of the above two claims actually settles (both are citations about the manual, not the manual's own imprint page).

verifies: question-werbos-tsp-manual-year-1975-vs-1977

Searches this session across WorldCat, HathiTrust, archive.org, Google Books, and general web search found no catalog record or digitized copy of "TSP in the Datatran Language" carrying its own title-page date. WorldCat's search interface returned HTTP 429 (Too Many Requests) on both WebFetch and archive_page attempts and could not be consulted directly this session. DTIC record AD-783626 ("The Cambridge Project: Computer Methods for Analysis and Modeling of Complex Systems," Feb. 1974, D.B. Yntema and Cambridge Project participants) — flagged in the original question note as a plausible adjacent record — again returned HTTP 403 Forbidden to extract_pdf on retry this session, consistent with the prior session's block noted in the question note; it remains a manual-consultation item, not a working fetch target. The question therefore stays open on its own terms, but the shape of the open question has changed: it is not "does Werbos's bibliography contain a slip," it is "which of two independent, non-trivial citations — the co-author's own repeated self-citation (1975, publisher 'MIT Cambridge Project') versus an unrelated 1989 paper's single citation (1977, publisher 'MIT Laboratory of Architecture and Planning') — matches what the document itself actually says," and that can only be settled by the document (or a library record of it), neither of which surfaced.

Provenance: negative finding, not a single citable source. Searches conducted 2026-09-08: WorldCat (search.worldcat.org, HTTP 429 both attempts), HathiTrust, archive.org, Google Books, DTIC AD-783626 (HTTP 403 on extract_pdf), and general web search for MIT Institute Archives / Cambridge Project holdings, none producing a locatable copy or catalog record of the manual itself. [unverified — could not confirm or deny after search]

Further leads

• WorldCat's search interface 429'd on both WebFetch and archive_page this session — worth a retry on a later session or a direct MIT Libraries (Barton) catalog search instead of the general WorldCat frontend.
• DTIC AD-783626 continues to 403 on extract_pdf (2026-09-08), consistent with the block already noted in question-werbos-tsp-manual-year-1975-vs-1977 and the original capture behind claim-werbos-tsp-manual-first-publication — a standing manual-consultation item, not a repeatable fetch target.
• The publisher-of-record discrepancy itself is unexplored: Werbos's 1988 citation names "MIT Cambridge Project" (distributed via MIT Information Processing Services Dept. and DOD/AFDSC) as publisher; Klensin & Romberg 1989 name "MIT Laboratory of Architecture and Planning." These could be the same imprint under different institutional labels (the Cambridge Project was itself housed within MIT's School of Architecture and Planning in this period) or two distinct releases — worth a dedicated look at MIT's Institute Archives Cambridge Project finding aid.
• Klensin & Romberg's own bibliography lists an adjacent, definitely-1975 MIT Laboratory of Architecture and Planning manual ("The Consistent System, Janus Reference Manual," Cambridge, MA, 1975) — a same-lab, same-era comparandum for how that lab dated its manuals, not itself evidence for the Brode/Werbos/Dunn date.

Entity candidates

• John C. Klensin — person — co-author of the 1989 Springer chapter that is the entire textual source of the vault's existing 1977 date; the discrepancy this question turns on runs through his citation practice, not Werbos's.
• Roselyn M. Romberg — person — co-author of the same 1989 chapter, alongside Klensin.
• John Brode — person — first-listed co-author of the TSP manual itself; no other vault material on him.
• E. Dunn — person — third co-author of the TSP manual; no other vault material.
• MIT Cambridge Project — concept — the DARPA/DOD-funded MIT research initiative under which the manual was apparently produced and distributed; distinct institutional identity from "MIT Laboratory of Architecture and Planning," and the discrepancy between the two names attached to the same work is itself unresolved.