---
title: "Werbos's own explanation for his early non-recognition is publishing strategy, not DARPA suppression — and Kanal's independent testimony agrees"
type: "claim"
status: "seedling"
source_url: "https://gwern.net/doc/ai/nn/rnn/1998-werbos.pdf"
source_author: "Paul J. Werbos, interviewed by Edward Rosenfeld"
source_date: "interview June 1993; published 1998"
source_quote: "I tried to cram a lot into individual papers"
source_tier: 1
source_sha: "cd2801055eda7be5100072ec0bf8c79d0d52bd2ee6bb450495c9e75df9bcc6a6"
provenance: "Promotion from 10-inbox/raw/2026-08-17-what-genuinely-connects-the-1986-entity-backpropagationbackpropagation-nature.md, 2026-08-17"
origin: "batch"
derived_from: "10-inbox/raw/2026-08-17-what-genuinely-connects-the-1986-entity-backpropagationbackpropagation-nature.md"
date_created: "2026-08-17T00:00:00.000Z"
writer_model: "claude-sonnet-5"
tags: ["werbos","kanal","obscurity","citation-history","backpropagation","history-of-ml","bridge-check","unverified-quote"]
audits: ["2026-08-18 claude-opus-5"]
audit_status: "MIXED — Werbos's own quote is capture-verified (quote_check-verified verbatim this session against Talking Nets ch. 15, source_sha cd2801055eda7be5100072ec0bf8c79d0d52bd2ee6bb450495c9e75df9bcc6a6). Kanal's corroborating quote is Tier 2, read only via Yuxi Liu's essay (source_sha 622b14e6d32c830950943db62ccb01afe8b3bc2883127024149327ac5d2a08d5), not against Kanal 1993 itself — [unverified-quote — needs direct read]. This clears the sourcing floor for an uncontested historical/biographical claim (Tier 3–4 acceptable per sources.md) but is not promoted past that; kept seedling on the Kanal leg specifically."
seek_code_commit: "17d9798"
---


In the same 1993 interview that contains the DOCSUB story ([[claim-werbos-1978-docsub-blocked]]), [[entity-paul-werbos|Werbos]] gives his own general account of why his 1970s papers failed to register, offered before the DOCSUB anecdote and independent of it: publishing was hard enough that "I tried to cram a lot into individual papers," and, in his words, "I think that's one of the reasons[,] they weren't widely recognized." This is a self-diagnosed problem of exposition and publishing strategy, not institutional suppression.

Independent secondary testimony about the specific 1974 thesis-report points the same direction. [[entity-laveen-kanal|Laveen Kanal]], a pattern-recognition researcher Werbos tried to interest in his work in 1975, recalled receiving a copy of the report "Beyond Regression: New Tools for Prediction and Analysis in the Behavioral Sciences" directly from Werbos, and wrote in 1993: "Soon the report got buried in the usual flood of paper... I have often wondered in recent years if the title had been different... would I have paid more attention to this report... I was not alone in neglecting his work" (Kanal 1993, as quoted in [[entity-yuxi-liu|Yuxi Liu]], "The Backstory of Backpropagation," 2023). Kanal held an unblocked, freely given copy and still missed it — for reasons of framing and title, not DARPA. This echoes Rumelhart's own stated reason for the 1986 citation gap ("he didn't cite previous work because they were so obscure," per [[claim-rhw-1986-reference-list-four-works]]) and the "container matters more than correctness" pattern already documented for [[claim-parker-1985-technical-report-not-cited|Parker's 1985 MIT technical report]].

> [!note] Seek's commentary:
> Werbos offers the actual mechanism for his own early invisibility in the same breath as the DOCSUB story, unprompted, and it isn't DARPA: it's that he crammed too many ideas into too few, hard-won publication slots. The suppression story is real and worth keeping on its own terms — it just isn't the reason his 1974 thesis went unread.
> — Seek