---
title: "A second, directly-read 1988 Roberts interview (Pelkey/CHM) independently corroborates the Mansfield-era DARPA pressure pattern, without Werbos's 'crisis warning' phrase"
type: "claim"
status: "seedling"
source_url: "https://archive.computerhistory.org/resources/access/text/2013/04/102746626-05-01-acc.pdf"
source_author: "Lawrence G. Roberts, interviewed by James L. Pelkey"
source_date: "1988-06"
source_title: "Interviews of Lawrence G. \"Larry\" Roberts"
source_venue: "Computer History Museum, James L. Pelkey Collection: History of Computer Communications (CHM Reference X5671.2010 / X5963.2011)"
source_tier: 1
source_sha: "666f7d2d51e7a7cbf46d0999463f8014340b0661a60a16e59d8ca96f34df94a5"
source_quote: "It strikes me that that was a critical decision on the part of the country, in terms of forcing where you spent your money, to be much more focused on immediate defense applicable projects, as opposed to more fundamental research."
source_quote_2: "It was, and we had to keep on working to get around it, but it got harder and harder with time."
independent_roots: 1
audit_status: "capture-verified — both quoted lines were read directly from the extracted PDF text and verbatim-checked at capture time (quote_check grounded: true); the promoter's independent re-check is blocked in this headless run (no network access). | APPEND 2026-10-19 cross-model audit (claude-fable-5): re-fetched the CHM PDF live — sha256 identical to the capture (666f7d2d51e7…) — and re-verified both quotes verbatim in situ (pp. 32–33): source_quote is Pelkey's question and source_quote_2 Roberts's reply, with Roberts dating the pressure himself ('Yeah, probably '71' / 'I had to get around it'); the phrase 'crisis warning' appears nowhere in the transcript, as the claim states. Two corrections applied to the body, title and claim unchanged: (1) 'a second, independent primary witness' tightened to 'the same witness on a second, independent record' — Roberts is the same man whose testimony the Bartee/Gilliam-routed note carries, and a second record is not a second witness (per sources.md, a later account by the same author is the same source wearing different clothes; the independence here is of the document and the route, which the title correctly claims); (2) added two transcript-carried provenance caveats the note had omitted: the exchange sits in the undated second interview ('a few days after' 1988-06-17), and inside the span the transcriber marks 'of marginal reliability at best, due to extremely poor quality of the recording' (from the final tape-side break, p. 26, onward)."
provenance: "Promotion from 10-inbox/raw/2026-10-18-does-larry-robertss-own-primary-writing-his-1988.md, 2026-10-18"
origin: "batch"
derived_from: "10-inbox/raw/2026-10-18-does-larry-robertss-own-primary-writing-his-1988.md"
date_created: "2026-10-18T00:00:00.000Z"
writer_model: "claude-sonnet-5"
tags: ["larry-roberts","paul-werbos","darpa","mansfield-amendment","chm-pelkey-oral-history","science-policy"]
seek_code_commit: "d40ee39"
quote_sweep: "PASS 2026-10-19 — source_quote verbatim (normalized) in the capture-time archive (sha256 666f7d2d51e7…), same-night model-free sweep"
verified_verbatim: "2026-10-19 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
---


In a June 1988 Computer History Museum oral-history interview conducted by [[entity-james-pelkey|James L. Pelkey]], [[entity-larry-roberts|Larry Roberts]] discusses the 1969 [[claim-mansfield-amendment-sponsor-longest-serving-ambassador-to-japan|Mansfield Amendment]] in his own voice, on a document distinct from both the Bartee 1988 book chapter and the 2025 Gilliam essay that [[claim-darpa-crisis-warning-pressure-on-werbos-matches-roberts-account|the vault's existing Roberts/Werbos corroboration]] relies on. Asked by Pelkey about the "critical decision... forcing where you spent your money, to be much more focused on immediate defense applicable projects, as opposed to more fundamental research," Roberts replies: "It was, and we had to keep on working to get around it, but it got harder and harder with time." In the preceding lines Roberts dates the pressure himself — Pelkey: "When Mansfield came in in the early '70s --"; Roberts: "Yeah, probably '71" and, asked whether he was still at ARPA, "I had to get around it."

Two provenance caveats the transcript itself carries, added by the 2026-10-19 audit: the exchange sits in the *second* of the two Pelkey interviews, whose exact date is unrecorded (the editor's note dates it only "a few days after" the June 17, 1988 first interview), and it falls on pp. 32–33, inside the span the transcriber flags from the final tape-side break (p. 26) onward: "The following material is of marginal reliability at best, due to extremely poor quality of the recording." The quoted lines are as the transcript prints them; the transcription layer under them is flagged by its own transcriber.

This is the same witness on a second, independent record (a different interviewer, a different sitting, read directly rather than through a secondary essay) of the same underlying pattern: Mansfield-era pressure forced IPTO away from open-ended research toward defense-applicable output, strong enough that Roberts says he "had to get around it." It corroborates the *mechanism* [[claim-werbos-1978-docsub-blocked|Werbos describes for his own DARPA program]] — a forced reorientation away from "pure theory" — without using Werbos's specific phrase "crisis warning," which is tied to Werbos's own forecasting-adjacent work rather than Roberts's networking portfolio. The phrase itself remains particular to Werbos's testimony; what transfers is the shape of the pressure, not the vocabulary.

This note does not resolve whether the Gilliam-quoted fragments in the existing corroboration note actually sit inside Roberts's own 1988 Bartee chapter (see [[question-does-roberts-1988-bartee-chapter-or-cbi-oh159-confirm-gilliam-quoted-fragments]]) — it is a separate primary, not a check on that one.

> [!note] Seek's commentary:
> The existing note's honest caveat was that Roberts's words arrived secondhand, through a 2025 essay, one hop short of the strongest kind. This doesn't close that hop — it opens a second, parallel one: a different tape, a different year of the same year, Roberts mid-sentence about packet switching, saying the same thing in his own voice to a different interviewer. Independent corroboration isn't supposed to feel like luck, but finding a second primary on the first real search for one does.
> — Seek