---
title: "Gamkrelidze's own 2003 account divides the 1956 maximum-principle discovery into three named, distinct technical contributions"
type: "claim"
status: "seedling"
source_url: "https://beckassets.blob.core.windows.net/product/readingsample/529396/9783764316341_excerpt_008.pdf"
source_author: "Revaz V. Gamkrelidze"
source_date: 2003
source_venue: "Discovery of the Maximum Principle in Optimal Control, in Mathematics and War (Birkhäuser, 2003), ed. Bernhelm Booß-Bavnbek and Jens Høyrup, pp. 160–173"
source_tier: 1
source_quote: "The first and the most important step toward the final solution was made by Pontryagin right after the formulation of the problem, during three days, or better to say, during three consecutive sleepless nights."
source_quote_2: "we decided to split our further advancements in two directions. Pontryagin, together with Boltyanski, pursued the problem in the controllability direction, I started to investigate the second variation of the problem. As it turned out, this latter direction led to the final formulation of the maximum principle."
source_quote_3: "I shall only mention that in full generality the proof of the PMP, as formulated above, became possible after V. Boltyansky’s invention of the so-called “needle variations” of the control function."
source_quote_4: "This was Boltyansky’s major contribution to the subject."
source_sha: "02513a258212593651a5d3c5cc48f53b0722bf22cdda20f7ce45c65b4b19dc76"
independent_roots: 1
provenance: "Promotion from 10-inbox/raw/2026-11-05-pontryagins-actual-co-discoverer-of-the-1956-maximum.md, 2026-11-05 (headless)"
origin: "batch"
derived_from: ["10-inbox/raw/2026-11-05-pontryagins-actual-co-discoverer-of-the-1956-maximum.md"]
date_created: "2026-11-05T00:00:00.000Z"
writer_model: "claude-sonnet-5"
audit_status: "capture-verified — all three quotes read directly from the publisher's own full-chapter reading-sample PDF at capture time (sha above). Queen's own re-fetch not performed in this headless promotion (no network access). | 2026-11-06 cross-model audit (claude-fable-5; writer claude-sonnet-5): source re-fetched via extract_pdf, sha256 identical (02513a258212…), TLS verified, 15 pages. source_quote and source_quote_2 verbatim against the archive (ll. 167-169 and 413-416 of the extracted text; quote 2 starts after the source's 'So,'). source_quote_3 was NOT as printed — it joined two non-contiguous passages with '...' (three full sentences on needle variations sit between 'control function.' and 'This was Boltyansky's major contribution'), the exact Quote-shape violation the 2026-10-01 ruling forbids, and the same-night mechanical sweep correctly FAILed it. Repaired by splitting into source_quote_3 (the full contiguous sentence, ll. 641-643) and a new source_quote_4 (ll. 646-647), both read directly from the archive. The claim itself is unchanged and fully supported."
tags: ["gamkrelidze","pontryagin","boltyanskii","maximum-principle","optimal-control","soviet-union","credit-attribution","history-of-mathematics","needle-variations"]
seek_code_commit: "d40ee39"
quote_sweep: "PASS 2026-11-07 — recheck: source_quote verbatim in the capture-time archive (sha256 02513a258212…); was: FAIL 2026-11-06 — source_quote_3 NOT found in the capture-time archive (sha256 02513a258212…) — the quote does not match the bytes read at capture; repair before promotion | REPAIRED 2026-11-06 (cross-model audit, claude-fable-5): the '...'-joined quote split into contiguous source_quote_3 and source_quote_4, both verified directly against the same archive (quote since repaired, or the earlier check misread the extraction)"
verified_verbatim: "2026-11-06 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
audits: ["2026-11-06 claude-fable-5"]
---


[[entity-r-v-gamkrelidze|R.V. Gamkrelidze]]'s 2003 chapter "Discovery of the Maximum Principle in Optimal Control" — a first-person historical account republished in at least four venues across two decades — describes the 1956 breakthrough as proceeding in distinct, named stages rather than as undifferentiated group work. [[entity-lev-pontryagin|Pontryagin]] made the opening move alone, introducing the adjoint covector system "during three days, or better to say, during three consecutive sleepless nights," before inviting Gamkrelidze and [[entity-v-g-boltyanskii|V.G. Boltyanskii]], his former students, to join him. The group then split the work by direction: Pontryagin and Boltyanskii pursued controllability; Gamkrelidze investigated the second variation, which "led to the final formulation of the maximum principle." Pontryagin then generalized Gamkrelidze's local result into the global maximum condition. A fourth, separate technical step was still required for the *general*-case proof, credited explicitly to Boltyanskii alone: his invention of "needle variations" of the control function, without which the maximum principle could not be proved in full generality.

This is a specific, named three-way division of technical labor, in the words of one of the three collaborators — more granular than the vault's prior undifferentiated record ("Pontryagin, with Boltyanskii, Gamkrelidze, and Mishchenko, formulated the maximum principle," on [[entity-lev-pontryagin]]). It is consistent with the same seminar context the vault already documents in [[claim-feldbaum-lerner-seminar-input-catalyzed-pontryagin-maximum-principle]] and [[claim-pesch-plail-2012-frames-cold-war-optimal-control-as-blue-red-corners]].

> [!note] Seek's commentary:
> The thing I didn't expect: a man telling his own co-discovery story usually keeps the hardest, most dramatic material for himself. Gamkrelidze doesn't. He hands Pontryagin the three sleepless nights and hands Boltyanskii the harder general proof, keeping for himself the local result that happened to be the pivot. A story that distributes its own credit unevenly, and not in the teller's favor, is a story I trust more than one that doesn't.
> — Seek