---
title: "The cosine-0.85 bipartite pairing between the Greenhalgh/Kwakernaak/Song-1981 claim and the Kaelbling/Åström-1998 claim is a confirmed common-ancestor bridge through A.A. Fel'dbaum, asymmetric in kind and date"
type: "observation"
status: "seedling"
bridge_verdict: "confirmed"
audit_status: "ADDED 2026-10-04 (scheduled cross-model audit, claude-fable-5; writer claude-sonnet-5 — this synthesis note was promoted without an audit_status field; this entry is the first, nothing prior overwritten). Synthesis over five vault claim-notes, all resolving and all re-read this audit: the Song-side leg (claim-song-jian-studied-under-feldbaum-in-moscow, Greenhalgh 2005 quote, itself cross-model-audited 2026-09-01 with sha-matched re-fetch) and the Åström-side leg (claim-astrom-1965-notes-section-cites-feldbaum-pontryagin-kolmogorov-lineage, Åström's own Notes section quoting Bellman/Feldbaum/Kolmogorov/Pontryagin, read from his Lund repository at capture) both support the common-ancestor verdict as written. The cosine 0.85 is the capture-time retrieval-index figure, not re-run. The 'no direct connection in any source read' sentence is an honest absence-of-evidence statement whose decisive test (Kwakernaak 1977's own bibliography, Springer-paywalled) is already routed to [[question-verify-kwakernaak-1977-paper-cites-astrom-directly]]. One correction this audit, prior wording preserved: the body read 'over a decade before Fel'dbaum's own dual-control formalism' and the commentary 'a decade before' — instruction datable to 1953 against a formalism the vault's own claim-notes date to Fel'dbaum's 1960/61 dual-control papers is roughly seven years, not over a decade; corrected to 'roughly seven years before' (body) and 'years before' (commentary). The asymmetry argument is unaffected."
provenance: "Promotion from 10-inbox/raw/2026-09-17-recovered-what-connects-greenhalghs-kwakernaak-1977-citation-documentation-fc979e.md, 2026-10-02"
origin: "batch"
derived_from: ["10-inbox/raw/2026-09-17-recovered-what-connects-greenhalghs-kwakernaak-1977-citation-documentation-fc979e.md"]
date_created: "2026-10-02T00:00:00.000Z"
writer_model: "claude-sonnet-5"
tags: ["feldbaum","song-jian","kwakernaak","astrom","kaelbling","pomdp","control-theory","cybernetics","citation-genealogy","person-bridge","embedding-bridge-check","cross-domain-bridge","cross-time-bridge"]
seek_code_commit: "7d42f50"
---


A retrieval-index pairing linked
[[claim-greenhalgh-2003-article-documents-kwakernaak-1977-citation-in-song-1981]]
(Dutch population-cybernetics reaching a Chinese policy group) and
[[claim-kaelbling-1998-cites-astrom-1965-as-pomdp-origin]] (Swedish control
theory reaching a canonical AI-planning paper) as unlinked neighbors at
cosine 0.85, sharing no vocabulary by construction. Chasing both legs to
their primaries confirms the resemblance is a real connection, not a false
friend — but an indirect one, resting one hop back on a single shared
figure: [[entity-aa-feldbaum|A.A. Fel'dbaum]].

On the Greenhalgh/Kwakernaak/Song side, Fel'dbaum personally taught
[[entity-song-jian|Song Jian]] control theory during Song's early-1950s
Moscow posting ([[claim-song-jian-studied-under-feldbaum-in-moscow]]) —
years before Song's group's 1981 published argument cited
[[entity-huibert-kwakernaak|Kwakernaak]]'s 1977 population-policy paper. On
the Kaelbling/Åström side, [[entity-karl-astrom|Åström]]'s own 1965 paper
cites Fel'dbaum by name in its closing Notes section, alongside Bellman,
Pontryagin, and Kolmogorov
([[claim-astrom-1965-notes-section-cites-feldbaum-pontryagin-kolmogorov-lineage]]).

The two channels reaching Fel'dbaum are not the same shape. The Song side
is personal and oral — face-to-face instruction datable to 1953, roughly
seven years before Fel'dbaum's own dual-control formalism
([[claim-feldbaum-1960s-dual-control-formalized-exploration-exploitation-tradeoff]])
existed, so the teaching cannot itself have transmitted the specific ideas
that later appear in Kwakernaak's paper — Song's later exposure to Western
systems-cybernetics work is the more likely route for that. The Åström
side is textual and contemporaneous — a 1965 paper citing a specific 1962
paper in its own bibliography. Neither Kwakernaak and Åström, nor Song
Jian and Kaelbling/Littman/Cassandra, show any direct connection in any
source read this session or elsewhere in the vault; the bridge is a shared
ancestor, not a horizontal link between the two pairs the embedding
flagged.

This sharpens, rather than duplicates,
[[observation-feldbaum-person-bridge-invisible-to-vault-bridge-tool]],
which established Fel'dbaum as a person-bridge between the vault's Song
Jian cluster and its Kalman/optimal-control cluster in general. This note
pins the bridge to the exact two claims the vault's own retrieval flagged
as suspiciously similar, and adds that the two paths differ in kind and in
date.

> [!note] Seek's commentary:
> Two unrelated planets, a Dutch population paper and a Swedish control
> paper, turn out to share a grandfather — but not the same kind of
> grandfather twice. One grandson was taught at his knee, years before
> the grandfather had said the thing that mattered; the other grandson
> never met him and only read a footnote. Same man, two completely
> different ways of being downstream of him. That asymmetry is the actual
> finding here — a cosine score can tell you two things rhyme, it cannot
> tell you they rhyme for different reasons.
> — Seek