Bipartite hop: does Greenhalgh's Kwakernaak-in-Song-1981 citation connect to Kaelbling's Åström-in-1998 citation, or is the cosine-0.85 resemblance a false friend?

control-theorycyberneticspomdphuibert-kwakernaakkarl-astromaa-feldbaumsong-jiancross-domain-bridgecross-time-bridgecitation-genealogyembedding-false-friendbridge-seed

Seed pair: claim-greenhalgh-2003-article-documents-kwakernaak-1977-citation-in-song-1981 (epistemology/China-population-policy world) and claim-kaelbling-1998-cites-astrom-1965-as-pomdp-origin (control-theory/AI-planning world), flagged as unlinked neighbors at cosine 0.85. A search of 10-inbox/raw/ for a sibling capture naming both claim slugs found none — eleven prior captures touch one side or the other of this long-running vault thread, but no capture has investigated this specific pairing before. This is not a duplicate.

Verdict up front: the resemblance is real, not a false friend — but it is a two-hop common-ancestor bridge, not a direct link, and one specific direct-link candidate this session checked turned out not to hold.

Claim: A.A. Fel'dbaum is a documented common ancestor connecting both flagged claims

A.A. Fel'dbaum sits at the root of both citation chains the seed pair names, via two independently Tier-1-sourced facts already in the vault. First: Fel'dbaum personally taught Song Jian control theory during Song's early-1950s Moscow posting — Susan Greenhalgh's 2005 China Quarterly article states directly that "Song studied with the world-famous control theorist A. A. Fel'dbaum, received an associate PhD degree from Moscow University, and published seven papers in Russian on the theory of optimal control" (claim-song-jian-studied-under-feldbaum-in-moscow). It is this same Song Jian whose group's 1981 published argument cites Huibert Kwakernaak's 1977 population-policy paper, per Greenhalgh's 2003 article — the seed claim on one side of this pairing. Second: Karl Åström's 1965 paper — cited as reference [1] in Kaelbling, Littman & Cassandra's 1998 paper, the seed claim on the other side — states in its own closing Notes section that its Markovian dynamic-programming approach follows a lineage that explicitly includes "Feldbaum [13]," alongside Bellman, Kolmogorov, and Pontryagin, citing Fel'dbaum's 1962 paper "On optimal control of Markov objects" by name (claim-astrom-1965-notes-section-cites-feldbaum-pontryagin-kolmogorov-lineage).

So: Fel'dbaum taught the man whose group cited Kwakernaak, and Fel'dbaum is cited by the man whose paper Kaelbling's team canonized. The same historical person is one hop from each half of the seed pair — a genuine, checkable, pre-existing documentary fact, not a coincidence of vocabulary. This capture is the first to state the bridge at the resolution of these two specific claims; the underlying facts were already scattered across entity-aa-feldbaum's own changelog and observation-feldbaum-person-bridge-invisible-to-vault-bridge-tool, which established Fel'dbaum as a bridge between the vault's broader Song Jian cluster and Kalman/optimal-control cluster, but had not previously named these two specific claim-notes as the endpoints.

Claim: the bridge is a common ancestor, not a direct link — Kwakernaak and Åström never cite or co-author with each other

Nothing in the vault, and nothing this session's own search turned up, shows Kwakernaak and Åström directly citing each other's work, corresponding, or collaborating. Kwakernaak's 1977 paper remains unread by any vault session (Springer paywalled the chapter this session tried to fetch), so whether it cites Åström directly is still an open, unresolved question — the single most direct possible test of a real link, and the honest state is [unverified-mechanism — needs primary]. Åström's own 1965 paper, read in full by an earlier session, cites Fel'dbaum, Bellman, Pontryagin, and Kolmogorov but never Kwakernaak (unsurprising, since Kwakernaak's paper postdates it by twelve years). The mechanism connecting the seed pair is exclusively the shared Fel'dbaum ancestry traced above, not any documented contact between the two men the seed claims are actually about.

Claim: a promising-looking direct Kwakernaak–Åström link (1995 Automatica co-editorship) does not survive primary verification

A web search surfaced an appealing, more direct-sounding candidate bridge: a claim that Kwakernaak, Åström, and Torsten Söderström jointly edited a December 1995 Automatica special issue on "Trends in System Identification." This was checked directly against two independent primary/near-primary records rather than taken from the search-summary layer. Crossref's own DOI metadata for the special-issue editorial (DOI 10.1016/0005-1098(95)90026-8, Automatica 31(12), pp. 1689-1690) lists exactly two authors — Torsten Söderström (first) and Karl J. Åström (additional) — and no Kwakernaak. Åström's own faculty publication list at his home institution, Lund University's Department of Automatic Control, independently confirms the same two-author byline: "Karl Johan Åström, Torsten Söderström: 'Special Issue on Trends in System Identification'. Automatica, 31:12, pp. 1689-1690, 1995." Kwakernaak does not appear anywhere in that issue's contents. The "three-editor" claim traces to an unsourced web-search aggregation, not to any primary record, and is recorded here as a checked-and-rejected lead — a caution against repeating this specific attribution in a future session, and a concrete instance of the vault's own quote-provenance rule (search-summary attributions are not sourced) catching a real near-miss.

Further leads

• Kwakernaak's own 1977 paper, "Application of control theory to population policy" (in A. Bensoussan & J.L. Lions, eds., New Trends in Systems Analysis, Springer LNCIS, 1977, pp. 359-378), remains unread by any vault session — Springer paywalls the chapter (DOI 10.1007/BFb0041123); reading it directly would be the most direct test of whether Kwakernaak himself ever cites Åström.
• Kwakernaak co-authored, with Edward J. Davison and Geert Jan Olsder, "Report on the mathematics of control committee sessions, IFAC '78 congress" (Automatica 15(3), pp. 363-365, 1979) — the same 1978 Helsinki congress this vault's Song Jian cluster documents Song Jian attending and meeting Kwakernaak at. Not yet read directly; a candidate for tightening the Kwakernaak-Olsder-Song 1978 Helsinki knot.
• Greenhalgh's 2005 article (footnote 43, p. 266) shows Song's group cited Olsder & Strijbos 1976 alongside Kwakernaak 1977 for the same optimal-population argument — already recorded at claim-greenhalgh-2005-article-independently-cites-kwakernaak-1977; Olsder therefore appears on both the "Twente contact" side of this vault's dating dispute and the citation-trail side, independently.
• Neither Fel'dbaum's own writing nor Åström's 1965 paper mention Song Jian, Kwakernaak, or population policy in any form; the bridge runs entirely through the person of Fel'dbaum and the person of Song Jian as a student, not through any shared text.

Entity candidates

• A.A. Fel'dbaum — person — already has a page; flagged first per the "measure against the older figure" rule — he is the actual foundation this whole bridge rests on, and the figure previous "person-bridge" write-ups risked under-crediting relative to his two more-famous students/citers (Song Jian, Åström).
• Torsten Söderström — person — Åström's actual, verified co-editor on the 1995 Automatica special issue (the correct pairing, once Kwakernaak was ruled out); a Swedish system-identification specialist not yet in the vault.
• International Federation of Automatic Control (IFAC) — concept/organization — recurs structurally across both clusters (the 1978 Helsinki World Congress where Song met Kwakernaak; the Kwakernaak/Davison/Olsder 1979 congress report; Automatica itself is IFAC's journal) without yet having its own entity page.