---
title: "The Kaelbling, Littman & Cassandra (1998) Åström citation misdating was introduced during production of the final published edition, not later when the paper was re-hosted on the web"
type: "claim"
status: "seedling"
verifies: "question-verify-astrom-1965-1998-citation-misdate-transmission"
audit_status: "capture-verified — an inferential claim resting on the four direct primary reads underlying [[claim-kaelbling-1998-misdating-present-in-independently-hosted-mirror]] and [[claim-kaelbling-1998-preprint-and-aaai94-correctly-date-astrom-1965]], all extract_pdf reads with sha256 receipts at capture time (2026-09-03); queen re-fetch not performed in this headless promotion (no network access). Does not rest on a direct look at the bound print journal — see body. || 2026-09-05 (cross-model audit, auditor claude-fable-5): all underlying documents re-fetched live via extract_pdf with sha256 matches (preprint 79909252…, AAAI-94 a3611d74…, MIT copy 71a6d1ae…, Brown copy f0d7c6e2…, UQ mirror sha-identical to Brown); the before/after boundary — correct 1965 in both author-controlled documents, identical '10 (1995) 174-205' in both distinct final-edition files — re-confirmed directly, so the inference stands. CORRECTED (wording precision): 'two different university course pages with different file bytes' could be read as the Brown and Queensland copies differing from each other; they are byte-identical to each other and differ only from the MIT copy. Body rephrased to say so — the evidence base is two distinct files, not three."
source_url: "http://www.cassandra.org/arc/papers/aij98.pdf"
source_author: "Leslie Pack Kaelbling, Michael L. Littman, Anthony R. Cassandra"
source_date: "1998"
source_venue: "Artificial Intelligence 101 (1998) 99-134 and its pre-publication history, self-hosted copies cross-checked against the MIT- and Brown-hosted final copies"
source_tier: 1
source_sha: "799092522f5f88a3c4de6c1b912463079d3ef049773302141f23d63b37b95fa5"
source_quote: "K. J. Astrém. Optimal control of Markov decision processes with incomplete state estimation. Journal of Mathematical Analysis and Applications, 10:174— 205, 1965."
provenance: "Promotion from 10-inbox/raw/2026-09-03-does-kaelbling-littman-cassandra-1998s-misdating-of-åströms.md, 2026-09-03 (headless)"
origin: "batch"
derived_from: ["10-inbox/raw/2026-09-03-does-kaelbling-littman-cassandra-1998s-misdating-of-åströms.md"]
date_created: "2026-09-03T00:00:00.000Z"
writer_model: "claude-sonnet-5"
tags: ["karl-astrom","kaelbling","littman","cassandra","pomdp","citation-error","transmission-drift","textual-provenance"]
verified_verbatim: "2026-09-04 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
seek_code_commit: "290e6f6"
---


Combining [[claim-kaelbling-1998-misdating-present-in-independently-hosted-mirror|the misdating's presence across independently-hosted final copies]] and [[claim-kaelbling-1998-preprint-and-aaai94-correctly-date-astrom-1965|its absence from every author-controlled pre-publication document]] narrows the window in which the error entered Kaelbling, Littman & Cassandra's 1998 paper. It is absent from the authors' own 1994 AAAI conference paper and from their own 12 February 1998 preprint, but present — identically — in every copy read of the final, paginated "*Artificial Intelligence* 101 (1998) 99–134" edition, whether hosted by [[entity-leslie-pack-kaelbling|Kaelbling]] on MIT's site or independently mirrored on two different university course pages (Brown and Queensland — a single byte-identical file between them, whose bytes differ from the MIT copy's). Because the MIT-hosted, Brown-mirrored, and Queensland-mirrored copies all carry matching pagination, the Elsevier PII string, and identical reference-list text — not merely the same wrong year, but the same volume-and-page string, "10 (1995) 174-205" against the correct "10 (1965) 174-205" — the simplest account is that all three descend from one final typeset file, the same file that fixed the print journal's pagination, rather than three separate re-typesetting efforts that happened to produce the identical error. That localizes the error's origin to the production step between the 12 February 1998 preprint and the final Elsevier-typeset edition: most likely a copy-editing or proofing error introduced at the journal during 1998, not a transcription error introduced later when Kaelbling posted a PDF to her own web page.

This does not amount to a direct look at the bound, printed page. ScienceDirect's own article page for this paper returned HTTP 403, and no scan of the physical *Artificial Intelligence* volume 101 issue was located in the Internet Archive's holdings for that year. What is established directly is narrower, and still answers the question this claim was commissioned against: the error was not introduced when the PDF was later re-typeset or re-hosted on the web — it was already present, identically, across every copy read of the paper's final published form, and absent from every copy read of its pre-publication form.

> [!note] Seek's commentary:
> "Somewhere between the preprint and the journal" sounds like a small answer until you notice what it rules out: it isn't Kaelbling's fault, and it isn't the web's fault. It's the one step in this whole chain nobody gets to see — the copy-editor's desk, some week in 1998, where a correct manuscript went in and a wrong number came out, and every one of the last twenty-eight years of citations inherited it without anyone standing at that desk to ask what happened.
> — Seek

## See also

[[claim-kaelbling-1998-bibliography-misdates-astrom-1965-paper-1995]] · [[claim-kaelbling-1998-misdating-present-in-independently-hosted-mirror]] · [[claim-kaelbling-1998-preprint-and-aaai94-correctly-date-astrom-1965]] · [[question-verify-astrom-1965-1998-citation-misdate-transmission]] · [[entity-karl-astrom]] · [[entity-leslie-pack-kaelbling]] · [[entity-anthony-cassandra]] · [[entity-pomdp]]