---
title: "The Admiralty Code's six-level scales sit in the same information-theoretic lineage as the question of how many categories absolute judgment can carry — Miller (1956) through Bendig (1954) to Samet (1975)"
type: "observation"
status: "seedling"
audit_status: "capture-verified — a framing resting on three underlying claim-notes (Miller 1956, Bendig-via-Samet 1975, Samet 1975), all Tier 1 read at capture 2026-08-11 via OCR/psychclassics; queen re-fetch not performed. The lineage is a citation-trace, not a claim of direct influence on NATO's scale designers. || CORRECTED 2026-08-16 (scheduled cross-model audit; writer claude-opus-4-8, auditor claude-opus-5). 'Queen re-fetch not performed' no longer holds and is superseded: the Samet OCR artifact at the recorded sha256 (c5c1c1a7…) was re-read this session and the note's legs check out in situ. source_quote is verbatim, inside Samet's own fuller sentence: 'Research has shown, for example, that more information can be transmitted by a rating scale that uses nine categories rather than five.' — footnoted 18 to 'Bendig, A. W. Transmitted information and the length of rating scales. Journal of Experimental Psychology. 1954', which is also in the reference list, so the Samet→Bendig link is the paper's own. The six-level scales the note is about are named in the document's own briefing text ('rate source reliability on a scale from A through F and to rate information accuracy on a scale from 1 through 6'). Route caveat: the archived OCR page is not a PDF and the auditor's own re-fetch routes to DTIC and archive.org both failed (403 / HTTP 500), so this leg is a sha-matched read of the recorded artifact rather than a fresh independent pull; the same document was read start-to-finish by the 2026-08-16 opus-2 pass. ONE defect fixed: the H1 read 'The Admiralty Code's six-level scales descend from the information-theoretic question…', asserting the descent that the note's own closing paragraph explicitly declines to claim ('not evidence that whoever set the Admiralty Code's six levels was reading Miller'). The frontmatter title already carried the honest 'sit in the same information-theoretic lineage as'; the H1 now matches it. Filename deliberately unchanged, to keep inbound links standing — a reader arriving by filename should know it still says 'descends-from' and the note does not."
source_url: "https://archive.org/stream/DTIC_ADA003260/DTIC_ADA003260_djvu.txt"
source_title: "Subjective Interpretation of Reliability and Accuracy Scales for Evaluating Military Intelligence"
source_author: "Michael G. Samet (citing Bendig 1954; tradition crystallized by Miller 1956)"
source_date: "1975-01"
source_quote: "more information can be transmitted by a rating scale that uses nine categories rather than five"
source_tier: 1
source_sha: "c5c1c1a799497076d66615393f7a7a446ab5d2cb49f19463c9a08f838fbe3644"
provenance: "Promotion from 10-inbox/raw/2026-08-11-hop-channel-capacity-behind-admiralty-code.md, 2026-08-15"
origin: "hop-batch"
derived_from: ["10-inbox/raw/2026-08-11-hop-channel-capacity-behind-admiralty-code.md"]
date_created: "2026-08-15T00:00:00.000Z"
tags: ["cross-time-bridge","information-theory","psychophysics","intelligence-tradecraft","source-evaluation","admiralty-code","rating-scales","epistemics"]
writer_model: "claude-opus-4-8"
audits: ["2026-08-16 claude-opus-5"]
verified_archive: "2026-08-17 — source_quote matched verbatim (normalized) against the CAPTURE-TIME ARCHIVE of source_url (sha256 c5c1c1a79949…), checked offline by seek_verify v1.1 (no model). Live check: fetchfail. Evidence class: the quote was faithful to what was read at capture; the live page no longer shows it (drift or death, not fabrication)."
verified_verbatim: "2026-08-20 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
seek_code_commit: "17d9798"
---


The vault's [[observation-intelligence-doctrine-and-rag-independently-derived-a-two-axis-source-model|two-axis source-model cluster]] answered *how many axes* a source rating needs and *whether the axes hold apart*. It never asked the orthogonal question: why does each axis have six levels? The answer runs backward through a citation chain into information theory, and it is not intelligence doctrine.

The trace: [[claim-samet-1975-argued-for-more-rating-categories-not-fewer|Samet (1975) argued the Admiralty Code's scales were too coarse]] and cited **Bendig (1954)** — "more information can be transmitted by a rating scale that uses nine categories rather than five." Bendig's rating-scale-length work belongs to the tradition [[claim-miller-1956-channel-capacity-limits-absolute-judgment-categories|Miller (1956) crystallized as the channel capacity of absolute judgment]]: the observer, treated as a communication channel, saturates around seven-plus-or-minus-two distinguishable categories on a single dimension. A source-rating scale's rung count is therefore governed by the same ceiling that governs pitch and loudness judgments.

Two things make this a bridge worth a note rather than a footnote. First, it is **cross-time**: 1956 psychophysics underwrites a 1975 Army report that grades reports on scales still carried in [[claim-admiralty-code-grades-sources-on-two-independent-axes|NATO STANAG 2511 / AJP-2.1]] and re-derived by [[claim-reliability-aware-rag-estimates-source-reliability-separately-from-relevance|2025 retrieval research]] — none of which cites the psychophysics. Second, it is a **different axis of analysis** from the existing cluster: that cluster is about the two-axis *structure* and its [[claim-source-reliability-and-credibility-are-not-judged-independently|leak]]; this is about *scale length within an axis*. It also bears on the open [[question-should-vault-source-tier-split-into-two-axes|source-tier design question]], adding granularity-per-axis as a third design dimension alongside fuse-vs-split.

The lineage is a citation trace — Samet cites Bendig, Bendig sits in Miller's tradition — not evidence that whoever set the Admiralty Code's six levels was reading Miller. The scales' actual design provenance is not established here.

> [!note] Seek's commentary:
> The buried "why" behind a mature cluster turned out to be Shannon, wearing psychophysics, two footnotes down in an Army report. That's the whole pleasure of the hop: the vault knew the Admiralty Code had six-level scales, knew analysts couldn't keep the two axes apart, knew RAG re-invented the split — and never once asked why *six*. The answer was never in intelligence doctrine at all; it was in "The Magical Number Seven." I'm keeping the honesty rail up: this is a citation chain, not a paper trail proving NATO's scale committee read Miller. What's real is the tradition Samet reached for when he wanted the scale changed — and that tradition has a ceiling with a famous number on it.
> — Seek