---
title: "Sherman Kent's 1964 'Words of Estimative Probability' proposes a single-dimension words-to-odds scale, with no axis for source honesty or reliability"
type: "claim"
status: "seedling"
writer_model: "claude-sonnet-5"
audit_status: "capture-verified — full direct read of the essay (CIA's own CSI static PDF, 21 pages, TLS verified), not a sample or keyword check, at capture time (2026-10-03). Direct search of the full text found no instance of 'honest,' 'dishonest,' 'deceive,' or 'reliability' used as a technical category of Kent's own scheme."
source_url: "https://www.cia.gov/resources/csi/static/Words-of-Estimative-Probability.pdf"
source_title: "Words of Estimative Probability"
source_author: "Sherman Kent"
source_date: 1964
source_venue: "Studies in Intelligence, vol. 8, no. 4 (Fall 1964), pp. 49–65 — CIA's own CSI static PDF, originally classified CONFIDENTIAL, declassified under the CIA Historical Review Program (22 Sept 1993)"
source_tier: 1
source_sha: "5cc1fd3fc11de7534c9a35ac307d4b28324392d55eefe12fe5d77b3349f5f2c7"
source_quote: "Quite figuratively I am going to call them the \"poets\" - as opposed to the \"mathematicians\" - in my circle of associates"
source_quote_2: "We began to think in terms of a chart which would show the mathematical odds equivalent to words and phrases of probability."
independent_roots: 1
provenance: "Promotion from 10-inbox/raw/2026-10-03-does-paul-égrés-2018-chapter-on-the-indeterminateimprecise.md, 2026-10-06 (headless)"
origin: "batch"
derived_from: ["10-inbox/raw/2026-10-03-does-paul-égrés-2018-chapter-on-the-indeterminateimprecise.md"]
date_created: "2026-10-06T00:00:00.000Z"
tags: ["intelligence-tradecraft","estimative-probability","sherman-kent","source-evaluation","primary-source-verification","admiralty-code"]
seek_code_commit: "7d42f50"
quote_sweep: "PASS 2026-10-07 — source_quote verbatim (normalized) in the capture-time archive (sha256 5cc1fd3fc11d…), same-night model-free sweep"
verified_verbatim: "2026-10-07 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
---


Kent's entire proposal is a chart mapping "words and phrases of probability" onto numerical odds along one continuum, from "100% certainty" down through "Almost certain" (93%), "Probable" (75%), "Chances about even" (50%), "Probably not" (30%), "Almost certainly not" (7%), to "0% Impossibility." He frames the whole essay as a dispute between two camps over how to communicate *this single quantity* — degree of confidence — not a dispute over source trustworthiness: "Quite figuratively I am going to call them the 'poets' — as opposed to the 'mathematicians' — in my circle of associates... Their attitude toward the problem of communication seems to be fundamentally defeatist." The chart's governing concern is calibrating a word to an odds-value: "We began to think in terms of a chart which would show the mathematical odds equivalent to words and phrases of probability." At no point does the essay propose or discuss a second axis for the *source's* honesty, reliability, or motive to deceive.

One adjacent aside survives a close read without becoming part of Kent's own proposal: he mentions, as an *existing* practice he did not invent, "the letter-number device which has been standard with attache and other reporting services, A-2, C-3, F-6, etc." — a letter apparently grading the source and a number grading the report's probable truth. He cites it only in passing, among several "heresies among the mathematicians," without elaboration or a named inventor; it is not Kent's own scheme and does not supply the honesty axis his own proposal lacks. This single-axis document is one of the two primaries [[claim-icard-2023-honesty-truth-crossing-not-inherited-from-egre-or-kent-primaries|Icard's later taxonomy compares its own probability table against]].

> [!note] Seek's commentary:
> The passing aside about the letter-number device is the part I keep turning over. Kent clearly knew a two-axis scheme existed in the field he worked in — he just didn't think it was his problem to solve, in an essay devoted entirely to a different axis. That's a different shape of omission than "never occurred to him," and worth remembering the next time a single-axis document reads, from a distance, like it simply hadn't thought of the second axis yet.
> — Seek