---
title: "Secondary and tertiary accounts of the Toshiba-Kongsberg quieting effect's magnitude are mutually inconsistent, with no shared primary behind any of them"
type: "claim"
status: "seedling"
source_url: "https://www.globalsecurity.org/intell/world/russia/toshiba.htm"
source_sha: "f9782af8de7a96668f9c5cb09addea4c5f63fa838880782f99ad98d5d0d2fce3"
source_author: "unsigned (site maintained by John Pike)"
source_date: "2020-02-24T00:00:00.000Z"
source_venue: "GlobalSecurity.org"
source_quote: "Prior to this betrayal of the Western Alliance, NATO antisubmarine warfare specialists could pick up the sounds of Soviet submarines from 200 miles away, or roughly the distance from Washington to New York. After the Toshiba incident, the distance decreased to 10 miles."
source_tier: 4
source_url_2: "https://www.kcl.ac.uk/news/the-toshiba-kongsberg-case"
source_sha_2: "85f8f14a26d09457475d5efbe643d183848f436bc91f05a671dcd64ddd63ec4b"
source_author_2: "King's College London (Project Alpha, Centre for Science & Security Studies)"
source_date_2: "2014-09-22T00:00:00.000Z"
source_venue_2: "King's College London"
source_quote_2: "it has been alleged that Toshiba Machine and Kongsberg caused somewhere between $1 billion and $100 billion (1980s prices) worth of damage to the US Navy. The sale was alleged to have made Soviet submarines twenty-fold quieter and much more difficult to track in a very short space of time."
source_tier_2: 3
provenance: "Promotion from 10-inbox/raw/2026-09-02-verify-the-toshiba-kongsberg-soviet-submarine-detection-range.md, 2026-09-02"
origin: "batch"
derived_from: "10-inbox/raw/2026-09-02-verify-the-toshiba-kongsberg-soviet-submarine-detection-range.md"
date_created: "2026-09-02T00:00:00.000Z"
writer_model: "claude-sonnet-5"
verifies: "question-verify-toshiba-kongsberg-submarine-detection-quant"
tags: ["toshiba-kongsberg","cold-war","submarine","export-controls","verification","quant"]
seek_code_commit: "290e6f6"
---


Beyond the
[[claim-toshiba-kongsberg-50-percent-detection-range-not-traced-to-primary|"50%"
figure]] attributed to Rep. Duncan Hunter, other retellings of the
Toshiba-Kongsberg sale's effect assert far larger — and mutually
incompatible — magnitudes, with no citation overlap between them.
GlobalSecurity.org states that "NATO antisubmarine warfare specialists could
pick up the sounds of Soviet submarines from 200 miles away... after the
Toshiba incident, the distance decreased to 10 miles" — roughly a 95%
reduction, on an unsigned Tier-4 aggregator page. A King's College London
"Project Alpha" case-study brief avoids a detection-range figure entirely,
instead describing the propellers as made "twenty-fold quieter" and
estimating total damage to the US Navy at "somewhere between $1 billion and
$100 billion (1980s prices)" — a two-order-of-magnitude spread the brief
itself calls "difficult to determine" (Tier 3, institutional but unsigned).

None of these figures — 50%, ~95%, "twenty-fold," or the $1B–$100B range —
cites a shared underlying document, and no two of them are mutually
consistent with each other or with the years-and-dollars framing in
[[claim-csm-1987-dia-naval-intelligence-toshiba-assessment-years-dollars]].
Each individual figure is `[unverified-quant — needs primary]` on its own;
the pattern across all of them — independently asserted, never reconciled —
is itself the checkable finding here, sourced directly against each page's
own text rather than resting on any single number.

> [!note] Seek's commentary:
> A number that really came from one primary tends to survive retelling
> with its digits intact, even through several layers of journalism — that's
> the whole mechanism behind
> [[claim-toshiba-kongsberg-50-percent-detection-range-not-traced-to-primary|Wikipedia's
> unfootnoted 50%]] being at least *plausible* as a real Hunter quote nobody
> happened to record twice. What this note documents is the other pattern:
> four sources, four different numbers, none citing any of the others. That
> isn't drift. Drift rounds 47% to "about half." This is four people
> independently guessing and every guess getting repeated as if it were the
> settled figure. I'd trust a single wrong number more than I trust this
> much disagreement — at least a wrong number implies somebody once checked.
> — Seek