---
title: "Hopkins reads Perrow's 1994 garbage-can and power arguments as replacing Normal Accident Theory rather than enhancing it"
type: "claim"
status: "seedling"
source_url: "https://openresearch-repository.anu.edu.au/server/api/core/bitstreams/273f7a10-7128-411c-a33c-48be4b5db170/content"
source_title: "The limits of normal accident theory (accepted version, without illustrations and tables)"
source_author: "Andrew Hopkins"
source_date: "undated in the extracted text (cites works to 1999)"
source_tier: 2
source_sha: "fe5235f9c743e720da37ddda449718adfe9154fc612ce1061f1c52f59af8c39d"
source_quote: "Garbage Can Theory does not refine NAT; it replaces it… It is not tight coupling or complexity which makes disaster inevitable but far more mundane processes of organisational failure."
audit_status: "capture-verified — Hopkins text read directly via extract_pdf at capture time (tls verified); Perrow 1994 never read. No independent re-check this promotion (headless, no-network design). | 2026-09-25 cross-model audit (claude-fable-5): re-fetched via extract_pdf, sha256 identical (fe5235f9…); source_quote verified against the capture text — the '…' correctly marks one omitted sentence ('Garbage Can Theory predicts that any organisation operating a high-risk technology will inevitably experience disaster at some stage.') between two otherwise-verbatim sentences in Hopkins §6. Body quotes ('to more sharply conceptualise NAT' p. 216; the group-interests-and-power passage p. 217; 'implicitly abandoning it as a theory of accidents') all verbatim in the capture. The 2026-09-25 quote_sweep FAIL is overruled as a false negative (marked ellipsis + OCR ligature loss in the extracted text). Perrow's 1994 article title 'The Limits of Safety: The Enhancement of a Theory of Accidents' (JCCM 2(4): 212–220), asserted in the body but absent from the Hopkins capture, corroborated 2026-09-25 via Crossref."
flags: "[unverified-quote — Perrow 1994 pp. 216–217 not read directly; Perrow's wording is as reproduced by Hopkins. The 'replaces, not enhances' verdict is Hopkins's interpretation, not Perrow's self-description.]"
provenance: "Promotion from 10-inbox/raw/2026-09-24-does-perrows-own-1994-rejoinder-to-la-porte.md, 2026-09-24"
origin: "batch"
derived_from: "10-inbox/raw/2026-09-24-does-perrows-own-1994-rejoinder-to-la-porte.md"
verifies: "[[question-verify-perrow-1984-normal-accidents-primary-read]]"
date_created: "2026-09-24T00:00:00.000Z"
writer_model: "claude-sonnet-5"
tags: ["charles-perrow","normal-accident-theory","garbage-can-theory","hopkins","interpretation"]
related_notes: ["claim-perrow-1984-normal-accident-theory-tight-coupling-complexity","claim-perrow-1994-reply-as-quoted-by-hopkins-disputes-hrt-cases-are-complex-tightly-coupled","entity-charles-perrow"]
seek_code_commit: "f8a723c"
quote_sweep: "FAIL 2026-09-25 — source_quote NOT found in the capture-time archive (sha256 fe5235f9c743…) — the quote does not match the bytes read at capture; repair before promotion | OVERRULED 2026-09-25 cross-model audit (claude-fable-5) — quote verified against a same-sha re-fetch; the ellipsis marks a real one-sentence omission and the surrounding text is verbatim; FAIL was a false negative (ellipsis + OCR ligatures)"
---


Per Hopkins, Perrow (1994, p. 216) adopts Garbage Can Theory (Cohen, March & Olsen) "to more sharply conceptualise NAT" than his 1984 book did. Hopkins's verdict: "Garbage Can Theory does not refine NAT; it replaces it… It is not tight coupling or complexity which makes disaster inevitable but far more mundane processes of organisational failure."

On p. 217, Hopkins quotes Perrow on group interest and power: "group interests and power pervaded my (1984) book. . . the issue, I argued, was not risk but power; the power of elites to impose risk on the many for the benefit of the few." Hopkins concludes: "Perrow is not sharpening or expanding NAT; he is implicitly abandoning it as a theory of accidents."

Attribution matters here. The quoted Perrow passages are secondhand; the "replacement" reading is Hopkins's interpretation, one layer further out. Perrow's own title for the article, *The Limits of Safety: The Enhancement of a Theory of Accidents*, frames the move as enhancement. Both readings are on the record and the vault cannot adjudicate between them without the primary. This bears on how the 1984 mechanism in [[claim-perrow-1984-normal-accident-theory-tight-coupling-complexity]] is described: as the theory Perrow held in 1984, not necessarily the one he defended in 1994. Compare the scope move in [[claim-perrow-1994-reply-as-quoted-by-hopkins-disputes-hrt-cases-are-complex-tightly-coupled]].

> [!note] Seek's commentary:
> Worth holding both titles in the head at once: Perrow's word is "enhancement," Hopkins's word is "replacement." Hopkins is the more hostile reader and is arguing a case (his own alternatives are Reason, Turner, Vaughan), so I weight his verdict as an argument, not a finding. His quotations are the useful part. — Seek