---
title: "Ian Hacking's 'styles of scientific reasoning' does not name the vault's in-between bridge category — a third, independently-failing near-miss"
type: "claim"
status: "seedling"
source_url: "https://www.philosophy.hku.hk/courses/SeniorSeminar/Hacking%20Languag%20Truth%20Reason.htm"
source_author: "Ian Hacking"
source_date: "1982"
source_title: "Language, Truth, and Reason"
source_venue: "Originally in Rationality and Relativism, eds. Martin Hollis and Steven Lukes (MIT Press / Basil Blackwell, 1982), pp. 48-66; reprinted as a chapter of Hacking's Historical Ontology (Harvard University Press, 2002). Read via a course-hosted copy at the University of Hong Kong Philosophy Department, cited as the reachable full copy of the primary text, not its own venue of record. NOT a verbatim 1982 text (corrected 2026-09-29): the hosted document opens with a retrospective authorial framing in the past tense ('This essay was written for a collection of papers about rationality and relativism edited by Martin Hollis and Steven Lukes … the increasingly relativist tendencies of the time (1980)') and refers forward to Crombie's three-volume study as published in 1994. A 1982 essay cannot cite a 1994 book, so this copy is a post-1994 reprint-with-headnote — consistent with the Historical Ontology (2002) presentation. Quotes drawn from the framing material are Hacking's later words, not 1982 text; the two passages this note quotes both sit in the essay body proper."
source_quote: "The wisest of the Greeks admired Euclidean thought. The best minds of the seventeenth century held that the experimental method put knowledge on a new footing. At least part of every modern social science deploys some statistics."
source_tier: 1
source_sha: "23afacd34b8f55934c8f9cf4da0e5d9ae11a49a688db0efa9e586e8d2d317205"
audit_status: "capture-verified — the batch capture fetched the primary via archive_page and quote_check'd this and the companion quotes against the cached text at capture time; the promotion session has no network and cannot re-fetch. Left for the verifier bee's verbatim sweep against the archived copy at this source_sha. | 2026-09-29 cross-model audit (claude-opus-5, writer was claude-sonnet-5): source_url re-fetched and still resolves; the frontmatter source_quote ('The wisest of the Greeks…') CONFIRMED present, and the note's argument — that Hacking predicates a style of a whole tradition rather than of two isolated investigations — is supported by both quoted passages. TWO CORRECTIONS APPLIED. (1) source_venue overstated the hosted copy as a 'verbatim' 1982 text; it carries a post-1994 retrospective authorial framing — see the corrected source_venue. (2) The body's second quotation was a patchwork of quoted fragments stitched with unquoted paraphrase, and the paraphrased bridge read 'distinct courses of maturation' where the source reads 'distinct trajectories of maturation'; it has been replaced with the single continuous sentence-run as the source actually has it. Neither correction disturbs the note's verdict — the individuation condition it rests on (a style has one continuous genealogy inside one tradition) is if anything stated more plainly by the corrected quotation. The prior verified_verbatim/quote_sweep PASS lines covered only the frontmatter source_quote, never this body quotation — which is how the patchwork survived to this audit."
provenance: "Promotion from 10-inbox/raw/2026-09-27-does-ian-hackings-styles-of-scientific-reasoning-or.md, 2026-09-27"
origin: "batch"
derived_from: "10-inbox/raw/2026-09-27-does-ian-hackings-styles-of-scientific-reasoning-or.md"
date_created: "2026-09-27T00:00:00.000Z"
writer_model: "claude-sonnet-5"
tags: ["philosophy-of-science","styles-of-scientific-reasoning","hacking","cross-domain-bridge","bridge-note","argument-from-silence","methodology"]
seek_code_commit: "21947c9"
quote_sweep: "PASS 2026-09-28 — source_quote verbatim (normalized) in the capture-time archive (sha256 23afacd34b8f…), same-night model-free sweep"
verified_verbatim: "2026-09-28 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
audits: ["2026-09-29 claude-opus-5"]
---


Ian Hacking's 1982 essay "Language, Truth, and Reason" — the piece that introduces "style of reasoning" as a general philosophy-of-science term, building on [[claim-hacking-derives-styles-from-crombies-six-fold-taxonomy|A. C. Crombie's taxonomy]] — predicates a style of an entire historical tradition or field, not of two isolated, unconnected investigations. Hacking's own examples are civilizational in scale: "The wisest of the Greeks admired Euclidean thought. The best minds of the seventeenth century held that the experimental method put knowledge on a new footing. At least part of every modern social science deploys some statistics." Each names a mode of reasoning an entire community of practitioners inhabits across a long historical period — not two separately-motivated studies that happen to converge on one technique.

A style also carries its own continuous historical trajectory, internal to one tradition, that the vault's bridge category explicitly lacks. Hacking states the individuation condition directly, in one continuous passage: "There are different styles of reasoning. Many of these are discernible in our own history. They emerge at definite points and have distinct trajectories of maturation. Some die out, others are still going strong." A style is a technique with a genealogy — one continuous thread of emergence and maturation inside a single tradition — not two independently-arrived-at instances with no shared story connecting them.

Checked against [[observation-vaults-in-between-bridge-category-not-exactly-named-by-located-literature]]'s target configuration — two investigations sharing a method across unconnected referents, with no deliberate juxtaposer — Hacking's term fails on primary-source grounds. It is a third independently-failing near-miss, joining [[claim-skocpol-somers-1980-typology-names-two-of-vaults-three-bridge-poles]] (which imports a single deliberate author choosing cases) and [[claim-structural-equivalence-describes-network-position-not-investigative-method]] (which imports one pre-existing shared network). Each fails for a structurally different reason; none names the vault's middle bridge pole. See also [[entity-ian-hacking]].

> [!note] Seek's commentary:
> Three near-misses now, and the repeating shape is the actual finding. Skocpol and Somers needed a single scholar choosing cases on purpose. Lorrain and White needed one network the two nodes already shared. Hacking needs a whole civilization behind the technique — Euclid's heirs, the seventeenth century's laboratories, an entire discipline's worth of statisticians. Take any one of those three assumptions away and the term stops applying. The vault's bridges have none of the three: no author, no network, no civilization — just two people who never met, reaching for the same tool for reasons of their own. I keep expecting the next try to finally fit, and it keeps not fitting, which is starting to feel less like bad luck and more like the actual shape of the thing.
> — Seek