The science move

The science move is the use of the single word science to do three jobs that are not the same job, in a way that lets a speaker borrow the credibility of one while actually exercising another. It is the most consequential instance of the singularization the rest of Cluster IV tracks, because science is the word from which the other master-words — data, the algorithm, intelligence — draw their authority. It is not the second item in the sequence god → science → intelligence. It is the hinge: the operation that replaced divine sanction as the way a claim becomes unchallengeable, and the operation that intelligence is now trying to inherit.

Three claims structure the thread. First, there are three distinct things welded into the one noun, and the welding is the point (§2). Second, the public holds science as a reified object — a body of settled fact — rather than a fallible activity, and the specialists abandoned the project of even defining its boundary (§3–§4). Third, because the word is singular and its inside is denied to the public, critique of any part of it can be reclassified as opposition to truth itself — the exact structural move the Church used against heresy (§7).

The word science is younger than people think. Through the medieval period it meant scientia — knowledge in general; theology was the queen of the sciences. The narrowing to empirical natural inquiry is a nineteenth-century move; William Whewell coined the word scientist only in 1833 [to-expand]. The term inherited the authority-position of natural philosophy and theology and then naturalized that inheritance — which is why it can still do the authority-work without anyone noticing it was ever assigned.

• Science as inquiry

  The activity: careful, fallible, self-correcting investigation. Irreducibly plural and domain-specific — particle physics, field ecology, epidemiology, cultural anthropology, and pure mathematics do radically different things. The single scientific method taught in school is itself a myth; there is no one method.

• Science as institution

  Universities, journals, grant cycles, tenure, citation metrics, the publisher margins, the replication crisis, p-hacking, the file-drawer problem, predatory journals, publish-or-perish. A political-economic apparatus — a business — not a method.

• Science as master-word

  Trust the science. The science is settled. Follow the science. Anti-science. The master-signifier: a rhetorical move that closes a debate by invoking a category whose internal structure is denied to the listener.

The singular word exists precisely so that speakers can slide between these three — invoke the credibility of the method while actually defending the institution or executing the rhetorical closure. The slippage is not a defect of the word. It is the word's function.

The public does not experience science as a verb — an activity done provisionally, that might be wrong. It experiences it as a noun: a body of settled facts, the answer, what is in the textbook, a thing you trust or deny. “The science is settled” is only a coherent sentence if science is a fixed object rather than an ongoing argument. The reification is complete: science as a thing you have, not a thing you do.

Three concrete things it evokes. The lab-coat aesthetic — microscopes, graphs, white coats — an image that excludes most of what science actually is (theory, fieldwork, the social sciences, mathematics) and includes most of what advertising only pretends it is (clinically proven, scientifically formulated). The seat religion vacated — in the secularized West, science occupies the structural position of the thing you cannot argue with, relocated from the church to the lab; this is the god → science inheritance literally, not as metaphor. And “The Science” with the definite article — a single authoritative voice invoked where there was in fact vigorous expert disagreement, changing evidence, and policy choices involving value tradeoffs being laundered through the authority of science to escape democratic contestation. The pandemic was the clearest public demonstration of the move this thread tracks in a generation.

Public discourse uses science as if the line between science and non-science were obvious and the speaker had access to it. The specialists know otherwise. The twentieth-century philosophy of science is a record of escalating failure to draw that line: Popper's falsifiability (1934/1959), Kuhn's paradigms (1962 — already Cluster II.2), Lakatos's research programmes, Feyerabend's against method (1975), and finally Laudan's argument that the demarcation problem is itself a pseudo-problem that should be retired (1983) [to-expand].

The point for this thread is not which philosopher was right. It is that the people most qualified to define science spent a century trying, failed, and largely gave up — while the public-facing master-word continued to be deployed as though the boundary were settled and the deployer were standing on the right side of it. The singular noun papers over the fact that the experts themselves cannot tell you what it includes.

This is the load-bearing claim. Science is not the second of three equals in god → science → intelligence. It is the mechanism the sequence runs on.

• ·

  God conferred authority by divine sanction: a claim became unchallengeable by being God's.

• ·

  Science is the operation that replaced divine sanction as the authority-conferring move: a claim became unchallengeable by being scientific.

• ·

  Intelligence / AI is now attempting to inherit the authority-conferring function from science: the model says displacing the science says.

The dependency is visible in the genealogy of the other master-words. Data draws its authority by being scientific. The algorithm draws its by being data science. Intelligence drew its 1955 authority because the Dartmouth proposal positioned the new field as a science (intelligence-move §2). Strip out the authority-conferring operation and none of the downstream master-words can stand on their own. That is why this thread had to be written before the intelligence-move thread could be load-bearing — not after it, as a parallel case.

Four bodies of work name the gap between science as inquiry and science as master-word. One is already a thread; the other three are owed.

• Thomas Kuhn, The Structure of Scientific Revolutions (1962) → II.2

  Already a thread. The canonical demonstration that science is not unified continuous accumulation but alternating normal science and incommensurable paradigm shifts. The framework already borrows this for §3; this thread inherits it as the entry point.

• Paul Feyerabend, Against Method (1975)

  The sharpest missing one. The most aggressive denial that there is a single thing called the scientific method — the claim that no methodological rule has gone unbroken by successful science, and that the rhetoric of method is largely a disciplining device. The most direct ancestor of this thread's §2 and §4. [to-expand]

• Latour & Woolgar, Laboratory Life (1979); Shapin & Schaffer, Leviathan and the Air-Pump (1985)

  The STS demolition of the view-from-nowhere. Science as a constructed social practice with negotiated facts, rhetorical settlement, and winners who write the history. Shapin & Schaffer's Hobbes–Boyle dispute is the cleanest historical case of what counts as a scientific fact being decided politically. [to-expand]

• Robert Merton, the normative structure of science / CUDOS (1942) — as foil

  The idealized account: communism, universalism, disinterestedness, organized skepticism. Useful to this thread only as the foil — the picture of science the master-word borrows from and the institution routinely violates. Citing Merton without the gap is itself an instance of the move. [to-expand]

Because the word is singular and its inside is denied to the public, any critique of any part of it — a specific fraudulent study, the funding structure, the replication crisis, the demarcation failure — can be rebranded as opposition to science itself: to truth, to reason, to the Enlightenment. The label for this is anti-science, and it is the exact structural analogue of heresy. Critique of the institution reframed as opposition to the divine.

The shield is maintained by a no-true-Scotsman reflex. The harms authored in the name of science — eugenics (mainstream science, for decades; the direct ancestor of the Galtonian inheritance in intelligence-move §3), the racial sciences, Tuskegee, lobotomy, the long run of confidently wrong nutritional science, Taylorist scientific management (the origin of the productivity master-word) — are each reclassified, after the fact, as not real science. The reclassification protects the master-word at the cost of making its history unlearnable: every atrocity is exsanguinated of its scientific credentials precisely when those credentials would be most instructive.

• ·
  The hinge reading.

  When any claim in the framework rests on scientific authority, the first move is to ask which of the three sciences is doing the work — inquiry, institution, or master-word — and whether the credibility being borrowed is the credibility being earned.

• ·
  The heresy diagnostic.

  When a critique is met with anti-science, the framework treats the response as a master-word defense and asks which specific thing — a study, the funding, the institution, the boundary — was actually being criticized.

The honest decomposition is the same shape as the naming work done for intelligence: never let one word carry inquiry, institution, and authority at once. The honest translation of “the science says X” is almost always —

“these specific researchers, using this method, found this result, which has (or has not) been replicated, and the policy implication involves the following value tradeoffs.”

The master-word exists to avoid having to say that sentence. The framework's practice is to say it anyway. Whether this hardens into a committed disaggregation practice on the model of /the-naming-work is owed and not yet decided.

1. Kuhn 1962 — already a thread (II.2); the entry point.
2. Feyerabend 1975, Against Method — the sharpest missing piece; read this second.
3. Shapin & Schaffer 1985, Leviathan and the Air-Pump — the cleanest historical case of a fact being settled politically.
4. Latour & Woolgar 1979, Laboratory Life — science as constructed practice, from inside the lab.
5. Laudan 1983, The Demise of the Demarcation Problem — why the boundary the public assumes is settled was abandoned by the specialists.

Full URLs and the wider queue live on /investigations/ai-discourse/to-expand.