The Making of Law
Watching the Conseil d'État work
Description
Somewhere in the Palais-Royal in Paris, a handful of counsellors in a wood-panelled room are turning over a grey cardboard file. One of them, the rapporteur, has spent weeks with it; another, the reviser, has read behind him; a third will represent the administration being sued. They are the Conseil d'État, France's supreme administrative court and, for two centuries, the body that decides whether the state has acted within the law. And what they are doing, mostly, is talking about paper — which stamp is missing, whether a deadline was met, whether this case looks enough like that older one to be treated the same way.
Between 1993 and 2002, the anthropologist Bruno Latour was allowed to sit in on this. He had spent his career watching scientists in laboratories, following how facts get made at the bench. Now he pointed the same patient, slightly naive gaze at judges. Not to expose them, not to celebrate them — simply to describe what legal reasoning actually consists of when you refuse to take the word "law" as an explanation. The result, published in French in 2002 and in English as The Making of Law, reads a court the way one might read a colony of microbes: closely, and without reverence.
What comes out of that grey file is stranger than any theory of justice. Law, watched from inside, is not the majestic application of principles to cases. It is a slow, material, oddly humble craft of moving files, weighing hesitations, and above all keeping something intact that has nothing to do with the truth of the facts.
The question we’re asking : What is legal reasoning actually made of, once we stop taking the word "law" as its own explanation?What we’ll see : An ethnographer sits inside France's highest administrative court and watches how a judgment is built, file by file — and how that craft parts ways with science.
Table of contents
01 Chapter 1 — An anthropologist in the corridors of the Palais-Royal
Latour came to the Conseil d'État with an unusual credential: he had already made his name watching people who were supposed to be too rigorous to be watched. In the 1970s he had installed himself in a biology lab in California and described, in almost embarrassing detail, how researchers pipetted, argued, and gradually turned a mess of readings into a stable fact. The lesson he drew was that truth is not found ready-made; it is assembled, and you can see the assembly if you stay long enough. He proposed to do the same thing with the law, and the Conseil, to its credit, let him in.
The institution he entered is peculiar and very French. Founded in its modern form by Napoleon in 1799, the Conseil d'État is both an adviser to the government on draft legislation and the court of last resort when a citizen sues the administration — a dismissed civil servant, a refused building permit, an expelled foreigner. The same body that helps write the rules also judges whether the state has broken them. Latour was interested less in this constitutional oddity than in the daily gestures: how counsellors greet a file, how they pass it around, how they speak in a shorthand that outsiders barely follow.
02 Chapter 2 — Following the file, not the principle
The unit of the Conseil d'État is not the case in the dramatic sense — a wronged person, a scandal, a right to defend. It is the file: a physical dossier of documents that arrives, circulates, and eventually leaves as a decision. Latour follows it the way he once followed a sample through a laboratory. A rapporteur is assigned. He reads everything, writes a draft opinion, and proposes an outcome. A reviser reads behind him and quarrels. Then the matter is argued and voted in a small collegial formation. At each step the object being worked on is not "the truth" but the file's readiness to become a judgment.
What strikes the observer is how much of the discussion is about admissibility rather than substance. Before anyone asks whether the citizen was actually wronged, the counsellors ask whether the claim can even be heard: Was it filed in time? Against the right authority? By someone with standing? A great many files die at this gate, and the counsellors feel no anguish about it. A claim that is right on the merits but late by a week is simply not the court's business. To an outsider this looks like heartlessness or pedantry. To the institution it is the condition of being able to function at all.
03 Chapter 3 — What holds a legal argument together
If a judgment is not deduced from principles, what actually binds it? Latour's answer is that law moves by a chain of connections he calls, borrowing the counsellors' own instinct, the passage of means. A legal argument advances by linking texts to texts, precedents to files, qualifications to facts, each link having to be authorised by another. The lawyer does not ask "is this true?" but "is this connected to something already accepted, by a route the court recognises?" The whole art is in the routing.
This is why precedent does such heavy lifting. When counsellors reach for an earlier decision, they are not invoking an authority that settles the matter from above; they are checking whether the present file can be attached to a line of already-decided ones without straining it. Two cases are "alike" not because reality made them alike but because the court decides to treat them as continuous. Continuity is manufactured, carefully, and everyone knows it is manufactured — which is exactly why they are so cautious about each new link.
04 Chapter 4 — Two ways of being objective
Having watched both laboratories and courts, Latour can finally say what he came to say: law and science are two entirely different ways of being reliable, and confusing them does damage to each. In the lab, a statement earns its truth by pointing outward — to the specimen, the instrument, the world it claims to describe. Latour called this the chain of reference: a fact stays true only as long as you can travel back along that chain to the thing itself. Break the chain and the fact collapses. Science is forever reaching beyond its own texts toward an external reality it did not make.
Law does the opposite. A judgment earns its authority not by pointing outward to the facts — those are already fixed and, in a sense, no longer the court's concern — but by pointing inward, to other judgments, to texts, to the coherent body of decisions it belongs to. The legal chain does not connect a statement to the world; it connects a decision to the whole apparatus that lets it count as a decision. That is why a scientist can be delighted to overturn a predecessor, while a judge overturns precedent only reluctantly and with elaborate justification. The scientist gains by breaking with the past; the jurist survives by remaining continuous with it.
05 Conclusion
The grey file eventually leaves the room. It goes out as a decision — a few dense paragraphs that a citizen may never read, joined to the two centuries of decisions before it, ready to be reached for the next time a case looks enough like this one. Nobody in the room discovered anything about the world. What they did was decide, with great care, that the state had or had not stayed inside the law, and to say so in a form that the whole edifice could bear.
Related books
