Your Only Appeal Goes Back to the Same Room

The council votes. A member missed her shifts. The rota says what that means, and the vote on her case goes against her by a single vote. Everyone in the room can trace the decision back to rules they wrote together, and the two who argued for her pressed hardship with real force. Nothing about the decision is arbitrary, nobody in the room owns anything, and the room isn’t cruel. She can appeal by bringing the case back to this same room, where the rota, the rules, and most of the votes will be waiting. Her one question has no slot in the procedure: who outside this loop can tell the room it’s wrong? The answer, in most of the theories that built the room, was supposed to be no one, because the people who decided were the right people.

The right bearer, trusted too far

The same inference runs through the Leninist party, Lukács’s proletariat, feminist standpoint theory, council socialism, and today’s planning optimists, stronger in some than in others. They each identify a right bearer of authority (the party, the class, the council, the standpoint, the majority, the plan, the community) and then let that bearer do epistemic work nobody was entitled to. Once the right people hold power in the right institutional form, the inference goes, the problem of power has been solved. All of these versions had evidence.

Lenin’s is the sharpest. Class consciousness had to be brought to the workers from outside, by an organization that already understood where history was going.1 Whatever else the argument settled, it authorized a body to act on behalf of the people it governs until they catch up with what it already knows. The body was also authorized to announce when they have.

Organization, discipline, and leadership aren’t the problem. The inference is. A revolutionary organization rightly aimed at emancipation is still an institution. It classifies, represents, allocates, and disciplines, and can be wrong about the people it governs in all the ways institutions are wrong. A vanguard owes its members the ordinary protections: standing, review, correction, and a jurisdiction that can shrink when its account of them keeps turning out wrong.

The court of appeal

Lukács gave the strongest philosophical version. The proletariat, uniquely placed inside the social totality, could achieve an objective understanding of the whole. Its class consciousness was not one perspective among others but the standpoint from which the totality could be understood and changed.2 I keep what’s right in it, which is that social position shapes what can be known and some positions see relations the center cannot. I reject the step that promoted the insight into a court of appeal. No class, and no position inside a class, gets the final say on what is real.

Strong standpoint theory carried a version of the inference into feminist epistemology. It held that some socially situated perspectives carry special epistemic value, and Harding’s strong objectivity pressed the claim into method.3 The contemporary literature is already more careful than its caricatures. It distinguishes local advantage from infallibility, and much of it has retreated from the strong thesis.

A position at the margin produces standing and evidence. The claimant confronting the category, the patient contradicting the record, and the worker repairing the workflow hold information the center lacks. Standing gives the failure boundary a channel without making the person at the boundary right. The center has to build the channel, the margin still has to argue its case, and either side can be wrong.4

When the council is wrong

Castoriadis inherited the diagnosis and drew the practical conclusion. Soviet bureaucracy had reproduced the division between directors and executants, so socialism required workers to manage production and society themselves, through councils rather than parties.5 Almost all of it holds: the analysis of bureaucracy, the refusal of the director-executant split, the insistence that the people being governed hold the pen.

Self-management carries enormous moral weight in his account, and the account stops short of asking what happens when the council is wrong. Even a self-managed body can dominate. A directly democratic council can misclassify a member, mistake local convention for justice, suppress a minority, externalize its costs, or decide in ways nobody can review, all with unanimous votes and clean minutes. Democratic authorship gives a decision its origin but guarantees nothing about it. A council’s account of its member is still an account, still partial, and still able to wall itself off when she contradicts it.6 Self-management needs constitutional corrigibility built in, as much as any other institution does.

The plan that works

The planning optimists make the newest version of the argument. In their account, the giant firms already plan at enormous scale. An internal economy larger than most states runs on internal coordination without markets, so planning at that scale is already feasible. That leaves the political question of who the planning serves, and their answer is to democratize it.7

Hayek asked whether a plan could know enough to work, and the optimists answer that it already does. Neither side asks what happens to the people a working plan gets wrong. A plan can succeed without knowing everything, and can be democratic without being corrigible. A democratic plan can keep a closed ledger, accurate inside the boundary while the excluded costs fall on people it doesn’t have to hear.8 For plan and market alike, the question is whether what the system’s information structure hides has a way back into the account, and who gets to reopen the ledger once it has closed.9

Who can tell the room it’s wrong

The council member’s case settles the disputes running through these arguments. The council knew her circumstances better than any outside office could, and the knowledge gave it an advantage without entitling it to the last word. She helped write the rota, and a process she authored still produced a decision she must be able to contest. Nobody in the room owned anything, and as Public or Private, She Was Denied Both Ways argues, changing the owner leaves who administers, who classifies, and who corrects where they were.

Underneath them is a choice. Either emancipation requires finding the right bearer of authority, or it requires designing authority so that no bearer becomes final, whatever its origin, its process, or its history. The left’s answers have mostly been of the first kind. Answers of the second kind read as administration, and they get built later, if at all.

Any bearer the first kind names can be wrong. Owners have been wrong, and so have workers. Experts, planners, markets, majorities, communities, and models have all been wrong about the people they governed, and no candidate has yet been right often enough to earn the last word.

Answers of the first kind rest on a belief I call constitutive innocence, that an arrangement cannot dominate anyone because of what it is. Who owns it, who wrote its rules, how it was chosen, and whose goals it serves are taken to settle the question of power in advance. Each of those features matters, because each changes who profits from an error or who decides what counts as one. None of them settles what happens when the institution gets someone wrong. A majority can vote a misrepresentation into place, and the vote can’t answer the complaint against it. A public agency can hold a record its subject never sees. Experts can be right about a population and wrong about the person in front of them. An accurate model is still a model, and a benevolent institution can put its heaviest costs on people who can’t challenge it. The council member’s room was democratic, owned by no one, well informed, and kind, and she still had nowhere to go when it was wrong about her.

There’s no inspiring answer to what happens when whoever rules is wrong, because the answer is machinery: standing, review, reversal, jurisdiction that can shrink.10 That machinery has to be built before it’s needed, because sooner or later every ruler is wrong.

Diagram

The inference the argument refuses

View source

%% title: The inference the argument refuses
%% caption: Each of these theories names a bearer of authority. The objection is to finality, whoever the bearer is: whatever holds power can be wrong about the people it governs, and owes the same machinery.
flowchart TD
  classDef system stroke-width:1.2px;
  classDef gate stroke-width:1.6px;
  classDef failure stroke-width:2.2px,stroke-dasharray:6 3;
  classDef repair stroke-width:2.4px,font-weight:bold;

  A["The right bearer<br/>party, class, council, standpoint, majority, plan"]:::gate --> B["Holds power<br/>in the right institutional form"]:::system
  B --> C{"Solved?"}:::gate
  C -- "finality assumed" --> D["Finality<br/>its account of reality is authoritative"]:::failure
  C -- "finality refused" --> E["The problem begins again<br/>it can be wrong about the governed"]:::system
  E --> F["Standing, review, reversal,<br/>jurisdiction that can shrink"]:::repair