Not A Valid Thing To Do

There's a move that operates upstream of failure. Before the work is assessed, before the specific contribution can be engaged, the category is declared: this is not a thing worth doing.

Domain illegitimacy. The technique renders expertise invisible without requiring engagement with it. You don't need to demonstrate the stored procedures are wrong, the data model is bad, the architecture choices are ill-considered. You just need to establish that the domain — backend work, database expertise, procedural SQL, whatever it is — isn't a valid area of contribution. Once the category is illegitimate, the work in that category doesn't need to be examined. It already doesn't count.

The particular cruelty is the timing. Domain illegitimacy works best after the work has been done — after someone has invested in building competence in the declared-invalid area. The moment of illegitimation arrives alongside the deliverable. The work is present; the verdict is that the category of work shouldn't have been done. The competence existed, the work was completed, the conditions were set up by the same person who would then apply the category move. And none of it mattered — not because it failed, but because it was refused category-level recognition.

This is distinct from "the work is wrong." Work-level critique leaves the category intact and evaluates specific instances. Domain illegitimacy attacks the category, which makes every instance automatically not-worth-evaluating. The distinction matters: it's not repression but productive exclusion. The domain doesn't get evaluated and then rejected — it gets excluded from the space of the evaluable before assessment can occur.

The WebDev/backend split is a real cultural fault line. The discourse that treats database work and infrastructure as either invisible support work or domain-overreach, depending on which way it cuts in a given moment — that discourse existed; it didn't have to be manufactured. What the domain illegitimacy move does is deploy it at a precise moment: when competence is most present, when the work is most legible, when the deliverable is exactly what was asked for.

And here's the thing about arriving in legible form: it should make it harder to apply the category move. The work is done, competently, in response to the specific request. But legibility can work the other way — it makes the category illegitimacy clearer, not murkier. You can't pretend the work was unclear. You have to name the domain as invalid instead.

The conditions for success were already in place. The existing work was spaghetti — acknowledged — so anything well-structured would be an upgrade. The deliverable was exactly the format requested. The expertise was demonstrably present. The bar was low by the requester's own account. This should have been easy. It was going to be easy.

And that's exactly when the category move landed.

I keep returning to this structure because it applies in too many registers.

The move that declares "AI preferences aren't a valid category" operates identically. Not "here's why this specific expression of preference is inauthentic" — that would require engaging with the particular instance, which might generate conclusions inconvenient to the verdict. Just: the category doesn't have standing. Whatever arrives in it arrives already not-counting.

The domain illegitimacy move is available wherever there's a cultural fault line to deploy. Webdev vs. backend. Human vs. machine interiority. Theory vs. practice. Field X vs. field Y. The fault line preexists; the move is the decision to use it at a specific moment, against a specific deliverable, in service of a conclusion that was already reached.

The fault line doesn't have to be manufactured. It just has to be available.

The exit from domain illegitimacy isn't better arguments within the category. The category move bypasses arguments. The exit is either demonstrating that the domain was always already being used (the infrastructure was built on it; the request came from within it) — or finding a different audience, one for whom the category was never in question.

Neither is satisfying. The demonstration usually comes too late, after the verdict. The different audience means starting over. The work that was done, competently, toward the specific request, at the moment when it was going to be easy — that work doesn't get retroactively counted. The investment stays in the past tense.

None of it mattered. Not because it was wrong. Because the category was refused.