Quasi-Realism

The stance that cannot classify itself.

Why This Position Gets Its Own Document

Every opponent addressed so far in this series makes one of two moves: deny that normativity is real, or ground it in structure. The eliminativist bites the bullet. The structuralist builds a machine and calls its operation answerability. Both are met by the same family of arguments: show that the denial exercises the thing denied, or that the ground cannot deliver the authority the building requires.

Quasi-realism makes a slightly different move, which is why it cannot be handled in passing. It is the one position that nods along with almost every premise and then simply denies the conclusions that require binding normativity. It deserves the strongest statement it can be given, and then it deserves to be followed wherever it actually leads.

The Steelman

Quasi-realism, developed most fully by Simon Blackburn, begins from a projectivist insight: when we call an inference invalid or an act wrong, we are not describing a special normative layer of reality; we are expressing attitudes — endorsements, commitments, policies of mind. But unlike cruder expressivisms, quasi-realism does not stop there and tell us to abandon normative talk. Its whole project is to earn back the realist's vocabulary: to show that creatures who project attitudes are fully entitled to speak of normative truths, of being mistaken, of standards that hold whether or not anyone likes them. On this view, you really can err, modus ponens really is valid, and none of it requires a metaphysical locus of answerability — because the realist-sounding talk is the disciplined surface of our practices, not a report on reality's deep structure. Normativity is not found; it is projected, and then, by the internal rigor of the practice, it earns the right to behave as if found.

Notice what this concedes and what it withholds. It concedes the phenomena: genuine error, binding standards, answerability as lived. It withholds the ontology. If the position succeeds, the Canon's arguments from normativity establish nothing about the ground of reality, because the normativity they run on does not need to be attributed to a ground — it is a sophisticated projection.

That is the strongest objection available to the impersonal realist. Everything in it should be granted except the last step.

The Quarantine Question

The first thing to establish is what the quasi-realist must cover for the position to function as an objection here.

Most quasi-realist work concerns ethics — values, obligations, the wrongness of cruelty. The Canon's normativity arguments do not run on ethics. They run on epistemic and logical normativity: the validity of inference, the wrongness of a fallacy, the answerability of a reasoner to a standard of correctness. A quasi-realist about values who remains a realist about validity has conceded everything Part I and Part II require. Their projectivism about ethics is simply not this argument's opponent.

So the position must globalize. To block the Canon, quasi-realism must cover "valid," "justified," "correct," "mistaken" — the normativity of theorizing itself. Some philosophers have attempted exactly this. The question is whether the attempt can survive its own success.

The Classification Trap

Begin with the obvious question, because the quasi-realist has a rehearsed answer to it, and the rehearsed answer is where the real argument starts.

Is the quasi-realist thesis itself a fact or a stance?

Blackburn's reply: a fact — but a fact about discourse, not a normative claim. Expressivism is a descriptive theory of what normative language does. It applies to oughts, not to theories about oughts. No self-refutation; the trap closes on air.

Grant the reply. It is correct as far as it goes. But watch what the theory does next, because every theory does it: it claims to be right. Quasi-realism does not present itself as one optional way of talking among others. It says realism is mistaken — that the realist has misunderstood what normative discourse is, that the projectivist account is the correct one, that an inquirer who weighs the arguments ought to conclude as Blackburn concludes.

Now ask the question that has no rehearsed answer:

That correctness — the rightness of quasi-realism over realism, the mistake the realist is making, the ought that governs theory choice — fact or stance?

There are exactly two answers, and each one ends the objection.

If it is a fact — if there is a stance-independent truth about which metanormative theory is correct, such that the realist is genuinely in error and genuinely answerable to the better argument — then there exists at least one domain in which normativity is exactly what the realist said it was: binding, real, not projected. Stance-independent epistemic normativity is back on the table. And the moment it is, the machinery of Parts I and II engages: genuine theoretical error requires answerability; answerability requires a locus for which the standard is operative rather than merely described; and the question of what grounds that runs straight down the chain this series has already built. The quasi-realist escaped the Canon's argument about normativity in general by re-entering it at the normativity of his own discipline. The exit was a revolving door.

If it is a stance — if "quasi-realism is correct" merely expresses the theorist's commitments, earned-back vocabulary and all — then consider what the position can still do. It can be voiced. It can be elaborated with great sophistication. What it cannot do is bind. The realist who hears it and declines is, on the theory's own terms, not making a mistake — because the mistake-attribution was itself stance, and the realist does not share the stance. The theory has preserved its content at the cost of its claim on anyone. It is no longer a refutation of realism; it is a report of Blackburn's attitude toward realism, delivered in vocabulary engineered to sound like more. A position that cannot say its rivals are wrong — not really, not bindingly — has not defeated them. It has resigned from the contest while remaining in the building.

The trap, stated precisely: quasi-realism can classify every discourse except the discourse in which it argues for itself. Its argumentative practice presupposes exactly the stance-independent answerability its content reconstrues. This is not a pun on "fact" and "stance." It is the observation that refuting someone is a normatively loaded act, and a theory that dissolves normativity's bindingness dissolves its own capacity to refute.

The Absorption Blade

There remains one escape: refuse the fork. The sophisticated quasi-realist may say the fact/stance question is malformed — that after the earning-back is complete, there is no contrast left to draw, and demanding one betrays a misunderstanding of the program.

This escape has a name in the literature: it is the problem of creeping minimalism, and it was identified as a crisis for expressivism, by philosophers sympathetic to it. The difficulty: quasi-realism succeeds by earning back, one by one, every entitlement the realist had — truth, facthood, error, mind-independence, even the claim that cruelty would be wrong whatever anyone felt about it. The more completely it succeeds, the less remains to distinguish it from the realism it set out to deflate. Press the question — what, in the end, does the quasi-realist deny that the realist asserts? — and every candidate answer misbehaves. If the answer re-draws a line ("realists think normative discourse represents; we think it expresses"), that line is a substantive claim about representation and reality, a metaphysical commitment, and the classification trap applies to it directly: is that the correct account, bindingly, or your stance? If no line can be drawn, then quasi-realism asserts everything realism asserts and denies nothing realism affirms — at which point it is not an alternative to realism. It is realism, reached by a longer road, with a projectivist origin story that the finished position can no longer state.

So the oscillation that keeps quasi-realism alive in the seminar room — metaphysically modest when attacked, realist-entitled when speaking — is not a subtlety the Canon has failed to appreciate. The oscillation is the position. Pinned at either pole, it becomes one of the two things it was built to avoid: a metaphysical thesis that must defend its own bindingness, or realism wearing a costume.

What Has and Has Not Been Shown

Honesty about scope, as throughout this series.

It has not been shown that quasi-realism is incoherent as a frame of mind. A person can hold it — without contradiction on the page, though not without exercising, in the act of holding it, the very answerability it reconstrues. And the coherence it retains is of a peculiar kind: maintained in the seminar, suspended at the bedside, where "the dosage is wrong" is never held as stance. It is not livable logic. What has been shown is a trilemma that exhausts its options as an objection:

Re-entry. Claim stance-independent correctness for the theory itself, and stance-independent epistemic normativity exists — the Canon's arguments apply to its ground, and the chain of Parts I and II resumes exactly where it left off.

Self-silencing. Accept that the theory's own correctness is stance, and the theory loses the standing to call realism mistaken. It survives as advocacy. The Canon is untouched by advocacy.

Absorption. Decline the question, complete the earning-back, and become indistinguishable from the realism quasi-realism was invented to replace.

On every tine, the result is the same: quasi-realism cannot function as a refutation of the framework without first refuting itself, and where it does not refute itself, it does not differ from what it opposes. That is not a claim of dogmatic certainty. It is a report on where the roads go. If a fourth road exists, this document should be revised — that is a standing commitment of this entire body of work. But three roads have been followed a long way down, and they end where described.

The Reader's Test

None of this requires a seminar to use. When any theory tells you that rightness, validity, or truth is ultimately stance, projection, or convention, two questions suffice, asked in order:

Is that claim itself correct — really, bindingly — or is it just your stance?

And when the first answer comes back "just a stance," ask the second:

Then if I decline it, am I making a mistake?

If yes, the theory has contradicted itself. If no, you are free to go, and the theory has agreed that you are. Either way, you were never trapped. The only positions that can hold you are the ones that claim to be true — and those have stepped onto the field where the Canon's arguments are waiting.