​Chapter 7. Deviation

The man dropped to all fours, and the act has causes, and a caused event is not free. You have not escaped the causal order. You have swapped the cause "norm" for the cause "revolt against norm," and it is as bound as the conformity it rebels against. Breaking the pattern is just more behavior that runs on causes like everything else.

The answer lies in when the assignment is made. Out of a thing's residue, what is derivable in advance is the normative behavior — the pattern doing what patterns do. Give me the constitution of the timid man and I can tell you, before the occasion arises, that he will flinch. That is what a nature is for: it makes the future readable, it lets the conditioned be inferred from its conditions. This is the whole power of residue and the basis of the determinist's confidence.

This is the working faith of the entire project of knowledge. Laplace drew the limit case: an intellect given the full state of the world at one instant, and the laws it runs by, would have the entire past and future open before it. Nothing would be left to find out, only to compute. Every law of nature is a smaller version of the same promise — that the behavior of a kind of thing follows from what the kind is, and so can be read off in advance. The promise has paid out so reliably, across so much of the world, that it has hardened into the assumption that it must pay out everywhere: that whatever has not yet been predicted is just not yet computed, and that prediction has no outer edge.

Kant took the determinist's confidence at full value and tried to keep freedom anyway. The empirical character — the whole of what a person does, seen from inside the causal order — is determined, every act derivable from the prior state. Freedom lives in the intelligible character, the same person seen from outside the series, where the will is its own origin and time does not reach it. Two orders, one world: the act is caused when read from below and free when read from above.

The structure is the nearest anyone in the Western tradition came to what we hold. A will caused in its origin and acausal in its operation, an act that sits inside the causal net and exceeds it — Kant saw this and built a house for it. The house is the difficulty. The intelligible character is a noumenal self, a thing-in-itself behind the appearances, carrying its freedom the way a substance carries its attributes. Svabhava, relocated behind a transcendental partition: unreachable by experience, unknowable in its content, and still a nature — the nature of the subject as it is apart from every meeting. Kant kept freedom by giving it an owner no one can inspect; the old self-standing core, set past the reach of critique.

There is no second floor. The act occurs through a carrier, in a meeting, leaving a residue, and exceeds its conditions there, where the encounter can read it. No hidden subject owns it from behind a veil. No nature, open or walled off, exhausts it.

The deviation is what does not follow from the residue. The timid man who, this once, does not flinch — nothing in the prior pattern yielded that. Had it followed from his constitution, it would not have been a deviation; it would have been the nature expressing itself, more residue. The very thing that makes it a deviation is that it could not be read off in advance. The norm is forecastable because it is inertia. The break is not, for it is the interruption of inertia, and you cannot derive an interruption from the thing it interrupts.

So watch what the determinist actually does. He waits. The man drops to all fours, and only then does the determinist produce the cause — the urge, the motive to seem free. The explanation always arrives after the act, never derived before it. And this retrospective fit is the structure we have a name for. The cause assigned after the event is residue settling onto the act. Before the act, that cause was not in the nature waiting to produce it; if it had been, the act would have been forecastable. The cause is a trace left by the deed, read backward as though it had been a source. The determinist sees the deposit the act left behind and calls it the spring the act came from.

What was called the nature before the act is quietly replaced by another nature after it. A moment ago he was the ordinary man, the reasoning creature whose track was known. Once the deviation occurs, the ledger is rewritten: he was always the sort of man carrying a motive to revolt, always secretly disposed to the very act that no one had derived. This is classification after the fact, residue laid over the break and then read backward as though it had been waiting there all along.

Psychology is the modern embarrassment of this problem. No discipline has worked harder to turn human response into law, and few have made the difficulty more visible. Its replication crisis is not an accidental scandal beside the point; it is the empirical shadow of the same issue. A stone repeats cleanly enough to become physics. A person repeats often enough to invite psychology, and breaks often enough to keep the law unstable. The failure marks the kind of object in which retrospective explanation constantly outruns prospective derivation.

This is the same shape we stated about freedom from the start, now turned into a test. An acausal act can be described after it is done and cannot be derived before. That sentence sounded like a definition; here it does work. The norm fails the decisive half: it is derivable before and described after, fully residue. The deviation is describable in retrospect, underivable in prospect. The asymmetry between forward and backward is the signature of an act that exceeds its conditions, the mark we were owed — the thing visible in behavior by which a break in residue can be told from the surrounding residue.

The determinist has one move left. He says: underivable-in-advance is a fact about your knowledge, not about the world. The cause was there all along; you simply could not compute it. Perhaps a perfect predictor, with the full state of the brain, would have derived the drop to all fours. Asymmetry is epistemic, a limit on the forecaster, not a gap in the causal fabric.

There is no proof against this, and we do not pretend to have one. Proof runs inside the causal frame, deriving the conditioned from its conditions; to prove the act exceeds the frame, one would have to derive it, which is to lose at the outset. The determinist's claim and ours meet right here, and neither can be established by argument, because the dispute is over whether the frame is total, and every proof presupposes the frame it was meant to adjudicate. He assumes the gap is only in our knowledge. We assume the gap is in the world. Both are assumptions; neither is a fact. The choice between them cannot be derived; it is itself an act made and not proven, the very thing one side claims and the other denies. To settle it would require deriving what was defined as underivable, so it stays where it began, in the enacting. The argument ends where the discussion with Nagarjuna ended: at the point where pressing further would itself be a deviation, described afterward, underivable before. We let the objection stand as the illustration it has become, and stop again.