"I Decided" Without Free Will

I have argued elsewhere that free will does not survive examination — that every version of it collapses into either ordinary causation or randomness, and neither is what anyone wanted.

That leaves a puzzle I did not address there. If there is no free will, what are we reporting when we say "I decided"?

The usual answer is that we are simply wrong — the sentence is a relic of a false theory, like saying the sun rises. I think that is too quick, and that ordinary usage is tracking something real.

The Distinction We Actually Draw

In an interview with Lex Fridman, Annaka Harris made the standard point: we tend to think I decided that, when in fact it was brain processing. True enough.

But look at when people actually use the phrase, and a pattern appears.

Suppose someone holds your parents at gunpoint and orders you to jump. You jump. You do not say you decided to jump. You say you were forced. Nobody finds this confusing, and nobody thinks you are making a metaphysical claim about causal chains.

Now suppose you jump into a lake on a hot day. You say you decided to.

Both events are equally caused. Neither involved anything uncaused. Yet the distinction people draw between them is immediate, universal, and — I want to argue — accurate.

What the Distinction Tracks

We say "I decided" when our own internal monitoring reports that we were not compelled by an external agent.

That is the content of the claim. Not "an uncaused event occurred in my skull," but "this action arose from my own processing rather than from someone else's coercion."

And that distinction is not a philosophical nicety. Any organism operating among other organisms needs to detect whether it is being manipulated or acting on its own assessment. A creature that could not tell the difference would be exploited immediately and permanently. The capacity is functionally necessary, and it long predates anyone's theory of the will.

So the vocabulary of decision is reporting the output of a real detection system. The system was never measuring metaphysical freedom, and the fact that it does not detect metaphysical freedom is no criticism of it.

What This Does and Does Not Rescue

It does not rescue libertarian free will. There is no capacity to have done otherwise, no uncaused choosing, nothing exempt from the causal order. That remains gone.

What it rescues is the ordinary distinction, and quite a lot rests on that distinction. Coercion is different from choice in a way that matters for law, for ethics, and for how we treat one another. A confession extracted under threat is worth less than one given freely — not because one involved a break in the causal chain, but because one was produced by an external agent's interest rather than by the speaker's own assessment.

That difference is real, detectable, and morally relevant, and it survives the loss of free will entirely intact.

The Middle Cases

And of course the distinction admits of degrees, which is exactly what you would expect of a detection system rather than a metaphysical category.

Partial coercion. Manipulation you half-noticed. Social pressure you could have resisted at some cost. An offer you could technically refuse. Addiction, where the compulsion is internal but does not feel like yours.

These are genuinely hard cases, and they are hard in the right way — they are hard because the underlying quantity is continuous, not because our concept is confused. A binary metaphysical notion of freedom cannot even represent them. The functional account handles them naturally: the question is how much of the outcome was determined by an external agent's design, and that admits a range.

So: no free will, and a perfectly good account of what "I decided" means. The sentence was never about metaphysics. It was about who was operating the machinery.