A difference of one part in ten thousand, and a different weather.
Determinism
That the state of the world at any moment, together with the laws, fixes every later state.
- Theme
- Cause, time and change
- Tradition
- Western
- Term fixed
- 1677
In plain terms
Determinism is a claim about the world, not about knowledge. It says the present state plus the laws fixes the future - whether or not anyone could work it out.
That distinction is constantly lost, and losing it produces bad arguments in both directions. A system can be completely determined and utterly unpredictable, because tiny differences in the starting state grow into large differences in the outcome, and no measurement is exact.
So the discovery that weather cannot be forecast beyond a couple of weeks is not evidence against determinism. It is evidence about measurement.
And conversely, if the world turns out to be indeterministic at the smallest scale, that does not deliver free will. An undetermined choice is a random one, which is not what anyone wanted.
As a philosopher would put it
Determinism is standardly stated as the thesis that a complete description of the world at a time, together with the laws of nature, entails a unique description at every other time. Laplace's demon is the epistemic image of it, and the two are often conflated.
Its relation to free will divides the field. Hard determinism accepts determinism and denies free will; libertarianism denies determinism to save it; compatibilism denies that the conflict arises. The consequence argument, sharpened by van Inwagen, is the strongest incompatibilist case: if our acts follow from the laws and the remote past, and we control neither, we do not control our acts.
Quantum mechanics complicates rather than settles it. On standard interpretations the theory is indeterministic; on Bohmian mechanics it is deterministic and nonlocal; on Everettian readings the wavefunction evolves deterministically and branching produces the appearance of chance. So whether the world is deterministic is currently an open question about interpretation rather than about data.
Where you meet it
In the winter of 1961 Edward Lorenz was rerunning a weather simulation on a small computer and re-entered a value from a printout as 0.506 instead of the stored 0.506127.
The run diverged completely from the original within a simulated couple of months. The model was entirely deterministic - the same equations, the same machine, no randomness anywhere - and a difference of one part in ten thousand produced a different weather.
His 1963 paper set out sensitive dependence on initial conditions, and the talk title asking whether a butterfly's wings in Brazil set off a tornado in Texas supplied the image.
The result established a practical limit: atmospheric forecasts lose skill after roughly ten days to a fortnight regardless of how good the model or the observations become. That limit is not a gap in determinism. It is what determinism looks like from inside a system you can only measure approximately, which is the only position anyone occupies.
What is actually disputed
Whether determinism threatens anything worth wanting.
Incompatibilists press the consequence argument: if every act follows from a past and laws nobody controls, then no act is up to anyone, and calling that freedom is a redefinition. Compatibilists reply that the freedom at stake was never freedom from causation but freedom from compulsion, and that an uncaused act would be random rather than free.
The majority of philosophers are compatibilists, and the disagreement has not narrowed.
A separate question is whether determinism is even true. The physics does not settle it, since the leading interpretations of quantum mechanics differ on exactly this point and are empirically equivalent. That is an unusual situation: a thesis with enormous consequences for responsibility, punishment and self-understanding, whose truth turns on a choice between interpretations that no experiment currently distinguishes.
Commonly confused with
- Fate c. 250 BC Fate says an outcome will occur however you act. Determinism says your actions are part of the causal chain producing it, which is a different claim.
- lack of free will Whether it follows is the whole dispute, and most philosophers hold that it does not.
Defined against
- Chance If any event is genuinely undetermined, determinism is false, which makes the pair exhaustive and the physics decisive in principle
Philosophers
- Baruch Spinoza Necessitarianism, in which nothing could have been otherwise
- David Hume Determinism reconciled with liberty as absence of constraint
- Immanuel Kant Determinism in the phenomenal world with freedom in the noumenal
- Gottfried Wilhelm Leibniz Every predicate contained in the subject, and the appearance of contingency
Works
- Ethics The necessitarian system
- A Treatise of Human Nature The compatibilist reconciliation
- Critique of Pure Reason The two-standpoint resolution
- Monadology Complete concepts and their consequences
- On the Nature of Things The atomist swerve, introduced to escape it
Schools
- Spinozism Where determinism is total
- British Empiricism Where it is reconciled with liberty
- German Idealism Where freedom is relocated
- Atomism Where the swerve is introduced to break it
- Vienna Circle Where the question is recast as one about laws
Related concepts
Causation c. 330 BC · Fate c. 250 BC · Chance c. 330 BC · Free will c. 397 · Necessity c. 330 BC
Spinoza's necessitarianism is the strongest version - nothing could have been otherwise at all; Laplace's 1814 image of an intellect that could compute the whole future from a complete present state is the formulation everyone cites.