Bluntly put, it is absurd to imply, as you report Björn Brembs doing, that chaos is a "rudimentary sort of free will" (19 May, p 16). This chaos is in fact a sort of determinism, which is antithetical to the strong sort of free will and irrelevant to the weak one.

The strong sort is known as origination: the assertion that decisions can be originated by the free will of an agent such that those decisions have no causes.

Thus, as an "uncaused" source of causation, origination is the province of faith, and its existence is widely doubted by scientists and philosophers alike.

The weak sort, called voluntariness, merely asserts that an agent has free will if it can behave in a way that suits its own preferences.

Arguably the most familiar conception of free will, voluntariness reduces to subjectively experienced freedom of choice.

Consequently, it can only be ascribed to entities believed to possess sufficient subjectivity. While voluntariness is easily constrained, by imprisonment for example, few doubt its existence.

The mathematically defined chaos to which Brembs refers cannot provide evidence for free will as voluntariness, because it tells us nothing about internally experienced freedom, or any other aspect of subjectivity.

The verified presence of mathematical chaos actually excludes free will as origination, because by definition this chaos conforms to a predetermined pattern, while origination admits no predetermination.

Conflating origination with voluntariness is a common error in unsophisticated discussions of free will. Suggesting that deterministic chaos is evidence for any conception of free will, however, propagates far worse confusions.