In 1983, Benjamin Libet asks subjects to flex their wrist whenever they wish, and to note the exact moment they become aware of their intention to move. Simultaneously, he records the electrical activity of the motor cortex. The Bereitschaftspotential, a motor readiness potential, appears on average 550 milliseconds before the movement. Awareness of the intention appears only 200 milliseconds before. The gap is 350 milliseconds. The brain has begun before the subject decides.
The offset has been confirmed by fMRI (Soon et al., 2008) and by single-neuron recording (Fried, Mukamel & Kreiman, 2011). The scales vary, from 1.5 seconds to ten seconds depending on the method, but the direction is constant. Predictive activity precedes awareness.
First reading: free will is an illusion. The conscious decision is an epiphenomenon, a narrative the brain constructs after the motor process is already underway. Wegner (2002) formalizes this position: the experience of will is a retrospective causal attribution.
Second reading: Schurger, Sitt & Dehaene (2012). The Bereitschaftspotential is not a decision signal. It is stochastic neural noise crossing a threshold. There is no hidden decision before the conscious decision. There is an accumulation of noise that resolves into action.
Third reading: the protocol has two boundary conditions. The initial state of the brain, and the movement actually performed. The offset between the readiness potential and the awareness of intention could be an artifact of projecting a single-edged description onto a process constrained by both.
Doctrine
The same dataset. Three frameworks. Three incompatible conclusions. The data does not settle it. The framework settles it.
Vecteur ouvert
350 milliseconds. The brain prepares, the subject does not yet know. The usual question: who decides? Another: what if the offset were not between a cause and its effect, but between two descriptions of the same event?
