Libet & the Readiness Potential

Basic bibliographic details

• Full title: "Time of conscious intention to act in relation to onset of cerebral activity (readiness-potential). The unconscious initiation of a freely voluntary act."
• Authors: Benjamin Libet, Curtis A. Gleason, Elwood W. Wright, Dennis K. Pearl.
• Journal: Brain, Volume 106, Issue 3, September 1983, pages 623–642.
• DOI: 10.1093/brain/106.3.623.

Core experimental setup

Libet et al. recorded scalp EEG to measure the readiness potential (RP), a slow negative potential that begins up to about one second before self-paced voluntary movements. Subjects made spontaneous, self-initiated finger or wrist movements while observing a rapidly rotating clock and later reported the angular position at the moment they first consciously "wanted" or intended to move — what Libet called W-time.

The design isolates three time markers:

• RP onset — when the brain's preparatory negative shift begins (measured directly from EEG).
• W — when the subject becomes consciously aware of wanting to move (reported via the clock).
• M — when the actual movement occurs (measured by EMG).

A separate control used a cutaneous stimulus (S) to calibrate the subject's ability to time perceived events on the clock with accuracy.

Key temporal findings

In type II RP series — spontaneous, non-preplanned acts — the onset of the main negative shift in the readiness potential preceded the mean reported W-time by about 350 ms on average, with a minimum of roughly 150 ms. In type I RP series — with some preplanning — RP onset preceded W by about 800 ms on average, or about 500 ms when RP onset was defined more conservatively (90% area criterion).

The brain's preparation for a "freely voluntary" act began, on average, 350 milliseconds before the subject was aware of the intention to act.

Subjects consistently distinguished the time of wanting to move (W) from the time of actual movement (M) and from the time of perceiving a cutaneous stimulus (S), with W occurring earlier than M and S in their reports. Libet interprets these results as showing that cerebral initiation of a spontaneous voluntary act can begin unconsciously, before any recallable conscious awareness of deciding to act.

Conceptual upshot for free will and consciousness

Libet concludes that preparatory neural activity for a "freely voluntary" act begins prior to conscious intention. This places constraints on the idea that consciousness initiates voluntary acts ex nihilo: at the moment one becomes aware of "deciding," the brain has already been preparing the action for several hundred milliseconds. Subsequent decades of work have used this result to argue, variously, that:

• Free will is illusion (Wegner, Harris, Sapolsky — the conscious sense of authorship is post-hoc rationalization of an already-determined neural process).
• Free will is veto, not initiation (Libet's own later proposal — consciousness cannot start an action but can stop one already begun, the so-called "free won't").
• The experimental design measures something other than what the debate needs (Schurger and colleagues, 2012 — the RP may be a stochastic accumulation that crosses threshold rather than a deterministic command; the W-report itself is unreliable to the very millisecond precision the argument requires).
• The compression gap is the architecture of choice (the 40-bit filter argument: the brain processes ~11 million bits per second but conscious awareness operates at ~40 bits; the 300-millisecond gap is the necessary lag between distributed neural preparation and integrated conscious recognition — not evidence against agency but the condition of its existence).

The honest reading thirty years on: the experiment shows something real about the temporal structure of voluntary action. What it shows is not a settled deletion of free will. The conscious "now" arrives some hundreds of milliseconds after the neural preparation that supports it, and the question of whether that preparation is deterministic or indeterministic, and of who or what the preparation is rendering for, remains open.

Methodological cautions

Subsequent work has flagged several reasons not to over-read the 1983 result:

• W-time accuracy. Subjects are asked to report the moment they "first wanted" to move by reading a clock that moves at roughly 25 ms per position. The reported W-time has its own confidence interval, plausibly larger than the RP-to-W gap.
• What the RP actually is. Schurger et al. (2012) showed that the RP can be modeled as the average of stochastic neural fluctuations that happen to have crossed a movement threshold. On this reading, the RP is not a "command" the brain issues before conscious intention — it is the recorded trace of noise that incidentally happened to be on the upslope when the action occurred. Read the Schurger explainer for the full leaky-accumulator model →
• The "free" act is paradoxical. The task instructs subjects to act spontaneously, without planning, at a random moment. This is a strange definition of "free will" — an unconstrained twitch, not a considered choice. Whether the RP for an arbitrary finger flick generalizes to genuine deliberation is itself an open question.

None of this dismisses the finding. It clarifies what the finding can and cannot bear.

Why this matters for the trilogy

The readiness-potential result is one of the empirical anchors of the modern free-will debate, and the trilogy engages with it explicitly. In Numen, Alex and Alma's conversation about Sapolsky and the 300-millisecond gap is built directly from this literature; in Limen, the same gap is reframed structurally.

The receiver model offers a striking re-reading. The standard productionist interpretation is: the brain decides, then consciousness arrives late and takes credit. The receiver model interpretation is: the local neural preparation is what the field is rendering through, and consciousness — the experience of being the one who decides — arrives when the rendering integrates into a coherent "now". The 300-millisecond gap is the latency of the render, not the proof that there is no one watching.

This is more than a verbal trick. Lucía Reyes's cymatic work in Numen shows the same 300-millisecond gap on the other side: in her experiments, geometric coherence patterns begin forming in water about 300 milliseconds before the chord is played. The field arrives before its cause. If the same gap appears at both ends of the chain — before the action in the brain, and before the cause in the cymatic medium — then the gap is not a deficiency of consciousness. It is a structural feature of how the field couples to local biological substrates across time.

Senna Park's compression thesis in Anima makes the same point in a different register: the 40-bit filter is not a limitation. It is the architecture that produces the conditions for genuine choice under uncertainty. Determinism is the view from omniscience — the view that has all 11 million bits. Freedom is the view from exactly the right amount of not-knowing. The Libet gap is the empirical signature of that architecture at work.

The reader who takes Libet as a final argument against free will is reading the data through one ontology. The reader who takes Libet as the visible edge of how the field renders local agency through biological tissue is reading it through another. Neither interpretation is forced by the data. The trilogy adopts the second because it absorbs the result without dismissing the felt fact of choice — and because it predicts the symmetrical pre-event gaps the cymatic literature now also reports.

For Libet's full paper, see the PubMed entry above and the canonical text in Brain 106(3). For the methodological critique, see Schurger et al., PNAS 109(42), 2012. For the philosophical framing, see the Chalmers explainer; for the productionist program Libet's result is usually deployed in service of, see the Crick & Koch explainer; for the synthesis that weaves all of these together, see What the Evidence Shows So Far.