Neuroscience and the Libet Experiments

In the 1980s, Benjamin Libet and colleagues performed experiments that became central to popular discussions of free will.

The basic result sounded startling:

the brain appeared to begin preparing an action before the person reported becoming consciously aware of the decision.

Did neuroscience show that conscious will is an illusion?

Not so fast.

The Basic Experiment

Participants were asked to perform a simple movement, such as:

flexing a wrist

whenever they felt like doing so.

At the same time, researchers measured:

• brain activity,
• muscle activation,
• reported time of conscious intention.

Readiness Potential

Before voluntary movement, EEG often shows a slow buildup of electrical activity over motor-related regions.

This is called the readiness potential, or RP.

It can begin hundreds of milliseconds before movement.

The Famous Timeline

In Libet’s experiments, the average readiness potential began roughly:

several hundred milliseconds before movement.

Participants reported becoming consciously aware of the urge or intention later.

This created a temporal ordering like:

brain preparation → reported conscious intention → movement.

The Popular Interpretation

A common conclusion was:

“The brain decides before you do.”

This phrase is rhetorically powerful.

It is also conceptually sloppy.

Your brain is not obviously something separate from you.

Brain vs Self

If the self is realized by the brain, then:

brain activity preceding conscious awareness

does not imply:

someone else made the decision.

It may show only that decision formation begins unconsciously.

Conscious Awareness Is Late

Much cognition occurs before conscious access.

We already know this from:

• perception,
• language,
• memory.

There is nothing surprising in principle about unconscious stages of decision formation.

Timing Conscious Intention

Libet needed participants to report when they first became aware of wanting to move.

This was often called W-time.

But subjective timing is difficult to measure.

The Clock Method

Participants watched a rapidly moving clock marker.

After moving, they reported where the marker was when they first felt the intention to move.

This requires:

• attention,
• memory,
• temporal judgment.

Each introduces uncertainty.

Introspection Is Not a Stopwatch

A person may not have precise access to the instant when an intention begins.

Intentions can emerge gradually.

Asking for one exact timestamp may oversimplify the process.

What Is an Intention?

Does intention begin when:

• motor preparation starts,
• a desire becomes conscious,
• commitment forms,
• action becomes inevitable?

Different definitions produce different timelines.

Simple Finger Movements

Libet tasks are intentionally minimal.

Participants choose:

when to flex.

They usually do not choose:

whether to save a life, which career to pursue, which argument to accept.

The ecological gap is large.

Arbitrary vs Deliberate Decisions

An arbitrary movement has little reason structure.

A meaningful decision may involve:

• values,
• evidence,
• long-term goals.

Neural dynamics may differ substantially.

Readiness Potential Is an Average

The classic readiness potential is often obtained by averaging many trials backward from movement onset.

This matters.

A smooth buildup in the average does not necessarily mean every individual trial contains the same deterministic ramp.

Schurger’s Accumulator Model

Aaron Schurger and colleagues proposed that the readiness potential may partly reflect stochastic neural fluctuations accumulating toward a movement threshold.

On this view, the RP need not represent a fully formed unconscious decision.

Threshold Crossing

Imagine ongoing neural activity fluctuating.

When activity reaches a threshold:

movement occurs.

Averaging many such trials backward from movement naturally produces a ramp-like signal.

This can resemble the readiness potential.

Spontaneous Action

Libet’s task asks people to act without an external reason.

In such a setting, spontaneous fluctuation may play a large role.

That may be very different from reason-guided deliberation.

Prediction Studies

Later experiments tried to predict choices from brain activity before participants reported deciding.

Some studies found weak above-chance prediction seconds before conscious awareness.

These results attracted enormous attention.

Above Chance Is Not Certainty

If a classifier predicts left vs right at:

60%

accuracy,

the result is statistically interesting.

It does not mean the decision was already fixed.

Prediction strength matters.

Soon et al.

Studies by Chun Siong Soon and colleagues used fMRI patterns to predict simple left/right choices several seconds before reported awareness.

The effect was above chance but far from perfect.

Interpretation remains contested.

Predictive Bias vs Completed Decision

Early neural activity might encode:

• bias,
• tendency,
• predisposition.

A bias is not necessarily a finalized decision.

The decision process can remain revisable.

Evidence Accumulation

Many decisions unfold through gradual accumulation of evidence.

Neural precursors can therefore appear before conscious commitment.

This does not imply one hidden instant where “the brain decides.”

Decision as Process

A better model may be:

distributed preparation → competition → threshold → conscious access → execution.

The exact ordering can vary by task.

No Single Decision Point

The popular narrative assumes:

first the brain decides, then consciousness learns about it.

But there may be no single discrete neural event corresponding to “the decision.”

Decision-making may be temporally extended.

Conscious Veto

Libet himself did not conclude that all free will was disproven.

He suggested a possible role for conscious veto:

an action may be initiated unconsciously but consciously stopped before execution.

He called this idea “free won’t.”

Free Won’t

The proposal is:

unconscious processes generate an action tendency.

Conscious control can inhibit it.

This preserves some role for conscious agency.

Is Veto Uncaused?

Critics immediately ask:

What causes the veto?

If the veto itself arises from prior neural processes, libertarian freedom is not restored.

The idea may support control, not metaphysical independence.

Inhibition

Neuroscience clearly shows humans can inhibit planned actions.

Stop-signal paradigms study this.

Response inhibition is a real cognitive capacity.

Whether it is “free will” is a philosophical question.

Stop-Signal Task

Participants prepare a response.

On some trials, a stop signal appears.

Successful inhibition requires rapid control.

This provides measurable evidence of inhibitory mechanisms.

Control Is Distributed

Action control involves:

• prefrontal regions,
• basal ganglia,
• motor systems.

No central inner commander is required.

Agency can be implemented by distributed control architecture.

Readiness Potential and Consciousness

The RP is not itself consciousness.

It is a neural signal associated with movement preparation.

Interpreting it as:

“the unconscious mind has made the decision”

adds theory beyond the raw measurement.

Correlation vs Commitment

A precursor correlated with future action does not prove the future action is inevitable.

A weather forecast can predict rain without causing or fixing it.

Predictive information and determination differ.

Reversibility

If participants can cancel or modify an action after early neural preparation begins, then the early signal cannot represent an irreversible decision.

This is an important empirical distinction.

Point of No Return

Researchers study when action becomes difficult or impossible to cancel.

The boundary appears closer to movement than the earliest RP onset.

This suggests extended preparation rather than early final commitment.

Deliberate Decisions

Experiments involving meaningful choices often show more complex neural dynamics.

Reasons and values shape the process.

Libet-style arbitrary timing tasks may not generalize cleanly.

Neuroscience Does Not Bypass Philosophy

Suppose every action has measurable neural precursors.

Would that refute compatibilist free will?

No.

Compatibilism expects free choices to have neural causes.

Libertarian Challenge

For libertarianism, deterministic neural precursors might be more threatening.

But current data do not establish that every human choice is fully predetermined long before awareness.

The evidence is more modest.

Hard Incompatibilism

Hard incompatibilists do not need Libet.

Their arguments concern:

• self-creation,
• causal history,
• luck.

Neuroscience may reinforce a naturalistic view of agency but is not foundational to the skeptical case.

Conscious Will as Constructed

Some researchers argue the conscious feeling of intention may be constructed retrospectively.

This is plausible in some cases.

But a constructed experience can still correspond to real control processes.

Wegner

Daniel Wegner argued that the experience of conscious will can be inferred from correlations among:

• thought,
• action.

His work showed that perceived agency can be manipulated.

Illusions of Agency

People can sometimes feel:

they caused an event

when they did not.

They can also fail to feel agency over actions they caused.

This shows the sense of agency is fallible.

Sense of Agency vs Actual Agency

We must distinguish:

feeling in control

from:

having causal control.

The two often align.

They can dissociate.

Alien Hand Syndrome

In alien hand syndrome, a limb may perform actions not experienced as intentionally controlled.

This shows action generation and sense of ownership can separate.

Utilization Behavior

Certain neurological conditions can produce stimulus-driven actions with reduced reflective control.

These cases help map components of agency.

Voluntary Action Is Layered

Voluntary action may involve:

• goal selection,
• motor preparation,
• inhibition,
• monitoring.

No single component equals free will.

Consciousness May Guide Higher-Level Control

Conscious awareness may be especially important for:

• novel situations,
• conflict,
• long-term planning.

It may not initiate every muscle movement.

That is consistent with a naturalistic account of agency.

Timing Does Not Equal Causation

Even if event A occurs before conscious awareness B, it does not follow that B has no causal role later.

Earlier processing can initiate.

Later processing can modify.

Causal chains are extended.

The Wrong Question

Instead of asking:

“Did consciousness start the action first?”

a better question is:

“What role does conscious processing play in the full decision-control system?”

This is more scientifically useful.

Agency Without a Magic Moment

Free will need not be located in one millisecond.

Agency may be the capacity of the whole person to:

• deliberate,
• inhibit,
• revise,
• learn.

That capacity unfolds over time.

The Philosophical Lesson

Libet’s experiments show that neural preparation can precede reported conscious intention in simple voluntary movements.

They do not show that:

• every decision is completed unconsciously,
• conscious control is causally irrelevant,
• free will has been scientifically disproven.

The experiments are important precisely because they force us to replace the myth of a single inner decision moment with a distributed process.

The Next Question

If a decision is a process rather than an instant, then we need to understand the components more carefully.

What is the difference between:

• choice,
• intention,
• agency?

That is the next essay:

Choice, Intention, and Agency.