Who are we really? Are we active agents making decisions and choices? Or are we passive observers of our actions with decisions made for us by autonomic or deterministic processes that allow us no real capacity for conscious decision?

Some contend that we are exactly those passive observers. Free will skeptics argue that forces far outside our control determine our actions and identities. The physical traits and capabilities that define us as humans do not originate with us as individuals, but with our ancestors long before we were born. They acquired those traits through natural selection in a process that played out over millennia, determining how our brains work and how we make decisions. The language that shapes how we think was invented 200,000 years ago and taught to us by prior generations. We are born into families, cultures, and civilizations—all of which define us before we take our first breath. Geography, climate, and environment give us certain opportunities, but deny us others. All these forces and events are themselves constrained by the movements of molecular and atomic and subatomic particles and fields that specify the range of possibilities available to every entity in the universe. In truth, there is much to support the view that we have little or no control over our actions, that free will and self-control are useful illusions.

The famous Libet “delay”

The proof often cited to support this view is the data generated by neuroscientists such as Benjamin Libet.[1] Libet conducted a series of famous experiments in which he took EEG readings of subjects asked to make random hand movements and record the time of their conscious decision to make each movement. The EEG readings showed a build-up of brain activity beginning before the subject was aware of the conscious decision to move. On average the readings showed electrical activity as much as half a second before the conscious decision, a build-up described as the “Readiness Potential”. Libet and others pointed to the onset of the Readiness Potential as the moment when the brain makes the decision to act, in this case by flicking a hand.

The experiments have been interpreted as showing a measurable delay between the beginning of the chosen action (assumed to be the onset of the Readiness Potential) and the conscious decision to take the action. In other words, the action seems initiated by a process other than the conscious decision itself. Apparently, our actions are determined by a physical process other than conscious decision-making. Conscious awareness seems to record the process after the fact, not initiate the process.

Free will skeptics cite these experiments as evidence that we do not have the agency we imagine we do. Our conscious decisions do not cause actions. We are instead passive observers of decisions driven by processes over which we have no control.[2]

Does the “delay” disprove free will? No.

The Libet findings and their progeny have been much discussed and debated over the decades since. The “delay” has become almost an accepted phenomenon in scientific and popular circles.

That broadly accepted view, however, is apparently wrong. The findings have been brought into doubt, even debunked,[3] by more recent neuroscientific research. They appear now to be artifacts of Libet’s analytical model rather than objective evidence of decision-making processes in the brain.[4] Contrary to the popular view, the “delay” data does not negate free will nor prove that conscious decision-making occurs after the fact.[5] New research interprets the data quite differently and offers a more robust explanation of the processes that govern decision-making in the brain.

The Libet experiments are not about meaningful decisions

Libet acknowledged that his findings do not apply to actions involving conscious deliberation,[6] the most common type of decision-making associated with free will or conscious awareness. His experiments focus on spontaneous decisions that are consequence-free for all practical purposes. Subjects are asked to flick a hand at a random time chosen by them on a whim without prior planning and without meaning attached to the movement. The experiments examine what occurs in the brain just prior to a movement made in the spur of the moment without deliberation.

Humans and other organisms make many kinds of decisions. There are decisions with grave consequences. There are decisions with almost no consequences, such as random choices about immaterial things. And there are decisions with many degrees of consequence in between. The choice to make a random, consequence-free hand movement is close to one extreme, almost an autonomic action relying on reflexive muscle movement more than thought or planning. The brain process for making such hand movements may be very different from the process for making decisions with consequences, which may require consideration over a period of time. For example, a decision to migrate from one region to another in search of food. Or a decision to take revenge on a murderous uncle, so elaborately over-thought that an entire dramatic production may be built around one lonely prince’s lengthy process of decision-making.[7]

Consequently, even if the “delay” were objective evidence of unconscious decision-making, it would be impossible to extrapolate from the data to a general model of decision-making for humans or other organisms.

The findings rely on faulty assumptions about how and when the brain decides to act

The brain is a living soup of electrical and chemical activity, with neurons firing constantly as they receive and respond to bits of information feeding into many parallel decision processes. The continual firing of neurons causes our brains to pulsate with electrical waves, which may help produce the state of watchfulness and preparation for action that the brain is designed to achieve.[8] Organisms must be ready to respond immediately to a threat or opportunity in the environment. Recurrent electrical waves with repeated peaks and troughs allow the organism to rely on the natural build-up of energy as a rapidly recurring launching point for action, helping the organism move faster in response to stimulus.[9]

Not every stimulus requires an immediate response, however, nor does every electrical wave result in action. Most electrical activity in the brain does not trigger awareness or response at all. The waves dissipate because they do not accumulate enough information or intensity to cross a threshold for action.[10] It is the waves that reach a certain point, perhaps in conjunction with waves or firings from multiple sources in the brain, that surpass a minimum threshold of attention and result in action. The brain decides to act, and the action starts to occur, when the threshold is crossed. Prior to that point, the wave is simply one of many recurring bursts of electrical activity that happen constantly without resulting in action.

The ”delay” is an artifact of the design of the experiment

Libet’s experiments by design focus only on electrical waves that precede spontaneous action. They measure the wave from onset to action, including the point prior to action when the subject becomes aware of an intent to act. The experimenters assume that the entire wave from trough to peak involves preparation for the spontaneous action and that the onset of the wave is when the brain decides to act. They overlook the possibility that what is identified as the Readiness Potential is not preparation for a specific action, but a general state of readiness which may or may not result in action. The action occurs in the experiments because they look only at waves that precede a spontaneous decision to act.

In fact, waves similar to the Readiness Potential can precede even actions that are not spontaneous, but prompted by the experimenter.[11] When subjects in Libet-style experiments are interrupted with a random click that cues them to initiate a hand movement immediately, faster responses tend to occur in conjunction with electrical waves similar to a Readiness Potential, even when the electrical wave began before the click cued the subject to act.[12] In other words, the movement by the subject, which could not have begun prior to the click, seemingly takes advantage of a pre-existing wave, as though hitching a ride to respond faster to the cue.

The same piggybacking may occur in the Libet experiments. A random hand movement may be just the sort of action to rely naturally on a recurring build-up of electrical activity in the brain. Asked to perform a voluntary movement with no particular reason to choose one time or another, the subject may rely on ongoing fluctuations in brain waves as a cue, in effect allowing the electrical waves to guide the choice of when to make the movement.[13] The result is that the movement happens near a natural peak of brain activity. It then appears to the experimenter as though the decision to move occurred at the trough of the wave rather than when the subject decided to act.

What ultimately causes action is whatever results in the wave increasing in intensity until it passes the threshold for action. That does not happen at the onset of the wave. It is more likely that the threshold is crossed when the conscious decision to take an action is made.[14] That may be exactly what pushes the wave over the threshold and results in action.[15] In other words, there may be no “delay” at all, because the time when the subject becomes aware of the intention to move may correspond with the time when the threshold for action is crossed.[16]

The Libet experiments make the classic mistake of building unchallenged assumptions into the structure of the analysis. The observed “delay” is likely a result of the assumptions underlying the experiment and an artifact of the data analysis, not something that occurs in the brain itself.

The experiments are premised on a dualist model of decision-making and consciousness

The concept of the “delay” is also founded on historical assumptions about the separation of “mind” and “body” that do not reflect biological processes of decision-making in the brain.

The experimenters assume a single point in time when decisions occur in the brain. They interpret the data as showing that the single point is not when we thought it was—at the time the subject becomes aware of an intention to act. Instead, they argue that the decision point is earlier—at the onset of the electrical wave which they identify as the Readiness Potential. The “delay” is then measured as the difference between the two points—the time between the unconscious decision and the later conscious awareness of the decision.

The assumption of a single decision point echoes the Cartesian notion of a central control room where “mind” resides and where decisions are made that control the “body”.[17] We imagine ourselves as having a center of consciousness in our brains where all decision-making occurs. The control room takes in data gathered by the senses, interprets the data, and makes decisions about how to respond. Libet-style experimenters accept this underlying notion that decisions occur instantaneously in the brain. They simply dispute that conscious awareness is where the decision occurs. Instead, they move the decision forward and calculate the “delay” between that presumed decision point and the point of conscious awareness.

Free will skeptics go further and argue that because awareness is “after the fact”, consciousness does not control decisions and free will is an illusion. They assume that a conscious and free human decision can be made only at an instantaneous time and place in something like a center of conscious awareness, the imaginary central control room.[18] If the decision or any part of the decision is made elsewhere, or made unconsciously or as an autonomic response of the body and brain, then the decision is not an act of free will. It is not controlled by “us” because “we” sit only in the central control room.

The straw man premise underlying the “delay”—and the free will arguments relying on it—is the magical central control room of mind-body dualism.[19] In fact, we know that humans and other organisms do not make decisions in that way.[20]

Decisions are processes

Decisions are iterative processes.[21] They do not happen instantaneously in one single place inside the brain. Some decision processes are fast, resulting in action within milliseconds. Some decision processes are slow, extending over a great many cycles of electrical activity.

Sensory information comes into different areas of the brain that process sight, sound, smell, touch, etc. Neurons fire in multiple areas. More information comes in. More neurons fire and signal other parts of the brain. Waves of electrical and chemical activity build and die out. Sometimes waves cross thresholds for action. More neurons fire.

All of this activity takes time. Time for information to flow around the brain. Time for information to be processed. Time for neurons to fire and communicate with other neurons. Time for thresholds of decision to be crossed and time for signals to travel to cells to trigger movement.[22]

Every one of these decision processes involves some autonomic or unconscious activity. Some processes are entirely unconscious. Some are partially autonomic or rely on a combination of autonomic and conscious processes. Some are deliberative and highly conscious. But even very deliberative decisions rely on biological and sensory processes that happen beneath the surface of our awareness.[23]

“We” are the entire process

Our cells and neurons, our tissues, our organs are what we are, but we have no ordinary conscious control over what they do.[24] The role of conscious awareness is not to manage autonomic or unconscious processes in the body, but to glean meaning from incoming information, to deliberate, to interpret external or internal events in ways that require conscious consideration.[25] When pre-programmed unconscious decision-making is insufficient to address a threat or an opportunity, that is when information comes to conscious awareness.

Lack of total control does not equate to zero freedom

Free will skeptics therefore are correct that conscious awareness does not drive all aspects of our decision processes. They are also correct that much of what we are as humans has been influenced or determined by forces and events far outside our individual control. We have neither total awareness nor total control.

But it does not follow logically that all our actions are autonomic and outside conscious control. We do not have zero control.[26]

We are imperfect and constrained decision-makers, but we choose nonetheless

Like other organisms, we exist in a state of uncertainty. Our knowledge of the external environment is filtered by sensory processes that have evolved through natural selection, but are imperfect. Conscious awareness of our own internal processes is limited. The “self” that we rely on for day-to-day survival is sometimes nebulous and even more uncertain than the external world. In fact, if Sam Harris and many spiritual mystics are correct, the entire concept of “self” is something of an illusion.[27]

We are not built for complete awareness of ourselves or our environment; we are built for uncertainty. Our senses and our conscious awareness have been tuned at a rudimentary level to distinguish between “us” and “not us”.[28] We use that limited knowledge of “self” to make decisions that have guided the development of our species over millennia. We take in sensory data, process it, and respond to the best of our capacities as organisms evolved to decide and act.

Both internally and externally our knowledge, our capacity, and our behavior are constrained. Yet those constraints make us who we are. They make us human instead of not human. They guide our behavior. As neuroscientist Kevin Mitchell has put it, “Selfhood … entails constraint. It is only constraint. The freedom to be you involves constraining the elements that make you up from becoming not you.”[29]

Despite every constraint and every uncertainty, we are designed by natural selection to be decision-making machines. That is who we really are.

---

[1] Libet (1983).

[2] Harris (2012), pp. 8-9.

[3] Gholipour (2019).

[4] “The RP [Readiness Potential] is generated by sampling only epochs that culminate in movement. In Libet-like tasks we never observe what happens when movement is not triggered. This raises the possibility that the RP is due to biased sampling, an artifact of the analysis process.” Schurger (2021), p. 562. “[T]he readiness potential is not in fact a signal of the intention to move that occurs long before subjective awareness but rather is an artifact of the way the data are analyzed.” Mitchell (2023), p. 185.

[5] Neuroscientist Kevin J. Mitchell describes the Libet experiments as “one of the most widely misinterpreted set of findings in human neuroscience….” “[T]he implications of these findings have since been widely extrapolated, way beyond the bounds of the actual experiment, to suggest that we never really make decisions at all, that our brains just do the deciding for us, and that we later make up stories to ourselves to rationalize our actions in some kind of post-hoc narrative. Indeed, these experiments are often cited as conclusive evidence that neuroscience has shown free will to be an illusion. This is, to put it mildly, a drastic overinterpretation.” Mitchell (2023), pp. 181, 183.

[6] “In those voluntary actions that are not ‘spontaneous’ and quickly performed, that is, in those in which conscious deliberation (of whether to act or of what alternative choice of action to take) precedes the act, the possibilities for conscious initiation and control would not be excluded by the present evidence.” Libet (1983), p. 641.

[7] Imagine the unconscious “readiness potential” that might appear in EEG readings of Hamlet’s brain over the course of his many days of doubt and indecision.

[8] See Mitchell (2023), Dennett (2003), and Dennett (1991) for general descriptions of brain behavior evolved through natural selection.

[9] Imagine a tiger crouching in the jungle, body and brain alert, muscles contracted and ready to respond to any sign of opportunity or threat, before it finally accumulates enough sensory information to cause it to spring, or alternatively, to relax its guard and move away.

[10] “[]he signal clearly fluctuates noisily up and down all the time. Sometimes it goes back down again, and the person does not move; other times it happens to reach a threshold, and then a movement is initiated.” Mitchell (2023), p. 184.

[11] “Indeed, simulations show that when the model is interrupted at random times and forced to produce a speeded response …, the fastest responses are preceded by a slow amplitude deflection (in the direction of the threshold) that long precedes the interruption itself, whereas the slower responses are not. Hence, even sensory-cued responses can be preceded by a readiness potential.” Schurger (2012), p. 2.

[12] “Presumably the increased (negative) electrical potential preceding faster responses cannot reflect preparatory neural activity, because the clicks [cues] were unpredictable.” Schurger (2012), p. 4.

[13] Mitchell (2023), p. 184.

[14] To return to the tiger analogy, the leap does not begin from the moment when the tiger’s muscles tense for action. Muscle contraction may occur over and over again before the leap. The leap occurs when the tiger has enough information to determine the time is ripe for the attack. That is when the tiger decides to move.

[15] “Indeed, if the decision to move is marked by the time of threshold crossing, then awareness of conscious intention to move coincides with the decision point, as common sense would suggest.” Schurger (2021), p. 566 (emphasis in original).

[16] “We propose that the neural decision to move coincides in time with average subjective estimates of the time of awareness of intention to move… and that the brain produces a reasonably accurate estimate of the time of its movement-causing decision events.” Schurger (2012), p. 7.

[17] See Dennett (2003), pp. 227-242.

[18] Dennett (2003), p. 242. “[T]here no such place in the brain. As I never tire of pointing out, all the work done by the imagined homunculus in the Cartesian Theater has to be broken up and distributed in space and time in the brain.” Dennett (2003), p. 237-238 (emphasis in original). (The “imagined homunculus in the Cartesian Theatre” is, of course, the magical central control room.)

[19] Dennett, after reading the 2019 article in the The Atlantic titled “A Famous Argument Against Free Will Has Been Debunked”, tweeted “The Libet results on free will and their many descendants are crumbling now, and there is more to come. A nice case of science exposing hidden dualist assumptions in neuroscience.” Tweet dated September 12, 2019, https://x.com/danieldennett/status/1172159910286680064.

[20] “In reality, mind and brain cannot be separated like this. A more accurate conception of the mind is as an interlocking system of cognitive activities that are necessarily mediated by the functions of the brain. In humans, some of these cognitive activities are associated with conscious mental experience, but they don’t all have to be to be effective.” Mitchell (2023), pp. 208-209.

[21] See Dennett (1991), pp. 134-135, for a description of what he calls a “multiple drafts” model of brain processing.

[22] “The brain processes stimuli over time, and the amount of time depends on which information is being extracted for which purposes.” Dennett (2003), p. 238.

[23] “We don’t experience the firing of our neurons or the flux of ions or the release and detection of neurotransmitters. What we do experience is what patterns of neural activity mean, at the level that is most relevant and useful and actionable for the organism as a whole.” Mitchell (2023), p. 209 (emphasis in original).

[24] And as Daniel Dennett reminded us, they are not even aware of our existence. “Not a single one of the cells that compose you knows who you are, or cares.” Dennett (2003), p. 2.

[25] “We do not need or want complete information for optimal oversight: what we want is the right information, at the right level. The key to control is precisely the selectivity of conscious awareness. We are configured so that most of our cognitive processes operate subconsciously, with only certain types of information bubbling up to consciousness on a need-to-know basis.” Mitchell (2023), p. 262.

[26] “Even if we can sometimes be primed by external factors, this does not mean that we never make our own conscious decisions for our own reasons.” Mitchell (2023), p. 251.

[27] Harris (2014).

[28] Mitchell (2023), p. 75.

[29] Mitchell (2023), pp. 247, 279 (emphasis in original).