Science and Free Will
There is a trope, a consistent assertion, that science has disproved Free Will. I have heard many people from Jerry Coyne to Sam Harris say conclusively that Free Will cannot exist because physics, or neuroscience, or some other area of study has shown that “free will is ruled out, simply and decisively, by the laws of physics.”1
Free Will is even attacked from a tautological point of view (meaning, disproven by definition) independent of science, with assertions that its existence is logically incoherent.
It’s extremely important to know whether they are correct. Our economic, justice, and social systems all operate on the basic assumption that humans are independent operators with the ability to make free will choices. If a significant portion of participants in those systems stop believing in Free Will (or really, ACT like they don’t believe in Free Will), then these systems will experience cataclysmic failure, with dire results. Further, it is hard to imagine a functional replacement for these systems that does not assume human Free Will, except those that most of us would find horrifically dystopian.
For instance, if we really did think that humans have no Free Will, then what is the point of treating people with respect in a justice system? If people fundamentally don’t have any choice in how they act, then people that do wrong are considered fundamentally defective, a permanent drag on society. The justice system would have no reason to respect the rights of accused criminals, but only see them as problems to be solved2. The entire system would logically devolve into a generator of carrot and stick actions, meant to sway the behavior of the rest of the population. If the goal is lower crime, extreme and draconian punishments would be administered as quickly and publicly as possible for even small crimes, in order to sway the obedience of the choice-less masses3. Cuban or Soviet or Chinese gulags come to mind. Thought crimes would become real crimes because if everyone is simply a meat machine, then one of the best predictors of future bad behavior is present bad thoughts. Why not get ahead of the game? Why not send the children of “wrong-thinking” dissidents to gulag “re-education” camps, like the Soviet Yagrinsky Labor Camp for children of enemies of the people, Russia, Arkhangelsk Region?
In fact, our current justice systems in many Western countries don’t respect the human ability to make Free Will choices enough in my opinion. If we had more faith in humans to make unexpected choices, our system would be more focused on rehabilitation than retribution. A strongly Free Will basis of understanding people would hold out the hope that no one is irredeemable, that anyone has at least a chance to become a better version of themselves.
In other arenas, a lack of belief in Free Will would lead to similar concerns associated with coerced labor in economic systems and general dehumanization in societal situations (“it’s better for everyone if we shun the Johnsons rather than my children be around a family that thinks differently than mine”). Of course those things still happen in systems that presume Free Will. But in systems that presume no Free Will, it is hard to see how such ugliness would not become the norm. Because if you think that people are products of experience, and you have no control over their experiences, then you have no hope that they can choose better things. You think of them not as free agents that can surprise you, but rather forces of the negative parts of society that you see as a problem, not as a person.
Let me paint the picture more explicitly and exactly on what happens when people fully embrace that Free Will doesn’t exist.
a. If humans don’t have Free Will, then their thoughts and actions are ultimately controlled by external experiences.
b. Generally, people don’t have control of their external experiences.
c. Therefore, I am not responsible for my thoughts and actions.
d. So then I can do whatever I want, because I am not responsible for actions.
e. If there are negative consequences for other people, then it’s really the fault of the Universe of external experiences I’ve had, not any choice I’ve made, because I can’t make choices. I just am what the Universe has made me.
This thought process might be inevitable for anyone that actually fully embraces that Free Will doesn’t exist.
If you think your neighbor is in essence a robot, then why should you care about their rights, feelings, or welfare? Further, propaganda (which is already a constant in all societies) would increase drastically. If we all believed that external influences are the only determinant of human actions, then we would also believe that we can deterministically control those around us in order to make political, social, or economic gain.
So it matters greatly whether Free Will exists, and perhaps even more so that the vast majority of people think it exists.
As I have partially demonstrated above, it matters deeply whether people really do internalize and start acting as though Free Will doesn’t exist. I don’t have time to exhaustively prove that this would be a much, much worse world…but I certainly assert that it would be a dehumanizing, ugly “society” in which to live. So we should be very careful to assert that science proves Free Will doesn’t exist.
It also appears that belief in Free Will is positively correlated with pro-social behavior. An Italian study comparing the lives of believers in Free Will versus believers in Determinism “found positive and moderate correlations between beliefs in BFW and Self-and Other-Representations of adult attachment, contributing to support the general assumption that a prosocial orientation and interpersonal security are associated to beliefs in individuals’ autonomy.”
Multiple studies across many different cultures show a significant correlation between fatalistic determinism and depression. Worryingly, many of these studies also found a strong correlation between determinism and a lack of resilience in teenagers and adults.
It is hard to imagine any other single thesis that, if widely espoused and truly, fully embraced, could have a stronger negative effect on the world.
So, what does Science have to say about Free Will? With any philosophical discussion, it is extremely important to define terms.
Free Will - The ability of a person to choose to think or act at least partially independently of the sum of inputs into their experience, and/or in a way that is untraceable (incalculable) from those inputs4. Free Will is the number one evidence of Agency, the ability to determine one’s own path. Persons with Free Will think and act as if they have choice in matters, and will sometimes vacillate between options.
Science - The pursuit of knowledge driven by observation, hypothesizing, and testing. Knowledge gained from science must be repeatable and observable. There is knowledge outside science, like that derived from logic, math, and philosophy. There are also observations of unrepeatable instances, like the assassination of Julius Caesar or the nailing of 95 Thesis to the door of a Wittenburg church. There is also created knowledge, like the emotional truths found in a piece of poetry. Many humans assert that there is also Revealed Truth held in religious texts and practices. Yet science has a special place for truth-finding in society precisely because its methods of precisely quantifying objective, repeatable truths is extremely reliable and useful.
Materialism - The assertion that observable physical processes are all there is.
Many people conflate Science and Materialism. But that is an error. Science can only perform its work on Materialistic systems, because Science requires objective observables that are also repeatable. Science cannot be used to prove or disprove whether Julius Caesar was stabbed by 7 or 8 knives, or just one. Nor can Science quantify how your first kiss felt. But Science is very good with Materialistic systems such as planetary orbits, chemical reactions, and the optics of a prism. But as demonstrated above, Science is simply the wrong tool to describe whether a poem or a sunrise is beautiful.
Just as you cannot use a car to drive to the Moon and you cannot use a hammer to view DNA, you cannot use Science to determine whether or not the Universe is Materialistic. You cannot use a tool that only works in one domain to test whether areas outside that domain exist. Science cannot say whether Hamlet is worth studying. Science cannot comment on whether a logic proof is true or untrue, unless one can implement that logic proof using materials (which most of the time one cannot5). Science cannot predict whether you will want a dark or milk chocolate mousse for dessert.
Science can be used to disprove some metaphysical assertions, but it cannot comment conclusively on whether metaphysical things like love, truth, or even souls exist. It can only provide some data on the physical manifestations of these things.
Very many people think that because Science is limited to testing Materialistic relationships, then this alone proves Materialism is all there is6. This is obviously false. Materialism might be true or false, but you cannot use Science to determine that because Science only operates inside Materialistic frameworks.
Of course, if Materialism is NOT all there is, then clearly we have mechanisms for the existence of Free Will, such as the soul that most humans strongly profess they have.
But I can prove Free Will is allowed even within the framework of Materialism.
First, we must note carefully the definition of Free Will. I’ll state it here again for convenience:
Free Will - The ability of a person to choose to think or act at least partially independently of the sum of inputs into their experience, and/or in a way that is untraceable (incalculable) from those inputs.
Does it say that no human interaction is in any way dependent on external experience? No. Clearly we are creatures capable of being influenced to change our minds. I might be on a diet, but if you offer me free donuts, that will likely change my course of caloric intake that day.
For Free Will to exist, only some decisions must be independent of (or untraceable to) causal effect. Now, I am not using ”untraceable” as a wiggle-word. It is extremely important to note that I don’t just mean “untraceable” by current human science. I mean that the dependent causes of many human choices are fundamentally, physics-wise, ultimately untraceable, even if one had perfect observation of all particles involved and an infinite amount of computation power to calculate such dependencies.
We find ourselves in a very strange Universe. All the energy and matter fields that make up our Materialistic universe behave in ways that are fundamentally probabilistic. At the scales up to the nanometer or so range, particles in low energy states don’t act like billiard balls bouncing off each other. If they did, we could possibly calculate many bounces forward in time, calculating how a whole system acts perfectly for many seconds, hours, or even years.
But particles such as electrons are really just wave packets, localized pinches in what is called the electron field (not to be confused with the electric field). The electron field permeates all of space in 3 dimensions (at least), and is one big field. Electrons are literally just an excitation in that field, as if the field were plucked at some point to produce a local 3D disturbance in that electron field. That little disturbance IS an electron. Now, because it has a charge, it interacts with the electric field permeating the universe, but the electron and electric fields are different.
And what is even more odd is that we have mathematical equations called wave functions that appear to perfectly describe not how these fields and their pinchey particles act, but rather describe how those particles will probably act. A big part of the uncertainty comes from the fact that the particles are a combination of 3D waves in a larger field, but fundamentally the wave function doesn’t describe those waves. It describes the probability that those waves will manifest a particle in a certain area.
That’s right. Our very best physics tells us in what region a particle probably is, but not exactly where it is. Also, it can’t exactly predict what the particles momentum is. Or spin. Or really even its exact mass or charge or velocity. The clouds in this carbon atom simulation below aren’t thousands of electrons. They are the possible areas in which carbon’s six electrons likely will be manifest if you hit the atom with a laser. Until that laser hits, each of the electrons do actually probabilistically exist in the entire allowable cloud areas. (Animation by Jeremy Mallin.)
And we are pretty sure at this point that this is not some failing of the theory. Our best physicists are pretty much unified in telling us that reality itself is fundamentally probabilistic. There is no way to make absolutely sure an electron will be found on the left side of its hydrogen atom and not on its right side at a certain time. It’s not just incalculable. The electron fundamentally, actually inhabits both the right and left side of the atom simultaneously in some cases.
So the fundamental interactions of the universe are untraceable.
On the scales of basketballs and planets and galaxies, this quantum fuzziness averages out and collapses away. We may not know where a certain electron exactly is, but a basketball contains about 10^27 electrons, so the probabilities tell us half are on the right side of the hydrogen atoms and half are on the left, and that works for knowing what is going on with a basketball. Planetary orbits can be calculated to within feet for thousands of years in simple one-star, one-planet systems.
But because human synapses operate at this nanometer level, they are subject directly to this quantum un-traceability. There aren’t enough electrons involved in each synaptic exchange to completely average out all the quantum probability weirdness.
Our brains are these extremely weird levers. The way a single synapse fires (or misfires) can actually change how you interpret or calculate something. And that change in perception can change an action of your body, this giant macroscopic thing that should experience no direct effects of quantum fuzziness. We are more massive than a basketball. But basketballs don’t have quantum levers like we do. Our neural synapses are probes that actually pinch down to touch the fuzziness of the quantum world and thus we can be directly affected by it. We are affected by it, constantly.
Most of the time, that quantum weirdness probably does average out, because we do have almost a quadrillion synapses. But our brains aren’t synapse averaging machines. Even a single different synapse fire can change things for us. So unlike a basketball or rock or a planet, human beings are in some sense fundamentally quantum-probability-touching creatures.
And anything involving quantum probability is fundamentally, actually untraceable7.
There is no way, in our Universe, to trace a human’s inputs down to the synapses and then back up from the synapses in a way that is deterministic. Because of our synapses, human cognition is non-deterministic.
Give me a lever large enough, and I can move the world. Give me a lever as small as a synapse on one end, and humans become decoupled from determinism.
So even if you do think that atoms and fields is all there is, you still have to believe that at least some of your choices are untraceable to any singular set of events. Physics proves you are an actor independent from perfect determinism, at least some of the time.
I re-emphasize: averaging out the random contributions does not work if you have a method of amplifying very small quantum interactions into the macro world. That is what synapses and neurons and brains do, at least sometimes. This is why a rock is not affected by quantum probabilistic indeterminism, but humans are.
In summary,
There is an assertion that Science demands Materialism. This is not true. Science is merely limited to operating on materialist systems. Science is agnostic on truths, knowledge, and information beyond materialism.
Even if Materialism is true, there is an assertion that Materialism requires determinism. This is not true, because synapses and neurons provide access to and amplification of quantum non-determinism. Thus human existence is not fully deterministic. So some Free Will is allowable to exist in humans, according to Science and even according to Materialism.
There is one other “scientific” attack on Free Will that I must mention briefly, via an excerpt from an excellent Scientific American article on the matter:
Many people believe that evidence for a lack of free will was found when, in the 1980s, scientist Benjamin Libet conducted experiments that seemed to show that the brain “registers” the decision to make movements before a person consciously decides to move. In Libet’s experiments, participants were asked to perform a simple task such as pressing a button or flexing their wrist. Sitting in front of a timer, they were asked to note the moment at which they were consciously aware of the decision to move, while EEG electrodes attached to their head monitored their brain activity.
Libet showed consistently that there was unconscious brain activity associated with the action—a change in EEG signals that Libet called “readiness potential”—for an average of half a second before the participants were aware of the decision to move. This experiment appears to offer evidence of Wegner’s view that decisions are first made by the brain, and there is a delay before we become conscious of them—at which point we attribute our own conscious intention to the act.
However, if we look more closely, Libet’s experiment is full of problematic issues. For example, it relies on the participants’ own recording of when they feel the intention to move. One issue here is that there may be a delay between the impulse to act and their recording of it—after all, this means shifting their attention from their own intention to the clock. In addition, it is debatable whether people are able to accurately record the moment of their decision to move. Our subjective awareness of decisions is very unreliable. If you try the experiment yourself—and you can do it right now, just by holding out your own arm, and deciding at some point to flex your wrist—you’ll become aware that it’s difficult to pinpoint the moment at which you make the decision.
An even more serious issue with the experiment is that it is by no means clear that the electrical activity of the “readiness potential” is related to the decision to move, and to the actual movement. Some researchers have suggested that the readiness potential could just relate to the act of paying attention to the wrist or a button, rather the decision to move. Others have suggested that it only reflects the expectation of some kind of movement, rather being related to a specific moment. In a modified version of Libet’s experiment (in which participants were asked to press one of two buttons in response to images on a computer screen), participants showed “readiness potential” even before the images came up on the screen, suggesting that it was not related to deciding which button to press.
Very few scientists today claim Libet’s readiness potential findings are a scientific argument against Free Will. So here is where we find ourselves:
Science tells us that humans are not always bound by any fundamental determinism. Thus Science tells us that Free Will is not forbidden by Science.
Now, that does not mean that Science has proven Free Will does exist. I do find it extremely curious that we happen to inhabit a Universe that seems customized at the fundamental physics level to allow for Free Will, but that is a more theological/metaphysical consequence.
Science has not yet discovered or even really postulated a mechanism of Consciousness (the ability to have subjective experience which includes self-direction and a sense of self). In fact, we are so far from solving “the Hard Problem” of Consciousness that we don’t even have a universally agreed-upon definition of it. The parenthetical definition I give above is only the simplest one I have come up with.
We must first make progress on the mechanisms of Consciousness before we can figure out what mechanism generates our choice functions and thus Free Will. But we do have one extremely strong set of evidence that Free Will exists.
Everyone thinks and acts like they have Free Will. Almost everyone you talk to knows that they make their own choices. Even if you are a Free Will skeptic, you have spent most of this article arguing with me in your head that Free Will is an illusion, because you want me to choose to not believe in Free Will. What is ever the point of debating with someone if you don’t believe in Free Will? If Free Will doesn’t exist, that person will always mechanistically think and act in whatever way they are destined to think and act. And what is the benefit of exerting influence to get them to not believe in Free Will?
Every time we fail a diet or decide to keep a diet, every time we choose a latte type, every time decide to look on our phone instead of finishing a task, our lives scream the true existence of Free Will.
You have Free Will. Science allows it, and experience all but proves it.
But if you still choose to not believe in Free Will, please do us all one favor:
Don’t act like it.
Jerry Coyne, quoted here: https://www.theguardian.com/news/2021/apr/27/the-clockwork-universe-is-free-will-an-illusion
In many times and places our current justice system falls into such errors even with a respect for free will and self-rehabilitation, but in many countries justice systems usually have some respect for rights of the accused and also convicted criminals.
The fact that this strategy doesn’t work with the War on Drugs very well is either proof that Free Will exists or proof that addictive drugs effectively reduce peoples’ levels of Free Will. I think both is true.
This is my best definition of Free Will that skips a few steps of thought. Most definitions talk about Free Will as having control of your thoughts and actions. But what is control? It is basically saying you are able to act independently from the forces and information external to you if you choose. But then others will say that of course we can provide evidence that ALL people are influenced by their surroundings to some extent, which is true. But common sense Free Will doctrine mere says that to some extent our choices are not fully controlled by external influences, that we have some choice in the matter. This is how I landed on this shortest possible somewhat-comprehensive definition of free will.
And even then, if you could instantiate a proof using physical objects, a scientific method of using real world objects to repeatedly prove your proof would likely be inexact and not as robust as a fully logically or mathematically proven proof.
Science isn’t just limited to studying what is Material. It is also limited to studying what is repeatable.
Semiconductors are an interesting edge case, as transistor gates are now small enough to interact with quantum indeterminism. But billions of dollars are poured into semiconductor research and engineering in order to average out all this quantum weirdness, in order to make their operations as deterministic as possible. The industry actually has measures of how often they think a bit or voltage will flip due to quantum indeterminacy, and they work to minimize that (with amazing success). Work on quantum computing does the opposite.