The Pseudoscience of Free Will
How scientific defenses of free will resemble creationism and astrology
I’ve always found scientific defenses of free will deeply frustrating. Typically, such books throw around complicated ideas like quantum dynamics, chaos theory, and emergence in a way that is confusing and begs obvious questions. I am left with the befuddled sense that the writer is not being straight with me. One of the many strengths of neuroscientist Robert M. Sapolsky’s Determined: The Science of Life Without Free Will is that it analyzes scientific defenses of free will clearly and shows their emptiness. Sapolsky does not use the term pseudoscience to describe these theories, but he provides the material to make the argument that they are. In addition, he helps the reader to understand the motivations behind this pseudoscience.
The position on free will that Sapolsky is defending, called “hard determinism,” is that people are wholly and deterministically formed by biological and environmental forces beyond their control. It’s the same position that Sam Harris defends in his famous (or notorious) book Free Will, but Sapolsky’s account is much richer and more satisfying. The processes that make us up are overwhelmingly deterministic, meaning that given the same starting point, they will always have the same outcome. Although Sapolsky recognizes that the universe is non-deterministic at the atomic level, he thinks that such quantum effects play no significant role at the level of human behavior. If two people had exactly the same genome and experienced exactly the same environments over the course of their lives, they would end up as essentially the same person.
Against Sapolsky are ranged the majority of people and philosophers who believe that we have a free will that enables us to transcend the deterministic processes of our biology. We can call such people “scientific anti-determinists” if they try to defend their theory of free will through scientific theories. There are many free will debates, and I do not wish to attack all notions of free will here, merely scientific anti-determinism.
The Bio-Environmental Mesh
The first part of Sapolsky’s book is a rich and engaging survey of the deterministic biological and environmental forces that make up who we are. He draws on research in psychology, neurobiology, endocrinology, genetics and other fields to give fascinating examples of how people are shaped by these forces over different time scales.
For example, our testosterone levels vary across the course of the day. If your testosterone levels are high, you will be more inclined towards risk-taking and impulsivity, as well as more likely to interpret facial expressions as being threatening.
Over the longer term, environmental influences modify the brain. If you experience chronic stress in your childhood, your frontal cortex development will likely be impaired, leaving you less able to control your impulses in adulthood. If you grow up in an environment of clement weather, you will likely be more individualistic and extroverted than if you grow up in a harsher environment.
At the bottom of all of our behavior is our DNA interacting with our environment in complex and surprising ways. A gene variant that produces aggressive behavior can result in someone’s becoming a murderer or a very competitive chess player depending on the environment that affects the gene’s expression.
Sapolsky’s opening is effective not only because the science is engaging in its own right but also because it seems naïve to believe that there is any part of ourselves that could escape the influence of this encompassing and multifarious mesh of deterministic influences. Also, Sapolsky illustrates the power of scientific research based on deterministic assumptions to explain who we are in rich, concrete, and credible detail.
A Science of Free Will?
Many scientific anti-determinists have tried to build a case for free will on quantum mechanics. The most common view among physicists is that the movement of quantum particles is non-deterministic. In a deterministic system, you can, at least in theory, predict the current state of a system from a past state, but you cannot predict the behavior of quantum particles in this way. Weirdly, they have no single location. Rather, they are distributed probabilistically across a number of locations at once. And this is only one of the weird, non-deterministic features of quantum mechanics.
Sapolsky dismisses most quantum free will writing as “mystical gibberish”1 that is published in non-peer-reviewed journals. However, he evaluates the merits of the most careful and reasonable such theories. His conclusion is that no theory of this type presents a plausible mechanism by which quantum indeterminacy could lead to free will.
The first reason is the phenomenon of quantum decoherence, which explains why quantum effects disappear when many particles interact. If you’re building a bridge, you don’t need to worry about quantum mechanics because quantum indeterminacy has no consequences for objects of this size. While neurological entities like neurons or neurotransmitters are much smaller than bridges, they are still large enough for decoherence to operate. Quantum indeterminacy has been discovered in single atoms of the brain, but because of decoherence, quantum effects are not likely to have a significant influence on behavior. Addressing research on the quantum effects on glutamate, a type of neurotransmitter, Sapolsky writes:
The dendrite in a single glutamatergic synapse contains approximately 200 glutamate receptors, and remember that we’re considering quantal events in a single receptor at a time. A neuron has, conservatively, 10,000–50,000 of those synapses. Just to pick a brain region at random, the hippocampus has approximately 10 million of those neurons. That’s 20–100 trillion glutamate receptors (200 x 10,000 x 10,000,000 = 20 trillion, and 200 x 50,000 x 10,000,000 = 100 trillion). It is possible that an event having no prior deterministic cause could alter the functioning of a single glutamate receptor. But how likely is it that quantum events like these just happen to occur at the same time and in the same direction (i.e., increasing or decreasing receptor activation) in enough of those 20–100 trillion receptors to produce an actual neurobiological event that has no prior deterministic cause?2
Even if there were a plausible theory of how quantum indeterminacy could meaningfully affect our behavior, there is no reason to think that the effect would look anything like free will. Since quantum indeterminacy is random, it could seemingly affect behavior only in random ways. However, no one has ever defined free will as random behavior. Rather, philosophers have conceived of free will as being rooted in our character and principles. Sapolsky reviews various tortured efforts to explain how quantum indeterminacy might be harnessed by our brains to generate a plausible theory of free will, but finds them unconvincing.
Some other scientific anti-determinists have tried to root free will in chaos theory. Chaos theorists have discovered that some systems are unpredictable because their outcomes are sensitive to starting states. In chaotic systems, applying the same rule to different starting states results in radically different outcomes. Sometimes you get a delicate crystalline pattern, other times a different pattern, other times no pattern at all. There is no way of predicting which outcome will occur from a given starting state. However, Sapolsky points out that unpredictable does not mean non-deterministic. Even in chaotic systems, what is happening now is caused by what happened before according to deterministic rules. A given starting state will always give you the same end state if you apply the same rule to it.3
Another candidate for a cause of free will is emergence. Simple rules practiced by a large number of entities can lead to phenomena that are different in quality from those manifested by their components. A group of ten ants will not demonstrate any specialized roles. But if you combine thousands of ants together, you will get a society with job specializations and complex functionality. Colonies of ants can build bridges, tunnel waterproof passageways lined with leaves, and construct nests with rooms designed for specific purposes, like fungi farming or brood rearing. No ant knows the rules for building a whole colony. Rather, the colony emerges out of a small repertoire simple behaviors. As an example, ants choose good sites to build nests through the operation of two simple rules. An ant who discovers a good site simply stays there. Other ants know that if they find an ant immobile at a given spot, they should inspect the spot as a possible colony location. This leads to an ant swarm forming at the spot. This quality of emergence is crucial for many aspects of human biology as well, including the formation of neurons and brain tissue.
Some free will defenders have made the case that free will is an emergent property like this. They allow that small groups of individual neurons behave in a deterministic fashion but argue that when one combines billions of them together, they behave non-deterministically. Not so fast, says Sapolsky. While they have qualities that their components do not have, emergent systems do not have qualities that cannot be explained by the simple deterministic rules that govern their components. When building blocks combine to create an emergent state, that state does not change the nature of the building blocks. As Sapolsky wrily puts it, “The mistake is the belief that once an ant joins a thousand others in figuring out an optimal foraging path, downward causality causes it to suddenly gain the ability to speak French.”4
Scientific Anti-Determinism as Pseudoscience
The philosopher Paul Hoyningen-Heune elaborates a theory of pseudoscience in his book Systematicity: The Nature of Science. A genuine science makes progress by continually explaining more and more facts about its subject matter and resolving theoretical problems, contradictions, and mysteries. However, pseudosciences make no or little progress.5 Commonly, the only reason for the existence of a pseudoscience is to cast doubt on a genuine science.
Consider the pseudoscience of intelligent design, a form of creationism. Intelligent design advocates argue that there exist examples of irreducible complexity in nature that cannot be accounted for by the unguided process of Darwinian selection and must be the work of an intelligent being. As philosopher of science Massimo Pigliucci notes:
the idea of intelligent design in biology has made no progress since its last serious articulation by natural theologian William Paley in 1802. Compare that to the stunning advances of evolutionary biology since Darwin’s publication of The Origin of Species in 1859, and one has a good picture of the difference between science and pseudoscience.6
While evolutionary biologists continually make positive progress by expanding the range of facts that their field explains and formulating more precise and internally consistent explanations, intelligent design is an entirely negative discipline. There are no intelligent design experiments that offer positive support for the theory that the universe had an intelligent designer. Rather, a typical intelligent design tactic is to point to a mystery in biology and attempt to show that evolutionary theory cannot solve it.
Another characteristic of pseudoscience is that it makes very weak claims. Pseudoscientists often tell us that it is possible that their theory is true, a fallacy known as the “appeal to possibility.” A related pseudoscientific fallacy is the appeal to ignorance. Pseudoscientists make the case that because we don’t know everything about evolution or the universe or whatever, there is no reason to think that their theories will not some day find support, no matter how implausible they are now. Real scientists, however, show that their theory is not only possible but plausible and that it offers a better explanation of phenomena than competing theories do.
Pseudosciences typically make plentiful use of what is commonly known as “mumbo-jumbo.” Pseudoscientists often use key terms confusingly to impede rational evaluation of their theories. Consider the esoteric complexities of astrology with its houses of planets, phases of the moon, cusps, exaltations, and benefics.
By these criteria, it seems that scientific anti-determinism is a strong candidate for the label of pseudoscience. Ever since the 1980s, anti-determinists have been writing that it is possible that quantum indeterminacy is the basis of free will, but they never get any further than that. The theory never makes any progress towards becoming a plausible, empirically supported explanation of human behavior that might convince a rational evaluator like Sapolsky. Just as the sole purpose of intelligent design theory is to cast doubt on the theory of evolution, so the sole purpose of scientific anti-determinism is cast doubt on determinism.
Scientific anti-determinists also commonly make use of appeals to possibility and ignorance. As an example, philosopher Alfred R. Mele writes that there is a “possibility” that free will is made possible by “quantum ion clouds moving toward nerve terminals.”7 He never gives us any details. Later, he writes, “For all we know, the universe and our brains leave room for deep openness,” the indeterminacy he needs for his theory of free will.8 This is an appeal to ignorance. A serious thinker would explain to us why their theory was plausible and not merely possible.
And then there is the mumbo-jumbo factor. Scientific anti-determinism is always dependent on conceptual confusions. Sapolsky discusses a number of these. We have seen one example: to make their case that chaos theory might be the basis for free will, anti-determinists must mix up non-determinism with unpredictability. Real scientists are more careful.
With all these weaknesses, why does scientific anti-determinism persist? One could ask the same of intelligent design or astrology. The answer is that these pseudosciences have important psychological, social, and political functions that the pseudoscientists don’t want to give up. The functions of free will beliefs are multifarious, but one of the most important is that this concept provides us with a theory of moral responsibility on which we can base a system of justice. As Sapolsky goes on to show in the last part of his book, our theory of moral responsibility, and with it the concepts of legal innocence and guilt, are founded on the notion of free will. Someone is guilty on this theory when they choose of their own free will to commit some crime. If lawyers can prove that a defendant was not acting out of free will because they were forced to commit the crime by some other person or were not in their right minds, the defendant may be found not guilty or their sentence may be reduced. All of the free will defenses that I have encountered also end up defending some version of free will moral responsibility. I have dealt with the issue of free will and moral responsibility here.
Sapolsky and other hard determinists are threatening because they debunk the very notion of moral responsibility. If we are all formed by factors outside of our control, we can never be meaningfully responsible for our actions. One of the main reasons for the persistence of free will pseudoscience is that it gives us an excuse not to face the consequences of hard determinism for the theory and practice of morality and justice. And these are only some of the profound consequences of hard determinism, which I hope to explore in future essays.
Robert M. Sapolsky, Determined: A Science of Life Without Free Will (New York: Penguin Press, 2023), p. 214.
Sapolsky, p. 220.
Figure from Sapolsky, p. 137.
Sapolsky, p. 200.
Paul Hoyningen-Huene, Systematicity: The Nature of Science (New York: Oxford University Press, 2013), pp. 203-07.
Massimo Pigliucci, Nonsense on Stilts: How to Tell Science from Bunk, 2nd ed. (Chicago: University of Chicago Press, 2018), p. 29.
Alfred R. Mele, Free Will: An Opinionated Guide (Oxford: Oxford UP, 2022), p. 48.
Mele, p. 142.
I got a question about my view of Kevin J. Mitchell. I actually deleted a long paragraph about him from the essay because it was getting too long. Here it is:
I did not enjoy the recent book Free Agents: How Evolution Gave Us Free Will by neuroscientist Kevin J. Mitchell. In the chapter “Harnessing Indeterminacy,” it seems as though Mitchell might address one of the fundamental questions that critics have been posing to anti-determinism for decades: how might quantum processes get harnessed by our brains to create free will? But no such theory is forthcoming. Mitchell tells us that the brain is full of “noisy” and “random” processes. However, the meaning of these terms fluctuate, and Mitchell nowhere makes the case that these processes are genuinely non-deterministic as opposed to being merely unpredictable. He writes that neurotransmitters get released probabilistically, which seems to be an example of what he means by “noise,” and that this process is somehow influenced by “quantum indeterminacy.” But he comes up with no testable theory that would establish the relevance of quantum indeterminacy to decision making, nor does he try to distinguish among the different causes and effects of different types of noise. Later in the same chapter, he tells us that the behavior of animals is also often “random,” but it isn’t clear that random brain processes and random behavior have anything to do with each other. I find Mitchell's book a murky stew of ideas.
> “hard determinism,” is that people are wholly and deterministically formed by biological and environmental forces beyond their control
If that's the definition, I think there's a problem right there that makes the discussion much less clear than it appears. Assuming physical materialism, who is the "they" in the phrase "beyond their control"? In this view, the person itself is none other than a set or pattern of collected physical patterns and forces, mostly stored in the physical configuration of their body. I'm not sure it even makes sense to say these forces are beyond their control when they literally *are* the person.
It only makes sense to say, for example, that the circulation of blood in my veins is "beyond my control" if we assume some kind of "me" that somehow matches the experiential feeling that I have, of an embodied consciousness which can voluntarily tell the arm to go up, but not the blood to go here or there. If physical materialism eliminates this very "me", the whole question goes up in flames.
Which could be a good reason why, when people try to bridge the gap and come up with theories about free will residing in such things as quantum indeterminacy, it comes out mostly as soothing nonsense.