Two critiques of Sam Harris's 'Free Will'

Sam Harris's book Free Will is a brief synopsis of the question of whether humans have free will. We feel that we are making decisions, but we also know that we are at least partly determined by natural causes. On a pool table full of billiard balls hitting other billiard balls, the path of each ball is predetermined, and no ball moves of its own free will. We do not like to think of ourselves in that way, yet this is the implication of saying that all action is initiated by the physical brain, which is itself acted upon by other physical influences. Harris leans toward the side of saying there is no free will.

Sam Harris's book will not be summarized here. Instead, we will look at two critiques.

Less persuasive to me

Free Will: Sam Harris Has It (Wrong) by Barry L. Linetsky (2013)

Linetsky says that "modern philosophy" is "dangerous" to its "victim[s]," as those philosophers who "reject reality" may cause "detachment" and "confusion." (Why specifically modern philosophy, I cannot say.)

Linetsky champions free will by saying, "we know it to be obvious by means of introspection," and that to deny it is "ludicrous and beyond rational comprehension." Everything from a small act of paying attention to being able to take personal responsibility for one's actions requires free will. He points out that Harris accepts the existence of the universe and human consciousness by saying that they are obvious to us. So why, he asks rhetorically, doesn't Harris accept the existence of free will on the same basis? He also suggests that if there is no free will, then people have no control over anything they think; this, for some reason, means they cannot have certainty about any of their conclusions, and that Harris's argument is therefore "invalid."

My problem here: Just because we assume the existence of some things (at least provisionally) does not mean we should decisively and permanently accept the existence of everything. Also, to the extent that certainty is a feeling that arises when we are confronted by evidence or logic that seems persuasive to us, there is no reason that the lack of free will should deprive us of this feeling. It is simply that our feeling of certainty would be determined, along with the thoughts and experiences that bring it about. Furthermore, in either a determined or a free universe, whether Harris feels certain about his own conclusion has nothing to do with whether his argument is logically valid.

Linetsky goes on to say that if our careful deliberation about our many choices – significant and insignificant – is just an illusion, because the outcome of our thought process is already determined, and we are "self-deluded about all of these things all of the time," then we can have no true learning, knowledge, or understanding. The deterministic view must be rejected on the basis, it seems, that this idea is unpleasant.

Pleasant or no, that has nothing to do with whether it is true. Maybe abandoning belief in our own free will does not reveal learning, knowledge, and understanding to be impossible, but rather provides an opportunity for us to redefine what they are. Today there is much investigation into "artificial intelligence." If it is possible (at least in theory) to program a robot to learn, know, and understand, perhaps our own thought and experience is rightly understood along the same lines. This may seem terrible, but it is only redefining and reinterpreting the meaning of human intelligence. What we originally thought about our own intelligence might turn out to have been an illusion. This does not mean that there is no such thing as intelligence.

Linetsky does bring some important insights to the debate. He says it prejudices the investigation to emphasize neuroscientific observation of physical brain activity and to de-emphasize behavioral observations of decision-making in real-life contexts. Moreover, just because the origins of thought are dark and mysterious and that the chain of events stretches back indefinitely far into the past does not enable us to assume that the germ of "free will" can never be found. Linetsky comes down too strong on these points, however. He says that a person's behavior is itself "sufficient evidence to demonstrate the existence of free will." (It is not; the explanation of the person's behavior is the very question at issue.) He also tries to dodge the question of the infinite regress of causes by saying that "free will" is contained in the moment of decision and that the unmanageably long list of causes predating that moment can be waved away. (They cannot; the question of whether the moment of decision is truly "free" hangs on whether it was controlled by other influences.)

More persuasive to me

Free Will: A Response to Sam Harris by Kurt Keefner (2012)

Keefner observes that Harris divides the human mind into neurological consciousness and neurological unconsciousness. Harris claims that we identify our sense of self with our consciousness, as opposed to our unconsciousness. In focusing on the brain, as Keefner critiques him, Harris seems to entirely neglect the role of the rest of the body. More pointedly, Harris seems to take a purist view that consciousness can be separated from unconsciousness, when there is no such firm dividing line.

Keefner writes: "The quality of consciousness as such does not exist by itself any more than the roundness of a ball exists apart from the ball....to try to understand the roundness of a ball apart from what the ball is made of, its weight, its rigidity or bounciness, etc., would be a serious error."

Consciousness is thus a way we describe something else that is real, but it does not have reality by itself. Consciousness should not be reified – it should not be understood to have more reality than it does. Consciousness is not a "freestanding entity," but the "activity of an integrated entity."

Keefner describes four "levels of functioning": "non-conscious" (e.g. physical vital signs and perception), "unconscious" (e.g. "knowing how to speak our own language"), "preconscious" (e.g. "remembering my sister's name when I'm not thinking about her"), and "conscious" (e.g. "thinking about free will"). These levels of functioning form a "nested hierarchy," such that, when one acts consciously, one also acts preconsciously, unconsciously, and non-consciously. "Physiologically we have to talk about separate processes, but philosophically, the right approach is to own the whole package and feel yourself permeate all of your processes rather than dividing the self up."

He pinpoints a problematic passage in Harris’s book: "Consider what it would take to actually have free will. You would need to be aware of all the factors that determine your thoughts and actions, and you would need to have complete control over those factors." He rebuts this definition of a free being as one with “no unconscious processes at all within it, a completely pure consciousness all the way down.”

He says:

This is a terrible straw man. There is no reason why unconscious forces could not shape part of our mental lives while we consciously exercise some kind of decisive control...True, there are unconscious processes within us, but they are us. The atoms or neurons we are made of don’t cause us to be conscious. Our atoms and neurons are conscious. Our atoms and neurons make choices. It’s not that individual atoms and neurons are conscious and make choices, but in the right organization, i.e. me, they become a whole that is conscious and does make choices.

He concludes that Harris's "philosophical case collapses" because there is no clean divide between unconscious and conscious processes and therefore the conscious ones cannot be purely controlled by the unconscious ones. He also provides a compelling critique of the particular neurological experiments that Harris chose as examples, explaining why they do not demonstrate what Harris says they demonstrate.

Free Will by Sam Harris

Free Will: Sam Harris Has It (Wrong) by Barry Linetsky

Free Will: A Response to Sam Harris by Kurt Keefner

Image: Taken in Zaragoza, España, 2012 and uploaded by 'Juanedc' to Wikimedia Commons. Creative Commons Attribution 2.0 Generic license.

---

The Illusion of Conscious Will
By Daniel M. Wegner
The MIT Press
DOI: https://doi.org/10.7551/mitpress/3650.001.0001
ISBN electronic: 9780262285896
In Special Collection: CogNet
Publication date: 2002

---

If you'd like to learn more about my work, I've published books. Also, I write for Medium. There, readers with a paid membership don't have to worry about the paywall.

Are there correct answers to moral questions? Yes, says atheist Sam Harris

Atheist neuroscientist Sam Harris's book The Moral Landscape: How Science Can Determine Human Values (2010) argues for a consequentialist and realist moral theory where the good is whatever promotes human well-being. Harris says that insofar as there are objective facts about what makes humans thrive, there are objective facts about how we ought to treat other people. In any one situation, there may be multiple relevant types of well-being to be considered, it may be difficult to gather data and assess risk, and there may be dissenting opinions, but this does not mean the moral question has no correct answer. Moral facts may exist even if we are untalented at perceiving them and tend to disagree about them.

There are facts about morality, he says, because there are facts about well-being. "If we were to discover a new tribe in the Amazon tomorrow, there is not a scientist alive who would assume a priori that these people must enjoy optimal physical health and material prosperity," Harris writes. Scientists would want to study the tribe's health and prosperity before passing judgment. On the other hand, he suspects that

"news that these jolly people enjoy sacrificing their firstborn children to imaginary gods would prompt many (even most) anthropologists to say that this tribe was in possession of an alternate moral code every bit as valid and impervious to refutation as our own...The disparity between how we think about physical health and mental/societal health reveals a bizarre double standard: one that is predicated on our not knowing - or, rather, on our pretending not to know - anything at all about human well-being."

We do indeed know some things about individual and social well-being, Harris says, such as that killing children doesn't contribute to it. "But this notion of 'ought' is an artificial and needlessly confusing way to think about moral choice," he qualifies. "For instance, to say that we ought to treat children with kindness seems identical to saying that everyone will tend to be better off if we do."

He bravely acknowledges some of the problems with consequentialism. For example, he quotes philosopher Patricia Churchland who has pointed out that "no one has the slightest idea how to compare the mild headache of five million against the broken legs of two, or the needs of one's own two children against the needs of a hundred unrelated brain-damaged children in Serbia." He also mentions the problem of calculating well-being as raised by philosopher Derek Parfit: If total well-being is our aim, it must be better to fill the Earth with hundreds of billions of people who have the barest glimmer of joy than to make the current seven billion people very happy. If, on the other hand, we are to focus on the average well-being per capita, we should euthanize our unhappiest people and we should prefer large numbers of mostly miserable people over one exceedingly miserable person. We would not normally intuit these to be moral goals, so perhaps something is amiss with these approaches to consequentialism.

Harris also acknowledged John Rawls's criticism that fair play and optimum outcomes sometimes seem at odds with each other. What if enslaving a few people would make the rest of humanity very happy? In a consequentialist moral theory, why can't we treat people as means to the ends of others? Harris doesn't try to tackle too much in his response to these objections - which was probably the right thing to do within the scope of a book that is already quite ambitious - but it is important to realize that these questions are left open. This kind of radical uncertainty doesn't lend support to his idea that moral questions have right and wrong answers. It isn't just that the calculation is too hard to crunch; it's that it looks suspiciously like it can't be run at all.

Harris explores the evolution of morality and its roots in cooperation and language, referring to the work of William Hamilton on kin selection, Robert Trivers on reciprocal altruism, and Geoffrey Miller on sexual selection. He talks at some length about how the brain's medial prefrontal cortex (MPFC) is associated with both mathematical and ethical beliefs; people with damage to that area are more likely to remove emotion from moral analyses. He correlates a healthy amount of fear with moral understanding.

He points out that, as Adam Smith noted, we tend not to be swayed by anonymous and distant suffering, even if we know it involves large numbers of people. He mentions of the work of psychologist Paul Slovic who has pointed out that this leads to "genocide neglect." Harris concludes: "Clearly, one of the great tasks of civilization is to create cultural mechanisms that protect us from the moment-to-moment failures of our ethical intuitions." Another gap in our moral reasoning is that we suffer from poor intuition on risk analysis. We tend to gravitate toward the power to "save" some and shy away from the complementary choice of "losing" the rest, even when these amount to the same thing and the only difference is in the linguistic expression of the glass being half full or half empty. Harris explains: "Another way of stating this is that people tend to overvalue certainty: finding the certainty of saving life inordinately attractive and the certainty of losing life inordinately painful."

He doesn't believe in free will and thinks we trump up the idea of it more than we actually feel a sensation that represents free will. First of all, from a neurological point of view, unconscious brain activity associated with an action has been documented to precede our awareness of having chosen to do anything. It's as if the car starts before we turn the key. The sense of having "chosen" may be an illusion. Secondly, although we have the freedom to emphasize certain facts, we cannot, given a body of available evidence, freely choose to believe something that contradicts the evidence. He concludes that scientific and ethical judgments have something in common; beliefs about facts and beliefs about values are not very different from a neurological perspective.

The book's shortcoming lies in his attempt to straddle the fence on the question of how much we should listen to our moral intuition. For example, he says that the fact that his moral theory fits our commonsense definition is a reason in favor of accepting his theory: "While moral realism and consequentialism have both come under pressure in philosophical circles, they have the virtue of corresponding to many of our intuitions about how the world works." This does not quite dovetail with his list, mentioned above, of all the ways in which our moral intuitions tend to let us down due to rational and emotional flaws in our brains. If we're so fallible about the details of calculating consequentialist outcomes, why should we assume that consequentialism itself is a correct theory? Who cares what seems right to a species that is so often wrong?

A specific application of this problem occurs when he takes anthropologist Scott Atran to task for claiming that the religious motives articulated by violent Muslim jihadis may not be what really causes them to become violent; they may be radicalized by a lack of social integration, for example. Harris seems outraged as he insists that "given the clarity with which they articulate their core beliefs, there is no mystery whatsoever as to why certain people behave as they do." This is an odd comment, coming from him; isn't one of the main arguments of his book that people frequently are mistaken about their own motives? As mentioned above, Harris did say that we should create "cultural mechanisms" that would, in essence, trick us into behaving better. There seems to be a trace of unexamined bias here: is it well-meaning atheists who would benefit from paternalistic cultural guidance designed to draw out increasingly better behavior that reflects their authentic inner goodness, while militant Muslims aren't expected to have psychological depth beyond what they announce publicly and are in this respect unreformable?

A related criticism is that he's unclear about the specific roles of reason and emotion in our moral judgments. Mentioning the work of philosopher Jonathan Haidt who has argued that moral decisions are usually made on closed-minded gut instinct rather than on open-minded careful reasoning, Harris insists that just because we're not always rational doesn't mean we shouldn't set rationality as a goal. But wasn't his mention of the MPFC's role in injecting emotion into moral decisions meant to imply that emotion is an inseparable part of moral decision-making? "Of course," he says, "it is now well known that our feeling of reasoning objectively is often illusory. This does not mean, however, that we cannot learn to reason more effectively, pay greater attention to evidence, and grow more mindful of the ever-present possibility of error." One wonders whether he would be willing to apply this charge of an "illusory feeling of reasoning" to his own "intuition about how the world works" as mentioned above.

Skepticism of any kind is a slippery slope to radical skepticism. Perhaps we should be radical skeptics about many of our moral intuitions, but Harris doesn't let us know exactly when and why he jumps off the skeptic's bandwagon. Alternatively, rather than a problem of skepticism, it could be a problem of subjectivity that confronts all scientists, and perhaps especially those scientists that study the human mind. How can a person use his or her own subjectivity to study other subjective beings and call it an objective process?

The conflation of reason and emotion, and of the metacognition that chooses which method of judgment to apply in a given circumstance, also plagued Jonah Lehrer's recent book on the neuroscience of decision-making. In How We Decide, Lehrer claims that reason and emotion are interrelated phenomena, and he prods this blurry area without defining (or successfully redefining) these terms. Lehrer's book is not listed in Harris's prodigious bibliography. Might Harris have avoided making the same error had he studied Lehrer's presentation?

Overall, this is a provocative book worth reading, especially for those with prior exposure to these popular topics in philosophy and neuroscience. Harris sticks to the narrow path and, for the most part, avoids wading into the treacherous waters of bitter diatribes against religion. He effectively demolishes the allegation of a metaphysical basis for morality that is defended by most religions, not by making sport of religious traditions and ancient texts, but simply by offering a substantive, nuanced, functional, science-based alternative.

---

If you'd like to learn more about my work, I've published books. Also, I write for Medium. There, readers with a paid membership don't have to worry about the paywall.

Is v. Ought: A related paragraph based on something I wrote in 2008 and rediscovered in 2025

We form elaborate, rationalized opinions (we call them moral judgments) about how we should respond to certain facts or guide our behavior. These are sometimes shared and codified. If they encompass a wide scope, we may call them moral theories.

The moral opinion is not inherent to the fact, and so we can’t discover what we ought to do. Instead, we create our moral opinions. We may start with a fact and then develop an opinion about it, but we are not discovering our opinion inside the fact.

There’s a gap between “is” (facts) and “ought” (opinion), but it isn’t the kind of gap that can be bridged. Only the facts are moored to one side of the river. The opinions are not moored in the same way; though it may be a fact that we believe them or have asserted them, the opinions themselves do not represent facts. In this metaphor, the facts are a Point A on the riverbank, but while the opinions are on the other side of the river, they aren’t a Point B. The bridge can’t be built because there’s nothing to which to bridge.