In January of 2014 Daniel Dennett, a philosopher and cognitive scientist who is renowned as a world’s leading authority on free will, wrote a lengthy critical review of Sam Harris’s book Free Will (Reflections on Free Will: A Review by Daniel C. Dennett). Sam Harris, a neuroscientist, then wrote a lengthy and testy reply in February (The Marionette’s Lament: A Response to Daniel Dennett).

Some of Dennett’s criticisms are, though still necessary and valuable, not of major significance. I will not summarize those points here. Rather, I will focus on the heart of contention between Dennett and Harris and summarize those arguments that are most significant to that. Notably, on both points Dennett extensively cites experts and published work in the subject, something Harris much more rarely does. In fact in this respect, and in their actual statements, Dennett explicitly takes philosophy as a field of study far more seriously than Harris does. And consequently, I believe, Harris fails to ever become a good philosopher. Because you can’t ever “be good” at anything you hold in such contempt that you refuse even to study it.

But back to the main point…

Dennett’s overall thesis is clear:

[S]ome of us have long recognized [what Harris says] and gone on to adopt more reasonable, more empirically sound, models of decision and thought, and we think we can articulate and defend a more sophisticated model of free will that is not only consistent with neuroscience and introspection but also grounds a (modified, toned-down, non-Absolute) variety of responsibility that justifies both praise and blame, reward and punishment. We don’t think this variety of free will is an illusion at all, but rather a robust feature of our psychology and a reliable part of the foundations of morality, law and society. Harris, we think, is throwing out the baby with the bathwater.

In fact, he finds Harris’s book “a veritable museum of mistakes, none of them new.” Dennett notes that Harris is in camp with a minority view among actual experts. He points out that over 59% of leading experts concur with compatibalism, the view Harris inexplicably rejects. That is actually a poll of top philosophy professors. And in fact only 12% of that group side with Harris (“accept or lean toward no free will”). Of the rest, 7% take indecipherable views of the matter, 9% are undecided or said they don’t know how to answer, and the other 14% still believe in libertarian free will (and those are mostly Christians). That puts Harris in some extremely minority company. More philosophers believe libertarian free will exists, than share Harris’ view that no free will exists at all!

So how has Harris gone so wrong?

Harris Does Not Deal Responsibly with Folk Beliefs

Besides Dennett’s point about the actual consensus of experts (which Harris ignores), Dennett shows that studies have determined that, contrary to Harris’s armchair assertion to the contrary, large majorities of ordinary people accept a compatibilist view of free will, especially when that option is made available to them in questioning. He also points out that it wouldn’t matter if folk notions of free will were incorrect, because that’s commonly the case in science, and that doesn’t eliminate the concepts that common people incorrectly understand, it corrects them.

Hence, as Dennett says, “when social scientists talk about beliefs or desires and cognitive neuroscientists talk about attention and memory they are deliberately using cleaned-up, demystified substitutes for the folk concepts,” rather than declaring that beliefs and desires and attention and memory “don’t exist” simply because common folk understand and define them wrong. No, they exist. And our job is to explain in what way they exist, and thus correct the common misconceptions about them. Not to “agree” with a common people’s error and then tell them the thing they got wrong doesn’t exist at all.

Ironically, except when he doesn’t (see below), Harris fully admits that the “folk” belief about consciousness is false—that consciousness is not a dualistic Cartesian observer external to the brain—yet he admits this without then concluding we should tell people consciousness doesn’t exist. And he even uses this to try and argue against compatibilism (in a convoluted way shown by Dennett). Why is this ironic? Because Harris completely fails to do this with free will. When it’s “free will” and not “consciousness,” Harris inexplicably does exactly the opposite.

As Dennett explains, all Dennett is doing is replacing a “folk” belief about free will with a scientifically correct model of it, exactly what Harris wants to do with consciousness. Yet for some reason Harris switches scripts and insists we should deny free will exists at all, and that trying to replace an erroneous folk version of it with a scientifically accurate one is playing games and not to be tolerated. If Harris actually thought that was a sound approach, then he should follow his own advice: he should do the same with consciousness (and “memory” and “belief” and “love” and so on), and insist we should deny consciousness exists at all (and likewise “memory” and “belief” and “love”), and that trying to replace an erroneous folk version of it with a scientifically accurate one is “playing games” and “not to be tolerated.” Harris won’t even listen to his own advice. Maybe because if he tried to follow his own advice for consciousness and everything else, he’d realize how stupid his advice was. Why he doesn’t get this is bewildering.

Harris is Looking at the Wrong Component

Dennett’s second major argument is that Harris is looking at the wrong thing. In effect, it is as if Harris wants to write a book about how the accelerator on an old car works, but spends the whole time looking at the carburetor, concludes a carburetor can’t accelerate a car, and therefore that cars don’t have “the ability to accelerate.” Dennett corrects this mistake: free will is a discourse about ability, not about gaps in causation; so we shouldn’t be looking at the carburetor (or ‘for gaps in causation’), we should be looking at the acceleration train (the thing that actually causes the decisions we want to evaluate). As Dennett says, the question of whether a car ‘could have gone 40mph rather than 60mph’ is not asking whether a car can defy the laws of physics and go 40 when the accelerator is depressed to 60, but rather whether the car simply has the ability to go 40 instead of 60. Yes, to do that, the accelerator must be positioned differently, thus the state of the universe has to be different. But that’s the point. That’s precisely the thing we want to know if we want to control our car.

Thus, in fact, what we want to know is whether someone is of a bad or good character (and what it takes to get someone to move from one to the other). That’s the ability we want to know (the ability to be good or bad, go 40 or 60). And an action that is freely caused by that ability (and not interfered with, e.g. by being pushed or coerced) is for that very reason an effect of it, and therefore observable evidence that that ability exists. Just as we can test a car to see if it can go both 40 and 60, and just as the car’s accelerator needs to be free of interference for us to actually run that test (e.g. just letting it roll down a hill will not help us know what the car can do), so can we also test a person and see if they do something benevolent or malevolent, thus discovering whether they are a benevolent person or a malevolent one. This test requires determinism, not an escape from it. This also requires the tested person to have free will, as otherwise we won’t know whose will is being evinced (e.g. that of a coercer and not the test subject’s?) or whether their will is being evinced at all (e.g. if they were accidentally pushed, then it wasn’t even a choice they made).

Harris Relies on Folk Beliefs of Consciousness

Even though as a neuroscientist Sam Harris certainly knows better, Dennett points out that he nevertheless keeps slipping into a folk belief about how consciousness works whenever he wants to prove free will doesn’t exist. He must be doing this without realizing it. But he keeps doing it all the same. As Dennett says, “Harris shrinks the me to a dimensionless point, ‘the witness’ who is stuck in the Cartesian Theater awaiting the decisions made elsewhere. That is simply a bad theory of consciousness.” Indeed.

Harris inexplicably ignores the fact that the “me” is the whole apparatus (the actual engine receiving and processing the data and calculating the outcomes and making the decision) and sometimes mistakenly claims the “me” is some sort of external observer to all this, that the “me” is just the conscious awareness of what I did, and not the actual thing (the brain) that is making the decisions. That’s pseudoscience. Your consciousness is not you. Your consciousness is just a state of awareness of you. It is a causal product of you. It is in that respect a useful part of you. But it is in no way a separate entity or the whole of your identity. That’s why it is absurd to say the conscious state of making a decision didn’t make the decision therefore “you” didn’t make the decision. To the contrary, you are the thing (the brain) making the decision, of which your conscious state is simply an output report of what you did.

Once this mistake is corrected, as Dennett points out, a lot of Harris’s reasoning against free will dissolves. When you add in the other correction (looking at our actual abilities to do things rather than looking for a useless ability to violate causal laws of physics), we discover that, as Dennett points out, “We can improve our self-control, and this is a morally significant fact about the competence of normal adults.” And that requires free will (people have to be free of coercion, deception, etc., to do this; they, as a person, have to be in control of their own will). That we can knowingly choose to take steps to intentionally become more self-actualized, and more aware of and thus more in control of how we make decisions, demonstrates that all of Harris’s handwaving about our lacking such abilities is demonstrably false. Moreover, we can evaluate someone’s moral reliability based on how much or little they do this when it comes to developing themselves as moral agents. We can’t evaluate that if their free will is meddled with (by pushes, coercers, electrodes, cages, deceptions). Thus the importance of ensuring they acted freely in the case we wish to evaluate. Which requires “acting freely” to actually be a thing, not an illusion. And in no way does this require violating causal determinism.

At one point Dennett puts this succinctly: “Infants don’t have free will; normal adults do.” That Harris cannot explain this fact demonstrates that Harris is not working with a scientifically useful definition of free will.

Harris Has No Coherent Sense of Self-Influence

Harris says we can’t choose our desires. Because we can only do so under the influence of some desire or other that we don’t choose. But that is a self-contradiction: admitting that “we can do so” while denying that we ever can. The role of unchosen desires as motivators is being used too confusingly here for Harris to say anything useful about how we actually control our own behaviors and desires. And scientists should care about how we actually do things.

As Dennett says:

We use the same tools to influence our own desires as we use to influence other people’s desires. I doubt that [Harris is] denying that we ever influence other people’s desires. His book is apparently an attempt to influence the beliefs and desires of his readers, and it seems to have worked rather better than I would like. His book also seems to have influenced his own beliefs and desires: writing it has blinded him to alternatives that he really ought to have considered. So his obliviousness is something for which he himself is partly responsible, having labored to create a mindset that sees compatibilism as deliberately obtuse.

All correct. We do not need absolute freedom to have some measure of freedom. Nor do we need absolute control to have some measure of control. Harris seems fond of black-and-white thinking like this in other areas, too. It blinds him to the need to deal with the fact that we aren’t dealing with absolutes, but degrees. The question is simply not “Are we free?” The question is “How free are we, and when?” The latter is a question of vital philosophical and scientific—and legal and political—importance to answer. The other question is a waste of everyone’s time.

As Dennett says, “If [Harris] can be the author of his book, then he can be the author of his thoughts.” Semantics for the win.

Harris’s Attempt at a Reply

In his response, after 200 words of introduction, Sam Harris first burns 600 words complaining about Dennett being mean to him, treating criticism as an affront to his dignity that requires elaborating on for some reason. This is not a good start. It is usually a red flag for not having an actual defense. It’s the sign of a hack: If you can’t rebut content, complain about tone. Harris then burns 300 words attacking the weakest and most irrelevant of Dennett’s points (the effects actual indeterminacy could have), one which I didn’t even mention, and which a mere three sentences could have dispatched by in fact conceding the point—which is actually what Harris does. Though he again frames it as a defense of his dignity against an unjust attack rather than simply saying they agree and moving on. And since in fact they agree on the disputed point, there is no actual rebuttal to it.

Then Harris says Dennett failed to prove any of his points, and demonstrates this by completely failing to correctly understand any of Dennett’s points.

Harris does this most amusingly by ignoring Dennett’s strongest analogies (e.g. how we respond to mistaken folk beliefs about “consciousness” should be the same way we respond to mistaken folk beliefs about “free will”; or that “how we could have done otherwise” should be answered the same way as “how a car could have done otherwise”) and instead only complains about the weakest ones (spending 500 words on this). For example, Dennett’s “geocentrism corrected to heliocentrism” case fails, Harris says, because the sun exists, free will doesn’t. Of course, Harris doesn’t notice that he is begging the question, assuming the thing doesn’t exist that Dennett is telling him does (indignant reliances on circular argument being another red flag for a bad philosopher). But more importantly, Harris completely failed to get the analogy. He doesn’t realize that sun-rises exist, even though the sun doesn’t move, and it was sun-rises that Dennett was using as the analog to free will, not “the sun.” Which exists in both geocentrism and heliocentrism, and thus Harris should have seen the existence of the sun cannot have been the analogy Dennett intended…

Yeah. Harris is a really bad philosopher.

Not a promising start.

Harris then burns 600 words repeating his position, without adding any new arguments or responding to Dennett, before “Changing the Subject,” a section comprising 3500 words (compared to the previous section’s 2300), mostly consisting of Harris just rephrasing all the same claims of his book (some of which Dennett rebutted, others Dennet was wholly unconcerned about).

Wherein Harris doesn’t answer any of Dennett’s fatal four points:

  • That Harris is being inconsistent (by treating free will differently than consciousness and every other mental capacity, vis-a-vis how we respond to erroneous folk beliefs about them).
  • That Harris is looking at the wrong thing (hunting for an ability to defy causation, rather than the ability to make informed and reasoned decisions reflective of and thus informative of our character and intentions, which are the decisions we actually want to be able to make and judge).
  • That Harris relies on folk psychology about consciousness to build his case against free will (even though Harris knows that version of consciousness is false).
  • And that Harris has no usable model of self-influence and control (he is more concerned to deny free will exists than to explain when we have it and how much of it we have and what we can do with it and how we can get more of it).

That constitutes a decisive defeat of Harris by the rules of debate. Harris dropped every crucial point Dennett made, failing to respond to any of them. Dennett wins by default.

Missing Every Point

Instead, Harris tries to answer Dennett’s point about research showing people do intuit compatibilism when they are given the option to by claiming this same “research suggests that people find the idea of libertarian free will so intuitively compelling that it is very difficult to get them to think clearly about determinism,” completely missing all three of Dennett’s points: first, that that isn’t true of a lot of people (Harris is thus falsely generalizing; another mark of a bad philosopher), that it has to do with the way questions are asked and what options are given (Harris is thus ignoring crucial nuance; another mark of a bad philosopher), and that even when it’s true it’s just as true of consciousness and memory and belief and so on, yet inexplicably Harris does not treat free will the same way.

For example, Harris would certainly admit that “research suggests that people find the idea of dualistic Cartesian consciousness so intuitively compelling that it is very difficult to get them to think clearly about neuromechanics,” yet he would not then say “therefore consciousness doesn’t exist,” not even at all, much less simply because people’s intuitions go against the correct facts of the matter. Harris makes no rebuttal to this point. He doesn’t even seem to be aware that Dennett made these points. Yet he did. In detail. Harris stalwartly maintaining his inconsistency on this point remains inexplicable and bizarre.

Similarly, at no point does Harris describe a usable model of self-influence and control (even though Dennett has, and it accords with the scientific facts, and abundant empirical observations beyond). Thus Harris’s denial of free will is a useless dead end; whereas Dennett’s reconstructed model of free will (like, Harris admits, our reconstructed model of consciousness) is extremely useful and highly applicable. Nor does Harris ever seem aware of how we actually use free will in practice (such as in law, medical consent decisions, or our judging of others’ consent, character, and benefit or danger to us). He keeps looking for the wrong thing. Not the actual thing people use free will to look for in the actual real world. Harris is divorced from reality. Dennett is deeply concerned with reality.

Only with regard to Dennett catching Harris relying on a false folk psychology of consciousness to build arguments against free will does Harris attempt anything like a reply. But in doing so he completely misses any actual point Dennett made. Harris argues that these folk beliefs really exist and cause social problems and really are worth debunking—all of which Dennett conceded in his critique. So Harris is not even responding to Dennett here. What Dennett caught Harris doing was using the false folk Cartesian model of consciousness as if it were a fact to demonstrate that, because it is incompatible with free will, therefore free will doesn’t exist. That’s incoherent. Dennett said so. Harris made no reply.

The Point Being Missed

Harris does not seem to understand why Dennett is right about this: if in actual fact consciousness is a whole-brain phenomenon and not (as the folk belief has it) a Cartesian theater separate from the brain, then Harris cannot argue that because free will can’t exist on the Cartesian theater model, free will can’t exist. That’s to confuse falsehood with reality. The reality is, as Dennett keeps explaining to Harris, that consciousness—and thus identity—is a whole-brain product. Therefore, that free will can’t exist on the Cartesian theater model tells us nothing whatsoever about whether free will exists. Because the Cartesian theater model is false. As even Harris himself repeatedly admits.

What we want to know is whether free will is compatible with the whole-brain model. Because that’s what exists. And as Dennett shows, the answer is yes. Thus Harris can’t say that “we” are the conscious experience of ourselves (which is false), the conscious experience of ourselves isn’t making the decisions we are conscious of (which is true), therefore “we” aren’t making those decisions. That is logically valid, but unsound, because “we” are not the conscious experience of ourselves. That’s what it means to say the Cartesian theater model of consciousness is false: that “we” are not the conscious experience of ourselves; “we” are the brain that those conscious states are giving us feedback on.

But since the brain is making the decisions we are conscious of, and “we” are that brain, it follows that “we” are making the decisions we are conscious of. And that freedom to make our own choices can be interfered with, or bypassed, or employed more, or employed less. And so we do have more or less freedom. And you can’t have more or less of something if it doesn’t exist. So clearly it does. Causal determinism is true, but irrelevant. Obviously decisions are caused by information and circumstances and prior states of the person (their born and learned character and reasoning, their acquired skills and knowledge). Irrelevant. Yes, you are constrained by where you were born, what education you had, what you’ve been told or allowed to know. But you also have a lot of freedom within that constraint—and even, often, to escape that constraint, and enlarge your prison, expanding the space you can move in. More and more. That power clearly exists. Because it alone explains the whole of human science, literature, and civilization. We are the Robot’s Rebellion.

Because your internal processor, which evaluates what you do know and what’s best to do about it, is not someone else’s marionette. It can evaluate what’s going on from the top down, and make rational, informed decisions. It can change course, precisely to the extent that you use it to leverage the benefits of causal determinism—through the causally fixed effects of logic and evidence. It’s only really a slave to the facts you know and the desires you have. Everything else is escapable. Which means when no one and nothing else gets in the way of your own internal processor’s cogitation, it only ever does what you want it to do. Which is all you’d ever want it to do. And insofar as you’d want it to perform better, e.g. evade the manipulations of others or the limitations of your environment, violating causal determinism wouldn’t help. You need determinism, so you can reliably and causally interact with the world and yourself, to acquire more accurate knowledge of the world (and hence of your constraints), and more accurate reflection on your own desires (and hence of what you really want out of life).

As Dennett explains in well-sourced detail, this means we can have a useful model of personal autonomy, consent, responsibility, self-reflection, self-control, self-influence, and rational and informed decision-making. Which means we can have a useful model of free will. Because those are the things people in practice mean by free will. That they are sometimes confused about how this works is as irrelevant as the fact that they are sometimes confused about how consciousness works, or how memory works. Our job should not be to confuse them more by telling them that because their intuitions about consciousness, memory, and free will are wrong, therefore “consciousness, memory, and free will” don’t exist. To the contrary, our job should be to correct their misunderstanding of the facts and give them a factually correct model of what consciousness, memory, and free will are, and how they really work. Because determinism itself gives us plenty of Elbow Room.

Closing Assessment

Compatibilism is the strongest consensus view of actual experts. It’s coherent, scientifically and empirically verified, and useful. It entails that causal determinism has no effect on the existence or absence, or even degree, of human autonomy and self-control. Sam Harris has no coherent, verified, or useful alternative to propose. His model of free will is archaic pseudoscience, a Frankenstein’s concoction of the ivory tower, and not a real world phenomenon. Of course it doesn’t exist. Yet he has no alternative on offer. Dennett does. And Harris is too arrogant to even notice. He lost this debate. Catastrophically. Which only illustrates yet again why Daniel Dennett is the real expert here.

In fact, it becomes clear in all of this that Harris is a really, really, really terrible philosopher. He should probably just get out of the philosophy business, and leave it to folks who actually respect the field, and know what they’re doing. But this also reflects a perfect example of how a lousy armchair philosopher—like, in all honesty, most people—can get the concept and application of free will so wrong. Circular arguments, contradictory assertions, ignoring facts and logic, relying on pseudoscience, and being too stubborn to learn or listen. The example set by Harris in this whole debate is practically its own textbook for how not to do philosophy.

For further discussion on this debate, see my Neat Podcast Discussion on Free Will: Cameron Reilly for Sam Harris and I for Daniel Dennett.

-:-

I need to add that I do think it’s a shame. Because Harris often has some interesting or innovative ideas. He just won’t learn how to articulate or defend them well. See, for example, how I find he dropped the ball on what I think was otherwise a brilliant model of moral reasoning. Time and again he does this. There is a lot of crap in academic philosophy he is right to critique. But he still needs to take it more seriously than he does. They’ve already debated most of everything he ever aims to defend. He needs to learn from that. Not ignore it. Otherwise, he just ends up on the other end of an expert’s bludgeon. In this case, Dennett’s. Maybe some day Harris will realize all the mistakes he made here, and how he may be making them elsewhere too. The outcome will be marvelous.

§

To comment use the Add Comment field at bottom, or click the Reply box next to (or the nearest one above) any comment. See Comments & Moderation Policy for standards and expectations.

Discover more from Richard Carrier Blogs

Subscribe now to keep reading and get access to the full archive.

Continue reading