A disclosure: my own thinking about free will was strongly influenced decades ago by Dan Dennett’s defense of compatibilism, Elbow Room. Like almost 60% of professional philosophers, I find myself inclined towards some such view. For those unfamiliar with this expression, the basic idea is to dissolve the conflict between free will and determinism by reconceiving free will as not really in conflict with determinism. Once this view is on the table, we seem to have three menu items instead of two: those who believe that free will is incompatible with determinism and thus end up committed to the existence of free will on some non-natural basis; those who believe that free will is incompatible with determinism and, rejecting any possibility of a non-natural basis for free will, end up determinists who reject or want to radically reform our moral practices; and those who don’t think they are incompatible on closer examination. This lay of the land is well-known to philosophy people and has been for a very long time.
The only reason why I’m writing at all, is because of qualifications I would want to make to Dennett’s otherwise essentially correct, if exceedingly civil response to Harris here, and make some meta-comments about the discussion itself. First as to Harris: Harris is notable for dichotomizing the world into things that are pro-religion and anti-religion, in order to give the pathos of his opposition the widest possible scope. Despite the predominantly secular character of contemporary analytic philosophy, Harris has made an antecedent judgment that professional philosophers are basically indistinguishable from theologians because they don’t simplify, heighten contrasts, cut to the chase, demonize opponents, etc. as he does. This is to mistake self-possession and intellectual curiosity, which he largely lacks, for treason. Thus it seems obvious to him that a view which is not that far from his own must in fact be its opposite. He wants to get those who refuse to be pinned down by him because they think they are being confronted with a false alternative to shut up and embrace the false alternative, and take sides as his ally or enemy–either will do, as either are the source of some satisfaction. Thus were culture warriors always, regardless of their cause. By contrast, Dennett is a professional philosopher, but also something of a popularizer, and from the beginning, his work on free will was driven by a desire to understand and clarify.
The crux of their debate, really, is over an empirical question: what do ordinary people think? Dennett claims that ordinary people are implicit compatibilists (and, unlike Harris, marshals some actual empirical evidence to support this). Harris thinks that it’s obvious that ordinary people are incompatibilist indeterminists, his primary evidence for this being that people seem to say they are. Harris could avail himself of further evidence: that people make moral judgments, which one could argue implies a commitment to incompatibilist indeterminism, but for the fact that he too wishes to make moral judgments (boy does he!) and hopes to save that practice from criticism (just like a compatibilist–this appears to be Dennett’s main evidence for thinking that Harris is a compatibilist despite himself).
Two related things seem to have gone off the rails in this debate. First is the assumption that anyone has a view clear enough and explicit enough to be pinned down and criticized, whether we are talking about Harris himself, or the ordinary folk he hopes to Enlighten. The second is that what we need here is The Truth in anything like the sense that natural science furnishes truths. I think that Dennett largely gets this latter point, but unfortunately his pragmatist clarity is obscured by his residual desire to figure out what people really think, and what is correct to think. At the risk of being pegged a nihilist for being a more thorough-going pragmatist, I don’t think there is a fact of the matter as to what people really think in this area, and that there is no need to get people to think “correctly” about it. What philosophy can do here, at most, is provide clarity, or, if you prefer, alleviate an intellectual cramp induced by a certain illusion. But what it cannot and need not do is furnish “truth.”
Here’s what we know: we are animals, and presumably as animals we have no non-natural parts. This far, Dennett, Harris and I agree. We are not unusual in being social animals who seek to influence each other by various forms of encouraging and discouraging behavior. What is unusual about us is the degree of intelligence and foresight we display, and the fact that our sociality is pervasively shot through with linguistic activity. Consider the following episode.
“Which of you took the cookies? They’re almost all gone!” asked Mom of two children assembled. A long silence ensues, until one of them, with a hang-dog expression, replies “I did. I’m sorry. I know you told us not to.” “Go to your room,” Mom replies, “no cookies for you for the rest of the day.”
The whole discussion between Dennett and Harris begins from the assumption that everything in that tableaux is a physical system and thus deterministic. So far so good. So what’s the problem? Really, from a pragmatic perspective, there isn’t one yet, not at all. Mom’s doing her job, and let’s assume she’s doing it well. Mom only directs her attention towards what Skinner called “operant behavior,” what a physician would call “voluntary movement” and what the folks call “choices.” Classifying one set of bodily movements as falling under one of these labels, and distinguishing it from reflex movements, epileptic seizures, the effects of external jostling and the like, seems to carve nature at the joints. I have no doubt that operant behavior has a distinctive neurological signature that shows up on an MRI (though it may be the case that finer-grainedly described kinds of operant behavior such as “ordering a milkshake” may not have their own distinctive neurological signatures).
The crucial point is, everything is in order as it is. Mom’s task is not impaired by a fundamentally wrong or confused way of classifying bodily movements. She’s not raising her voice at epileptic seizures or reflexes. She seems to be doing just fine, and in the fullness of time, so will her kids, who will not only learn to be able to avoid cookies, but engage in some “herding” behavior of their own, not only of their own children, but of their peers as well.
Now Dennett and Harris have views about what, metaphysically, Mom believes. Dennett especially should appreciate this point, but I think Mom doesn’t believe a damn thing. She believes in the same way that a person, when pressed, “desires” what they ordered on the menu at a restaurant (this is Dennett’s own example, I don’t recall from where). In this case the pressing comes from a philosopher to pick from a menu of philosophers’ characterizations of what she thinks. Philosophers have been doing this since Socrates, and it always struck me as unfair to foist an interpretation on someone, get them to endorse it, and then attack them for holding it. Undoubtedly the second book of Plato’s Republic would read differently if Thrasymachus had written it.
Dennett wants to tell Harris that it is far from clear that Mom is committed to the Cartesianism Harris wants to foist on her the better to lump her in with the sermonizers he reviles; at most she is committed to the concepts necessary for certain action-ascription behaviors. I sympathize. Harris, by contrast, finds it unlikely that Mom has as complicated and sophisticated a view as Dennett attributes to her. But I also sympathize. Mom acts like a compatibilist not because she’s committed to it but because it describes her behavior. But Mom could probably be induced to assent to Cartesianism if pressed in the right way, which would tell us exactly nothing about what she really thinks and what, if anything, about that needs to change.
The better question is not what’s going on with Mom’s linguistic behavior, but rather, what’s going on with Harris’ and Dennett’s. Harris wants to flog Cartesians. Fine. They exist, and he’s probably right to think that they exist in certain religious communities especially, because religion, like philosophy, provides interpretations of what we do, including things we’re perfectly capable of doing without any interpretation at all. But to suppose that Mom is an implicit Cartesian, and to take as evidence of this that she would agree to a menu item at a metaphysical club, that you could induce her to say “then I guess I’m a Cartesian” tells us little about what she thinks, and tells us nothing about the value of her behavior in the cookie scenario, because her behavior there is not itself a truth claim in metaphysics. Unless Harris wants Mom to radically reform how she raises her children (and I don’t think he does, assuming that she is effective and humane) then it seems what he wants to do here is teach her how to say “I am not a Cartesian” before anyone else can get to her and teach her how to say “I am a Cartesian.” Seen from afar and with a pragmatist lens, this all seems rather silly.
So I’m with Dennett still? Not quite. Because I still think Dennett wants to tell us the truth about the will, and I don’t think there is a truth about the will to be told. I think there is a more helpful way to think about this, and that it is very close to what Dennett is doing in Elbow Room and Freedom Evolves. What is the free will-determinism problem at bottom? Is it a metaphysical question? I would submit that it is, instead, a funny kind of practical problem. The problem has the following contours. Among the things we do is reinforce behavior in accordance with rules, “punishing” “bad choices” as we call one form of it. But sometimes, following a kind of exception clause, we do not. The “bad choice” is “excused.” The Martian zoologist would say that it is pointless to lecture people for bodily movements caused by neurological disease, for example. The problem is that the language we use in these contexts sounds like “he couldn’t have helped it” and the like. And it is precisely this language that the incompatibilist determinist misuses, to act as if he is trying to induce the no-reinforcement response to behaviors which do not fall under the rule governing excused behavior. Sometimes he is quick to add that he isn’t trying to actually change our behavior, other times he insists that he is. The problem is that the blaming and excusing practices work just fine as they are if you view them not as the expression of theories but patterns of activities, and simply refuse to give them any interpretation of interest to a philosopher at all.
What Dennett and other compatibilists want to do, ultimately, is to restore the status quo ante of smooth social action before it was disrupted by the confusing effects of Cartesian and anti-Cartesian interpretations. But they think the best way to do this is through a theory of their own. If the goal is to participate in social conversation among philosophers, that’s fine, but that is not what lends interest to the topic. The better goal is to nullify the disruptive effects of the other philosophical interpretations, by whatever means available. For example, every time someone started to engage in philosophical interpretation, you could wrap their knuckles with a stick, like some Zen master trying to provoke inner silence. That might work. Or you could carefully retrace the confused footsteps of analogy and emotion that led to the interpretations, until the other sees the under-motivated and unhelpful character of that path, in something like the spirit of the later Wittgenstein, who insisted that he held no views at all.
Free will/determinism is what I call a “puzzle,” but it is unusual as a puzzle because it links to our moral practices and in particular confuses us about our excusing practices. Harris’ response to it is to enlist it into his culture war on behalf of secularism, by uncharitably attributing a certain interpretation to most people, the better to attack them. Dennett’s response is to charitably attribute a more plausible interpretation to most people, the better to justify leaving them alone. I propose instead that we leave them alone by not attributing interpretations to them at all, for the most charitable thing one can do in the face of the panoply of human conduct is to say nothing.