People can and have taken strong issue with the view, favored by Sam Harris and me, that free will as traditionally conceived is a complete illusion.  Many people have responded with diverse versions of “compatibilism”—the view that determinism is still compatible with some notion of free will. (I happily note, though, that almost nobody questions determinism itself, though some have urged us to keep it quiet lest it rile up the hoi polloi).

And the discussion, which I think has been fantastic  and instructive, has been almost completely free of invective or ad hominem arguments.  That is, until last Monday, when science journalist John Horgan published a nasty and mean-spirited attack on Harris at Cross Check, his Scientific American blog.  The post, “Will this post make Sam Harris change his mind about free will?“, is ugly and almost incoherent.  Horgan starts off with a few punches to the solar plexus:

But Harris keeps intruding on my thoughts, in part because he keeps emailing me about his writings, and especially his new book Free Will (Free Press, 2012). Also, I admit to a certain voyeuristic fascination with Harris. I wonder, what crazy idea is he going to peddle next? Some of his righteous rants give me a perverse pleasure. I’m simultaneously irritated and titillated. I get the same feeling listening to Rush Limbaugh or Rick Santorum.

First of all, Sam did not “keep emailing” Horgan about his writings, a statement that implies that he’s beleaguering Horgan with unsolicited emails.  In fact, as Sam has verified, Horgan signed up to receive Harris’s email announcements, which are almost always about Sam’s new posts.  Horgan is in fact one of 45,000 people who get these announcements—voluntarily.

Although I try not to derive psychological motivations for rants like Horgan’s, he makes it easy because he gives them at the outset:

Harris’s new book rates orders of magnitude higher on Amazon’s Best Sellers lists than my new book, The End of War (McSweeney’s, 2012), which concludes with a chapter called “In Defense of Free Will.” That rankles.

Indeed, Horgan tried to get me to publicize his book (now #27,446 on Amazon), which seems to have a message of increasing peace on earth similar to that of Pinker’s Better Angels of Our Nature (#1,426), but for some reason I wasn’t motivated. Now I’m glad I wasn’t.

Insofar as Horgan tries to argue for free will, he fails. He says that he “loathes” the philosophy of determinism, and argues that “free will” (which he never defines) involves making real choices that are not determined but “constrained”:

Harris asks us to consider the case of a serial killer. “Imagine this murderer is discovered to have a brain tumor in the appropriate spot in his brain that could explain his violent impulses. That is obviously exculpatory. We view him as a victim of his biology, and our moral intuitions shift automatically. But I would argue that a brain tumor is just a special case of physical events giving rise to thoughts and actions, and if we fully understood the neurophysiology of any murderer’s brain, that would be as exculpatory as finding a tumor in it.”

Harris seems to be advancing a reductio ad absurdum, except that he wants us to accept the absurdum: there is no fundamental difference between me and a man compelled to kill by a brain tumor. Or between me and someone who can’t help washing his hands every 20 minutes, or someone who’s schizophrenic, or a babbling baby, or a newt, or a worm. I mean, if I’m not different from a guy who kills because a tumor provokes him into murderous rages, how am I different from anyone or anything with a brain, no matter how damaged or tiny?

Horgan’s mistake here is to assume that “normal” behavior is less physically determined than behavior mandated by obsessive-compulsive disorder or brain tumors that cause aggression.  But even Whitman had a choice: he could presumably have controlled his murderous rages, and obsessive-compulsive disorder can sometimes be cured if the sufferer seeks help. In that sense there are still apparent (but not real) “choices”.  But the real point is that all pf our behavior stems from our neurological conditions, be it Horgan having fries instead of chips or Whitman picking off students with a rifle in Austin.  In that way there is no fundamental difference in the degree of determinism of our behavior.   So how is Horgan different from Whitman?

Here’s the difference. The man with a tumor has no choice but to do what he does. I do have choices, which I make all the time. Yes, my choices are constrained, by the laws of physics, my genetic inheritance, upbringing and education, the social, cultural, political, and intellectual context of my existence. And as Harris keeps pointing out, I didn’t choose to be born into this universe, to my parents, in this nation, at this time. I don’t choose to grow old and die.

But just because my choices are limited doesn’t mean they don’t exist. Just because I don’t have absolute freedom doesn’t mean I have no freedom at all. Saying that free will doesn’t exist because it isn’t absolutely free is like saying truth doesn’t exist because we can’t achieve absolute, perfect knowledge.

To me this is incoherent.  “I do have choices, which I make all the time” says nothing different from what Whitman did, especially because Horgan avers that his choices are indeed “constrained” by things like the laws of physics.  But what does he mean by “constrained”?  Is there some real freedom beyond physics, genetics, and environment? Does “constrained” equal “determined”? If so, then Horgan is well and truly a determinist.  If not, then he’s a dualist.

Indeed, I think Horgan is a dualist, perhaps conditioned by his “loathing” for determinism.  Here’s a sign of the ghost in Horgan’s machine:

But the strange and wonderful thing about all organisms, and especially our species, is that mechanistic physical processes somehow give rise to phenomena that are not reducible to or determined by those physical processes. Human brains, in particular, generate human minds, which while subject to physical laws are influenced by non-physical factors, including ideas produced by other minds. These ideas may cause us to change our minds and make decisions that alter the trajectory of our world.

I find it amazing that a science writer can say this.  Yes, there are emergent phenomena (like the “wetness” of water) not predictable from a more reductionist analysis, but those phenomena are indeed determined by lower-level phenomena, with the laws of physics at the bottom. Higher-level processes might also not be predictable by human endeavor (chaos theory is an example), but they are still a) deterministic (absent quantum phenomena) and b) must be consistent with lower-level phenomena.  There is nothing we know about minds that implies that they’re not reducible to and determined by the physical processes in our brains.  And, of course, nobody—including incompatibilists like Sam and me—denies that minds can be changed by environmental influences (like the words I’m writing now), or that those changed minds can change the world. Indeed, later on in his piece Horgan claims that if Harris had changed his mind and actions, that would be real evidence for free will.  That’s balderdash.  Both compatibilists and incompatibilists agree that minds can be changed.

In the end, I conclude that Horgan really is a dualist: he thinks there’s something to our minds beyond the physical structure of our brains.  What finally convinced me is his penultimate paragraph:

We are physical creatures, but we are not just physical. We have free will because we are creatures of mind, meaning, ideas, not just matter. Harris perversely–willfully!–refuses to acknowledge this crushingly obvious and fundamental fact about us. He insists that because science cannot figure out the complex causality underpinning free will, it must be illusory. He is a throwback to the old behaviorists, who pretended that subjective, mental phenomena—because they are more difficult to observe and measure than planets and protons—don’t exist.

First, we consider free will illusory not because we can’t figure out how brains work to the tiniest neuron, but from both first principles (our brains are material and must obey physical laws) and the increasing evidence that our view of “agency” can be radically changed by neurological or psychological experiments.

And yes, we are just physical, for our mental phenomena—and that includes our so-called “choices,” and our “mind, meaning, and ideas”—are, and must be, the result of physical processes.  Those mental phenomena can differ only if the underlying physical substrates differ.  Yes, we can speak of “minds” and of “choices” as entities that are meaningful in human discourse, but in the end they all come down to neurons and molecules.  And “free will”—at least the contracausal form conceived of by many, including millions of religious people—is indeed an illusion.

Oh, and in his last paragraph (remember, this is on Scientific American) Horgan can’t resist a completely gratuitous slap at Sam’s intellect:

Dwelling on Harris depresses me. All that brainpower and training dedicated to promulgating such bad ideas!  He reminds me of one of the brightest students I’ve ever had, who was possessed by an adamant, unshakable belief in young-earth creationism. I did my best to change his mind, but I never succeeded. I probably won’t change the minds of Sam Harris and other hard-core determinists either, but it’s worth a shot.

Stay classy, John.  Ten to one Sam will respond to you (if he does) without anything like the invective you’ve heaped on him.

Alert reader Gregg has called my attention to a new study in neuroscience that overturns the Libet and Soon et al. studies by showing evidence for free will via brain scans.  Practical Ethics reports on the study, to appear in this week’s Science:

According to the authors of the study, previous neuroscientific studies have failed to detect free will because they were looking for causation in the wrong place, or at the wrong level. Most neuroscientific techniques are aimed at detecting patterns of activity at a physical level, whether macro-level, cellular, or atomic. For example, the common fMRI (functional magnetic resonance imaging) technique essentially measures differences in blood flow to various areas of the brain.  As a result, previous studies have only been able to detect the physical causes of our thoughts and actions. The group now publishing in Science has developed a new type of scanner called a Metaphysical Field Imager. Using functional metaphysical field imaging (fMFI), the researchers can detect energy patterns as they occur at sub-physical (i.e. metaphysical) levels. When superimposed over a map of the physical brain, fMFI is able to reveal the exact timing and location of flashes of free will in the brain, as people make decisions. The experimenters were able to show that, in their experiments, a flash of free will occurred in the prefrontal cortex immediately before a playing card was freely picked, strongly indicating that the free will there produced the relevant behaviour.

Here’s a figure from the paper—an fMFI scan showing the transitory signature of free will (the orange area of activity) in the prefrontal cortex of a human brain:

I’ve long espoused the idea that free will is a phantasm: a comforting fiction built into us, perhaps, by natural selection, but a good scientist knows when he’s licked.  I provisionally suspend my critiques of free will.  Indeed, the newest findings mandate not compatibilism, but true dualistic free will.

As I noted yesterday, the Chronicle of Higher Education has a special section in its latest issue devoted to the question, “Is free will an illusion?”  I’d like to give my brief (I hope!) reaction to each of the five essays (I don’t include mine).

Note first that every writer affirms that there is no dualism: our choices are purely products of a material brain. The ineluctable conclusion is that determinism reigns, so that contracausal free will is indeed an illusion. But many nevertheless try to save the concept of free will.  Now I know that compatibilism (the idea the free will is compatible with physical determinism) has a long history, even preceding modern neurobiology’s finding that there is no “spooky” soul that can direct our choices.

But many compatibilist philosophers try to save the “notion” of free will by separating it from its historical definition (“I could have chosen otherwise”).  Some of this effort is admirable, as when philosophers try to square determinism with how we think about “moral responsibility,” or how we might change our system of punishments and rewards in light of science.  But other thinkers seem to me to engage in rearguard philosophy—trying to save the notion that we are free agents, since that notion is so ingrained in us.  The latter tactic approaches the method of theology: the post facto rationalization of a comforting fiction that must be saved in the face of contrary scientific evidence.

Alfred R.Mele, “The case against the case against free will”.  Mele notes that he is director of a big Templeton Foundation initiative on free will. His first assertion is that, contrary to my claim, most people don’t have an intuitive view of free will that is dualistic; they don’t really think that their choices aren’t determined. His evidence is this:

But philosophers don’t own this expression. If anyone owns it, people in general do. So I conducted some simple studies.

In one, I invited participants to imagine a scenario in which scientists had proved that everything in the universe is physical and that what we refer to as a “mind” is actually a brain at work. In this scenario, a man sees a $20 bill fall from a stranger’s pocket, considers returning it, and decides to keep it. Asked whether he had free will when he made that decision, 73 percent answer yes. This study suggests that a majority of people do not see having a nonphysical mind or soul as a requirement for free will.

To me this doesn’t say much about whether most people view free will as contracausal.  They are told that “everything is physical” and that a mind is a “brain at work,” but it’s not obvious that they conclude from this that all actions are predetermined by physical processes.  Accepting the scientific scenario may just be something people do to go along with the experimenter; it’s not at all clear that they think deeply about what it means vis-à-vis their actions.  I am not convinced by this description of the study (n.b.: I haven’t read the study) that most people aren’t determinists. Drawing that conclusion, I’ve found, requires engaging in long conversations with people to explore their thoughts; I doubt it can be derived from a single quick question.

Mele then criticizes the experiments by Soon et. al. (and by extension that of Libet) showing that neurological scans can predict decisions before the decider is conscious of having decided.  He notes that the accuracy is only 60% (I think it’s up to 70%) in some experiments, and that he could do nearly as well with the 50% accuracy of a coin toss.  This is fatuous.  It’s early days yet, and we may be able to predict decisions in advance with substantially more accuracy when we know more about the brain.  He concludes that there is only a “slight unconscious bias toward a particular button.”  But that misses the point that the “bias” can be seen in the brain before the subject is conscious of it, and, as I recall, there were no biases towards pressing the left or right button, which renders Mele’s conclusion completely wrong.

Finally, Mele doesn’t define what he means by free will, which makes his essay useless.

Michael Gazzaniga: “Free will is an illusion, but you’re still responsible for your actions.“  I like this essay, probably because, by and large, Gazzaniga agrees with me. He’s absolutely insistent on determinism and thinks the concept of free will is incoherent:

Neuroscience reveals that the concept of free will is without meaning, just as John Locke suggested in the 17th century. Do robots have free will? Do ants have free will? Do chimps have free will? Is there really something in all of these machines that needs to be free, and if so, from what? Alas, just as we have learned that the world is not flat, neuroscience, with its ever-increasing mechanistic understanding of how the brain enables mind, suggests that there is no one thing in us pulling the levers and in charge. It’s time to get over the idea of free will and move on.

But he wants to save the idea of personal responsibility nonetheless. Well, I agree with him in the sense that people must be held accountable for their acts. Absent accountability, punishment, and the like, society couldn’t function.  But he sneaks in a sort of anti-determinism when he says that people are responsible in the sense that they can choose whether to follow society’s rules:

Holding people responsible for their actions remains untouched and intact since that is a value granted by society. We all learn and obey rules, both personal and social. Following social rules, as they say, is part of our DNA. Virtually every human can follow rules no matter what mental state he or she is in.

Thus Jared Loughner, who has been charged with shooting Representative Gabrielle Giffords, is judged to be insane. Yet he followed one kind of rule when he stopped to make change for the taxi driver on the way to murder and cause mayhem. Should society really allow that the act of not following another kind of rule (not to kill anyone) be accepted as an excuse for murder? Since responsibility exists as a rule of social interaction and not normal, or even abnormal brain processes, it makes no sense to excuse the breaking of one kind of social rule but not another.

The flaw here is this: if our actions really are predetermined, how can we “choose” whether or not to follow the rules? Loughner was predetermined to make change for the driver, but predetermined to kill as well.  What is the contradiction in that?  He followed the rules in one case but not the other, but in what sense was he making free “choices”?  And who on earth is “excusing” breaking one kind of social rule but not another?  Breaking a “no murder” rule is far more serious than not making change properly; neither is in the strict sense “excusable,” but flouting one rule is far more serious than flouting another, and society comes down far harder on murder than on not making change. This is proper, for punishing severe violations of the social order has more salutary consequences: it affects both the miscreant’s brain and those of onlookers, prompting the former, perhaps, to reform and the latter to not emulate Loughner. Excessive punishment for small infractions is damaging to society.

Outside of that, I agree with Gazzaniga’s conclusion that people are accountable but we need to take into account determinism when thinking over how we punish wrongdoers. The problem is that he uses “responsibility” in the sense of “that person did it”, while many conceive of the word as meaning “that person had a choice whether to do it.” I agree with the former but not the latter, which is what I see as the erroneous notion of “moral responsibility.”  To me, moral responsibility implies true choice.

Hilary Bok, “Want to understand free will? Don’t look to neuroscience.”  Bok, a philosopher, seems to be taking the theological route, trying to preserve our feeling that we’re free by defining free will in a way that allows it. She doesn’t want to go where the science takes us, I think, but prefers to engage in semantics:

The problem of free will for compatibilists is not to preserve a role for deliberation and choice in the face of explanations that threaten them with elimination; it is to explain how, once our minds and our choices have been thoroughly naturalized, we can provide an adequate account of human agency and freedom.

How can we reconcile the idea that our choices have scientific explanations with the idea that we are free?

But that “adequate account” of agency must come from science, not philosophy.  And what does she mean by “freedom”?  She never defines it (a flaw in some of these essays) unless she means “our feeling that we decide freely.”

In the end, she takes what I see as a philosophical cop-out, i.e., free will lies in our ability to make seemingly reasoned choices:

A person whose actions depend on her choices has alternatives; if she is, in addition, capable of stepping back from her existing motivations and habits and making a reasoned decision among them, then, according to compatibilists, she is free.

But what is it mean to be “free” if the decisions we make after “reasoning” about them (a process that of course is not free, but itself determined by physics), were determined in advance? That just gives us the feeling of freedom, but not actual freedom. If you want to take the former to be the latter, fine, but then admit that even one’s reasoning is determined.  In this sense humans have free will in the same way that crows or chimps do—animals that also “reason.”  If by “free will” we mean the Dennett-ian sense that “we are highly evolve beasts to take in and mull over inputs before coughing up an output,” then we should simply say “decisions arrived at after pondering” rather than “decisions made by free will.” It’s so much less confusing that way, and jettisons the historical and religious baggage associated with free will.

Owen Jones, “The end of (discussing) free will.”  I like this piece, first because of Jones’s explicit rejection of non-determinism:

The problem with free will is that we keep dwelling on it. Really, this has to stop. Free will is to human behavior what a perfect vacuum is to terrestrial physics—a largely abstract endpoint from which to begin thinking, before immediately moving on to consider and confront the practical frictions of daily existence.

I do get it. People don’t like to be caused. It conflicts with their preference to be fully self-actualized. So it is understandable that, at base, free-will discussions tend to center on whether people have the ability to make choices uncaused by anything other than themselves. But there’s a clear answer: They don’t. Will is as free as lunch. (If you doubt, just try willing yourself out of love, lust, anger, or jealousy.)

Too many philosophers (and others) aren’t willing to make a statement this strong and true. You’ve probably seen—on this website and elsewhere—people saying that we shouldn’t go around proclaiming determinism because it’s bad for people to think that their actions are all predetermined.  But if Mele is right, and most people don’t think that, then what’s the harm? (I don’t think Mele is right.)

And I like Jones’s clear statement that our understanding of determinism has real implications for the law, and for punishment. I agree with him 100% here:

Which brings us to law. How will insights from the brain sciences affect the ways we assess a person’s responsibility for bad behavior? Answer: only somewhat, but sometimes significantly. Many people assume that legal responsibility requires free will, such that an absence of free will necessarily implies an absence of responsibility. Not true, as many scholars have amply demonstrated. Full, complete, utterly unconstrained freedom to choose among available actions might be nice to have, but it is not in fact necessary for a fair and functioning legal system.

This is not to say that degrees of freedom are irrelevant to law. Science hasn’t killed free will. But it has clarified various factors—social, economic, cultural, and biological in nature—that constrain it.

My single quibble with Jones here: his statement that “science hasn’t killed free will.”  But he implied earlier that there’s no such thing as free will! If he thinks there’s a form of free will that remains viable, he needs to define it.  He doesn’t, which is a flaw in an otherwise admirable piece.

Paul Bloom, “Free will does not exist. So what?”  Bloom’s thesis is that we’ve long known that determinism rules, so there can’t be contracausal free will.  But his take is that it doesn’t matter:

Our actions are in fact literally predestined, determined by the laws of physics, the state of the universe, long before we were born, and, perhaps, by random events at the quantum level. We chose none of this, and so free will does not exist.

I agree with the consensus, but it’s not the big news that many of my colleagues seem to think it is. For one thing, it isn’t news at all. Determinism has been part of Philosophy 101 for quite a while now, and arguments against free will were around centuries before we knew anything about genes or neurons. It’s long been a concern in theology; Moses Maimonides, in the 1100s, phrased the problem in terms of divine omniscience: If God already knows what you will do, how could you be free to choose?

But he then falls into the same trap as Bok: true free will comes from the facct that our predetermined decisions are produced after a process of deliberation:

More important, it’s not clear what difference it makes. Many scholars do draw profound implications from the rejection of free will. Some neuroscientists claim that it entails giving up on the notion of moral responsibility. There is no actual distinction, they argue, between someone who is violent because of a large tumor in his brain and a neurologically normal premeditated killer—both are influenced by forces beyond their control, after all—and we should revise the criminal system accordingly. Other researchers connect the denial of free will with the view that conscious deliberation is impotent. We are mindless robots, influenced by unconscious motivations from within and subtle environmental cues from without; these entirely determine what we think and do. To claim that people consciously mull over decisions and think about arguments is to be in the grips of a prescientific conception of human nature.

I think those claims are mistaken. In any case, none of them follow from determinism. Most of all, the deterministic nature of the universe is fully compatible with the existence of conscious deliberation and rational thought. These (physical and determined) processes can influence our actions and our thoughts, in the same way that the (physical and determined) workings of a computer can influence its output. It is wrong, then, to think that one can escape from the world of physical causation—but it is not wrong to think that one can think, that we can mull over arguments, weigh the options, and sometimes come to a conclusion. After all, what are you doing now?

But exactly how is this claim mistaken: “We are mindless robots, influenced by unconscious motivations from within and subtle environmental cues from without; these entirely determine what we think and do. To claim that people consciously mull over decisions and think about arguments is to be in the grips of a prescientific conception of human nature.”? That seems pretty accurate to me; in fact, it’s what Bloom agreed with at the outset.

I agree that our predetermined decisions arrive after humans appear to think about stuff, but those decisions remain predetermined.  In what sense, then, is there any “freedom”?  To me, the claim that “conscious deliberation and rational thought” can affect our decisions is a form of snuck-in dualism, for those thought processes themselves are based on the laws of physics, and their outcome is determined.  External forces like other people, or insults to the brain, can affect the output, but the deliberations themselves cannot.  That is dualism.

***

In the sense that any of these thinkers agree that our will is “free,” they mean that some of our decisions appear to be made after conscious processes of deliberation—after thinking about them.  Of course, that’s a result of evolution, and many animals probably do the same thing.  Perhaps these folks will agree that humans aren’t unique in having this form of “free will,” for all beasts are evolved to absorb and process input before producing a behavioral output.  But where is the “freedom” in all this? What, exactly, are we free to do? We’re not free to think—that’s a result of evolution—and we’re not free in how our thought processes operate, or in what “decision” they produce.  Perhaps some readers can explain to me what the “freedom” is in this form of free will.  I don’t get it.  Sure, we can mull over things more than, say, a slug, but so what?  Our form of “mulling over things” is merely a more sophisticated version of what a slug does when it surveys its environment, and there’s no qualitative difference in either the determinism or in the way the nerves and ganglia work.

Pray where, oh where, is all that vaunted “freedom”?