Neuroscience gives the wrong answerby Daniel Dennett / October 16, 2014 / Leave a comment
Published in November 2014 issue of Prospect Magazine
“Many have thought that if the physics of our world is deterministic, free will is impossible whatever the details of brain activity prove to be.” © lillisphotography For several millennia, people have worried about whether or not they have free will. What exactly worries them? No single answer suffices. For centuries the driving issue was about God’s supposed omniscience. If God knew what we were going to do before we did it, in what sense were we free to do otherwise? Weren’t we just acting out our parts in a Divine Script? Were any of our so-called decisions real decisions? Even before belief in an omniscient God began to wane, science took over the threatening role. Democritus, the ancient Greek philosopher and proto-scientist, postulated that the world, including us, was made of tiny entities—atoms—and imagined that unless atoms sometimes, unpredictably and for no reason, interrupted their trajectories with a random swerve, we would be trapped in causal chains that reached back for eternity, robbing us of our power to initiate actions on our own. Lucretius adopted this idea, and expressed it with such dazzling power in his Stoic masterpiece, De Rerum Natura, that ever since the rediscovery of that poem in the 15th century, it has structured the thinking of philosophers and scientists alike. This breathtaking anticipation of quantum mechanics and its sub-atomic particles jumping—independently of all prior causation—from one state to another, has been seen by many to clarify the problem and enunciate its solution in one fell swoop: to have free will is to be the beneficiary of “quantum indeterminism” somewhere deep in our brains. But others have seen that an agent with what amounts to an utterly unpredictable roulette wheel in the driver’s seat hardly qualifies as an agent who is responsible for the actions chosen. Does free will require indeterminism or not? Many philosophers are sure they know the answer (I among them), but it must be acknowledged that nothing approaching consensus has yet been reached. What people seem to want—though articulating this idea usually causes them to backtrack in embarrassment—is to be a sort of god, perched somehow on the edge of the physical universe, neither a part of it nor remote from it, able to interfere “at will” with its ongoing streams of causation, without at the same time being caused by those very streams to choose which of the options to favour. What people don’t like, apparently, is the idea, borne in on them every day as science marches through their genetics and into their brains, that a person is merely a slub in the fabric of the universe, no more than a complicated and clever bulge amid the threads of causation, rather than a free-wheeling, free-choosing, autonomous, responsible initiator of deeds. How could such a mechanistic consolidation-station be the locus of moral authorship? (Warning bells should ring in the reader’s mind at this point. Note the weaknesses in the previous two sentences: a case of “rathering”—why couldn’t we be both enmeshed in causation and an autonomous chooser?—and a rhetorical question that discourages us from seeking an answer.) Can we reconcile the discoveries of contemporary science with the traditional presuppositions of law and morality? Our laws declare that those who sign valid contracts must choose of their own free will to sign them, and that one is responsible only for the effects of actions freely willed. People who “lack free will” have “diminished responsibility” or no responsibility at all, and in the case of the obviously impaired this is widely accepted as the only just policy. But what about those who are apparently normal and competent? Is our free will real, or have we been deluding ourselves all these years? Is science letting the cat out of the bag, and if so, what will be the repercussions? Read more from our philosophy section: Does the history of philosophy matter? The world’s most cerebral marriage Is there such a thing as the self? In recent years a growing gang of cognitive neuroscientists have announced to the world that they have made discoveries that show that “free will is an illusion.” This is, of course, momentous news, which promises or threatens to render obsolete a family of well-nigh sacred values: just deserts (for both praise and blame), guilt, punishment, honour, respect, trust, indeed the very meaning of life. Can it be that scientists have uncovered a secret quirk or weakness in our nervous systems that shows that we are never the responsible authors of our actions? Many have thought that if the physics of our world is deterministic (with no random swerves), free will is impossible whatever the details of brain activity prove to be. Philosophers have spent the last half century or so sorting themselves into two camps on this: “compatibilists” who claim that free will is compatible with determinism in spite of first appearances, and “incompatibilists” who disagree. In his new book Free: Why Science Hasn’t Disproved Free Will, philosopher Alfred Mele sets that apparently interminable wrangle aside and argues that whether you address the “modest free will” of compatibilism or the “ambitious free will” of those who hold out for indeterminism playing a role, the verdict is the same: these scientists have not come close to demonstrating that free will, in either sense, is an illusion. Their ingenious experiments, while yielding some surprising results, don’t have the revolutionary implications often claimed. Mele is not the first philosopher to reach this dismissive verdict, but it may well be that his tactical neutrality on the issue of compatibilism will win him the support of the vast majority of philosophers who have thought hard about free will: we philosophers disagree vehemently about whether compatibilist free will is the only kind of free will worth wanting (as I, for one, have argued for many years), but we agree with Mele that the scientists have jumped to unwarranted conclusions, for the reasons he presents so calmly and clearly in this little book. There are three main sources for the scientists’ unsettling “discoveries.” First, there is the putative timing of subjects’ decisions. This line of thought grew out of the work of the neurologist Benjamin Libet in the 1980s, which seemed to show that our brains made decisions before we, the conscious agents resident in those brains, could put in our oars. Second, there is the putative fallibility of subjects’ introspective judgements of their own agency, as reported by Daniel Wegner in his 2002 book, The Illusion of Conscious Will, which disclosed systematic errors in people’s judgements of their own role in actions undertaken by their limbs. Third, there is the unrecognised influence on subjects’ decisions of contextual factors that shouldn’t be decisive, growing out of Stanley Milgram’s and Philip Zimbardo’s notorious experiments into authority and obedience with college students back in the 1960s and 70s. More recent work exploring all three veins has come up with some new findings, and these have been seen by some to strengthen or even confirm the case from science against free will. In each area, Mele provides accurate, jargon-free accounts of the experiments and what they do and don’t show. And in each case he locates what, in my opinion, are the most fundamental flaws in the reasoning by those scientists. The mistakes are so obvious that one sometimes wonders how serious scientists could make them. What has lowered their threshold for careful analysis so catastrophically? Perhaps it is the temptation of glory. What a coup it would be if your neuroscience experiment brought about the collapse of several millennia of inconclusive philosophising about free will! A curious fact about these forays into philosophy is that almost invariably the scientists concentrate on the least scientifically informed, most simplistic conceptions of free will, as if to say they can’t be bothered considering the subtleties of alternative views worked out by mere philosophers. For instance, all the experiments in the Libet tradition take as their test case of a freely willed decision a trivial choice—between flicking or not flicking your wrist, or pushing the button on the left, not the right—with nothing hinging on which decision you make. Mele aptly likens these situations to being confronted with many identical jars of peanuts on the supermarket shelf and deciding which to reach for. You need no reason to choose the one you choose so you let some unconscious bias direct your hand to a jar—any jar—that is handy. Not an impressive model of a freely willed choice for which somebody might be held responsible. Moreover, as Mele points out, you are directed not to make a reasoned choice, so the fact that you have no clue about the source of your urge is hardly evidence that we, in general, are misled or clueless about how we make our choices. Similarly, Daniel Wegner’s case amounts to generalising the surprising discovery that in Ouija-board situations, people can often be made to feel they are the authors of acts that are in fact caused by the experimenter’s accomplice. Since in these rather artificial and strange circumstances we can be misled into thinking retrospectively that we chose to act when in fact we were manipulated into action, Wegner believes that it must follow (mustn’t it?) that we are never authoritative about the authorship of our acts. There are some complications to Wegner’s case, but this non-sequitur lies at the heart, and Mele has no difficulty providing evidence of cases in which our knowledge of our own reasoned choices is unassailable. It is a fact that when faced with actually tough decisions—about whether to intervene in somebody else’s crisis, for instance, or to go along with the crowd on some morally dubious adventure—we often disappoint ourselves and others with our craven behaviour. This sobering fact has been experimentally demonstrated in the Milgram and Zimbardo experiments and a host of milder, less traumatic experiments, but far from showing that we are always overwhelmed by context, these experiments invariably exhibit the capacity of a stalwart few to resist the enormous pressures arrayed against them. Is there a heroic minority of folks, then, with genuine free will, capable of being moved by good reasons even under duress? It’s better than that: you can learn—or be trained—to be on the alert for these pressures, and to resist them readily. Mele is careful not to overstate the case against the scientists’ reasoning. Some of the experiments that have captured the limelight have had deep flaws in their design and really ought to be ignored, and Mele points to some of these major oversights. Among those that have been well conducted we can indeed learn something about the wellsprings of our choices, but nothing that wasn’t already enshrined, less precisely and in less detail, in folk wisdom: people can be manipulated into doing things they know better than to do; people’s introspective access to their own thought processes is far from foolproof, and you shouldn’t play poker if you can’t maintain a relatively inscrutable poker face. People who don’t know these home truths are perhaps too benighted, too naïve, to be granted full responsibility for their actions, but the rest of us, wise to these weaknesses in our own control systems, can take steps to protect our autonomy and be held responsible for doing just that. Science may someday come up with some further line of investigation that does indeed show we are deluded about our capacity to make responsible choices, but to date, the cases made are unimpressive, and that is all that Mele modestly attempts to show. This review could similarly end on the mild, modest verdict that Mele has done his job and done it well. But there is a larger context worth considering. Suppose you were reviewing a scientific report that drew the conclusion that a diet without fat was in fact unhealthy, and that butter and cream and even bacon in moderation were good for you, and suppose further that the science was impeccable, carefully conducted and rigorously argued. Good news! Yes, but the author acknowledges in fine print that the research was financed by a million dollar grant from the Foundation for the Advancement of Bacon. We would be entitled—obliged—to keep that fact in the limelight. The science may be of the highest quality, honestly and sincerely reported, but do remember that the message delivered was the message hoped for by the funder. This is not reporting a finding contrary to the goals of the fact-seekers. So it is important to note that Mele’s research, as he scrupulously announces, and not in fine print, is supported by the Templeton Foundation. In fact, Mele is the director of a $4.4m project, “Free Will: Empirical and Philosophical Investigations,” funded by the Templeton Foundation, almost certainly the most munificent funding of any philosopher in history. The Templeton Foundation has a stated aim of asking and answering the “Big Questions,” and its programmes include both science and theology. In fact, yoking its support of science with its support of theology (and “individual freedom and free markets”) is the very core of its strategy. The Templeton Foundation supports, with no strings attached, a great deal of excellent science that is otherwise hard to fund. The Foundation supports theological and ideological explorations as well, and it uses the prestige it garners from its even-handed and generous support of non-ideological science to bolster the prestige of its ideological forays. It could easily divide itself into two (or three) foundations, with different names, and fund the same research—I know, because I challenged a Templeton director on this score and was told that they could indeed, but would not, do this. Alfred Mele is in an unenviable position, and there is really nothing he can do about it. Was his decision to stay strictly neutral on the compatibilism issue a wise philosophical tactic, permitting him to tackle a more modest project, demonstrating the weakness of the scientific argument to date, or was it a case of simply postponing the more difficult issue: if, as science seems to show, our decision-making is not accomplished with the help of any quantum magic, do we still have a variety of free will that can support morality and responsibility? The Templeton Foundation insists that it is not anti-science, and demonstrates this with the bulk of its largesse, but it also has an invested interest in keeping science from subverting some of its ideological aspirations, and it just happens that Mele’s work fits handsomely with that goal. And that, as I persist in telling my friends in science whenever they raise the issue, is why I advise them not to get too close to Templeton.