Review of Doyle: Free Will

Free Will
Larger Image

Free Will: The Scandal in Philosophy
by Bob Doyle
450 pgs.

at Harvard Book Store
Free Will Review
by Tom Clark 1

Bob Doyle’s Free Will: The Scandal in Philosophy offers students of the free will debate a great deal: a detailed history of the debate with extensive quotes and commentary on all the major players; a relational taxonomy of names for the various positions taken over the ages and their distinctive conceptual components; chapters on contemporary philosophers and their arguments; an extensive glossary of frequently used philosophical terms; and an equally extensive bibliography. The book is derived from Doyle’s pages on free will at his Information Philosopher website.

The free will debate grips us because many suppose human freedom is at odds with deterministic natural laws, which might seem to swallow up our capacity to originate actions for which we can be held responsible. Doyle offers his own position on the debate, which he claims should end hostilities between two major factions, compatibilists and incompatibilists. The former argue that free will is compatible with determinism, the latter that it requires some sort of indeterminism.

Both sides could perhaps agree that to act freely I have to have the capacity and opportunity to control my behavior according to my wishes, such that my actions are up to me. Both sides would also probably agree that to be justly held responsible (and just about everyone agrees we have to hold each other responsible to maintain moral and civic order), I have to be free in this sense.

The bone of contention, however, is about the self and its capacity for choice and control. Compatibilists think so long as I’m acting on my own recognizance, that is, sanely and without being coerced, that’s enough for me to be free and held responsible. The controlling self doesn’t have to include or partake of anything indeterministic that breaks the chain of causation going back to the Big Bang. Incompatibilists think that more is required: that for actions to really be up to me, and to be justly held responsible, the choice process has to partake of something that breaks the causal chain.

We can see that there’s a factual claim at stake here about the nature of human action and decision-making. Does it involve something indeterministic or not? But the other question at issue is whether indeterminism is required for responsibility. Incompatibilists say yes, compatibilists no. This doesn’t seem to be a factual question about the world, but rather a normative question about our responsibility practices. Compatibilists say that even if determinism is true our responsibility practices are, overall, fair and just. Incompatibilists disagree. How do we adjudicate this dispute?

On the first question, Doyle’s view is that human decision-making involves selecting among alternatives that are irreducibly indeterministic by virtue of being generated by quantum noise in the brain. This should satisfy incompatibilists, he says, since the deterministic causal chain is broken. Whether Doyle’s quantum noise hypothesis holds up is of course a matter for the evidence to decide; the neural mechanisms which generate the options that occur to me (consciously or unconsciously) are not yet mapped out, so the claim that quantum noise makes them irreducibly indeterministic is speculative at this point. But research is underway.

Related to the second question, Doyle argues that causal responsibility (as distinct from moral responsibility, see below) only accrues if behavior is controlled by the agent, and indeterminism doesn’t help with that, rather it hurts. The selection among alternatives has to be determined by the agent’s character, motives and rational capacities; only then does it make sense to hold him responsible as a causal factor that perhaps needs restraint or rehabilitation. Anything which breaks the deterministic causal chain of the agent’s selection process among the alternatives would make him less, not more, responsible, since it wouldn’t be the agent’s doing, but rather attributable to chance. This view of responsibility should satisfy the compatibilists, Doyle says, since after all it’s compatible with, indeed depends on, what we might call agent determinism. We can and must hold people responsible as identifiable causes of socially and morally consequential effects.

So Doyle’s view is a two stage model of free will, where the first stage introduces freedom from global determinism, and the second stage makes the agent responsible by preserving a type of local determinism. He calls it a “comprehensive compatibilism” since it’s compatible with both determinism and indeterminism. Again, whether the brain actually generates random possibilities for consideration, and then suppresses quantum indeterminism such that the selection process is fully deterministic, are open, empirical questions.

Doyle is careful to document how his view is anticipated and nearly replicated by several philosophers, notably William James and now by Daniel Dennett and Al Mele (closest of all)). But he suggests that most philosophers have simply missed the boat (some by inches!) when it comes to free will, which is why this “scandal” in philosophy has persisted for so long.

And it persists in very concrete ways, for instance with regard to our responsibility practices. On Doyle’s view, are they fair? Compatibilists and libertarians (incompatibilists who believe we do have indeterministic free will) generally hold that retributive punishment – punishing someone because he deserves it, whether or not it produces any good consequences – is fair and justified. Retribution is an accepted rationale for punishment in many, perhaps most societies, and is often (not always) justified by saying that an agent could have done otherwise in an actual situation by virtue of indeterminism: the agent wasn’t fully determined to act as he did, so the act is up to him in a buck-stopping way that it isn’t on the compatibilist view.

But one might ask: is it fair and just to exact retribution against an agent who has deterministically selected among randomly generated alternatives? Does the indeterminism in the first stage, what Doyle says should satisfy libertarians, help to justify retribution? I didn’t see where Doyle addresses this question (I may have missed it), although he does express distaste for retribution, so perhaps his answer would be no.2

The real scandal as I see it is the compatibilist commitment to the idea of moral desert, in which punitive responses to wrongdoing are somehow obligatory whether or not they produce any good outcomes – the essence of retributivism. A naturalistic understanding of human agency (no spooky supernatural controller or ultimately self-caused self), combined with the humanitarian principle of minimizing suffering, makes it difficult to justify inflicting punishment independent of consequentialist considerations. At least I haven’t come across any justifications that strike me as plausible. Bruce Waller’s forthcoming book, Against Moral Responsibility, targets compatibilists on this score very convincingly.

Doyle does say that the ethical, normative question of moral responsibility (as distinct from causal responsibility, see above) is independent of the scientific question of free will, so that we shouldn’t just define free will as whatever it is that makes someone morally responsible, as some philosophers do. Fair enough, but if by being morally responsible we mean being deserving of punishment independent of consequentialist considerations (this might be the common understanding, I’m not sure), what would make someone morally responsible? Compatibilists say moral responsibility is (somehow) entailed by being uncoerced and sane in one’s decisions and actions; incompatibilists say that in addition behavior must be exempt from determinism in some respect. But of course these are their respective definitions of free will. The connection between the debate about free will and our actual responsibility practices, as shaped by our perhaps antiquated and untenable notion of moral responsibility, is arguably why the debate matters so much, for instance in criminal justice.

What matters mostly for Doyle – the significance of the free will debate as he sees it – is that we not be machines, not just the working out of a pre-determined fate. It’s very important that our futures be open by virtue of quantum indeterminism, even as the past gets fixed by our deterministic deliberations. The concern that “new information” come into being, so that the universe can be truly creative through human choices, palpably motivates Doyle’s as yet unconfirmed quantum noise hypothesis.

Now, he may be right, but why does it matter? Speaking for myself, discovering facts about the world, having novel and enjoyable experiences, living out a complex and demanding social and political life with one’s partner, friends and peers (Zorba the Greek’s “full catastrophe”), and making the world a better place (according to my lights), is enough. Even if all this were the inexorable working out of causal laws such that the future is fixed, not open, the fact that I don’t know what the future holds makes the game of life compelling. My ignorance is what drives my concern about the future, for myself and others; determinism be damned. For Doyle, this sort of epistemic uncertainty isn’t enough; there has to be a constant injection of new information into our lives via indeterministic quantum noise, otherwise …what? Well, we’d be merely machines.

This new information, should it play a role in our behavior, doesn’t help make us responsible, or morally responsible, since, as Doyle himself argues, responsibility comes from agent determinism. Indeterminism serves only to break the causal chain so that we can say our behavior is in some respect free from being pre-determined. If indeterminism in the brain affords us an open future, that’s fine with me. If it doesn’t, and we’re all just machines, that’s fine with me too. Either way, we can and must hold each other responsible, and it’s perfectly possible to hold mechanisms responsible. But the question is, what responsibility practices are justifiable? I would have liked to see more on this.

My other caveat is that Doyle’s eagerness to make his case results in considerable repetition concerning his two-stage model of free will; some pruning is perhaps in order for future editions. However, the current version affords the reader a clear picture of the model and how it fits with his ideas on information, physics, biology, neuroscience, and cosmology. And whether you buy his view or not, Doyle’s guided tour of the free will debate and its history is a very useful addition to the literature on human action in an age of science.


1. Tom Clark is the director of the Center for Naturalism and founder of Philosophy Cafe.
2. On page 260 Doyle says “There are excellent stand-alone reasons for preferring rehabilitation and education to retributive vengeance.”

One Response to Review of Doyle: Free Will

  1. JenniferL says:

    I agree — I appreciated this book for the guided tour of the philosophical debate about free will and human behavior. The book is primarily written for serious students of philosophy, but is accessible, though challenging, for a more casual philosopher. (Challenging to read on the subway — usually my favorite place to get absorbed in a book!)

    I enjoyed thinking about the implications of Doyle’s two-stage model, with the random, “quantum noise” in our brains providing the first stage of decision making, and an intelligent, conscious choice forming the second stage of decision making.

    I’m interested in thinking more about how this model could apply to social science, psychology, and public policy. Right now I’m particularly intersted in how public policy and public education can (or cannot) affect people’s behavior, and therefore interested in this as a potential application of Doyle’s model and of the free will debate in general.

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>