Local vs. ultimate control

When hearing about determinism, or the idea that we are fully caused to be who we are, people often jump to the conclusion that if this were true they would lose power and control. This is demoralizing. The intuition seems to be that if I am entirely the product of genetics and upbringing, if I don’t create myself in some ultimate respect, then I’m the mere working out of impersonal circumstances, I don’t really contribute anything to the world. As the quote above suggests, determinism seems to imply I don’t really choose, decide, or make things happen. The agent-me gets swallowed up or displaced by the causal story that explains who I am and what I do. My will – my desires, intentions, and ambitions – doesn’t count for anything since we can trace it to antecedent circumstances I didn’t choose. So therefore my current choices aren’t really my doing, they are the doing of omnipresent, impersonal cause and effect.

This conclusion, of course, is mistaken. People and their wills aren’t disempowered when we explain them in terms of antecedent causes. Just as my antecedents, genetic and environmental, had the causal power to create me in all my glory, I too have causal power to influence the world. So don’t forget about me. You can’t logically attribute power to the world and not to the agent, which is what the previous paragraph does: it concedes the causal efficacy of what created the person, but denies that the person plays a role in how things unfold in her immediate neighborhood, and sometimes well beyond. Although we don’t have ultimate control over ourselves – there’s no evidence we are self-created in a way that can’t be traced back to non-self factors – we have plenty of local, proximate control and power: our actions, controlled by our wills, often have the intended effects. This control and power doesn’t go away when we admit that the will itself has causal antecedents, that it didn’t create itself.

Getting clear about this is crucial, since science is in the process of dismantling the myth of the ultimately self-made self. How we present the naturalistic truth about being fully caused creatures makes all the difference. If we forget the causal powers of the person and the control on behavior exerted by her will, if we suppose that lack of ultimate control means lack of local control, then we might indeed become demoralized. As the word gets out about our complete causal connection to the world, we must make sure that this doesn’t get misunderstood as disempowering the agent.

The purported dangerousness of determinism

A recent Psychological Science paper by Kathleen Vohs and Jonathan Schooler illustrates this point. As has been widely reported (see here, here, here, here, here and here), they found that students seem to cheat more, compared to a control group, after reading passages by scientists to the effect that we don’t have free will. This seems a perfect example of demoralization by determinism.

Although skepticism about the methodology, and therefore the findings, is certainly warranted (see here and here and especially here for useful analysis and discussion), I’m going to proceed on the assumption that the students were in fact morally compromised (at least temporarily) by acquiring the belief they didn’t have free will. Although they offer some well-taken caveats about the significance of their findings, this is certainly what Vohs and Schooler think they have shown. As a result they caution that “If exposure to deterministic messages increases the likelihood of unethical actions, then identifying approaches for insulating the public against this danger becomes imperative” (p. 16 of the pdf).[1] In another paper which cites these findings (it’s a book chapter available here, written with Azim F. Shariff), they say  

Having one’s traditional understanding of free will disturbed by the determinist argument seems to encourage a form of moral laxity. Contrary to the view that discussions of free will are largely academic, this work suggests that the belief in free will, be it justified or mistaken, affects behavior. Although it may be true that free will is an impotent epiphenomenon, the belief in free will can have real and potent consequences. Should the illusory free will position advanced in academic circles enjoy popular support among the lay public, it may be accompanied by larger social implications. The message that there is no free will may go from being understood as “nothing is controllable” to “everything is permitted.” Again, regardless of the actual status of free will, a scientifically backed repudiation of it may encourage debauched behavior.

Vohs and Schooler (henceforth V&S) worry that hearing about determinism might be understood as “nothing is controllable,” which in turn might lead to moral laxity since “everything is permitted.” Their concern buys into the scenario sketched at the start: if we don’t have ultimate control of the sort being self-caused exceptions to determinism makes possible, then maybe nothing is controllable; human agency is for naught. They concede we may not have this kind of free will, but if we stop believing we do, we might be in trouble. This is what Daniel Dennett calls in his book Breaking the Spell “belief in belief.”

It’s clear from their papers that V&S themselves think that determinism robs agents of power and control. Here are passages from their Psychological Science paper which characterize the “deterministic worldview” of science, all of which highlight the power of external influences or internal sub-personal mechanisms, while downplaying person-level control and responsibility (emphasis added):

Yet the view from the scientific community is that behavior is caused by genes underlying personality dispositions, brain mechanisms, or features of the environment (e.g., Bargh, in press; Crick, 1994; Pinker, 2002).

What would happen if people came to believe that their behavior is the inexorable product of a causal chain set into motion without their own volition?

If reducing people’s sense of control also reduces the amount of effort they put toward improving their performance, then advocating a deterministic worldview that dismisses individual causation may similarly promote undesirable behavior.

Although some people have speculated about the societal risks that might result from adopting a viewpoint that denies personal responsibility for actions, this hypothesis has not been explored empirically.

The really dangerous idea

It looks as if V& S believe, wrongly, that a “deterministic worldview” obviates the possibility of local control, that it undermines personal power and responsibility because genes, environment and brain mechanisms are what’s really in control, not the agent. Now, if people come to believe they don’t have ultimate control, and if they have something like the authors’ (mis)conception of what not having it entails, then indeed they might become demoralized. This could explain the results of the study. But it’s important to see what’s demoralizing isn’t the empirically and logically well-supported conclusion that we don’t have contra-causal, libertarian free will, that we are not ultimately self-created, but the inference that if we are not free in this way then we aren’t causally efficacious agents. In other words, the dangerous idea isn’t that we are fully caused creatures, it’s that in being fully caused we lose our own causal powers, lose control, can’t be held responsible, and thus are at liberty to flout ethical norms because “my causes made me do it.” If we suppose people will draw this conclusion when hearing about determinism, then we might think, as do V&S (and Israeli philosopher Saul Smilansky, see here) that people must be shielded from science. But it’s patently infeasible and counterproductive, not to say paternalistic, to hide fundamental truths about human nature from ourselves, given our thirst for self-knowledge and the practical advantages such knowledge gives us.

V&S would disagree with this allocation of dangerousness, since they think being ultimately self-created selves really is a necessary condition of moral responsibility, so to deny contra-causal free will really is to invite moral collapse. In their book chapter, they admit that local control compatible with determinism gives us a lot, but not everything necessary for responsibility. What’s missing on the mainstream scientific view of ourselves, they say, is the capacity of the undetermined, freely willing conscious self to intervene in the causal stream, and thus be an uncaused first cause of behavior (emphasis added below):

The new [scientific] approach [to free will] dissolves the conscious self into the larger “I” and “the me that does the thinking” is embedded within the whole brain. “I” still control my actions, but the “I” is reconceived to be the coalition of my brain processes. And although I am still fully in control of my actions—although my will is still free—the nature of conscious free will changes. Instead of saying that people have conscious free will, we must instead say that people are conscious of their free will. Instead of saying that my consciousness (me) is making the decisions, we need to say that I am conscious of the parts of my brain (still me) that are making the decisions. Instead of saying the “I” moves the machinery of my brain, “I” am the machinery of my brain, and “I” consequently move myself.

Although this distinction may initially seem to be just semantic— rhetorical sleight of hand—closer inspection yields important differences, both philosophical and pragmatic. Most importantly, it seems that one loses the ultimate sense of free will that Kane (1996), and many others, believe worth wanting. Giving up control from the conscious, agentic self to the holistic “you are your brain” controller means just that. The “I” that one usually identifies with— one’s conscious self—loses its claim on causation. Insofar as people continue to identify with the former conscious agent, they do not have free will in the way that Kane defines it: “the power of agents to be the ultimate creators (or originators) and sustainers of their own ends and purposes…”.

Again, for V&S there has to be some sort of ultimate control for us to count as real agents, otherwise the self “loses its claim on causation.” But it’s telling that they themselves cannot articulate a clear, scientifically defensible account of how contra-causal, libertarian free will might work. In a personal sidebar, Schooler adverts to quantum indeterminacy as its possible basis, and claims that our first-person experience of conscious choice counts as good evidence that we have such free will (a highly suspect methodological assumption, see here). [2] But when it comes to solving what they call the “hard problem of free will,” that is, transparently explaining how we might be ultimate creators of our own ends and purposes, and how the non-physical and free conscious will moves the body, V&S admit to coming up short: “For the libertarian position to be taken seriously, its adherents need to demonstrate its viability, and not just its possibility.” And as Schooler puts it in his sidebar: “Ultimately, the viability of the libertarian perspective will depend on the generation of accounts that are both phenomenologically compelling and scientifically tractable.”

Countering the meme of libertarian freedom

Given the empirical and conceptual difficulties of defending libertarian free will (which is why few philosophers these days attempt it), one wants to know what ultimate control gives us that local control can’t. This, after all, is why V&S suppose we must maintain at least the fiction of such control, even if we can’t explain how it works. What belief in ultimate control gives us is the idea of buck-stopping agency, something which can’t appeal to antecedent factors to explain itself, and thus excuse itself. Libertarian free will blocks the appeal to universal causation as a universal excuse by setting up an entity that somehow causes itself. But once we understand that we can be morally responsible without being self-caused, then this empirically forlorn and conceptually tortuous strategy becomes unnecessary. So the key to avoiding V&S’s untenable position – their recommendation that we shield popular belief in ultimate control from science – is to show decisively that we don’t need to be contra-causal agents to be held responsible.

This can be done, and in fact has been done repeatedly by philosophers over the ages, so here I’ll simply footnote some recent accounts of how moral responsibility (and therefore moral and social stability) is compatible with our not being ultimately self-created.[3] The problem, though, is that often such accounts simply bounce off the heavily defended meme of libertarian freedom resident in our culture, one which celebrates the triumph of self-created will. It takes a lot of patient, hard, repeated drilling into the assumption of contra-causal agency before people see the light, and even then they often don’t (some, a distinct minority, see it right away). Schooler is a good example of someone fully in thrall to the libertarian intuition, someone who may never cross over into a transparent naturalism about the self, freedom and responsibility:

In principle, the compatibilist perspective is ideal as it allows us to have our material cake while freely choosing to eat it, too. The problem is that I, for one, simply cannot get my head around compatibilism. I understand that determinism does not rule out the opportunity to make deliberate decisions that are free from coercion, I acknowledge that I can redescribe the control of my actions as being completed by my brain as opposed to my mind, and I admire Dennett’s (2003) argument that the evolution of culture and the human brain’s capacity for rational analysis has enabled individuals to make reasoned decisions. I just don’t understand the following: If any noncoerced reasoned decision that I am about to make must necessarily be carried out in a specific manner based on a preexisting causal chain, then how can I be free to choose otherwise? And if I really have no option but to do exactly what I end up doing, how can it be said that my choice was free? On this point, I find myself in agreement with Greene and Cohen (2004) who argue, “ . . . contrary to legal and philosophical orthodoxy, determinism really does threaten free will and responsibility as we intuitively understand them.” (p.1780)

Schooler can’t, it seems, conceive of how we could hold people responsible unless they could have done otherwise in the exact same conditions, that is, acted contra-causally. His intuition that we must have libertarian freedom to have responsibility resists any compatibilist deconstruction, which is why he argues at such lengths in his sidebar for the bare possibility that a coherent libertarianism might someday be forthcoming. This mindset is widespread in the West: the ultimately self-created self is the existential, metaphysical bedrock to which all other beliefs must be accommodated, no matter what mainstream science says, and no matter what seems logically and conceptually most plausible.

The upshot is that V&S will likely persist in their attempts to show that disbelief in contra-causal free will is destabilizing, since according to their lights it has to be. Such efforts, of course, will simply reinforce the meme of libertarian freedom in the culture and the perceived importance of protecting it. Thus do some memes (god, free will) survive: by presenting themselves as necessary and irreplaceable foundations for morality, as indeed most of the news stories about their study pitched it. They also survive by becoming central to our identities: no one likes to change their mind about fundamental beliefs.

Instead of protecting the meme of the self-caused self, what if we countered it head on, and had students read debunkings of contra-causal free will that made it clear moral responsibility isn’t dependent on ultimate control? (See note 3 for some readings.) They would start to question the belief (if indeed they possessed it) that being morally good depends on being causally insulated from the world, on being a “moral levitator” as Daniel Dennett so wonderfully puts it. It isn’t rocket science to realize that even if we don’t ultimately choose ourselves, that practically speaking we must hold each other responsible to keep cheating and other dishonesties, immoralities and crimes in check. This sort of responsibility – being responsive to the prospect of moral and legal sanctions, as most sane adults are – is our moral back-up when temptation starts to get the better of us. High school and college kids can understand this.

Our students might even come to see that there are, as Spinoza, Einstein and Darwin believed, along with many other smart thinkers past and present, considerable moral and epistemic virtues to be gained from giving up the idea of libertarian free will;[4] that for instance we become less retributive and contemptuous, more compassionate, and more cognizant of what actually makes people tick.[5] V&S give this possibility short shrift in their papers, suggesting that it’s our dark side which will emerge should people be disabused of belief in free will. But this could well be a self-fulfilling prophecy. If people are taught that we can’t be good without being ultimately self-caused, then it’s more likely they’ll use debunkings of libertarian free will as excuses to behave badly, which will of course prove V&S’s hypothesis. At which point we’ll have to conduct a wholesale rebunking campaign to make sure people are safely, shall we say, abused of belief in libertarian free will. This will be difficult since there’s no plausible basis for it.

To avoid such difficulties, it’s better to take the high, open road of scientifically plausible accounts of human nature, and tell it like it is. This is a piece of cake compared to keeping the truth about ourselves under wraps. Cheaters beware: there are good reasons to hold you responsible if we find you out, namely so that only those who actually know the coursework get the good grades, and thus get the jobs for which they are qualified. What indispensable role does ultimate control play here? None. So long as we don’t forget about the me, the agent that’s actually on the scene, contributing its causal share to local outcomes, we won’t be demoralized by the realization it isn’t a little god. We’ll be content with the causal powers and rational capacities we have. These are considerable, and all we need to hold each other responsible and so keep debauchery at bay.

TWC, March 2008

Notes