Skip to content

The will is caused, not free

By Brian Earp

See Brian’s most recent previous post by clicking here.

See all of Brian’s previous posts by clicking here.

Follow Brian on Twitter by clicking here.


The will is caused, not free

Everyone is talking about free will these days. Sam Harris has a new book out. Eric MacDonald has weighed in on that. Jerry Coyne, Paul Bloom, and some philosopher-types have a debate going on in the Chronicle of Higher Education. And way back in 2009 the Society for Personality and Social Psychology hosted a “showdown” between psychologists Roy Baumeister and John Bargh on the topic: What does the ‘free’ in ‘free will’ really mean? [A video of Bargh’s half can be seen here. Baumeister is here.]

The SPSP conference led to a fiery exchange of blog posts between the two principles, and then to a more sedated pair of papers in the society’s newsletter, Dialogue. Baumeister enlisted Kathleen Vohs to co-author his piece, and Bargh (for some reason) enlisted me. Here is what Professor Bargh and I had to say — after this delightful FoxTrot comic by Bill Amend.

We welcome the opportunity to summarize our main points from the SPSP debate; first though we will respond to the additional arguments by Baumeister and Vohs in this issue concerning determinism and causality. We see no problem with the assertions that psychologists need not be strict determinists to practice their science, and that determinism and causality are not the same thing.

However, neither of these points is relevant to the basic question of free will. The ‘free’ in free will means freedom from causation, either by external forces (in the political sense of the term) or internal ones (in the psychological sense); and in our view it is just as problematic to claim that the will is uncaused as it is to argue it is not determined.

Free will may be defined as an agent’s ability to act on the world by its own volition, independently of purely physical (as opposed to metaphysical) causes and prior states of the world. The folk notion of free will is laden with the concept of a soul, a non-physical, unfettered, internal source of choice-making … in other words, an uncaused causer.

“The soul” may have gone out of fashion, and “the mind” taken over many of its functions and connotations, but the intuitive notion of free will has stayed much the same: there is something inside each of us that allows us to make “real” choices — choices that even an omnipotent being, one who knew every environmental influence, and every physical fact leading up to the choice-making event, could not foretell with perfect confidence and accuracy.

Determinism, if it were true, would indeed rule out this sort of free will, or shunt it into the realm of total redundancy. But indeterminism (of whatever flavor) isn’t any kinder to the notion. Just because some event is not strictly determined by prior physical data doesn’t mean it is caused by a free will. It may be simply indeterminately, probabilistically, or (to whatever degree) “randomly” caused by prior physical data.

If one wishes to use the existence of error variance as evidence for the existence of free will, we can only point out that our business as scientists is to strive to reduce this unexplained variance by replacing it with explanation. Calling it ‘free will’ and walking away satisfied rather misses the point.

But let us assume that there is a free, internal source of control that guides our behavior and is ultimately responsible for ‘real’ choices. To attribute human behavior to this mystical source is to place one’s bets on an increasingly shrinking sphere. The project of social psychology, after all, has been to identify (a) external-to-the-individual causes of judgment, motivation, and behavior, such as situational influences, and (b) internal-to-the-individual causes, which research has shown increasingly to operate outside of awareness and conscious intention — not “freely chosen” in any sense of the term.

Are there some human behaviors that are possible only if free will exists and is a true causal source of action? There may be. But let’s not give up on the search for non-mystical causes just yet.

This brings us to an area of agreement revealed in the debate: that a belief in free will is important for human strivings. People cherish their sense of control over the world and their own behavior. In the debate, we noted recent empirical articles by Vohs and by Baumeister [see here and here] showing negative consequences (cheating, aggression) of informing participants that free will does not exist. Our response to these ‘new’ articles is that our field revealed the existence of such positive illusions decades ago, and we already know how essential they are to normal functioning.

Clearly it is motivating for each of us to believe we are better than average, that bad things happen to other people, not ourselves, and that we have free-agentic control over our own judgments and behavior — just as it is comforting to believe in a benevolent God and justice for all in an afterlife. But the benefits of believing in free will are irrelevant to the actual existence of free will. A positive illusion, no matter how functional and comforting, is still an illusion.

And we must caution against drawing conclusions from such research findings (implicitly or explicitly) that we should either (a) not make findings against the existence of free will known to the public or (b) stop doing such research altogether. The belief in personal free will is a deeply rooted aspect of human phenomenal experience, and is so powerful that even those who do not subscribe to it intellectually still feel it in their personal lives as much as everyone else.

It is not uncommon for one’s first-person experience to be at odds with physical reality: 500 years after Copernicus we still see a morning sunrise, not the earth (and ourselves) tilting towards the sun, even though we know better scientifically. As Dan Wegner, Paul Bloom, Dan Dennett, and others have argued, there are strong natural supports for the belief in supernatural entities, just as there are for free will — and sunrises too, for that matter.

And if, as countless recent surveys show, the prodigious evidence in favor of evolutionary theory accumulated over the past 150 years has done little to erode the popular belief in a creator-god, then we can rest assured that the relatively nascent research on unconscious causes of motivation, judgment, and behavior will not result in anarchy or the collapse of social norms and moral behavior.

We should also not forget past social psychological research demonstrating that the belief in personal free will is selective: people routinely make self-serving attributions about the causes of their behavior. We take credit for the positive things we do (free will), but not for our misdeeds and failures ( “I had no choice”, “I was abused as a child”, “I was angry”). This suggests to us that much of the emotion surrounding the issue of free will is not about freedom per se but about self-esteem maintenance.

We take personal pride in our ancestors, our blue eyes or rich brown skin, our height or birthday or name (as in the name-letter effect) — none of which we chose or had any control over. Accordingly, we [unpublished data] analyzed hundreds of individuals’ spontaneous self-descriptions, and indeed 34% of their first-to-mind completions to the stem “I am _____” were such non-chosen aspects of self.

It seems that people do not possess a consistent belief in free will so much as they strongly wish to take credit for the good things they are and do (regardless of whether they caused them), and to distance themselves from the bad things (even if they caused them). Evidently, the belief in free will is not principled, but socially strategic in nature.

So what, then, if one’s will is not ‘free’ of internal causation? It is still your will and my will and each is unique: a confluence of genetic heritage, early absorption of local cultural norms and values, and particular individual life experiences. After all, one can claim personal ownership of one’s will just as much as one claims ownership of one’s name, eye color, and birthday, and be as proud of one’s will and its products as one is proud of the exploits of great-great-Grandma the pioneer, even though one’s ‘free will’ played no role in any of these.


Bargh, J. A. & Earp, B. D. (2009). The will is caused, not ‘free.’ Dialogue, Vol. 24, No. 1, 13-15. Invited commentary for the Society for Personality and Social Psychology. Originally published online at Psychology Today, at this link. Also available here.


See Brian’s most recent previous post by clicking here.

See all of Brian’s previous posts by clicking here.

Follow Brian on Twitter by clicking here.

Share on

9 Comment on this post

  1. Sam Harris does feel that free will is mostly an illusion. I believe we can make choices, but seldom freely. In my (free) ebook on comparative mysticism, “the greatest achievement in life,” is a chapter called “Outside the box.” Here are three paragraphs from it:

    What if you had to make all your decisions about living while detained in a jail cell? The cells may be open for brief periods each day, but the prisoners are still surrounded by walls. There are also walls around cells of everyday life. We are restricted by our ability to control our emotions, mind and body. Even with full command of our “self,” we must live within the restraints of Nature and society. Freedom is relative.

    “Free will” is really quite limited, despite belief that we control ourselves and our lives. We think we have endless choices…until we try to make them. Each decision must not only be based on what we “want to do,” but also on our own capabilities and what is expected of us. Nature and society imprison us, whether we like it or not. The key to release is mystical realization. All in One and One in All, the divine unity, opens the gate between a universal consciousness and most people’s constrained awareness.

    Outer walls are the boxes of Nature and of society. Inclement weather, lack of sunlight, gravity, and/or other natural phenomena may restrain our movements. Our own natural aptitudes, practiced talents and learned skills are always lacking in some areas. Human nature is controlled mostly by society. What we believe that other people expect of us greatly influences how we feel, think and act. Considering the reactions of our family, friends, business associates, community, and/or nation determines much of what we do. Those “laws” of Nature and society govern our lives, usually more so than we wish. Mystical awareness can allow us to obey divine law here and now.

    Sam Harris has written positively on mysticism and said “I see nothing irrational about seeking the states of mind that lie at the core of many religions. Compassion, awe, devotion and feelings of oneness are surely among the most valuable experiences a person can have.” Harris’ personal background reflects his own search toward that goal.

  2. Initially, I wrote a really ill-tempered response. Let me take it back (luckily I did not press post) and try to explain why this kind of post annoys people like me (a philosopher – down the hall from you, Brian – who works on free will). You take yourself to debunking ‘free will’, which to you means freedom from causation”. Now philosophers don’t mean anything like this by free will. They mean a power such that, by exercising it, agents are responsible for what they do. Does this power exist? Nothing you say is to the point. If you have an argument why we can only be responsible if we are free from causation, you might mention it (but don’t expect me to be impressed, since I doubt very much it will be original: we have been doing this for 2000 years).

    Well, perhaps it is not the philosopher’s conception of free will you are targeting, but the folk conception (maybe you think that philosophers have made up a notion of free will that is just a philosopher’s fiction). You say “The folk notion of free will is laden with the concept of a soul, a non-physical, unfettered, internal source of choice-making … in other words, an uncaused causer”. Now scientists often accuse philosophers of theorizing from the armchair, and not gathering data. So now I want to know how you know what the folk believe? You see, philosophers have gone out and asked the folk, and the results do not support what you say here. The empirical evidence is equivocal; it turns out that folk attributions of freedom are sensitive to the pragmatics in various ways, But there is absolutely no evidence that the folk think that free will requires absence of causation (though some evidence that it might require indeterministic causation – some evidence; again, the evidence here is hard to interpret), and none that it requires a soul. I am happy to provide a short bibliography of empirical work on folk conceptions of free will, some by (so-called) experimental philosophers and some by social psychologists.

    So your remarks are quite simply beside the point. You refute neither the philosopher’s conception of free will nor the folk (assuming there is a difference, which is unclear). Rather, you refute a fiction, the same one that Sam Harris and Jerry Coyne share with you. That’s not an exercise that I find particularly illuminating.

    1. Dear Neil,

      Thank you for your thoughtful reply – and for being so “tempered” in spite of your annoyance. I’ll first say that I would very happily accept a list of readings on this topic (I’ll send you an email in a moment asking for it) and that I expect I have a great deal to learn. This essay was first-authored by my dear mentor, the psychologist John Bargh at Yale, with some help from me when I was an undergraduate a few years ago – so anything I contributed that didn’t rise to the challenge of 2,000 years worth of philosophy, or even the latest x-phi research, is a reflection of my state of learning at the time. If you’re willing to share some of your expertise with me, I will be very grateful.

      I think anyone who talks about “the folk” conception of anything is getting into tricky territory. Perhaps it would have been better to talk about an idea of free will that “some folks” have, rather than “the folk” in general. But even the experimental philosophy data is not clear-cut on what “the folk” think … different folks think different things. And a lot of them think that they have a soul and that their actions emanate freely therefrom. If people with this belief aren’t 51% of the population, and so don’t dominate the surveys, I could accept that, but maybe I just grew up in weird circles.

      The context of our article is important — it came as a response to a debate at a social psychology conference. Baumeister’s point (if I remember) was that (a) error variance in psychology experiments is evidence for free will and (b) if people find out that their actions are influenced by forces outside of their awareness or control–such as unconscious factors–then they will behave in anti-social ways … with the implicit message being that maybe philosophers and psychologists should refrain from making their thoughts about the non-existence of a radical free will known to the public. I don’t think our point was so much to define free will, or refute any specific notion of it, but rather to say the following:

      1) It is useful for scientists to pursue explanations of human behavior that do not stop at “they emanated from a free will” because then they’d lose some motivation for looking at other causes, like unconscious factors, environmental factors, and so on. There might be some sort of free will (in the philosopher’s sense, or the folk sense, or the way we define it), but as an approach to doing research, looking at error variance in psychology experiments and saying “it’s due to a free will” is a really good way to stop thinking creatively about what other types of explanations could be involved.

      2) Doing research on unconscious forces influencing people’s behavior — which shows, if not that people do not have free will, that at least they do not act on totally on the basis of conscious motivations, and are therefore perhaps “less free” than they think they are — is not going to lead to the collapse of social norms, morality, and so on.

      Those are pretty limited points, and we probably could have done a better job of tailoring our rhetoric to just those arguments. I will say that if you find yourself annoyed at articles like the one I wrote with Dr. Bargh, and if Sam Harris and Jerry Coyne get under your skin as well, then there are probably a lot more people than just us few who aren’t up to speed on the more “illuminating” discussions about free will you’d like to be having … So, again, send me your readings — convert me to your view — and I will take up the charge of spreading the light!


  3. Dear Brian,

    For my benefit, as much as anyone else’s let me summarize the literature on the folk conception of free will, as it stands today (keep in mind we have only been doing this systematically for a decade, and much remains unprobed). Doing so will allow me to reply to the points you make.

    I think the available evidence suggests that there is a remarkably robust folk conception of free will (we even have cross-cultural data). The primary way in which folk conceptions of free will have been probed is by asking subjects about the freedom or responsibility of agents in deterministic universes (obviously, the experiments don’t use the word ‘determinism’; rather they typically cash it out in terms of 100% predictability of behavior by a supercomputer prior to the birth of the agent). One very consistent finding is that ordinary people have an incompatibilist theory of free will. That is, they hold a theory according to which people can’t be free if determinism is true (asked whether someone is free in a world in which all behavior is 100% predictable, they say ‘no’; in philosophical jargon, this is incompatibilist because it denies that freedom is compatible with determinism). Before I go on, let me note that this view – that incompatibilism is true – is not anti-scientific as it stands. To say that freedom is incompatible with determinism is not to say that free actions must be uncaused. In fact, most philosophers who are incompatibilists think that all actions are caused; they also think, however, that if there are free actions, they are indeterministically caused. And indeterministic causation is thought by most physicists to be a feature of this universe. So thinking that free will requires indeterministic causation is not thinking that it requires anything spooky. Moreover, the evidence – in psychology and neuroscience – is neutral with regard to whether behavior might be indeterministically caused. Firings of neurons are stochastic events; whether this is because quantum level determinacy has an effect at the level of the neuron (unlikely, in my view, but not impossible) or because the brain is such a complex system we have not yet been able to detect all the relevant variables, is at this stage unknown.

    So already we see that people like Coyne are wrong about what the everyday picture of free will requires, and about its compatibility with science. But the picture gets more complicated. If you ask not whether agents can be free in a deterministic world, but whether a particular agent in such a world performed a particular action, described contretely, of their own free will (or is responsible for it), the majority of subjects give compatibilist responses (see Nahmias, et al 2005; Surveying freedom: Folk intuitions about free will and moral responsibility, Philosophical Psychology 18; Nahmias et al. 2006; Is Incompatibilism Intuitive? Philosophy and Phenomenological Research 73). The figures are 76% for free will and 83% for moral responsibility.

    So the next step in the literature has been to explain the apparent divergence between folk theories and folk judgments. There have been two strategies. Nichols and Knobe (Moral Responsibility and Determinism: The Cognitive Science of Folk Intuitions, Nous 43) argue that the concrete description of cases causes an emotional response that overrides subjects’ cool theoretical judgments. They see the emotional response as interfering with the judgment; the idea, roughly, is that the compatibilist response is a performance error. They found that subjects were less likely to give compatibilist responses in low affect cases than in high affect cases.

    In response, Nahmias et al. 2007 (Free Will, Moral Responsibility, and Mechanism: Experiments on Folk Intuitions. Midwest Studies in Philosophy 31) argue that the description of determinism Nichols and Knobes used causes people to think that future actions had to happen no matter what anyone did. That is, the descriptions suggested fatalism, not determinism (where an event is fated if I can’t avoid it, like Oedipus’s killing his father – it doesn’t matter what Oedipus does, because the event occurs independently of what he does). Nahmias et al. therefore suggests that it is the incompatibilist response, and not the compatibilist response that is a performance error. They think that incompatibilist intuitions are caused by confusing determinism with bypassing mental states: people are not worried that their thoughts are determined; they are worried that their actions might be determined regardless of what they think. In support of this, they showed that if determinism was described in psychological terms (‘once specific thoughts, desires, and plans occur in the person’s mind, they will definitely cause the person to make the specific decision he or she makes’), incompatibilist responses fall dramatically.

    In addition to this X-phi literature on free will (of course there is a lot more; this simply describes the opening salvos which set out the landscape of the debate), there is also a literature which analyses the statements subjects made when invited to describe choices in their own life that were free, and even to give a description of what free will is. Content coding approaches were used. Surprisingly, people didn’t mention causation at all. Rather, they described free actions and choices as those free from constraints like compulsion and coercion and regarding which they had time to deliberate. As Malle and Monroe say, “No evidence was found for metaphysical assumptions about dualism or indeterminism”. The published studies I know of (I know of a third unpublished one) are Monroe and Malle (2010). From Uncaused Will to Conscious Choice: The Need to Study, Not Speculate About People’s Folk Concept of Free Will. Review of Philosophy and Psychology 1 (2):211-224 and Stillman, Baumeister and Mele (2011). Free Will in Everyday Life: Autobiographical Accounts of Free and Unfree Actions. Philosophical Psychology 24 (3):381 – 394.

    So the folk notion seems remarkably consistent and coherent, and nothing like what Coyne and Bargh imagine. The evidence I have seen is that most people are dualists (I like Jesse Berring’s work on this, and have published a related study), but their dualism does not seem to play a role in their judgments concerning freedom. That’s a psychologist’s fiction! Let me note the irony here: it is us philosophers who are trying to do data-driven work, while (frankly arrogant) scientists pontificate about what the folk believe from the armchair.

    Now let me turn to your other 2 remarks. You say that “it is useful for scientists to pursue explanations of human behavior that do not stop at “they emanated from a free will””. I think this remark displays a confusion about free will is supposed to be. To say that an action was performed freely is not an alternative to saying it was caused by … (insert a scientific explanation here: implementation intentions, thalamo-cortical networks, or whatever). It is a property that some actions have in virtue of satisfying some high-level description. It is a little like saying that an action was kind, say. It’s not an explanation of the action; it’s a description that might nevertheless be true. The challenge to a naturalistic philosopher is to try to say which scientific explanations of behavior are candidates for free behavior. Here is where science is useful, because my understanding the mechanisms we can begin to understand the properties of actions that make them candidates for freedom and unfreedom.

    As for the second point, I think it is an open empirical question whether denying free will will affect people’s behavior. It better be an open empirical question, because I am currently engaged in studying just this question. As a matter of fact, the hypothesis we are exploring is that Baumeister is wrong. But he does have empirical evidence:

    Baumeister, R. F., Masicampo, E. J., & DeWall, C. N. (2009). Prosocial benefits of feeling free: Disbelief in free will increases aggression and reduces helpfulness. Personality and Social Psychology Bulletin, 35, 260-268.

    Vohs, K. D., & Schooler, J. W. (2008). The value of believing in free will: Encouraging a belief in determinism increases cheating. Psychological Science, 19, 49-54.

    Empirical questions require empirical answers!

    Sorry for the length of this response.



    1. Thanks, Neil, for your lengthy reply. What you’ve said here makes a good deal of sense to me, so I’m not going to take a bunch of time to push back on details. I will say that I cited, in my post, the Baumeister/Vohs/Schooler findings about pro-social benefits of believing in free will, and I think their evidence isn’t so strong either (I’ve tried to replicate some of their findings, with little success). I’m engaged in a slightly different research project, in which I’m trying to find out whether belief in free will may have ANTI-social consequences, but I’m waiting on the last bit of data so I won’t share my hypotheses just yet. I’d be very interested to learn more about your work suggesting that denying free will may not actually influence people’s behavior in the way Baumeister and colleagues suggest – how can I get a notice when the results are out? Finally, I published a mostly-failed study on free will and responsibility in the New School Psychology Bulletin ( in one portion of which I asked participants what they thought free will meant (to get at their conscious notion), and the results were varied, inconsistent, and some of them invoked a soul. I didn’t go through those data as systematically as Malle and Baumeister and others did in their reports (since it wasn’t the point of my investigation) but maybe that adds one more drop to this more direct way of getting at “the folk” conception.


  4. I haven’t read all these publications in detail, but what parts I’ve read is all really old stuff: either they try to say there is no dualism (the last publication with any weight on dualism is “The self and its brain” from, IIRC, 1977) – why beat a horse that’s been dead for 40 years? Or they say the brain is deterministic, when we all know since 90 odd years that the world is indeterministic – are brains deterministic bubbles in an otherwise indeterministic universe? Finally, they say chance is not free – then what is free? Chance and necessity make evolution be neither – so is free will: a biological trait of the brain that’s composed of both chance and necessity. Have these authors never read a philosophical book that’s younger than 40 years or a physics book that’s younger than 90? If you can’t follow me, read this longer review:
    Let’s do real, modern science without ignoring the last couples of decades of development.

    My apologies if I have missed anything more modern, but I stopped reading when the same old slogans were being thrown about.

  5. Here are two sites I just found, the PhilPapers Survey, ,
    and AskPhilosophers ,

    I am just a layman new to this stuff in the last few months, actually just getting exposed to this debate on Jerry Coyne’s site, but i quickly became frustrated with the insistance of JAC and others that most people mean libertarian free will, or derive their opinions from religion. This is mostly irrelevant, in the case of everyday society, and patently wrong for both philosophers and the general population. The vast majority stipulate, or lean towards, the belief that humans exercise free will, or more precisely, responsibility for the choices and actions they take. Up to 84% for professionals, and 90+ % for everyone.

    I am getting frustrated with Jerry getting so much attention in all this. It is just my uneducated opinion, but he has little of worth or insight in this debate, and is relatively naive and unsophisticated in both his attitude, and finality of his pronouncements.

    The main thing, I think, is that behavioral psychology already employs the idea that our situation and history greatly influences our choices and ability to cope and adapt, so the arguments are mostly moot to begin with. People, generally, will always be religious and susceptible to mystical and pseudo-scientific thinking, not to mention lacking the increasingly more sophisticated critical thinking skills needed to keep up with the onslaught of manipulative politics and advertising already happening.

    It is far more pressing to educate people on the basics of humanitarian and co-operative understanding with education than to worry about who thinks what, why, about the physics and science of causation of our decisions.

    Empirically, everyone experiences the process of thinking and planning and deciding, and, I’m sure, the almost 100% correlation between deciding to take a specific action, and the carrying out of the behavior when they intend it.

    Thanks for letting me ramble!

  6. If I could pipe in here for a minute . . . The Stanford Encyclopedia of Philosophy says free will, “is a philosophical term of art for a particular sort of capacity of rational agents to choose a course of action from among various alternatives.” ( As a term of art, then, free will is merely an intellectual construct or device as distinct from an empirical phenomenon.

    I argue that free will is not and can never be a provable empirical phenomenon. We are the sum of our own personal histories as well as the sum of all of life’s histories going back some 400+ million years or so. Now, to the extent that our reality contains an arrow of time, and that the arrow is always going from past to future, and that our histories are always expanding as the future becomes the past, then, by definition, we cannot alter the past because that would be something akin to time travel, which, as we all know, is a paradox.

    Does this then mean we must accept the notion of determinism? No, says yours truly. This is true because, as noted, the arrow of time is always moving into the future and the future is always uncertain. But can the future be predicted? And is a prediction a good approximation of the future? If so, then we have to talk about causation, but not in the normal sense of science and mathematics. Free will includes the idea of making choices among alternatives, but doing so in such a way that causation plays little or no part.

    You know that if you choose to hold your hand over a fire for very long, you’re going to get burned. But that’s more of an intuitive thing; call it instinct. So, there’s not much decision-making going on there, much less any determinism. Well, how about robbing a bank? You’ve cased the joint, you’ve got a trustworthy( you hope) crew, and you’ve got what you believe is a workable plan. So will you execute the plan or not, and in either case, how much does free will play in your decision? I submit that it plays no role at all. Your personal history will drive your decision. You may predict you’re going to get away with it, but there’s that damn arrow of time. Uncertainty is the evil twin of causation.

    Thanks for your kind consideration in letting me express my thoughts here.

Comments are closed.