I'm trying to figure out whether compatibilism (SEP, Wikipedia) is falsifiable, a metaphysic, or something else. One way to get at this is to take a Popperian approach, and ask whether any conceivable observation could prove compatibilism to be false. This is different from proving libertarianism to be true; I would be interested in knowing if it has to be compatibilism or libertarianism, or whether there are other options.

One answer I've come across is that if we could show that there is no such thing as 'mind', then compatibilism would be trivially false. That is, suppose that there is too much overlap between my mind and your mind, then the thing that makes a decision is some combination of the two minds, putting into question the idea that individual minds make decisions. I'm skeptical of this type of answer, because it seems to merely beg the question of whether 'mind' was defined properly before getting to compatibilism.

It may be interesting to note that dualism does not rescue us from compatibilism, unless we allow [apparent] paradox to reside there, instead of insisting that the logic that applies to particles and fields applies to whatever is in the other realm. There is no reason to suppose that 'reasoning' and 'logic' and 'choice' would be thought of differently in a nonphysical realm. Either the choice was founded in reasoning and observation, or it was random. Or we could say that choices have rational and irrational components. Where would the room for libertarian free choice be found? Surely purely random choices cannot be considered 'libertarian'? Here I'm assuming a dichotomy between compatibilism and libertarianism, which may be in error. Consider this paragraph a search for what might falsify compatibilism, a search which failed.

  • Falsification is not quite the right word. Compatibilism, unlike libertarianism, is not a doctrine of first choice, it is a forced move to reconcile free will intuitions with apparent determinism of conventional biophysics and to cure them of apparent incoherence. Remove the motivation, e.g. experimentally detect meaningful influence of quantum effects on brain activity and come up with a model that coherently accounts for control of actions under indeterminism, and compatibilism will become pointless and unattractive. Like solipsism or radical skepticism, which are also unfalsifiable.
    – Conifold
    Dec 19, 2017 at 22:40
  • @Conifold Can you cite studies to back this up?
    – H Walters
    Dec 20, 2017 at 3:22
  • @HWalters It can not be backed up by studies, it is an observation about the structure of the arguments that compatibilists offer.
    – Conifold
    Dec 20, 2017 at 21:34
  • @Conifold Might I at least point out that your characterizations of what type of intuitions people hold regarding the subject of free will runs counter to multiple studies performed in experimental philosophy on the subject. The result of such studies actually suggest that most lay folk hold compatibilist intuitions, completely counter to your offered opinion. Such studies may be questionable on procedural grounds, but I think it's extremely bold to claim they're so flawed that you get to claim compatibilists work counter to intuitions...
    – H Walters
    Dec 21, 2017 at 2:36
  • I think your initial claims on what people's intuitions are, by itself, are bold enough. But the notion that this cannot be studied I find outright ludicrous. Unless you hold that all compatibilists conspire to lie about their intuitions, then studying this is as easy as handing out surveys and analyzing the responses. (Not to mention that your opinion that this cannot be studied does not proportionally align to the boldness to which you made the claims).
    – H Walters
    Dec 21, 2017 at 2:58

3 Answers 3


If one takes compatibilism to mean “both natural determinism and free will are true” then this can be falsified by noting empirical evidence against the assertion of natural determinism. This evidence has two main sources. (1) Observe the indeterminism in quantum theory. (2) Observe the empirical evidence of common sense when we assert responsibility for our actions.

However, falsifying compatibilism has not stopped people from talking about compatibilism as if it could somehow be true. This should make one question what the role of falsifiability actually is. Is falsifiability a rational criteria for science or is it a form of rhetoric that we happen to find culturally acceptable to discredit other people’s theories?

There are reasons for people to remain interested in compatibilism or determinism in spite of the falsifications. Here are some.

Omniscient God. See Linda Zagzebski’s “Foreknowledge and Free Will” for a discussion of “infallible foreknowledge”. If God knows everything, how can we be free? For what it is worth, my view of this is simple, perhaps too simple. Define a free act as something that “is not knowable”. Define omniscience as knowing everything that “is knowable”. The two sets, what is knowable and what is not knowable, are disjoint. One can now have both an omniscient God and free will.

Laws of Nature. Some people believe in the existence of external Laws of Nature that are deterministic and true throughout all space and time. Their aim for science is to discover these external laws through theories and experimentation. At least some of this belief system will have to be modified. They seem to think there are only two options: determinism or uniform randomness. The existence of free acts would provide a third option.

Rationalism and Intuitionism. What is a free act? Is it the result of a rational process in the brain (rationalism) or is it a snap decision (intuitionism)? I see it as a snap decision. I move my left foot, take a breath, blink or whatever without asking my brain to do anything more than implement the decision. I owe my understanding of this distinction between rationalism and intuitionism to Jonathan Haidt’s “The Righteous Mind” although I don’t know his views on free will. His lecture, “The Rationalist Delusion” is a quick summary arguing for intuitionism. Now add to this John Conway and Simon Kochen’s “The Strong Free Will Theorem" . They proved that “if indeed we humans have free will, then elementary particles already have their own small share of this valuable commodity.” These quantum particles do not have brains. Whatever a free act for us is defined to be, I expect it to relate in some way to what these particles are doing since its existence and this proof guarantee that quantum particles have free will also.

Artificial Intelligence. Although natural determinism has been falsified, artificial determinism is temporarily possible. Some people working in artificial intelligence hope to someday create a deterministic machine that they can claim is conscious. One way to show consciousness in their machine would be to show that it has free will. For this to be possible they will need some kind of compatibilism to be true. Or, if that is not possible, they will have to argue that consciousness need not involve free agency.


I'm not certain that "compatibilism" itself is something that can be falsified. This is because compatibilism, far from being a single theory, is really more of a classification that covers several different kinds of theories. A rough analogy (though inaccurate, as all analogies are to some extent) would be to try to falsify "quantum mechanics." Various interpretations of quantum mechanics can be offered and those specific forms of interpretation might themselves be open to falsifiability, but quantum mechanics itself does not seem to be falsifiable, in any rigorous sense.

Your suggested falsification of compatibilism, i.e., by showing that there is no such thing as mind, seems to rest on a notion that compatibilism requires dualism. In fact, most compatibilist accounts I'm familiar with actually are based not on a dualism, but on some account of mind that is inseparable from physical reality, i.e., some form of physicalism, epiphenomenalism or nominalism, rather than any version of dualism. Dualism seems, prima facie, to support libertarianism. Thus, even if one could prove that "mind" (whatever that might be) doesn't exist, many versions of compatibilism would still be applicable, since they do not rely on a robust account of mind as anything other than a physical or epiphenomenal or linguistic construct. In short, you are right to be skeptical of this as an answer.

For example---a version I was once told in conversation---one might hold that the question of "free will" really devolves into a question of whether or not we can assign praise or blame to someone for their actions. Prima facie, we do this all the time, without necessarily needing a particular theory of mind; we instead have a theory of when an action counts as praisable or blamable or neither. We can then call any action to which we can assign the actor "praise" or "blame" a "free action," regardless of either the status of causation or of the actors mind. This would be a version of what I'm calling a nominalist theory of compatibilism. Such a view is, I think, utterly unfalsifiable, though it isn't strictly metaphysical either.

In general, on the various positions responding to the free will problem, the best internet resource I am aware of is Honderich's DETERMINISM AND FREEDOM PHILOSOPHY WEBSITE. There are several options here that your discussion doesn't consider, including strict determinism, and the options which Honderich calls "neither compatibilism nor incompatibilism." (Honderich's main issue is that we do not have a sufficient account of what "origination" might be, i.e., how it is an action might be said to be the free choice of an actor, precisely what you are getting at with your last paragraph).

Finally, I'm not sure if "falsifiability" is the correct standard to use for philosophical rather than empirical problems (and Popper himself notes that there are parts of the scientific method that cannot be explained in terms of "falsifiability," i.e., unfalsifiable notions which science itself draws on).

  • 1
    (1) I like the connection of free will to moral responsibility; I have been recently dwelling on that possible connection myself. (2) Perhaps instead of 'falsifiable', I should ask whether there are imaginable phenomena which compatibilism would not well-model? I don't have a comprehensive schema for what makes a given metaphysics 'bad' on hand. :-)
    – labreuer
    Jul 23, 2014 at 17:42
  • I'm not sure if there are phenomena that compatibilism couldn't somehow model, though I do wonder what it is that compatibilist might be trying to save, apart from a folk intuition. The question, then, might be: what does strict determinism rule out? And, if something can be found, why is that worth saving?
    – ig0774
    Jul 25, 2014 at 17:28
  • 1
    Excellent questions, although I'm not sure this question is the place to address them. I've been reading about naturalism and whether it provides the resources for knowing reality (R. Scott Smith's 2012 Naturalism and Our Knowledge of Reality + Penelope Maddy's 2009 Second Philosophy: A Naturalistic Method); perhaps after reading these I would be able to write a coherent question.
    – labreuer
    Jul 25, 2014 at 18:30

There is no ‘test’ for free will. There have been psychology experiments with results that are supposed to strength the case for hard determinism or hard incompatibilism (no free will whether determinism is true or false due to quantum randomness etc.) rather than the case for libertarian free will.

It’s not possible to come up with a completely consistent definition or coherent definition of free will that isn’t either non-scientific or contradictory in some way. Can the laws of physics and nature be suspended so that we can go about our business and be moral actors? That’s a contradictory claim. There is no evidence for libertarian free will of this nature this but again no evidence against this idea (not incontrovertible anyway), and this can’t be experimentally verified. So you’re stuck in a loop. Both camps asking the other to ‘prove’ or ‘disprove’ free will which can’t be done. Not that we know of now anyway.

I prefer to think about free will as a level of control we have over our actions as coming up with a ‘scientific’ definition of free will isn’t very plausible. The onus is on those who believe we have free will to present evidence for this being the case rather than say that we intuitively know this or a deity gave it to us or they can’t ‘prove’ we don’t have it. Mind you, I don’t take the majority position and find hard incompatibilism the most convincing but still am skeptical. This position doesn’t try to prove the case for or against determinism, but it takes the position that either way we don’t have free will and should look at the consequences of our actions for not accepting this to be the case (having levels of control over moral actions is not the same as there being complete libertarian free will or compatibilism in this sense). This can create a society that is more empathetic and rational such as how we treat criminals. Again invoking free will and morality together is not necessary for such a philosophy as argued by Pereboom et. al.

In Living Without Free Will, I develop and argue for a view according to which our being morally responsible would be ruled out if determinism were true, and also if indeterminism were true and the causes of our actions were exclusively events. Absent agent causation, indeterministic causal histories are as threatening to moral responsibility as deterministic histories are, and a generalization argument from manipulation cases shows that deterministic histories indeed undermine moral responsibility. Agent causation has not been ruled out as a coherent possibility, but it is not credible given our best physical theories. Hence we must take seriously the prospect that we are not free in the sense required for moral responsibility. I call the resulting view hard incompatibilism. Furthermore, contrary to widespread belief, a conception of life without free would not at all be devastating to morality or to our sense of meaning in life, and in certain respects, it may even be beneficial.

Hard incompatibilism is similar to hard determinism so similar arguments stand as to what this means for morality. That is there are degrees of control we have over our actions but this can't be conflated to mean we can be totally free and be the 'most moral agents' that are possible. We are therefore constrained by our circumstances: psychological, economic, genetic, geographic, temporal, etc. as to how moral or immoral we are. In that sense how can we really have a truly compatible sense of morality with determinism? These deterministic views trace back to Spinoza. The fact that it seems so counterintuitive to say we have no moral sense or moral impact on others is difficult to accept. Maybe the paradigm is outdated and the free-will versus determinism dichotomoy is an outdated one. I suspect this is becoming more and more the case.

There is mounting evidence to support such a universe (Deterministic with quantum randomness) from what we know about the laws of physics, evolution, neuroscience, neurobiology, psychology and other sciences that free will isn’t a scientific concept and can’t ever be thought of as one. Keep in mind that free will is largely a Western religious Christian concept tied to the idea of sin. There are other religions that have a more deterministic view of the universe and don’t have the need for the concept of free will such as Buddhism. On the other hand, what we value is so central to our sense of who we are as people that extinguishing all intentionality and volitional control over our actions seems absurd. But if there is no evidence to say we are totally free agents that can suspend the laws of nature to have complete free will and be as moral as we choose seems just as nonsensical.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .