Debates on free will often raise the question of evidence: who will have the best evidence? However, before discussing evidence, I would like to know who bears this responsibility in philosophical and scientific discussions. Should it be borne by proponents asserting the existence of free will or by skeptics challenging its validity, given advances in neuroscience and psychology?

  • 8
    I don't debate with determinismists... they are so predictable. Commented Apr 30 at 21:22
  • 3
    There have been other questions about burden of proof. Why should the burden be different for this than any other issue?
    – Barmar
    Commented May 1 at 14:27
  • 1
    Neither has a choice.
    – rasher
    Commented May 2 at 21:49
  • Whoever makes any claim bears the burdern of proof of that claim.
    – Rabbi Kaii
    Commented May 7 at 23:59

9 Answers 9


Honestly most of the answers here are absolute waffle. Here's the actual answer. Both of them. Burden of proof applies if you are making a claim. That's it. Sometimes people will emphasise the point more by saying 'positive claim' but that doesn't actually mean anything different in this context.

People are used to thinking only one side has burden of proof from arguments where someone is trying to convince the other party of their claim and the other party is not believing. The important thing is that these situations only arise if the other party PURELY lacks belief in the claim, NOT if they also start arguing the opposite or for an alternative.

The most widespread example I think people will have picked up on this is arguments on the existence of God. The atheist shouldn't be expected to 'prove' God isn't real because they're not claiming anything, they're simply currently unconvinced by the claims of others. However, if this atheist then made the claim that God ISN'T real (not just that they currently don't believe it is), they would also have a separate burden of proof for this claim. You can flip the situation by removing the believer's claim. If a believer is minding their own business and someone just starts claiming to them that God isn't real, they should not be expected to prove that God is real, lest the atheists claim be treated as correct.

To finalise the point, an advocate for the existence of free will has the burden of proving their claim. The skeptic also has the burden of proving their claim. If one of these sides dropped their claim and was simply considering and critiquing the claim of the other side, THEN there would be one side whom had the burden of proof (who would be whichever side was making the claim).

  • My answer explains how that justification requirement already was met in the strongest way we can by free will based on pragmatic utility. Ted Wrigley’s answer points out how little justification that determinists provide.
    – Dcleve
    Commented May 3 at 9:17
  • If you are going to posit equivalence between someone arguing for the existence of a thing, and someone arguing that we should assume it doesn't exist until evidence is provided, you are going to waste an awful lot of time arguing against absurdities. In the context of free will, we can be reasonably certain that the vast majority of the 'stuff' in the universe doesn't have it. Therefore, the claim that humans do (uniquely, or nearly so) is the extraordinary claim, and should carry the burden of proof, IMHO.
    – Paul
    Commented May 6 at 11:19
  • @Paul I don't posit that equivalence. In fact you should find it clear that I was pointing out the crucial distinction between those 2 things as the decider of whether a party has burden of proof or not. Talking about the ordinary-ness of claims is something for your argument, not for meta discussions. It doesn't effect burden of proof. If you're claiming something, acting as if it were true without you proving it is unreasonable. Hence burden of proof. It's no more complex than that before it becomes inaccurate.
    – Blue_Crow
    Commented May 9 at 11:28

The Burden of Proof isn't an absolute property. There's no experiment you can perform on a person, a brain, or a position to show that that thing has the burden of proof.

The Burden of Proof is a social construct. Consider that in physics, there's a concept of inertia: objects at rest stay at rest, objects in motion stay in motion. In the realm of ideas and beliefs, the same concept is there: we have inertia with respect to our ideas. Just like it takes work to change the motion of an object, it takes work to change the beliefs of a person or of a group of people.

The Burden of Proof is a way of expressing the need for that work. If you want me to change my beliefs, you must put in some work to overcome my inertia.

If you don't have any desire to change someone's beliefs, then you also have no burden.


Human beings work on a presumption of free will; few people (outside of certain psychological disorders) believe they have absolutely no control over their actions moment to moment. Therefore the burden of proof lies with the determinists.

This is exactly the same as the belief that the sun rises and sets. This is how the world presents itself to direct sensory experience, and so the burden of proof fell on physicists to show that (in fact) the Earth rotates giving the appearance of the sun rising and setting. It may be that the presumption of free will is wrong, but until determinism is demonstrated through some empirical research we must (perforce) go with the way our minds present to us, as having a capacity for free will.

It's ridiculously easy to assert a counterfactual; every precocious adolescent does it with abandon. But counterfactuals are only as useful as the evidence supporting them. The argument against free will is (currently) a purely theoretical counterfactual that flies against what seems to us to be true. I won't quibble with the framework, but it lacks any connective tissue.

  • 2
    I like the first two paragraphs. A general discussion of counterfactuals causes me some concern. I think counterfactuals are inherent in the effort to reason. In science the ideal case, or in law the comparative case, is a counterfactual case. Galileo, for example, uses counterfactual reasoning to develop our best concept of motion (force changes momentum). He argues that in the absence of resistance, for example, a rock and a feather would fall at the same rate of acceleration; or a body projected at initial speed along a horizontal plane continues forever at that speed in a straight line. Commented Apr 30 at 19:06
  • 2
    @SystemTheory: Counterfactuals are marvelous as a starting point. We have to be able to imagine what doesn't seem to be if we want to understand what is. But they are just too easy; we can dredge up counterfactuals almost as fast as we can speak (e.g., the earth is flat, and the moon landing was faked, …). We have to privilege what appears to be true on face value, so that counterfactual propositions have to demonstrate some traction before being taken seriously. Otherwise we just get swamped with wild claims. Commented May 1 at 1:10
  • 1
    Having control over your actions does not imply free will, as that control could be utilized in a way that is not free. This becomes cleat when you try to answer the question "Why do people will what they will?"
    – rus9384
    Commented May 1 at 17:13
  • 2
    @rus9384: That would be a speculative counterfactual. What appears to us is that we have free control over our own actions. That appearance may be wrong, but the burden is on those who must demonstrate that it is wrong. The minute you use the phrase 'could be' you assign yourself the task of turning that 'could be' into an 'is'. Commented May 1 at 17:48
  • 1
    @rus9384: I think you're conflating the philosophical concept of 'free will' with the political concept of 'liberty'. Commented May 1 at 21:46

In dealing with our world, we have developed pragmatic models of a) selfhood, b) agency c) a physical world d) other mind are agents, etc for a large suite of conventional worldview assumptions. These models are developed in infancy and early childhood, and have the support of empirical pragmatism.

Denials of free will, agency, selfhood, time, consciousness, the external world, mathematics, morality or values, etc therefore have a burden of proof to overturn our cumulative pragmatically evidenced worldview.


From my perspective, both proponents and skeptics of free will should shoulder the responsibility of providing arguments. Proponents must offer compelling evidence demonstrating the compatibility of free will with scientific understandings of determinism, while skeptics should present thorough analyses of scientific findings in neuroscience and psychology that challenge the concept of free will.

  • 4
    Science shows our universe is not deterministic. philosophy.stackexchange.com/a/88579/29339
    – Dcleve
    Commented Apr 30 at 15:56
  • @Dcleve and yet the answers on physics SE (linked as a comment to your answer) dont say that at all...
    – JMac
    Commented Apr 30 at 17:20
  • @JMac That voting on PhysicsSE does not correspond with the physics references I provided or the voting among professional physicists arxiv.org/abs/1612.00676 is a QC issue for physics SE.
    – Dcleve
    Commented Apr 30 at 20:21
  • 1
    @Dcleve That's like using that philosopher survey to determine the nature of reality. Those questions ask about physicists opinions, and it doesnt seem like that suggests they are saying "science shows our universe isn't deterministic". At best I suspect many would say "I believe our universe isn't deterministic based on our current knowledge"; but we really dont understand the science enough to say it "shows" our universe is non-deterministic. The science isnt there.
    – JMac
    Commented Apr 30 at 21:01
  • @JMac Do you know what determinism means? It is that outcomes are knowable and not probabilistic. Neither are true of the “interpretations” endorsed by 97% of the physicists surveyed. Do you accept science/empiricism? That operates on the acceptance that expert consensus DOES reveal the nature of reality.
    – Dcleve
    Commented May 1 at 8:08

the burden of proof lies on proving existential claims: if one claims that "there exist (a) non-white swan(s)", then they better exhibit a concrete example

this happens because the logic of empirical matters and observations is not classical, but constructive: even if it's possible to somehow 'prove' ¬∀x¬φ(x), the proof may fail to explicitly provide an object a such that φ(a) holds, and in such cases asserting ∃xφ(x) becomes suspicious; dually, it's not in general possible to "prove non-existence", that is, ¬Exφ(x), for this entails ∀x¬φ(x): even if all swans were in fact (non-non-) white, how could we hope to prove so?

just think about it: if some purported object simply happens to not exist, how exactly would it leave traces or evidence behind it, and of what kind?

  • 1
    does the same go for moral responsibility?
    – andrós
    Commented Apr 30 at 14:14
  • @user66697 you mean as in existence of 'moral truths'? sure
    – ac15
    Commented Apr 30 at 14:15
  • 1
    @ac15 Unless you think that there are Schrodinger's stars, then either there are stars outside the observable universe or there aren't. Exactly what is the alternative? BTW I very much doubt you will find a reputable astronomer or cosmologist that isn't confident that there are stars outside the observable universe - it isn't at all controversial. Commented Apr 30 at 15:32
  • 1
    @DikranMarsupial "Such "non-classical" reasoning would make science impossible" That is quite a counter-factual: even just talking of mathematics, that most if not all of mathematics can be developed constructively is rather known. Commented May 1 at 13:06
  • 1
    @DikranMarsupial But indeed my point is that that does not even hold in maths, let alone more generally: that we need classical reasoning. Indeed, the scientific method par excellence is about experimental evidence as the criterion of truth, so rather belongs to the "(inductive and) constructive" approach if one has to look at the logic of it. Commented May 1 at 13:36

If you have free will, you can decide whether you want to give evidence for your opinion of not. If you don’t have free will, then you don’t have the choice. And of course your opinion may not agree with reality.

So this is a weird question. Most questions about free will are weird.

  • Only people that have free will can change their opinion, so the burden of proof lies with the determinists. Commented May 1 at 20:04
  • 2
    Determinism does not deny the ability to make decisions or provide evidence for an opinion. On the contrary, it argues that our decisions result from preceding factors and natural laws, but this does not necessarily imply that we cannot make decisions or act according to our will. It's more about determining the influences that shape our choices. Thus, while determinism may challenge the nature of our freedom, it does not mean that we cannot deliberate, choose, and act. Instead, it questions the process of making these "choices". Commented May 2 at 7:54
  • 1
    @LittleEndian Says who? People without free will can change their opinion. Not only can they change their opinions, they have no choice.
    – gnasher729
    Commented May 9 at 23:06
  • If someone's decisions and opinions are entirely a result of preceding events isn't it more like the decisions and opinions are being made for them? Therefore they can't "change their opinion" or "make decisions". Commented May 10 at 4:19

There are "existence" claims on both sides of the determinism/libertarianism debate. Both sides in some sense claim that something exists, either complete determinism or freedom of will (*).

One can certainly have a preference or disposition towards one position, of the debate, without necessarily having to justify this thesis to anyone. For example, one position might feel more natural in relation to the other. Certainly, in this sense, we all act under the unmistakable sense of having a degree of agency, and, denying the reality of this feeling is something of an extraordinary claim, in need of similar extraordinary justification.

In spite of the above, we can scrutinize the arguments one may use one way or another.

Determinists have to argue for the actual existence of complete determinism. Partial determinism, which is what science has observed and established up to now, does not mean complete determinism and failure of libertarianism. In fact, partial determinism is required by libertarian position. At most, established partial determinism can be used as a hint for the hypothetical existence of complete determinism that nevertheless needs to be verified.

Libertarians, on the other hand, need to argue for freedom of will. Lack of freedom or freedom that is not willed cannot do. Libertarians can point to established partial determinism as supporting free will by enabling both freedom, and a form of causality, both needed as prerequisites for willing. Furthermore they can point to various, objective and subjective, ways a free action can be attributed as up to the agent. For example, a) stated goals are indeed reached with such high probability that is substantially statistically significant than pure chance and/or complete lack of agency would make possible, or b) the existence of the irreducible sense of willing that cannot simply be explained away as global illusion, etc..

(*) Compatibilism in the sense of fully deterministic (unfree) will is not addressed separately but there are arguments addressing it here or there above. That said, will that does not actually will anything, since everything is bound to happen anyway, is rather a meaningless concept.

So, weight the arguments and feel free to choose..


Free will is not a theory or any matter of belief. Free will is a matter of definition. Free will is just a name given to a phenomenon that can be either real or imaginary.

Within the framework of one definition there are no advocates or skeptics. The definition tells, whether free will is a real phenomenon or an imaginary one.

In the absence of one "official" definition the debate is about what would be the most appropriate definition for free will. In any case free will exists, we just have to define what it is.

You must log in to answer this question.

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