Libertarian freewill is the position that we have some measure of metaphysical freewill. Per this position, a free agent at a given point in time is able to freely select a course of action among several possible courses of action, i.e. given the exact same set of initial conditions, a system with agency has more than one possible course of action. Libertarian freewill thus implies indeterminism.

This is in opposition to compatibilism, which defines freewill as the ability to act according to ones own motivations, but that physically speaking, there is still only one possible outcome for one given set of initial conditions. Based on this definition, freewill is compatible with determinism.

Per libertarian freewill, if I am at an intersection, I am free to turn either left or right, and this independently of my past history and my mental state at the time. However, even a libertarian would concede that if I am falling from the 20th floor of a high rise, I'm not free to go anywhere besides down. Whatever circumstance led up to the point where I fell off the building have put constraints on my freedom to move in whichever direction I want to.

This shows, that even for those that accept metaphysical freewill, they are still bound by the laws of physics in some situations, and they would need to define the boundary between those situations where the laws pf physics limit their freedom and the situations where they don't.

  • How do those who subscribe to metaphysical freewill define those boundaries?
  • Doesn't the fact that metaphysical freewill requires such a boundary between what is constrained by physics and what is not imply a form of dualism? Does metaphysical freewill necessarily imply dualism?
  • 2
    +1 I'm really interested to see an answer to this one. The times I have asked such libertarian freewill subscribers this question, their answer is typically in the form of, "well, this rock has no freewill, it's physical. You and I have freewill. We can choose." Then they look at you like you're a idiot for even having to need it explained to you. I look forward to a good answer with citations!
    – Cort Ammon
    Commented Dec 10, 2015 at 0:30
  • As to your last bullet, did you consider idealism in your phrasing? Metaphysical freewill may possibly prohibit physicalism, but idealism, where everything is mental, still meets Libertarian freewill needs.
    – Cort Ammon
    Commented Dec 10, 2015 at 0:31
  • 1
    I didn't consider idealism because I feel it falls in the same category as radical skepticism, as in "yeah it's possible,...but come on!" -- if it were true, wouldn't just be the mirror symmetry of physicalism? If idealism is true, and the laws of physics don't hold, then we're all in wonderland or the matrix or something. Yay!!! Commented Dec 10, 2015 at 1:15
  • I like to include idealism for the same reason I like radical skepticism. It is remarkably good at questioning the assumptions that are so fundamental to science that we often forget we are making them =) Personally, I find idealist worlds useful for challenging the status quo. We often make physical assumptions that were not justified, and then build layer upon layer on top of them. Idealism has a tendency to shortcut those layers, even if its just an approach being used in a physicalist world.
    – Cort Ammon
    Commented Dec 10, 2015 at 1:30
  • That and, if you're exploring theories for which we have no known scientific way to determine which is right, why would you artificially withhold one (unless you have a non-scientific reason to withhold it)?
    – Cort Ammon
    Commented Dec 10, 2015 at 1:31

4 Answers 4


There are ways to reconcile libertarian free will even with classical physics. One could say (as was common position in 19-th century) that the laws of nature are only approximations and do not prescribe future events in every detail, free will is constrained only as far as the metaphysical margin of error. Non-Lipschitz forces (gravity, resistance) produce equations with non-unique solutions, creating “causal gaps” that can be filled by something extra, this strategy was advocated by Boussinesque and Maxwell in 1870s. Today, chaotic dynamical systems, which complex systems like living organisms contain, can be used for the same purpose, because infinite precision, required to assert determinism in them, is unattainable in principle due to quantum fluctuations. Finally, one could embrace quantum theory directly and claim that all nature has are statistical regularities we call its laws, but nothing metaphysically deterministic underneath, and “free agents” co-determine reality through interactions that respect the regularities.

Many philosophers designed their own models that reconcile regularity and libertarian freedom, so-called two-stage models of free will. James is considered the earliest, his first (“free”) stage consists of assembling available alternatives for making a decision, and is subject to chance, while the second stage (“will”) is determined by the agent’s character, etc. Poincare came up with essentially the same scheme based on his own introspections. Kane later criticized Jamesian models because

the chance ("free") part is not in the control of the agent and the "will" part is fully determined by a combination of the chance part and other determining factors, so the final choice is determined by factors, none of which the agent has control over at the time of choice”.

Kane himself is probably most persistent in playing on quantum indeterminacy to accomodate free will, see his profile on Information Philosopher, and short self-account in On Free Will, Responsibility and Indeterminism. In particular he challenges the traditional definition of libertarian free will as being able to "act differently in the same circumstances":

"Where events are indeterminate, as are the efforts they were making, there is no such thing as exact sameness or difference of events in different possible worlds. Their efforts were not exactly the same, nor were they exactly different, because they were not exact. They were simply unique."

Compton in 1950s revived Boussinesque’s extra "guiding principle" with the roles of stages reversed:

A set of known physical conditions is not adequate to specify precisely what a forthcoming event will be. These conditions, insofar as they can be known, define instead a range of possible events from among which some particular event will occur. When one exercises freedom, by his act of choice he is himself adding a factor not supplied by the physical conditions”.

Popper was generally sympathetic, and Kane himself produced a version: “undetermined self-forming actions or SFAs occur at those difficult times of life when we are torn between competing visions of what we should do or become”.

Also, while I agree that libertarian free will has to contend with natural laws, there may not be any “boundary” as in the two-stage models. To give an analogy, quantum evolution has both deterministic and random aspects, but unlike classical stochastic evolution it can not be split into them or composed from them, that is the content of Bell’s theorems. Quantum uncertainty may or may not be related to free will, but it shows that the traditional argument that the absence of strict determinism implies intrusion of “pure chance” at some point, because we could split any mixture and reduce, is invalid.

  • 2
    I do not think that an advance can be made on conifold's answer, "One could say that the laws of nature are only approximations and do not prescribe future events in every detail, free will is constrained only as far as the metaphysical margin of error." Except to say that the "margin of error" may not necessarily be "metaphysical," (I love the term "metaphysical margin of error" though) but simply our current inability to sufficiently accumulate and "make sense of" the data.
    – gonzo
    Commented Dec 10, 2015 at 5:08
  • Fair enough. But if there is no (or no sharp) boundary, from your answer, I still get the impression that a world which is neither random nor deterministic implies some form of dualism. Indeed that could be the very definition of the dual "realm" : As that which fills the causal gaps. Commented Dec 11, 2015 at 21:13
  • 3
    @Alexander S King I was intrigued by your expanded comment in another thread that freewill "entails a certain intentionality about the future state of the agent. Saying that freewill lives in the gaps left by uncertainty and imprecision amount to a form of dualism, almost by definition..." But if Dennett can dismiss intentionality as illusion why can't Kane say that full formation of intent is contemporaneous with exercise of freewill, and dismiss "that which fills the gaps"? Or perhaps interpret reality not as causal with gaps but as flux with regularities, this may be Marx's view of matter?
    – Conifold
    Commented Dec 14, 2015 at 4:36
  • 2
    @Conifold, I'm still trying to wrap my head around this the thing that I'm struggling with is truly metaphysical: Is there or is there not a 2nd form of indeterminism, besides quantum randomness? Is more than one course of action possible at some given point in time? Maybe I'm not understanding Kane, but from what I see, it's either dualism or epiphenomenalism - and I'm kind of hoping to avoid both. Commented Dec 14, 2015 at 5:37
  • 2
    @Alexander S King From what I understand he objects to the usual modal formulation of the question in the presence of indeterminacy:"Where events are indeterminate, as are the efforts they were making, there is no such thing as exact sameness or difference of events in different possible worlds. Their efforts were not exactly the same, nor were they exactly different, because they were not exact". There is more detailed presentation of his views here informationphilosopher.com/solutions/philosophers/kane
    – Conifold
    Commented Dec 14, 2015 at 20:19

The problem of freewill results from the fact that our subjective experiences from the first persons's viewpoint are different than our scientific observations and experiences from the third person's viewpoint.

When going bottom-up the levels of physics, chemistry and neurobiolology of the brain we successfully operate on each level with deterministic laws. While on the top level of conscious mental experience we subjectively feel free within a wide range. The problem is to bridge the gap between these two kinds of layers.

To bring into play the indeterminism of quantum mechanics has not proved to be a solution. On one hand, freewill libertarians do not propagate that our decision are by chance. On the other hand, it seems that understanding how neuronal assemblies run, does not require to operate on the level of microphysics.

1) The strategy of freewill libertarians is to postulate a special actor causality as emphasized in the answer of @conifold. The original question hits the weak point of this approach:

  • How to determine the boundary between the two different types of causality?

I did never found a precise proposal. Actor causality seems to me an ad-hoc explanation, hence a pseudo explanation.

Freewill libertarian make an assumption which they have never tested: That we could decide other than we actually did. In order to test this claim one has to reproduce the original situation and to repeat the decision experiment. How could such experiment look like? Did anybody perform such an experiment? Hence freewill libertarian use an untested and possibly even untestable hypothesis. But first one needs the data of the experiment, than one can search for an explanation.

2) The following book of Henrik Walter deals with the whole subject on the base of neurophilosophy:


  • Chapter 7.2 (German original from 1997) also considers a monistic freewill libertarianism.
  • I never heard of Henrik Walter. This is helpful. Commented Dec 11, 2015 at 21:16
  • @Alexander Walter is from Germany, now at Charité in Berlin de.wikipedia.org/wiki/Henrik_Walter
    – Jo Wehler
    Commented Dec 11, 2015 at 21:20
  • Is your answer then, to the original end questions, that dualism is implied? At least to me that's what a "special actor" implies (whether ad-hoc or not, doesn't it illustrate the point?)
    – LightCC
    Commented Dec 22, 2015 at 18:01
  • @LightCC "Doesn't the fact that metaphysical freewill requires such a boundary between what is constrained by physics and what is not imply a form of dualism? Does metaphysical freewill necessarily imply dualism?" Yes to both questions, I think so. But the physical component of dualism seems to be defined much clearer than the other component.
    – Jo Wehler
    Commented Dec 24, 2015 at 16:41
  • 1
    Current libertarians no longer postulate free causality, they rather reject the "every effect has a cause" postulate of traditional metaphysics. This exposes compatibilism as vulnerable to the same objection it raises: there never was an experimental test of determinism, nor can there ever be. Not a single natural law that we know is deterministic, they are all approximate statistical regularities. Just because precise laws apply approximately does not mean they approximate any laws that apply precisely. QM is not needed either, classical physics to finite precision is already indeterministic.
    – Conifold
    Commented Feb 1, 2016 at 1:54

This question mixes the action of the will with the original desire or choice of the will to attempt to carry out an action over time.


  1. Humans can behave irrationally
  2. Irrational behavior by definition does not have constraints
  3. Behavior follows from willed choices

Then it follows that human will can make irrational choices that have no constraints (physical or otherwise).

The Answer

It seems you are trying to put the will in a box it doesn't fit in - limiting it to a choice at a particular point in time among physical actions (similar to how the compatibilists define it). The problem with this is that when the will makes a choice, it is assuming that it will endure to carry that choice out. It doesn't limit it's decision making to a particular point in time. Our mind can also think abstractly and thereby have access to non-physical possible solutions to our dilemmas.

The laws of the universe, to the extent that they truly do exist, certainly limit our actions, but they do not limit our will. Now, granted, we "learn over time" and limit our own will through the normal feedback process of seeing the results of our actions and creating mental models of how the world works. So, sometimes we "imprison ourselves", but there is no requirement that we do so.

Therefore, my answer is that there is no physical limitation at all on the choice/desire of the libertarian free will (though the apparent physical choices often heavily influence our choices...)

If you wish to talk limitation, perhaps we need to talk about the limits of what the human mind/will are able to conceive.

Edit: It is worth noting, based on the comments, that the will doesn't even have to choose to address a particular issue that is presented to it physically. For example, people that are catatonic don't respond to sense stimuli at all. In social situations, we often change the subject or outright ignore the question or statement that we don't want to address. Human beings have a huge array of issues/problems/items they are working on in life, and while most of us typically set a priority on immediate physical issues (I would guess because they are typically an impediment to the longer-range, larger-scope things we are working on), that certainly isn't true in all cases.

Several detailed examples:

Example 1 - Walk through the wall

I can choose to walk through the wall in front of me. The result of that choice may be that I end up with a bump on my head, but the point is that I can make that choice and follow through with attempting to walk through the wall. The mind, whether casual or not, acts in an abstract manner that allows one to choose to attempt to perform physically irrational behavior.

Example 2 - Falling from a 20 story building

Someone falling from a 20 story building, actually has lots of choices for how to spend their time.

  1. They can scream, or not.
  2. They can attempt to slow their fall or somehow direct their airflow to land on a ledge a few stories down and perhaps survive.
  3. They can flap their arms, like wings
  4. They can close their eyes and imagine they are back in Kansas and there is nowhere like home
  5. They can meditate
  6. They can attempt to will themselves out of their current situation by mental power alone.

Granted, most of these will have limited to no power to change the physical situation their body is in (based on general understanding of our universe - cannot truly leave out solipsism, idealism, etc.), but the point is that doesn't affect their will to try to take action, or not take action, in various rational or irrational ways.

One thing I do know - it is highly likely that the subject will soon find out if the mind/will survives the physical body or not. Well, if it does not, they really won't "find out", but you get the point...

Example 3 - lifting a 50 ton weight

A better illustration may be a situation from the comments - lifting a 50 ton weight. First off, by free will I can choose to ignore it - why do I have to lift it? Second, I can attempt to lift it, or decide that wouldn't be worth the effort. Granted, I could try for an unreasonable amount of time, but who says the will is rational? I could literally kill myself attempting to move it (beat myself to death against it, exhaustion, etc.). I could even sit down and try to use mental power to move it.

Are most of these irrational? Based on generally accepted understandings of the universe, yes, and yet, I'm sure we could come up with examples of real people who have done similar things. Some examples can be found in the "Darwin Awards" - google if you are not familiar with these.

But the point is not that I will physically succeed in my willed actions, but that I can choose to take a course of action independent of whether the outcome is or appears to be limited by physical constraints

Beyond the irrational, I could will myself to move it, and thereby work hard to get the money over my life to buy the land the weight was on, and afford a crane that was capable of moving it. Moving that weight might be my life work, and I may or may not succeed in the end - but the point is I can make a "free will" choice (if one believes the libertarian side) whether to decide to attempt to move it or not, regardless of the physical limitation you say (and we all may agree) exists.

  • Either I misunderstood the answer or it misses the question. Suppose someone is falling off a building. You are saying he is free to flap his hands or meditate or do whatever else he chooses to do, for example, to count the floors on the way down, provided it is possible in the physical circumstances. But per @AlexanderSKing question, you should ask further, doesn't the physical state of the falling brain constrain it to choose to count floors on the way down, or to meditate, or scream?
    – nir
    Commented Dec 23, 2015 at 8:31
  • I edited to hopefully address. The basic idea is that I never have to address a physical situation at all. I can ignore it and "work on" a different problem in my life. I'm not forced by a physical circumstance to utilize my will, it just provides sensory input that may provide choices for issues/problems to address, if I choose to take them up. Granted, most of us would address the "falling from a building" as our #1 priority, but some would not (catatonic, etc.).
    – LightCC
    Commented Dec 23, 2015 at 15:51
  • you write: "The basic idea is that I never have to address a physical situation at all." but what is that "I" that never has to address a physical situation. Does your conception of the "I" include nothing physical? you have an ongoing electro-chemical party of about 90B neurons in your head — is it not a physical phenomena? does it have nothing to do with what you call "I"? @AlexanderSKing asked where does the physical end and the libertarian, non-physical begins.
    – nir
    Commented Dec 23, 2015 at 17:19
  • 1
    @nir I don't read his question that way, but I really only answered his first question. In regard to his final bullet point, I would say that libertarian free will does imply a form of dualism, which I believe is along the lines of your point. Where exactly is the boundary? Is that a knowable piece of information?
    – LightCC
    Commented Dec 23, 2015 at 20:17
  • 2
    you write "Where exactly is the boundary? Is that a knowable piece of information?" — and it seems a legitimate answer to the OP question.
    – nir
    Commented Dec 23, 2015 at 20:43

My two cents.

Libertarian free will requires the negation of determinism. The negation of determinism is that "not all events are determined", leaving room for some events to be completely determined and also room for law-like processes.

Furthermore, libertarianism requires some form of causality, compatible with negation of determinism, which makes free will meaningful. It would be meaningless to be able to choose but this choice amounting to nothing that can be related to the goal motivating the choice. In other words, libertarian free will requires that a choice's outcome reflects and is related to its cause (ie the goal motivating the choice).

Known physical laws can play this role of law-like causality as long as they do not amount to complete determinism. In other words as long as the future state is under-determined to some extent (future is open). For example, conservation of energy acts as a form of causality since at any given state the possible outcomes must be compatible with that energy level and not others, restricting the outcomes and also providing a causal link since any outcome is uniquely related to that energy state. In this respect standard interpretations of basic physics (ie quantum mechanics) are completely in accord with libertarian free will since they provide a suitable form of "causal indeterminism".

Quoting my other answer

Causality (ie claiming effects have causes) is not identical to determinism (ie effects are uniquely defined by causes in one-to-one manner).

Causality can co-exist with indeterminism proper, at least in the sense that causes limit the set of possible outcomes and/or provide constraints which the possible outcomes must satisfy and provide a drive towards a set of possible outcomes (and not others). So, even though which outcome is finally realized is open, it is bound to be from this set of outcomes, circumscribed and driven by these causes.

  • The negation of determinism is that causes do not determine their effects with absolute accuracy and that not all causes are events. In reality, we have all kinds of inaccuracies, uncertainties, probabilities and alternative possibilities. We also have agent causation where agents can self-cause their own actions. Some people call agent causation by the name free will. Commented Mar 27 at 9:35
  • @PerttiRuismäki I would not like to restrict free will mechanisms to agent theories vs event theories or others. it is possible though.
    – Nikos M.
    Commented Mar 27 at 9:43
  • M Free will is not a mechanism or a theory. Free will is a concept in psychology, our ability to make decisions. Nothing to do with physics. Commented Mar 27 at 11:33
  • @PerttiRuismäki physics is a description about how processes function at some level. Psychology is about other level. So in a sense if a level is compatible or incompatible with free will, it has important consequences for other levels.
    – Nikos M.
    Commented Mar 27 at 11:47

You must log in to answer this question.

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