Free will? Free of what?

Credit: Jens Langner (

The concept of free will is intimately tangled up with the idea of responsibility.  Are you responsible for your actions?  To what degree are your actions predetermined?  If they are predetermined, how can we hold anyone accountable for their actions?  Does the idea of moral responsibility even make sense?

Libertarian free will

The classic definition of free will is a will free of the laws of nature.  That is, you are free to choose independent of the workings of the universe.  This version requires a belief in mind-body dualism, the idea that the mind is an independent non-material spirit free from physical constraints.  If you do believe in this kind of substance dualism, then you would probably believe in libertarian free will (the word ‘libertarian’ here has nothing necessarily to do with the political philosophy).

However, if you consider the mind to be the brain, or that it arises completely from the brain’s operations, then  the laws of physics make libertarian free will difficult, if not impossible, to accept.  The workings of your brain are completely controlled by the laws of physics and every choice you make is ultimately determined by those laws, even if it is unlikely it will ever be possible to predict them.

What about quantum effects?  If quantum uncertainty enters into mental processing, wouldn’t that mean that our choices are not really predetermined?  There’s no real compelling evidence that quantum uncertainty does enter into mental processing any more than it enters into other physical processes such as computer chip operations (which would fail if they weren’t deterministic).

Even if quantum uncertainty is a factor in mental processing, that still wouldn’t save libertarian free will.  The random quantum events might save the decisions from being predetermined, but they wouldn’t make the decisions free from that quantum randomness, free from the laws of nature.

I’ve already written on the evidence that the mind is completely contained in the brain.  So libertarian free will seems to be ruled out.

Are we done then?  Many purists will insist that we are.  However, most philosophers who reject dualism still see free will as existing.  Why?  Is it simply a lack of nerve on their part?  Or are there reasons for their position?


Our decisions may be metaphysically determined, but that is no guidance when we are faced with an actual decision.  We still have to evaluate different courses of action and make decisions.  The fact that we can evaluate different options and then make a choice implies a certain type of freedom.

Purists will insist that nothing has changed, that this freedom is an illusion, being ultimately only the workings of neurons and synapses effected by electrical inputs from the senses, in other words, of the laws of nature.  This is true, but only in the same sense that this blog post is an illusion.

Strictly speaking, this blog post doesn’t exist.  You are looking at patterns of pixels on a screen, which were formed from communications of magnetic patterns stored on a hard drive in a datacenter somewhere.  You can’t point to the post anywhere without someone insisting that you’re pointing at something else.

Of course, if the post doesn’t exist, then what am I writing, and what are you reading?  At some level, the post certainly does exist as a useful pragmatic concept.  (Perhaps you feel this post’s content lacks value, but it’s still productive for you to think of it as something that exists that lacks value.)  The post exists at a level of abstraction above transistor states and molecular magnetic patterns.  A blog post is an emergent concept that exists on top of those lower level constituents.

In the same manner, our freedom to weigh options and make mental determinations exist, at a certain level of abstraction, as an emergent phenomenon.  This recognition that we still have choices to make, that at a certain level of abstraction,  we are still free to choose among available options, leads to compatiblist versions of free will.

Compatibilism is, of course, controversial, most notably among neuroscientists.  But the controversy is ultimately a definitional one.  It boils down to the question, is the term ‘free will’ still a useful one once we’ve given up on the libertarian version?

Moral responsibility

Among the experiences that affect us is learning knowledge of the consequences of our actions.  I know that if I commit illegal actions, and I’m caught, that there will be consequences.  I also know that if I commit immoral actions that aren’t necessarily illegal, I could still face consequences for my reputation.  This knowledge, among other things, affects my choices.

Consider two cases.  In the first, we have Bart, a person who had a bad childhood and lives in poor economic conditions, which leads him into a situation where he commits a crime.  In the second, we have Joe, who has a brain tumor that causes him to have delusions, which cause him to commit a crime.

Which one, Bart or Joe, acted with free will?  Which one should be held accountable for their actions.  Most people would say Bart should be held accountable, should be punished, but that Joe should receive treatment.  But why?  Both are victims of their biology and environment.  Why shouldn’t both just receive treatment?

I think the answer boils down to the fact that it is productive to hold Bart accountable for his actions.  It will have a deterrent effect, at least to some degree, on others with Bart’s background.  Many people have bad childhoods and live in poor economic conditions.  Holding Bart accountable may have an effect on their decisions.

But it’s not productive to hold Joe, with the brain tumor, accountable.  Punishing Joe would be unlikely to affect the actions of others with brain tumors.  Therefore, we would consider Joe to have mitigating circumstances, that he did not act with free will.

Of course, in holding anyone accountable, we should remember our lack of libertarian free will, and show as much mercy as possible while being compatible with deterrence.  It’s only by the whims of chance that we weren’t born with their nature and lived their experiences.


Free will, from a pragmatic point of view, means simply a will free from coercion, physical restraint, brain abnormalities, or any other unusual constraints that would preclude the usefulness of holding someone accountable for their actions.  It doesn’t mean freedom from the workings of normally functioning brain or the laws of physics.  And it doesn’t mean freedom from our experiences.

The people who say that libertarian free will does not exist are right.  The people who say that compatibilist free will exists are also right.  Once libertarian free will is ruled out, the remaining debate is a definitional one.  And like most such debates, it tends to be endless and pointless.

It’s important to understand what the limits of our freedom are, but also to understand that those limits should not absolve us of responsibility for our actions, except in extraordinary conditions.

Further reading

Everyone has an opinion on this.  People like Jerry CoyneSam HarrisDaniel DennettSean Carroll, and many others have weighed in on it.  Coyne in particular has blogged extensively on it.  I tend to favor Dennett’s and Carroll’s take, but I would since I agree with them.

58 thoughts on “Free will? Free of what?


    1. In what all ways can astrology be of use to modern man to live a successful life ? Don’t you feel dependence on it will stifle initiative? Is there scope for free-will ? How ?

    2. “ Astrology is valuable to me because it balances the hint of fatalism with the reality of being able to read the forewarnings in a chart to trnscend them by right-minded action taken in the light of the chart. “ Discuss.
    3. Argue the case for free-will against Fate providing a place in it for astrology in proof of its validity.

    4. Does Astrology deny free-will ? What is part played by karma in relation to Astrology ?

    5. Does astrology recognise the importance of free will ? Or is it deterministic ? Give reasons.

    6. Is there any place for free-will in astrology ? Give your reasons.

    7. Does astrology have a place for the freedom of the will ? Explain.

    8. Astrology is not a doctrine of fatalism: Discuss.

    9. Does Astrology deny freewill?

    10. Explain the importance of Fate and Freewill?

    11. To what extent is fate ‘determined in advance’ and to what extent is there scope for freewill and Karma after birth?

    12. “Those who know Astrology can only indicate in a way what will take place in future. Who else except the creator Brahma can say with certainty what shall finally happen”. Discus fate and freewill,
    in the context of the above saying..
    13. Discuss Fate and freewill.

    14. Why belief in destiny? Illustrate your answer with two examples.


    Our experience shows that free-will exists but it is regulated by the karmic limitations. A good exmple is the calf tied to a tree with a piece of coir. It can move freely within the circle, created by the length of the rope, but it cannot move beyond that circle. Same is the case of a dog or a bird, shut in a cage. At any rate, man cannot live without doing some action and the results follow the actions automatically, like the echo of a sound or shadow of a figure. As a result, the present is conditioned by the results of past karmas and the results of the karmas that we do now, shape our future. As already explained, prarabda karma is the results of the actions which we are motivated to enjoy or suffer during the present life itself. Out of these, the results which come under adridha, and a portion of dridhadridha also, constitute the free-will portion.


      1. You are 100% correct. Karma is cause and effect theory and Indian astrology is based on it. But as to the accuracy of astrological predictions, I am
        disappointed. I cannot demonstrate it, because
        there is nothing to demonstrate! All is guess work.


  2. I wrote a comment just last night on how morality is affected by free will. (My conclusion is that the results are no different, minus an excess of righteous rage.) It’s the first time I’ve ever written down my thoughts, so I’d be interested in your opinion.

    [A lack of] free will, from my perspective, is not the view of everyone being an automaton. Free will is an illusion caused by our complexity – we can’t hope to keep track of all the factors which influence our being. Ultimately, we are a product of our genetic material and the environment we live and have lived in, from conception until now. At any given moment in time, a given input into the system which is you will produce a predictable (if you know all the variables) output. Intent comes into play when you consider that you (your consciousness) may affect you (your body as a whole). Your conscious awareness allows you to assess memories and intercept emotional responses, altering the output in a way that is identified as your personality.

    If you have a personality which is considered moral and brave, you may find yourself jumping onto rails to save someone. Society keeps moral norms so that individuals may develop personalities which will benefit the group as a whole (saving a stranger). Therefore, personalities which express the intent to harm are punished as morally wrong, in order to influence the personalities of that society toward the moral standard. We behave as if we have free will.

    With or without free will, the solution to good or bad intent is the same – a moral response. The difference, however, is that those of us who deny free will find it hard to hate those who defy morality. (Not really true, it’s easy as hell. Till you think about it. As per The Righteous Mind, intuitive response first, then reasoning.)


      1. Yes, your post covered the debate surrounding the nature of free will, but my comment was concerning the impact of a lack of free will on secular morality. My conclusion is that there is no difference in the practice of morality between the potential natures of free will.

        I guess I feel like that should raise eyebrows. But I suppose compatibilism could be viewed as determinism plus secular morality, so maybe we really don’t disagree on anything at all.


        1. I think we do agree.

          On something you said above about someone who doesn’t believe in free will should find it harder to be angry at the immoral. In a passage I had to cut from my post, I talked about the intense passion many of the people who deny free will have towards those who commit heinous acts. I noted that if someone pointed out that those villains were acting without free will, they would almost certainly say that the two subjects had nothing to do with each other, and that my response would be, that’s the point; metaphysical freedom is largely irrelevant.


          1. I followed that part by partially denying it’s true. Hatred is a moral intuitive response, which you then are forced to reason out. When we are faced with someone who defies morality, we have immediate moral responses. This is why I said hatred is easy – it literally comes easily. As a determinist, I can bolster my reason with the knowledge that this is a product of the environment we live in, which brings out compassion. Punishment is still in order, but hopefully with an eye on reformation.

            So, I can’t say that the debate is irrelevant. I believe that if we lived in a world of determinists, society would be much more fair and peaceful.

            If you’ve not read The Righteous Mind, you should give it a shot. It covers evolutionary morality. If you do, my advice is to take notes as you go, it’s a lot to wrap your mind around.


          2. Thanks. I’m actually a big fan of ‘The Righteous Mind’ and Haidt’s work in general.

            As I said above, we generally agree. Except that I don’t see the term ‘free will’ worth fighting over. It’s too pervasive in society and in legal doctrines.

            Determinism is true, at a certain level of emergence. It emerges as an approximation of large numbers of quantum events, which are not individually deterministic. Chaos theory shows that it effectively disappears for complex dynamic systems at the higher level.


          3. I just read the intro yesterday and I will be taking notes, but I’ve a skimmimg reread of Gazzaniga’s ‘Who’s In Carge?’ to do first(for the expressed purpose of taking notes – new habit as it helps cement stuff in my head 🙂


    1. Well said, though that may not be worth a lot as I’m not much of a writer – have a hard time getting thoughts into coherent series of sentences. I do agree though with most, if not all, of what you’ve expressed. We are our biology and experience.


  3. Moral relativism is the view that moral judgments are true or false only relative to some particular standpoint. By this guideline and in the light of our “fee” cultures, we have free will. In my estimation, we are not the experts as individual. I believe therefore that morality must be legalized. Since we do not have an agreeable absolute morality, laws must be made to keep us on a civilized pathway to cooperation that is necessary to sustain societies. In the future this will present more opportunities for nations to join together in fashioning global cooperation. The universe apparently supports a relativist morality. I do not. We must work to censor and allow… while relativist view will continue to emerge as well as our societies evolve.


  4. Thanks for the link, SAP!

    Nevertheless, I worry that your rejection of libertarianism in general was too hasty due to you attacking a straw-man version of libertarianism that no libertarian accepts. As such, you haven’t really created a dilemma that moves us toward compatibilism, because you haven’t refuted the version of libertarianism that libertarians posit.

    Libertarians certainly do think the laws of nature apply to and affect the brain, such that the libertarian has no qualms in conceding that far reaching causal chains, enforced by the laws of nature and physics, are manifested as causal forces in the social environment. Events are determined, but not fully determined, by past causes and effects, because at the moment of choice the agent chooses that one effect obtains for the reasons that the agent had for making that choice obtain. It’s worth noting that the libertarian is in no way committed to the claim that all of our choices are freely will, only the once in a life time freely willed choice need obtain to vindicate the possibility of libertarianism.

    This is where emergentism, also known as property dualism comes into play, because emergentism allows the libertarian to explain the instance of agency without violating laws of physics such as the causal closure of the physical realm or the laws of thermodynamics. In the moment of agency no new energy is created because lower-level subvening properties realize higher-level emergent properties of conscious agency, and because mental states of agency are realized by physical properties, the realized state is itself a physical phenomena and is thus not outside of the realm of the physical.

    All of this is to say that the libertarian is not committed to the claim that the mind is unaffected by the laws of nature. What they are committed to is the claim that not every event is fully determined by past events. I must admit, even with this weaker claim, that I think libertarianism faces a difficult up hill battle in advancing that claim. So while I think you are right in being skeptical of libertarianism, your argument does not refute the core idea of libertarianism, and the libertarian can deny that your argument applies to their theory.


    1. You’re welcome. Thanks, as always, for the conversation.

      I fear I’m having trouble parsing this snippet: “because at the moment of choice the agent chooses that one effect obtains for the reasons that the agent had for making that choice obtain”. I wonder if you might be willing to reword it.


      1. Haha! Yes that is pretty convoluted, my apologies. The statement was an attempt to synthesize the views of the two standard-bearers of agent-causal libertarianism in the 1990’s, Randolph Clarke and Timmothy O’Connor (though Clarke has since become a compatibilist, interestingly enough; Robert Kane was the other predominate libertarian, though he is of the event-causal variety, which is not very defensible because it is subject to the indeterminacy objection you advanced in relation to quantum effects). Here’s how I would state the view of each theorist:

        Clarkes’ view is that past events influence but do not necessitate choices, such that, when an agent acts freely, they are the source of acting for those reasons, even if past events have caused the current set of choices, and caused some choices to have a higher probability of being chosen. O’Connor holds that an agent acts freely when they cause the internal state of having an intention to carry out an act, but other internal factors of deliberation make it that those reasons do not determine that that action be done. As a result, “our prior reasons can explain our actions without causally producing them”.


        1. Thanks! You do a good job of describing this in your agent causal blog post, but I wasn’t sure if it was the same reasoning.

          I can’t really see how Clarkes’ and O’Conner’s views are compatible with determinism. You might be surprised to learn that I do think an argument can be made that determinism for complex dynamic systems isn’t the slam dunk most people assume, but I’m not sure even that rescues their position. But maybe I’m missing something.


          1. You flatter me! I copied and pasted my statement of their views from my blog, so yep, the reasoning is roughly the same.

            I am surprised to hear that you admit that! But by that do you mean that determinism isn’t certain on the macro-level, or the more extreme claim that indeterminism exists at the macro-level, not just the quantum level? I welcome the latter approach, put forth by thinkers such as John Dupre. Dupre has argued that indeterminism is misunderstood — it is not the thesis that all events are random, after all, that would be a higher-order determinateness. Rather, that indeterminism is a macro-level higher-order event means that it is indeterminate whether some lower-order events will be determined or not. Dupre thinks that this kind of indeterminism does allow for agent-causal libertarian free will, because there are some undetermined events, and it is in this vacuum that agents call introduce causal control into the universe.

            I think this notion fits with Clarke and O’Connors theory. So you are quite right that their theories are not compatible with strict determinism. Earlier I was attempting to show that libertarianism is compatible with the laws of nature, I didn’t specify that some of those laws of nature are made true by indeterminism, as well as determinism.


          2. Oops, sorry. I projected my own de facto conflation of determinism and the laws of physics onto your statement.

            I guess the question is, if true randomness somehow enters into the picture, such that some mental events are effectively uncaused, is that the same as libertarian free will? I personally can’t see how it could be, since such an event would be unearned, unsought after, and uncontrollable by the mind. But I’ll grant that that couldn’t be demonstrated. (Of course, all of this is metaphysics and indemonstrable. 🙂 )


  5. You are quite right about metaphysics!

    But I don’t think I made Dupre’s position clear, what you have just described certainly does not constitute “real” free will, but it is not what Dupre had in mind. Dupre thinks that it is the nature of indeterminism that some events have antecedent and far reaching causal chains such that they were predetermined by prior events to occur, and some events do not have these causal chains in such a way that it is not determined what will occur next. That some events do not have these far reaching causal chains provide the opportunity for agents to inject their causal power into the universe. So free mental events are caused, and they are caused by the agent when a prior to undetermined event becomes determined by their realizing their causal powers.


    1. Actually, I do think you present it well. But I can’t see that this is sound: “That some events do not have these far reaching causal chains provide the opportunity for agents to inject their causal power into the universe”, at least not without injecting substance dualism (or perhaps a stronger form of property dualism than I would find tenable).


      1. I see, so you’re saying that to be a complete causal origin the agent himself has to be uninfluenced by past causes, and the only way to account for this is to posit a substance dualism where physical effects do not touch the mental substance that is the agent? Hmmm. I think that’s a pretty fair point.


  6. I wouldn’t write off Libertarian free will. Do we know the laws of Physics? Not yet in full. And just because a brain is bound by physical laws, does that restrict its thinking? Turing showed that a universal computing machine can compute any computable sequence. The limitation here is Godel’s incompleteness theorem, not any laws of physics.


    1. I actually agree that there are limits to determinism (see my post for this morning). The question is whether unpredictability is the same as libertarian free will. I can’t see that they are. Of course, this is a philosophical discussion, and if you define libertarian free will as unpredictability, then you have a shot. I’m not convinced that’s the free will most are hoping for though.


  7. This likely wasn’t nearly so complicated until Anaximander started over-thinking it. Philosophy – bleh! “but what if … ?”, “why?”, “but what if … ?”, ad nauseam 🙂

    That said I’ll have to the following a read someday:

    ‘History of the Free Will Problem’

    … but I find myself leaning toward much of what’s expressed in:

    ‘Fully Caused’

    … in that we are our biology and experience – oh, and I’m starting my reread of ‘Who’s In Charge?’ today(not-so-subtle reminder).


        1. That’s an interesting write up on Gazzaniga. I definitely need to read his book, if nothing else than to get the first person account of his split brain experiments. Might be a while though. I’m still slogging through ‘Farewell to Reality’ and really want to read a few fictions books afterward that have also been waiting in my Kindle list.


  8. Here’s a few more references for those interested. The first link details the problem of free will. The second is critique of Sam Harris’ take on free will or lack thereof. The third link is a critique of the neuroscientific criticisms of free will.

    Feel free to comment. Enjoy.


  9. Love the way you are thinking about it as merely a vibration or feeling in the mind is not complete because expressions of love are very much intertwined with the behavior which accompanies it. There is no realistic separation of feelings only without the corresponding behavior. When you love having sex it is feeling coupled with the sexual act and when you are infatuated and constantly thinking about your loved one it has behavioral consequences such as buying gifts and making plans for the next meeting mentally. Love is not only very intense pleasurable vibrations but it is about something you care about and it is mixed with a desire or feeling of wanting to be with the loved object. Love is very intense emotion mixed with desire and almost always accompanied by some form of behavior. Best wishes. Uldis


  10. I think it is reasonable to assume that we are able to choose whatever alternative we deem best. When we consider what course of action to take and take the best course of action, it is exactly that deliberation process that determines the action. Hence, when in the middle of deliberation, you know that you (are determined to) choose whatever action you will decide to take, hence the assumption of freedom.


      1. I see the debate mostly pointless. The assumption of freedom is an useful way to think of situations where we are able to take the course of action we deem best. I see no point in turning this useful way of thinking into a metaphysical problem.


          1. Mathematics has very little to do with metaphysics. Metaphysics has more to do with the philosophy of mathematics, and most mathematicians do not bother themselves with philosophy of mathematics. Different schools of philosophy of mathematics manage to justify the exactly same body of theorems from very diverse metaphysical points of views, so metaphysics seems not to affect the actual practice of doing mathematics.

            The same goes with science. Science is tied with empiricism, and you cannot empirically test metaphysics. Hence, metaphysics seem to have very little to do with science.

            To return to the original question about the free will, it is important for me to get to decide, when I am informed on the subject, I have strong opinions, ethical values etc. about the subject. In these cases my decision is in a very large extent determined by the aforementioned things, and any free will of the metaphysical type plays very little role.

            When my decision is not determined in such a manner, but it is more arbitrary, it is usually a less important one, for example, which brand of tobacco to buy, and I cannot see the significance of such decisions.


  11. First of all, thanks for joining me on twitter. I found your website title instantly attractive and wanted to peek inside. On the left menu of top articles, this title “Free Will? Free of What?” caught my eye and I decided not to read further. I felt I never came across such a logical practical question about universally loudly debated concept of free will and I should think of it myself in the lines of the question you asked in this title. Really, free will makes no sense unless we know the will free of something. I will come back to it once I am prepared to grasp your ideas, but I want your question to haunt for a while. Glad to know you. I am Anand.


  12. Good point that quantum mechanical uncertainty principle does not help in the argument for free will against determinism. Random chaotic choices cannot be considered free will any more than choices 100% predetermined by laws of physics and previous conditions.


      1. Yes. I’ve heard of this case on NPR and thought that it’s a great example for the discussion on free will, moral choices, responsibility, and justice. It’s amazing that the lawyers were able to pull this affluenza defense in court causing, of course, an uproar among friends and families of victims. This seems to set a quite interesting precedent of treating the conditions of upbringing similar to conditions of mental health. I don’t think, your Bart with low-life background would be able to pull something like this in court.

        On the other hand, I don’t think it would be useful to send Ethan behind bars for 10 years. He doesn’t seem to be a criminal and a threat to society and the whole experience of killing 4 people and going through this court might be sufficient to “cure” his “affluenza”. I doubt he will ever drink and drive again. So, 10 years of probation may be a reasonable sentence.


Your thoughts?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.