Suppose libertarian free will exists. Say that a person is presented with choices A and B, and she chooses A. Then, her memory is wiped, and her brain, body, and surrounding conditions are reset to the exact same conditions as when she chose A the first time. The only difference is that she is at a different point in the timeline and that she, though unknowingly, has chosen A before. Will she necessarily choose A again? Will she possibly choose B?

I want to know what philosophers believe about the implications of libertarian free will in a situation like this. To me, there are problems with both views. If she necessarily chooses A, then free will would seem independent of time and fundamental to the person. The problem is that it doesn't seem as if she made a free choice the second time, since it had to be the same as the first time. On the other hand, if she could possibly choose B, then free will seems to merely be dependent on the point in time, and it is hard to assign praise, blame, reward, or punishment to her for choosing A, if the same person might have chosen B in the same situation.

To me, the former (always choosing A) is more palatable than the latter (maybe choosing B), but both are puzzling. I can perhaps explain the former situation by saying that she makes the choice once and for all outside of time in a spiritual realm, whereas the latter situation is simply disturbing and chaotic.

Is there an option philosophers believe is more harmonious with the concept of libertarian free will? Are they divided on this? Has this been discussed much at all?

  • 4
    Even if you sent the person to the same moment in time and recreated her exact brain state at that moment she may still choose B instead. That is what "freedom to do otherwise" means on libertarian conception. That the same person might have chosen otherwise does not prevent assigning praise or blame for what the person did choose. Each choice adds to the person, and it is the person after the choice that is praised or blamed. Overcoming fear when succumbing to it was an option makes her not less praiseworthy but more.
    – Conifold
    Commented Sep 17, 2022 at 3:13
  • 2
    Your question is completely useless speculation on the impossible. It is not logically possible to revert the Universe or even the test person's brain to a prior state. There is nothing we could learn from such a thought experiment. "The ability to choose otherwise" is a bad definition for free will. We can choose only one-wise, there is no other-wise. We can make a choice only once. The next time it's another choice. Commented Sep 17, 2022 at 3:49
  • 3
    @PerttiRuismäki You are mistaken. It is certainly logically possible to put qualitatively identical people into identical circumstances, even though it may be physically impossible. But we have our physical theories consider impossibly idealized situations, such as frictionless motion or absolutely rigid constraints, all the time, and gain important insights from that. Because there are possible situations that are close to them. And there is certainly nothing impossible about similar people deciding in similar circumstances, and nothing wrong with idealizing.
    – Conifold
    Commented Sep 17, 2022 at 5:29
  • 1
    That bit about making choices once and for all in a spiritual realm is interesting, it reminds me of Kant's view of this question. For he did say that we make all our choices outside of time and hence there is a sense in which we have already always chosen to do A at whatever local time we do it. OTOH Kant is one of my favorite dead horses so IDK if I should post up a reply based on his one-choice-for-all-time remarks (he does complicate the matter by bringing up that redemption, on his view, would require timelessly remaking our fundamental choices). Commented Sep 17, 2022 at 10:02
  • 1
    Unlike primitive reflexive stimulation of the subconsciousness, you may regard the advanced free will mark of consciousness as a highly delicate nonlinear system with many stable attractors, repellors and other extremely indeterminately vague non-stable states, not much different from the famous n-body problem. And once the free will capable agent fully conscious of the good and bad, even under determinism most contemporary moral philosophers would judge it's compatible with ethics... Commented Sep 18, 2022 at 20:02

3 Answers 3


According to the Stanford Encyclopedia of Philosophy (SEP) entry on Free Will:

nearly all libertarians agree that exercises of free will require agents to be reasons-responsive (e.g., Kane 1996; O’Connor 2009a; Clarke 2003, chs. 8–9; Franklin 2018, ch. 2).

In other words, most agree that even 'free' decisions occur not in some realm isolated from any and all influence, but in response to extant conditions.

It also typically maintains that the responses to these conditions are somehow in the hands of the actor; that given a choice between an apple and an orange for example, that a person may freely choose either the apple or the orange (or neither/both). This suggests that - were it possible for such an agent to re-experience an identical choice in identical conditions (including identical temporal conditions) - that they might be capable of making a different choice.

Noting that your description of 'a different point in the timeline' constitutes a non-identical set of conditions, this would seem to be irrelevant to an agent which possesses the kind of freedom of choice/action described here. The freedom that exists at point A would also exist at point B, if there is nothing about the choice 'apple or orange' which changes at point B.

Does libertarian free will necessarily choose the same thing every time?

It seems to fair to say that libertarian free will would never necessarily choose anything, lest it cease to become libertarian.

Even a coercive, 'gun-at-the-head' command allows for the libertarian agent to freely make a choice, whether it be self-preservational or self-sacrificial; a choice which, given an identical circumstance, could have been made differently.

Of course, we never - as far as we know - go back in time, and so the notion of 'could have done differently' is impossible to test, a circumstance which presents a seemingly insurmountable hurdle - in the absence of any new neurological discoveries - for any aspiration to figure out whether or not we have free will.

  • If reason itself has vague/hazy/indeterminate substructure, I wonder how this might affect reasons-responsiveness. Would it mean that some responses (those which are responses to vague/w/e kinds of reasons/subreasons) are themselves indeterminate? Commented Sep 17, 2022 at 12:15
  • 1
    My immediate response is that that kind of vagueness/haziness would be hazy/vague only to us; ie. that actual reasons - whether or not we could detect them - exist nonetheless and are (perhaps predictably) influential. But I'm not sure I've properly grasped your point. Commented Sep 17, 2022 at 12:23
  • Well, for one, it seems (to me at least) that the boundary between logic and mathematics is vague. This is not due to us, though, so much as to what the two structures are: they involve so much generalization and universalization and absoluteness that they both project into each other, and into science as well arguably, rendering notions of stable objects suspect. So much of apparent physical reality is abuzz with flux and randomness of circumstance, which scientific technology has solved in many cases, but then due to us, oddly enough (we were the ones who stabilized the world this much!). Commented Sep 17, 2022 at 12:31
  • @KristianBerry All men are born free but are everywhere stabilizing the world.
    – Scott Rowe
    Commented Mar 28 at 11:12

Here is a link to one of the most respected contemporary dualist and libertarian free will philosopher's (Richard Swinburne) take on free will: https://www.amazon.com/gp/customer-reviews/R18J8OJA7QPLKX?ref=pf_ov_at_pdctrvw_srp.

Swinburne's thinking would apply several principles to your thought problem:

  1. Our physical universe is not deterministic, because it is not knowable at the quantum level. The circumstances in any time can not be "identical" because they cannot be fixed, due to the Heisenberg uncertainty principle. Therefore, quantum effects in our brains, and our interface with the world, can amplify into different decisions even if we did not have will to influence a choice. We could have done A or B even if will were not exercised.

  2. Most choices are not based on free will, as they are not moral choices. Swinburne limits free will solely to morality.

  3. Swinburne applies virtue ethics to human behavior, which assumes we develop habits of behavior, reinforced by prior choices. The outcome of the first choice, if it were a moral one, would make an imprint on our moral character, which is spiritual. This means that there are different outcomes in his model for moral vs non-moral choices.

  4. For moral choices, the choice itself, and how we react to our making it, will have an influence on our moral character, which is part of our spiritual self. Reset the physical situation as best one can (within Heisenberg limits), and the probabilities of making the same choice will change -- they may go up, if we were morally satisfied with our prior choice, or go down, if we were dissatisfied. Repeat this thought experiment 100 times, and for a moral choice, we would settle into a consistent outcome (which may not be the one chosen first), and one will eventually have only one "possible" outcome morally, even despite the random inputs of Heisenburg on our brains.

  5. For non-moral choices, Swinburne's free will concept is less clear. Our habits of action are reinforcing in Virtue thinking, and this should not be limited to only moral habits. Most Cartesian dualists would put some significant part of those choices in the spiritual realm, and, if Virtue thinkers, would make the same inference to increasing consistency in all such choices. I am not sure that Swinburne would do this, due to his limiting free will to morality. Swinburne may be holding that for non-moral choices, our developing habits of thought may only be based on modifying our neural pathways, therefore a physical reset would have the same probabilities of A or B every time.

  6. I initially limited your thought problem to a physical reset, but Swinburne puts the spiritual realm within time not outside time, and the spiritual realm is therefore changeable, hence is malleable. So for his model, an omnipotent experimenter could ALSO reset the spiritual state of an agent. If that were done, then for any choice A or B, the same probabilities would apply to a revisited choice, for both moral and non-moral choices. So for a physical AND spiritual reset, we could do either A or B, and one can (by repeating the experiment enough times) figure out the likelihood of our doing either.

As I implied in my review, despite being a dualist virtue ethicist I do not agree with all aspects of Swinburne's thinking. I would say that all choices have a free will component, and that spiritual character would influence all of them. I also disagree that one's character becomes fixed, and mental inclinations are nearly as determining as Swinburne asserts. Specifically, people can regret even a lifetime of choices, and reverse course -- which Swinburne's deep channel model of character and overwhelming nature of inclinations prohibits. Swinburne implicitly admits to this, in his discussion of what he considers an incomplete exculpation for blame based on a moral course change late in life. But he does not then explain how such a moral course change is possible with his fairly rigid inclinations model of the mental world. These caveats I have to his thinking, would modify the above by collapsing 5 into 4, and for 4 would allow that even 100 times into your thought problem, our choices would not be fully predictably settled into one direction.

  • 1
    "Sow a thought and reap an act. Sow an act and reap a habit. Sow a habit and reap a character. Sow a character and reap a destiny."
    – Scott Rowe
    Commented Mar 28 at 11:07

One cent.

According to libertarians the person presented with same options under same conditions need not choose similarly as in previous time (ie ability to do otherwise).

Of course there is nothing wrong with the person choosing the same over and over, as long as one understands that all these are new choices made and not a copy of a predetermined choice (libertarianism).

Casually speaking, a person having some reasons to choose option A or B, will re-choose it under same conditions and same reasons, (there are cases where the same conditions and reasons may result in choosing differently everytime, eg choose a choice not chosen previously, etc) however one should make clear that all these do not predetermine nor fix the choice, it is (or should be) free.

Libertarianism requires the negation of determinism, that is "not all events are determined". In this sense, of not being pre-determined, a reincarnation of the choice situation need not, in principle, lead to the same outcome as in previous case. This is related to the ability to do otherwise principle and the open future principle of libertarianism.

  • 2
    I upvote this one every time I first encounter it.
    – Scott Rowe
    Commented Mar 28 at 11:14
  • 1
    @ScottRowe right on target..!!!
    – Nikos M.
    Commented Mar 28 at 11:14

You must log in to answer this question.

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