Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

If I Could Choose Again

  1. Apr 4, 2007 #1
    I have often posed this thought experiment in debates on the “free will” question, and rarely does anyone provide a coherent answer.

    The thought experiment is this :
    Firstly, imagine that you take a “free will decision” to do something (for example, you decide to have tea rather than coffee with your breakfast).

    And now imagine that the entire universe could be “reset” to a few moments before that “free will” decision, such that the properties of absolutely everything in the universe (including your internal brain states, desires, intentions, wishes, wants etc) are identical to the way they were the first time.

    And now you choose again.

    Would your choice necessarily be the same as it was the first time?

    If “yes”, then this is determinism pure and simple.

    If “no”, then what logical or rational explanation (apart from random or stochastic behaviour) could you possibly give for choosing differently?
  2. jcsd
  3. Apr 4, 2007 #2
    Even though I don't see how our will might be free (from what, everything?) this experiment seems flawed. If an experimenter sets your intentions, wishes and wants to some previous state, is he not setting your will? How can you then call it free? The second decision to have tea is set by the experimenter to match the previous one. Maybe this is your point.
  4. Apr 4, 2007 #3


    User Avatar

    Staff: Mentor

    I don't see that contradiction, but I also don't see a logical reason why making the same choice twice implies a lack of freewill. If the situations are exactly identical, then your logic in making the choice is identical, regardless of whether the choice is really free or not. So I don't see how this thought experiment shows anything at all.
  5. Apr 4, 2007 #4
    I think no, not if you have true "free will" to "think or not to think". That is, the second time around you may, using your free will to think about what to drink, freely decide not to think about either tea or coffee and decide to take a glass of milk. Unless milk is allowed this second time around, then you are correct, all is determined. If milk allowed, then the "logical explanation" for choosing different action second time is called "free will", when you accept that the definition of "free will" = volitional action of mind "to think or not to think".
  6. Apr 4, 2007 #5
  7. Apr 5, 2007 #6
    The problem is that if your brain and body does not control your will, then that assumes duality or metaphysical properties which are outside the physical, and thus not provable.

    Another problem is also that everything, even the metaphysical, must be deterministic.
  8. Apr 5, 2007 #7
    I totally disagree with the idea of free will to a maximum level available since we always make decision dependent on some other issues that guide us to do so. We may not realised it but all our decision are dependent on other experience that we have faced before!
  9. Apr 5, 2007 #8
    In a way, yes it is my point. The "experimenter" here is not actually doing anything to cause you to decide one way or another, he is just a passive experimenter who is simply replaying the outworking of the universe, he is not interacting with or affecting that outworking in any way.

    What is "flawed" imho is the idea that free will (of the libertarian kind) is a coherent notion - and this incoherency is what the thought experiment illustrates.
  10. Apr 5, 2007 #9
    If your choice is necessarily the same the second time around, then this is (as I said) determinism pure and simple. (determinism in simple terms is the thesis that "same past = same future").

    In what sense could you consider your will "free" in this case? Only in the compatiblist sense, certainly not in the metaphysical libertarian sense (the latter being incompatible with determinism).
  11. Apr 5, 2007 #10
    OK. If "milk is allowed" then it is allowed both times (this is the way the experiment is defined).

    And can you then offer a rational or logical explanation as to why you would choose milk the second time and not the first (given that all the antecedent factors are identical in both cases)? What rational explanation can we give for choosing differently, apart from random behaviour?
    Last edited: Apr 5, 2007
  12. Apr 5, 2007 #11
    The spiritualist might explain this by pointing out that if the will is indeed free then it is unrewindable. Since your experiment resets my will, it obviously assumes that it is not free, so you are proving your hypothesis by postulating it. You may be able to reset the material universe, but not the spiritual mind.
  13. Apr 5, 2007 #12


    User Avatar

    Staff: Mentor

    You're defining determinism and the scenario in such a way as to make it the only possibility. You're essentially saying 'if the universe is deterministic, is the universe deterministic?' So like I said, the thought experiment doesn't say anything about anything.

    Drop the word "necessarily" from the question and reconsider it....
    Last edited: Apr 5, 2007
  14. Apr 5, 2007 #13


    User Avatar

    Staff: Mentor

    "same past=same future" doesn't say anything at all about whether or not you have freewill. Consider the difference between hard and soft determinism. Basically:

    Hard determinism: You don't have a choice.
    Soft determinism: You do have a choice, but you'll make the same choice.


    Also, you may want to consider whether your premise is even scientifically possible. It may not be reasonable to assume it is possible to have identical states.
  15. Apr 5, 2007 #14
    Because humans are given no automatic means to decide such things, humans are constrained such that they must make "choices" about which specific "choice" of action they will take (eg. drink tea [T], coffee [C], milk [M], etc. the second time around) to reach an end state or goal--let us call it [E].

    So, for the simple example under discussion, let [H] = a human, [E] = goal (happiness from drinking) and [T], [C], [M] possible sets of actions to obtain [E]. Thus we have three possible trajectories:
    [H] --> [T] --> [E]
    [H] --> [C] --> [E]
    [H] --> [M] --> [E]

    So, at time (t)=0 suppose action [H] --> [T] --> [E] is selected. Now, we move forward in time, then rewind back to (t)=0, must [H] select same action ? I say no, not if we realize that the goal of the section process at (t) = 0 never was to select [T], [C], or [M] in the first place--but it was to select [E].

    In other words, whenever [H] has to make choice among many different actions to reach a goal, and by definition the trajectory of action is not unique, then what we have is the process of selection by a "Markovian Machine". So, while the reaching of a goal [E] by a Markovian Machine is "determined", (that is, [E] must be obtained), the way to obtain [E] must be "random" in that the specific trajectory to be taken at (t) = 0 is a type of trial and error selection process.
    Thus I suggest that human selection of any action = dialectic union of determinism + random ​

    and this then is my rational and logical explanation of why any human might select milk the second time around--in short, because humans are Markovian Machines.
  16. Apr 6, 2007 #15
    The distinction (imho) doesn't make any sense. If you will necessarily make the same choice then that is equivalent to saying that you have no choice (your choice is determined by the antecedent states). In other words, since you cannot choose differently to the way you do choose, you have no choice about your choice.

    This is not a "scientific" experiment - it is a logical (thought) experiment. Why should it not be possible? Can you suggest any reasons why it should not be possible?
  17. Apr 6, 2007 #16
    If the choice the second time around is not necessarily the same, then it follows that the choice could be different, in which case (if it is different) what is the rational explanation for it being different?

    If no rational explanation can be given for the choice being different, then what is the difference between this and random behaviour?
  18. Apr 6, 2007 #17
    Why should it be "unrewindable"? Why should it be impossible (in principle) to recreate the precise conditions that applied before? Apart from claiming by arbitrary fiat "it is impossible", I see no rational argument which leads to the conclusion that it is impossible.

    It makes no such assumption. I do not believe that Robert Kane (a leading defender of metaphysical libertarianism) would have any problem accepting the thought experiment as it stands, without claiming that it assumes the will is not free.

    Once again - why not? What rational, logical or empirical reason could you give for the belief that the spiritual mind could not be "reset"?
  19. Apr 6, 2007 #18


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Hi moving finger.
    Seems to me the thought experiment is better when viewed with respect to various assumptions.

    1. Assume the mind is computational: Then there is no free will in the common sense of the word. Computationalism is by definition, completely deterministic.

    2. Assume the mind operates on classical mechanics: Then free will is highly questionable. The only way around having anything like a non-deterministic outcome is to suggest that at the level of classical mechanics, quantum mechanical deviations from the averaging we otherwise use may affect something at the classical level in an indeterminate way. Note this also assumes QM is not deterministic.

    3. Assume the mind operates on a quantum level: Then free will may be real, depending on your view of whether it may be a deterministic QM world or not. If you insist QM is deterministic, realize what non-local means here. It means your decision was made for you either in the future or at some location in space that is so far away, no signal could have reached you unless it traveled faster than light.

    Personally, I don't think free will is all that interesting. It can't be explained any better than experiencing the color red. We have not theory for either.
  20. Apr 6, 2007 #19
    Constrained? Thus free will (the metaphysical libertarian sense ) does not exist?
    I have no problem with the notion that everything is either determined or random or some mixture of the two. What I do have a problem with is how such a union generates “free will” in the metaphysical libertarian sense.
    Agreed. But a Markovian Machine does not necessarily have free will in the metaphysical libertarian sense.
  21. Apr 6, 2007 #20
    Firstly, I'd say it depends on the person you are examining.

    Take for example a coffee junkie Joe. If you rewinded the sitation
    100 times, then the coffee junkie will take coffee 100 times with
    great probability.
    I say probability because there could be some random
    thought "popping up into his mind" that gives him the idea to drink
    tea instead of coffee.
    As to how and why this random thought pops up, I don't know.
    (back to this later)

    In other words, the choice of the person will depend on his previous
    experience and preferences ("I like coffee").
    There could also be another person Peter who drinks both coffee and
    and tea but decides randomly every morning.

    In other words, Joe and Peter's decision will depend on their experience.
    And the experiment has to be considered in the framework of probability
    ( Prob(Joe drinks coffee)=99%, Prob(Peter drinks coffee)=50%.

    So one could criticize the experiment because the person already could
    have some preference. One should choose a situation where
    the person has no preference, for example there are two doors which lead
    to a big hall. Which door will the person take if you repeat the situation
    several times?

    In my opinion it will not always be the same door if you repeat the experiment.
    (Apart from the preference thing, I think your thought experiment is interesting).


    Secondly, maybe we should make clear first, what you mean with determinism.
    In my understanding, determinism is to be considered in the sense of Laplace's
    demon. Here, you have a supercomputer with infinite processor power.
    You just plug in the information of all the particles in the world (position, velocity)
    and can calculate a simulation of the universe.

    We know however, that according to quantum mechanics there is no certain (inherent?) position and velocity of a particle, thus it is not possible to run the "correct" simulation. In fact, there is then no such thing as a "correct" simulation. At least that's my opinion.


    As to why random thoughts pop up (see coffee junkie Joe):
    I think that quantum mechanics delivers a good argument. It may be that some parts of the brain can be described quantum mechanically: You have a wavefunction which only allows predicting probabilites for particles.

    If we transfer this to the left-right door problem, then in the moment you
    decided to take the left door, some wavefunctions collapsed.

    Of course, the question is if quantum mechanics really applies to a macroscopic object such as the brain. And one could also criticize on how quantum mechanics provides a free will:
    Is a "will" still involved or is everything random? Do I have any control on my decisions if some random process is involved.

    This leads to the question: How is a "free will" defined?


    This article here may be interesting for you:
    Last edited: Apr 6, 2007
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: If I Could Choose Again
  1. I travel again (Replies: 5)

  2. I am traveling again (Replies: 23)

  3. I can not sleep again (Replies: 24)