Here goes. The argument is in steps. After each step there is a paragraph or more justifying the step.

Thesis: It is reasonable to believe in free will, if one perceives that one has free will.

Q: Does this argument prove free will exists?

A: No. I think such an argument would be impossible. Free will is epistemologically subjective, even though it can be ontologically objective. That means that there is no way for me to perceive other people's free will. If I didn't think I have free will there will be nothing to convince me that other people have free will. Indeed, it would be perfectly consistent for me to believe that nobody has free will. Any experience of free will can be attributed to delusion or some other psychological factor, perhaps evolutionarily evolved one. Therefore the aim of such a discussion must be different.

Q: So what's the aim?

A: The aim is to show that if one feels that one has free will, then it is rational for them to believe they have free will. Indeed, the argument will show that it is more rational to believe one has free will than the opposite.

Q: But isn't that obvious? What does such an argument achieve?

A: It's not obvious at all. There are many things that are perceived. I may be drunk, look at the floor and perceive an earthquake. That doesn't make it rational for me to believe there is an earthquake happening. It's much more rational to attribute my perception to my drunken state than to the state of the earth.

There are a lot of people who feel like they perceive their own free will, but think that this perception is illusory. The argument will attempt to convince those people that the perception should be trusted and one can do so while staying perfectly rational. That is a huge achievement.

Q: Some premises seem obvious. Why do you talk about them?

A: I've had anons critique every single premise in the argument. They might seem obvious to you but for many anons agreeing with a single premise is already a HUGE achievement.

Q: Why are the arguments below the premises so long?

A: I have argued with many anons about this issue and tried to take account of all of the arguments that I've heard. Naturally, there were many arguments against these premises and so the response to them all will be quite long. I strongly suggest you simply skip the points that you agree with and don't need to be convinced about.



Let's start with the argument

1. It's possible to talk and reason about a concept without having a precise, rigorous definition of the concept.


For example, mathematicians have largely correctly reasoned about the real numbers long before they were rigorously defined. The introduction of the definition served to clarify some points but people were still able to do fine without it. They understood what was meant by real numbers. For more examples, just think about what a definition entails. It defines a concept in terms of other concepts. In order for definitions to be meaningful, the concepts which are used in the definition are either meaningful or need definitions of their own. But such a chain of definitions is never infinite. There are always concepts at the end that are not defined in terms of others which one takes for granted.


2. It's meaningful to talk about free will.


Free will may be defined as the ability of a person to choose, the ability to have control over their future. Not complete control, just some control (we don't expect them to have the control to jump to the moon or to overcome their involuntary instincts like closing your eyes when someone claps in their face). However, even without settling on one definition of free will, it is still meaningful to talk about it. People understand what is meant by free will, which is why they are able to have meaningful discussions about it. A possible objection is that free will is not a physical thing or cannot be defined in terms of physical things. But does that mean that free will is meaningless? Surely not. Memory, ideas, concepts in mathematics like numbers, rings, vector spaces are all concepts that are non-physical and cannot be defined in terms of physical objects. But clearly they are all meaningful concepts that are used and reasoned with daily. In grade school you have probably reasoned about numbers like 1,2,3, and how they interact. Did the teacher provide you with a definition of these numbers? No. You intuitively understood what they meant from examples.

But perhaps the word "choice" is meaningless? There seems to be no good reason to believe it. In our daily lives we attribute decisions and choices to one another nearly every day and have no problem with the concept. Now perhaps you think that free will is just illusory, in which case you will think that all choice is illusory. But for a thing to be illusory, it must be meaningful in the first place. For if you're asserting that something is illusory, that something must have meaning, otherwise there's nothing to attribute the property of being illusory to. Xaweaoiso cannot be illusory because Xaweaoiso is not a meaningful word.

Perhaps you think the concept free will is logically incoherent. But I see no good reason to believe this that does not depend on the observation that it's illusory.


3. If some fact or statement is immediately perceived to be true then it's reasonable to believe the statement unless you have a valid reason to doubt your perception.

This has been the point of most contention, therefore my elaboration on it will be the longest.

Examples: I remember that yesterday I drank a cup of coffee. There are no valid reasons to believe my perception is wrong and that in fact I didn't drink any coffee yesterday. In absence of such counterevidence, it's reasonable to believe that I actually drank coffee yesterday. I do not need to come up with a mechanism for how my memory works to believe it, nor do I need to give precise technical/philosophical definitions of memory and the word "drank" for it to be reasonable for me to believe this fact. My immediate perception of this fact is enough for it to be reasonable, in absence of evidence or reasons to the contrary.

Now what about when our perception is wrong? Surely then it's unreasonable to trust it? Imagine a colorblind person seeing two different colors which he perceive to be the same. Is it reasonable for them to believe they're the same color? The premise does not immediately assert this. It asserts that it will be reasonable but only if the colorblind person does not know that he's colorblind. If he knows he's colorblind then clearly he should not trust his perception to tell him that the colors are the same. If he does, he's being irrational. Now what about if he doesn't know that he's colorblind, that his perception is flawed? Is it rational to believe the colors are the same? I think so. Indeed, he has no reason to doubt his perception. His perception in other areas has served well so it's natural to assume it serves well in this case too. If we want to assert that even when he has no idea that he's colorblind, his equating the colors is irrational, then we must also concede that most, if not all, of our own beliefs are irrational. For most of our beliefs it is possible that the perception by which we arrived at them has been flawed in one way or another. For example, I think most people think that the country Australia exists. We have lots of evidence from it that we have seen, some of us have been to the country ourselves. Now it's logically (even though it doesn't seem probable) possible that we have all been deceived. There is no such country as Australia, all of the people talking about it have completely confused it with some other part of the world, the world maps are all due to some technical error that produced a large blot on the map. Now does that mean it's irrational for them to believe that Australia exists? Surely not. We have no good reasons to doubt our perception of all things involving Australia. We are doing best we can to be rational about the belief in Australia, and in the same way the colorblind man is doing the best he can to compare the colors. It is rational for the colorblind person who has never had the fortune to learn that he's colorblind to trust his perceptions and think that the two different colors are the same, in the same way that it is rational for us to trust that Australia exists.

But shouldn't we only trust scientific evidence? Some people claim that perception is subjective and cannot be considered to be valid evidence. Instead, we must be objective and solely restrict ourselves to science and scientific evidence in forming our beliefs. But is that a reasonable position? I think not. After all, we learn to appreciate all the evidence in science through perception. I believe the moon is at such and such position at such and such time because I perceived it through my eyes with the help of a telescope to be there and then, and have no good reason to doubt my perception. Alternatively, I believe that some other group of scientists have observed the moon at such and such time and place because I perceived them telling me this information and I have no good reason to think they're lying. All scientific evidence and knowledge is learned and appreciate through perception.

But shouldn't be we be doubtful about everything we perceive? If all of our perceptions could be lying to us, isn't it more reasonable to doubt everything, as in the famous passage of Hume? I think that's unreasonable and works only in the imagination. We may think it's best to doubt everything but practically that's impossible. In order to make decisions in the world we act according to our beliefs. Some things we believe more than others, but we have to commit to believing some things in order to operate in the world (for example, beleiving that option A is more reasonable than option B). And sure, my perception could be wrong. But am I stupid for trusting it? Am I being more rational by doubting everything than those who just go about believing things? Clearly not. It's a fun scenario to think about but in practice everyone has things they believe in, they must in order to function.
Another thing to notice about this particular objection is that not only is it self-defeating (we should doubt that we should doubt everything) but also that it can work in favor of free will (we should doubt determinism, we should doubt that we don't have free will). Either the objector thinks that we should doubt all things equally (i.e. doubt that the earth is flat just as much as that the earth is round) or he thinks there are things we should doubt more than others. In the latter case, this argument works just as well with "reasonable" replaced by "less doubtful than the opposite". Indeed, then the premise becomes "If some fact or statement is immediately perceived to be true then it's less doubtful to believe the statement than its opposite unless you have a valid reason to doubt your perception more than you trust it". Then the conclusion of the argument becomes "If one immediately perceives free will, one should doubt that one doesn't have free will more than the opposite". But enough of this objection. We return to the normal use of the words reasonable/rational and ask the objector to replace them with his preferred nomenclature himself.

What if you perceive that you don't have free will? Same holds. If there aren't any good reasons to doubt your perception, it's rational for you to think you indeed don't have free will. I think it's likely that there are people who do have free will and those who don't (although nowhere in the argument do I use this. I never assume anywhere that free will exists. The argument works without the need to assume anyone has free will).


4. I immediately perceive my own free will.


I perceived it when I was a child and I still perceive it. In fact, my own perception in free will has been central to the way I live and the way I think about the world. I regularly spend time thinking about how to improve my own future which would make no sense in a worldview where I am not able to influence my own future (for example, in a worldview where everything is already predetermined).

You may ask how it's even possible to perceive one's own free will. By perceiving it I mean I immediately recognize the fact that I am in control of (some of) my decisions to be true. It does not come through any logical process. I do not ask myself "Hmmm. Could it be that I have free will?", evaluate all evidence and settle on an opinion. No, the proposition comes to me without any further evidence or elaboration as strikingly immediately obvious.

This perception does not come through the usual senses, no. I do not see, nor hear my free will. But it's a perception nonetheless. It's similar to other perceptions as my perception that I drank coffee yesterday. I think about it and immediately see this proposition as obvious. I don't hear, nor see, nor touch the fact that I drank coffee. I perceive it with the help of my memory and other cognitive faculties, without any intermediary logical steps. The perception of free will is of an analogous kind.

The perception might be flawed, there might be good reasons to doubt the perception, but I cannot doubt the fact that it's a perception. I cannot doubt the fact that I feel that I have free will.

Perhaps you don't perceive it. That's ok, I've interacted with anons before who didn't perceive their own free will before. Then this argument will not apply to you. It applies only to those who perceive their own free will.


-----ATTENTION!-----

Now the argument takes a turn. We look at the reasons why our perception of free will might be doubted and evaluate how effective they are.

Notice an extremely important point here. In this argument we have already achieved something that I've never seena anyone else arguing for free will achieve before, at least not explicitly. We have completely turned the tables of the argument. We made it so that the burden of proof is on the objector of free will to give a valid reason to doubt my perception of free will.
We do not search for mechanisms for free will. We do not look at how to account for free will using our known physical laws. One might try to attempt such a thing (which personally I don't think is likely to yield anything interesting) but the crucial point is the argument does not need it! It works just as well if we have no idea whatsoever about how to account for free will in physical terms or what is the precise mechanism of free will. This already eliminates a VAST majority of objections to free will that I have seen, i.e. all the objections like "You can't prove free will" (I don't need to for the argument to work), "How can free will arise from small particles?" (I have no idea and I don't need to have an idea for the argumen to work). This point cannot be understated and it's the point that confuses most people.

-------------------

Now we take a look at the supposed defeaters for free will.

The first and the most obvious one is determinism.


5. There is no rational reason to prefer determinism over indeterminism.


In the days of enlightenment, physics has had great success with laws which were all deterministic in all interpretations. Classical mechanics can be viewed through the lens of Lagrange, Hamilton or Newton. They're all different interpretations of classical mechanics but all are fully deterministic. Input the state at some time, and get the states at all other times.

With the success of classical mechanics and later discoveries in physics in describing parts of the universe with great accuracy in a perfectly deterministic way it seemed reasonable to believe that the whole universe was deterministic and that one day we would find the complete deterministic mathematical description of how the world works.

With the advent of quantum mechanics that hope is destroyed. Quantum mechanics does NOT show that the universe is indeterministic, but it does destroy all hope to prove determinism through laws of physics. There are equally valid interpretations that are indeterministic (Copenhagen interpretation) and deterministic (Bohmian interpretation, which posits some hidden variables that we simply have no access to). Now the point is this fact shows that as of now, there are no good reasons to prefer determinism over indeterminism stemming from laws of physics. Both positions give you equally valid answers, therefore physics can no longer be used to show that the world is deterministic.

An extremely common objection is: quantum effects only have effects on the small scale. People operate at such a large scale that all the quantum effects become negligible, and classical laws of physics still account for how we act and give a deterministic description of large-scale phenomena.

This is simply not true. Consider an experiment. It's possible in quantum mechanics to make a measurement of the spin of a particle with 50% chance of outcome A and 50% chance of outcome B. There is currently no known laws of physics that could deterministically predict the outcome of the experiment. Now suppose before performing the experiment you decide to go on a killing spree if the outcome of the experiment is A, and sit and home and do nothing if it's B. Suppose physics could predict such large scale behavior as what you're going to do. In that case, physics would be able to predict whether or not you went on a killing spree or stayed home, as going on a killing spree is obviously a large scale phenomenon. But if it were able to do that, it means it would have been able to predict the outcome of the experiment, contrary to our assumption that it's not possible. Thus we have a contradiction.

You might object by saying that still the machine could predict with 50% certainty what I'm going to do and that's good enough. But this objection won't do. Imagine instead of two choices you select 2^20=1048576 different locations on earth sufficiently far away that you can travel to in a day. Now imagine performing 20 such quantum experiments, and having the sequence of outcomes determine where the location where you will go. Now the objection that "knowing the probabilities is good enough" clearly looks ridiculous. The information that probabilities give you about where you are going to be are so low as to be negligible. Such a prediction has no right to be called a prediction. I might as well say I can predict what you're going to say next by assigning to to each combination of characters 2000 long or less an equal probability. Nothing about this is deterministic.

You might also object that such an experiment is an extremely specific and rare thing. In reality people do not perform quantum experiments and so quantum effects have little to no influence over our behavior. But I see no reason to believe this. Indeed, if we adopt purely materialistic, deterministic point of view, there is no reason to think that the moments when I'm performing quantum experiments are in any way special to other moments, as consciousness and experiment are not physical things. Otherwise there simply seems to be no good reason to think quantum effects have no influence over the daily life. The burden on proof would be on the determinist to demonstrate that they don't have any influence and that a (theoretical) deterministic machine/Laplace's demon can indeed be built for all cases where no one around is performing any quantum experiments.


Can arguments in favor of determinism come from other areas than physics? Some determinists think determinism is a logical fact following from the assumption that everything that happens has a sufficient reason for happening. If everything that happens has reasons for happening, that must include our decisions, which would mean that they are influenced by outside factors. But is this a good objection? I don't think so. There seems to be no good reason to preclude the possibility of the reason to principally be the choice itself. The choice could have influences, factors which made the choice more likely, and yet still the choice could be the sufficient and principal reason for why something happened. Obviously all decisions are influenced by something.The question is whether there are good reasons to believe those influences determine our choices. Notice that right now we're in good position. We don't need to prove that they're not (read the ATTENTION part again). Nor do we need to provide a mechanism for free will. The burden of proof is on the people making the objections. All we need is to call the objections into doubt. As far as I can see, there is no logical inconsistency between our decisions being influenced by outside factors and us still having free will, i.e. it still being the person that makes the decision. I am not aware of any strictly logical argument to this note (if you have one, please post it). If there is an explicit contradiction in here, then it must be well-hidden. On the other hand, it seems possible to at least imagine a situation where a choice is made and yet everything having reasons. Thus as far as we can tell, this is not a good objection either.


There is a simpler, more intuitive argument against determinism, but it's subtle in its own way and may not be easily believed by everyone. Imagine it were theoretically possible to build a machine that could predict the future. You ask it where you're going to be in one hour. Taking note of the response, you go elsewhere and falsify the prediction.


6. Answer to the objection "But determinism does not give you free will. It only gives you randomness, and free will cannot arise from randomness, nothing about randomness is free, it's just random."


First let me just note: if you concede that there's no reason to prefer determinism, you've already made a MASSIVE concession. Many people cling to determinism like their life depended on it. Another thing to note here that the objection is of the form "How can there be free will? Quantum mechanics is random!". It implicitly asks us to give a mechanism for free will or to explain how it's consistent with quantum mechanics. But read the ATTENTION part again. It's no longer our job to explain how it's possible or give a mechanism for free will. The burden of proof is now on the objector to prove that free will is inconsistent with quantum randomness. I've never seen anyone bring about the explicit contradiction. I will try my best to do it myself.

I assume the thinking is something like the following. The mind may not be physical but it is contained within the brain which is physical. Now the brain is composed of different parts. We can keep breaking it down until we get to the quantum level of atoms which constitute the brain. Now the atoms that constitute the brain behave randomly according to quantum mechanics. There do not seem to be any signs that the constituent particles of the brain are capable of thought or will. And since the brain can be reduced to its constituent particles, there is by extension no reason to think the brain, and by extension the mind, is capable of having free will.

This is more explicit but still not explicit enough. What is the reasoning behind the inference from atoms not having free will to the brain and the mind not having one? There are many properties which parts of the object do not posess yet the object posesses. Think about a digital image of a dog. The image is composed of bits, 0s and 1s. None of the bits have the property of portraying dog. Is it reasonable to infer that the image itself does not portray a dog? Of course not. So there must be something different in the case of the brain and the atoms. As we already discussed, the burden of proof is on the objector to bring about the explicit contradiction. If there is one, it's well hidden.

A theorem due to John Conway and Simon B. Kochen states that if we have a free will in the sense that our choices are not a function of the past, then, subject to certain assumptions, so must some elementary particles. Putting aside the question of whether the assumptions are reasonable enough to accept, is the conclusion a good reason to doubt our perception of free will? Are there good reasons to doubt elementary particles in a sense have free will? The only good objection I can see is that we know from quantum mechanics that particles behave randomly, and randomness is not free will. Let us address this objection.

Let us first clarify what is meant here by randomness. Indeterministic nature of quantum mechanics does not have to imply randomness. It only implies that determinism is false, i.e. the future cannot be predicted just by knowing the state of the universe at some point in time. But does that mean the universe is random, without order? Hardly so. We may not see any patterns in it if we look at it through the lens of laws of physics but that does not mean that there are no patterns. Merely that we haven't discovered them.

You might say, but we have laws that are statistical that describe what happens well, which show that they are truly random. But does it really show it's orderless? I don't think so. Imagine I don't know a word of spanish and have never encountered any spanish speakers. I might look at their texts and decide that they look random to me. I make a statistical map of what characters appear in the text and find that they conform to a specific distribution. The letter A shows up with frequency 11.72 %, B with 1.49 % and so on. Does that mean that the texts are completely random? No! It just means that I see no pattern in them. Perhaps it's because I'm just looking at it through the wrong lens. Instead of analysing the frequency of characters I should make the assumption that it's a language and try to figure out what the words actually mean.

Another example is as follows. Imagine I am playing a video game with a controller that has one button but you don't see me pressing the button (perhaps the game is flappy bird). You believe that I do not control the game because you observe that before every time the character in the video game jumps, a bird outside chirps. Now you say that it is not me, but the bird that's controlling the game. But later on you notice that the chirp goes out of sync with my jumps. And now you say well you can't predict my jumps anymore, they seem random to you. So you argue I you can't control the game from randomness, I can only get randomness from randomness, which is different from an ability to control the character. There's nothing random about being in control of the character therefore there is no way I have control of it. But clearly you would be unreasonable to assert this. You are just looking at the wrong place and don't see a pattern. Your inability to see a pattern or order does not imply that it's not there. The fact that you haven't seen me press the button doesn't mean that I'm not.

To conclude, I see no good reasons to regard this objection as any more reasonable than the objections given in the examples. And remember, it's not our job to prove we have free will or to give a mechanism of it. We merely have to look at the objection and see how reasonable it is. I think it is safe to conclude that this objection is not reasonable either. Perhaps the objector needs to work on it a bit more to bring out the explicit contradiction from physics' in ability to predict the future to free will being impossible (which is exactly what the objection attempts to do, when you unravel what the word means. Indeterminism means the opposite of determinism: it's simply the failure of a machine or physics being able to predict the future from knowing the past). I see no good reason to think that indeterminism is in any way incompatible with free will. To the contrary, it seems that if free will could theoretically/logically exist, it's only under an indeterministic (but not necessarily random/chaotic! Although parts of it could be orderless) universe.

-----SIDE NOTE: COMPATIBILISM-----

You may ask: what about compatibilism? Does determinism really pose a challenge to belief in free will? And that's a good point. I myself am not entirely sure how to bring about a contradiction to free will on the basis of determinism. But just because I can't do it, that doesn't mean it's not a good challenge. Indeed, intuitively it makes sense that determinism if not make it impossible to believe in free will, at least poses a formidable challenge to the belief. And without formally verifying what kind of challenge it is, it's still good to take it as the main challenge and try to refute it (as I think I have done above).

------------------------------------

7. Answering the Sam Harris objection about predicting our choices before we become aware of them.

Scientists have developed a way to scan the brain and investigate the brain states. They have performed an experiment where a person is asked to choose between different options. They have found that the computer can detect the choice about 30 seconds (let's be generous) before the participant of the experiment becomes aware of the choice. According to Harris, this shows that we don't have free will. It is our subconscious brain states making our choice for us way before we do. We think we have control over the choice but we don't.

Is this a reasonable objection? I don't think so. The central implicit assumption on which this objection depends is that the self is only that which is conscious. Everything outside of what one is consciously aware belongs outside of self. I don't think that's a good assumption. I, and I think many other people, regard my subconscious to be part of who I am. Just looking at the way we use language, we talk about our consciousness. Our consciousness is part of who we are, it is not necessarily the whole of us.
Imagine reframing the observation but using the language of choice. "I made the choice and after some time I became conscious of my making a choice. There has been a 30 second delay between my making a choice and my becoming consciously aware of my choice". The meaning stays exactly the same but now it seems much less likely that there's an actual contradiction involved.

<cont>