86 Comments
author

I had this conversation about the argument with Bing chat:

Me: I want you to evaluate an argument for me. This is the argument:

1. With respect to the free-will issue, we should refrain from believing falsehoods. (premise)

2. Whatever should be done can be done. (premise)

3. If determinism is true, then whatever can be done, is done. (premise)

4. I believe in free will. (premise)

5. With respect to the free-will issue, we can refrain from believing falsehoods. (from 1,2)

6. If determinism is true, then with respect to the free will issue, we refrain from believing falsehoods. (from 3,5)

7. If determinism is true, then the belief in free will is true. (from 6,4)

8. Determinism is false. (from 7)

Is that a good argument?

Bing: The argument you presented is known as the “Direct Argument Against Determinism” and it’s a valid argument. However, it’s important to note that the argument is not sound because premise 2 is false. Therefore, the conclusion that determinism is false does not follow from the premises. Here’s a table summarizing the argument:

# Argument

1 With respect to the free-will issue, we should refrain from believing falsehoods. (premise)

2 Whatever should be done can be done. (premise)

3 If determinism is true, then whatever can be done, is done. (premise)

4 I believe in free will. (premise)

5 With respect to the free-will issue, we can refrain from believing falsehoods. (from 1,2)

6 If determinism is true, then with respect to the free will issue, we refrain from believing falsehoods. (from 3,5)

7 If determinism is true, then the belief in free will is true. (from 6,4)

8 Determinism is false. (from 7)

I hope this helps!

Me: Why did you say that premise 2 is false?

Bing: Premise 2 is false because there are things that should be done but cannot be done. For example, it should be done that we prevent natural disasters from happening, but we cannot do so because we don’t have the power to control nature. Therefore, premise 2 is false and the argument is not sound. Is there anything else I can help you with?

Me: Do you mean that we should prevent all natural disasters, including the ones that we can't prevent, or only that we should prevent the ones that we can prevent?

Bing: I mean that we should prevent the ones that we can prevent. For example, we can prevent some natural disasters like floods by building dams and levees. However, we cannot prevent other natural disasters like earthquakes because they are caused by natural forces that are beyond our control. I hope this clarifies things. Is there anything else I can help you with?

Me: Then I don't see how you're saying that premise 2 is false. You're saying we should prevent the natural disasters that we can prevent. That's not a counter-example to the claim that if we should do X, then we can do X.

Bing: I apologize for the confusion. You are correct that if we should do X, then we can do X. However, premise 2 is still false because there are things that we should do but cannot do. For example, we should prevent all natural disasters from happening, but we cannot do so because some natural disasters are beyond our control. Therefore, premise 2 is false and the argument is not sound. I hope this clarifies things. Is there anything else I can help you with?

Me: Didn't you just contradict yourself?

Bing: I apologize if I was unclear. I did not mean to contradict myself. What I meant to say is that premise 2 is false because there are things that we should do but cannot do. However, I understand that my previous response may have been confusing. Is there anything else I can help you with?

Me: You said I was correct that "if we should do x, then we can do x". But you also said "there are things we should do but cannot do". Do you not see a conflict between those two statements?

Bing: You are correct that there is a conflict between those two statements. I apologize for the confusion. What I meant to say is that if we should do X and we have the power to do X, then we can do X. However, there are things that we should do but cannot do because we don’t have the power to do them. For example, we should prevent all natural disasters from happening, but we cannot do so because some natural disasters are beyond our control. I hope this clarifies things. Is there anything else I can help you with?

Expand full comment
Mar 26, 2023Liked by Michael Huemer

I just concluded a conversation with ChatGPT-4 regarding your argument. I thought you might find it amusing. Here is the transcript (I've added the "Q:" and "A:" to distinguish my questions from GPT-4's answers.)

Q: Can you evaluate the validity of reasoning through logical principles like modus ponens?

A: Yes, as an AI language model, I am capable of evaluating the validity of reasoning through logical principles like modus ponens. Modus ponens is a deductive inference rule that states that if a conditional statement of the form “If A, then B” is true and if A is also true, then it logically follows that B must be true as well.

For example, if we have the premise “If it rains, the ground gets wet” and the premise “It is raining”, then we can logically deduce the conclusion “The ground is wet” using modus ponens.

The validity of this reasoning is based on the fact that the logical form of modus ponens is sound. In other words, if the premises of the argument are true, then the conclusion necessarily follows. Thus, as long as the premises of an argument are true and the argument is structured in accordance with the principles of modus ponens, the conclusion can be considered valid.

However, it is important to note that the soundness of an argument depends not only on the logical structure but also on the truth of the premises. If the premises are not true, then the conclusion may not be valid, even if the argument is structured correctly.

Q: Here is a sequence of assertions, structured as a logical argument, concerning the existence of determinism versus free will. 1) Premise: one should only believe the truth. 2) Premise: if a person should do act A, then that person can in fact do act A. 3) Premise: if determinism is true, then if a person can do act A, then that person will do A. 4) We deduce from premise #2 and premise #3 that if determinism is true, then if a person should do act A, then that person will in fact do act A. Please evaluate his sequence of arguments.

A: The sequence of arguments presented is valid and follows the principles of deductive reasoning. Here’s a breakdown of the reasoning:

1. Premise: one should only believe the truth. This premise establishes a standard of truth-seeking and rationality that underlies the subsequent argument.

2. Premise: if a person should do act A, then that person can in fact do act A. This premise is a logical assertion that connects the ideas of moral responsibility and agency. In other words, if a person is morally obligated to do something, then that person must have the ability to do it.

3. Premise: if determinism is true, then if a person can do act A, then that person will do A. This premise is a statement of the deterministic view that all events, including human actions, are ultimately determined by prior causes. It implies that if determinism is true, then any given action is the necessary outcome of prior conditions, and therefore the person must do it.

4. Conclusion: if determinism is true, then if a person should do act A, then that person will in fact do act A. This conclusion follows logically from premises #2 and #3. If a person should do act A (according to some moral standard), and if determinism is true, then the person will inevitably do act A, since determinism holds that all events (including human actions) are determined by prior causes.

Overall, the argument is sound as long as one accepts the premises as true. However, the conclusion of this argument is not universally accepted, as some philosophers and thinkers argue that the existence of moral responsibility and free will are compatible with determinism.

Q: Let us continue the argument with the argument. 5) If determinism is true, then premise #1 and conclusion #4 imply that a person can believe only the truth. Is this valid?

A: The argument in premise #5 is not entirely valid, as it makes a leap in logic that is not fully supported by the previous premises.

Premise #1 states that one should only believe the truth, and conclusion #4 states that if determinism is true, then if a person should do act A, then that person will in fact do act A. However, premise #5 attempts to link these two premises by asserting that if determinism is true, then a person can only believe the truth. This conclusion does not necessarily follow from the previous premises, as it is not clear how the truth of determinism leads to this conclusion.

Furthermore, the conclusion of premise #5 seems to be at odds with the idea of moral responsibility, which is a key concept in the argument. If determinism is true and a person can only believe the truth, then it is unclear how that person can be held morally responsible for their actions.

Therefore, while the argument in premise #5 attempts to draw a logical connection between the premises, it ultimately fails to provide a clear and convincing argument for its conclusion.

Q: But argument #4, which you believe to be valid, asserts that if determinism is true, then if a person should do act A (in this case, act A is believing the truth), then that person will in fact do A, which means that the person will in fact believe the truth. That is the conclusion of #5.

A: You are correct that the conclusion of premise #5 does follow logically from premise #4. If determinism is true and if a person should believe the truth, then that person will in fact believe the truth, since determinism holds that all events (including human beliefs) are determined by prior causes.

Therefore, premise #5 can be considered a valid deduction based on the previous premises. However, as I mentioned earlier, the idea that determinism necessarily implies that a person can only believe the truth is still controversial and subject to philosophical debate. Some philosophers and thinkers argue that even if determinism is true, humans still have the capacity for moral responsibility and free will. So while the argument in premise #5 may be logically sound, it may not be universally accepted as a true or compelling argument.

Q: I agree that it may not be accepted.

A: Thank you for the discussion. Is there anything else you would like me to help you with?

Q: Yes, I want to continue the conversation. Suppose you accept premise #6 that I believe that I have free will. Does the following then logically follow: #7 If determinism is true, then it is true that I have free will.

A: No, the conclusion in premise #7 does not logically follow from the premises.

If you accept premise #6, which is the belief that you have free will, then it is not necessarily true that determinism must be true. In fact, one could hold the belief that they have free will and reject the idea of determinism, believing instead that their actions are not fully determined by prior causes.

Even if we assume that determinism is true, the conclusion in premise #7 does not follow logically. Determinism and free will are often seen as incompatible concepts, since determinism suggests that all events (including human actions) are determined by prior causes, while free will suggests that humans have the ability to choose their actions freely and without external constraints.

Therefore, the conclusion in premise #7 is not a valid deduction based on the given premises.

Q: You agree that #5 is a valid deduction. Thus, #5 implies if determinism is true, then I can believe only the truth. Since by #6 I believe that I have free will, does this not imply #7: if determinism is true, then I have free will?

A: I apologize for my mistake earlier. You are correct that if we accept premise #5 and premise #6, then it logically follows that if determinism is true, then you can only believe the truth, and if you believe you have free will, then you have free will even if determinism is true. Therefore, the conclusion in premise #7 does logically follow from the given premises.

So, in summary, if we accept premise #5 that determinism entails that one can only believe the truth, and premise #6 that you believe that you have free will, then it follows logically that if determinism is true, then you still have free will, even if your actions are determined by prior causes. Therefore, the conclusion in premise #7 is valid based on the given premises.

Q: It seems that this line of reasoning leads to the following conclusion: #8 Determinism is false. That is because we agree that #7 follows from the premises. And #7 says that if determinism is true, then it is true that I have free will. But free will and determinism are incompatible. So if I have free will, then determinism must be false. Is this a correct deduction?

A: Yes, you are correct that if we accept the premises and the conclusion in #7, it would follow logically that determinism is false. If we assume that free will and determinism are incompatible, as you have mentioned, then it follows that if you have free will, then determinism cannot be true. Therefore, the conclusion in #8, that determinism is false, logically follows from the given premises.

However, I should note that the relationship between free will and determinism is a complex and controversial philosophical topic, and there are different views and arguments regarding whether they are truly incompatible. Some philosophers argue that free will can coexist with determinism, while others argue that free will is incompatible with any form of determinism. So while the deduction in #8 may be valid based on the premises, it may not necessarily be accepted by everyone as a true or compelling argument.

Q: You are very correct about your final conclusion. Thank you!

A: You’re welcome! It was a pleasure discussing this topic with you. If you have any more questions or if there’s anything else I can help you with, feel free to ask.

Expand full comment
Mar 26, 2023·edited Mar 26, 2023Liked by Michael Huemer

Hello professor Huemer. This is a neat little argument. What you’re really arguing for is the conclusion that we have the ability to do otherwise (you explicitly say that you’re not trying to refute the compatibilist). If we substitute that in the appropriate places, and also accept the sort of modification to the first premise which you suggest in your response to objection 3 (which was my initial objection), we see that you’re really showing an inconsistency between the view that we cannot do otherwise and certain intuitively compelling premises concerning norms. I think you’re right that there is such an inconsistency. Here’s a variation on the argument.

1. We should believe only what is most reasonable (premise)

2. If S should do A, S can do A (premise)

3. If we lack the ability to do otherwise, then if S can do A, S does A (premise)

4. If we lack the ability to do otherwise, then if S should do A, S does A (from 2 and 3)

5. If we lack the ability to do otherwise, then we always believe what is most reasonable (from 1 and 4).

6. We do not always believe what is most reasonable (empirical premise)

7. So, it is not the case that we lack the ability to do otherwise (from 5 and 6)

Expand full comment
Mar 25, 2023Liked by Michael Huemer

The argument in this post is very good. Or at least, it confirms some of my long-held biases. I seem unable to choose not to believe it!

I think a basic problem is the self. Clearly, something is causing my actions, and external factors are not fully determinative, or no one would ever have an accident. When I drive my car, we don’t know ahead of time if I will have an accident. But usually I don’t. Something is choosing to make the car follow the rules of the road and making usually accurate judgements about what is safe, having the effect that I don’t crash. If external factors were fully responsible, the identity of the driver would not matter.

But I am not always conscious of all these choices and judgements. I’ve become habituated to them, and they no longer take place consciously. Instead, I might be thinking about free will arguments. So it is easy to think my conscious mind does not have free will. But is that me? Or am I also my subconscious? Where is the line between me and the external factors supposedly determining my behavior?

I suppose this is a reply to one of the interesting demonstrations of evidence supposedly refuting free will. Neuroscientists can demonstrate that the conscious mind is unaware of decisions that have already put actions into motion for a short period of time. But this only shows that some decisions are not fully conscious. Can we create a process that assures us that a decision will be made consciously? What sort of decisions should only be made fully consciously? This seems empirical, and likely to vary among persons. Hmm...

I have always thought that there is nothing at stake when we discuss free will. It is hardly relevant whether people do or don’t deserve punishment or blame if we think everything is determined. If so, whether we blame of punish is also determined. And so long as actions that we might label as blame or punishment have beneficial effects, we will and probably should continue using them. Determinism does not prove that blame and punishment have no effect, or a negative effect. This is a purely empirical question, which the argument does not address. If external factors fully determine persons' actions, and blame and punishment are external factors, why shouldn’t we use them, if they are effective? Why should we blame ourselves for blaming, if we can’t choose not to? I’m pretty sure evolution would have junked it if it was useless.

What is someone who disbelieves in free will doing discussing morality, anyhow? I suppose a sufficiently sophisticated robot could discuss altering its own program, even if it lacked the ability to upload it. But it would be pointless, unless it was designing a new sort of robot. This, in some sense, would all be predetermined by the robot's original program. But that doesn’t mean it could not discover an improvement. Determinists would have to say it is all being caused by the environment. Why does the mindless environment get credit, but the robot gets none? What sort of credit goes to the programmers, if they can’t understand the process they control?

Expand full comment

Reading the cute refutation reminds me of Anselm or some other philosopher who plays fast and loose with epistemic and metaphysical language in order to pull an impossibility out of a hat.

There are many ways to think about free will; its supposed opposition to determinism is just one way of looking at it. I have always found the free will vs determinism argument boring though, not because it has been discussed to death, but because the concept of free will has always seemed to smuggle in some kind of crypto-dualism where there is some unstated and unacknowledged homunculus viewing the world through the cartesian theater. I just don't see how the concept of "free will" could make any sense at all if one thoroughly rejects dualism.

How do you explain the fact that some physical movements can be considered "free" while others, such as the involuntary movements associated with Huntington's disease or other disorders, are not? Depending on the disease state, it may only be a question of a relatively small group of neurons dying that makes the difference. What does freedom mean to you exactly?

The only way I can think of preserving the concept of "freedom" from the biological-reductionist approach (which is the only approach that makes sense to me) is to think of it as a higher-order/emergent psychological phenomenon as opposed to some intrinsic feature of human action.

If "free will" is to have any meaning at all, I prefer to think about it like a self-regulatory capacity one develops rather than a statement about causality. It is closer to something like agency, or the feeling of integrity that comes with knowing that your thoughts and actions are in harmony. When people talk about not being "slaves to their passions", this is getting at the kind of freedom I am talking about. The ability to control rage, for example, makes us freer to respond to a situation in a wider and more useful varieties of ways. Often restraint and limitation leads to greater freedom later on, as in the case of saving money. This approach takes free will as a more "psychological" or ethical subject, but I think we'll get more mileage out of the concept if we think about it that way.

Another way of saying all of this is that the question of free will only comes up for us because the actual workings and mechanisms of our minds are opaque to us.

Expand full comment
Mar 26, 2023Liked by Michael Huemer

Like you, I've always thought determinism is self-refuting. You have to assume free will is true in order to refute determinism. This is the hallmark of an axiom: you have to assume it is true in order to attempt to refute it. The very fact that you can ask yourself if determinism is true or not tells me you are entertaining an alternative, a choice, and choice refutes determinism. Thanks for distinguishing between the hard determinists and the "compatibilists." I had never hear that term before. BTW, the impossibility of determinism is why I don't even like to watch most time travel moves, because I can't make them seem believable. ("Back To The Future" being a big exception!)

Expand full comment
Mar 26, 2023Liked by Michael Huemer

You can count me, Mike, I like this argument very much (:

Expand full comment
Mar 26, 2023Liked by Michael Huemer

"Each time, the referees said to reject it because they could think of some objections to my argument. (Note: Any philosopher can think of objections to any interesting argument.)"

I was wondering: when you're a referee, what do you think is a good reason to reject a paper? E.g., do you think "I can think of good* objections to the argument" is a good reason to reject a paper?

* I think context, I'm taking "good" to mean "I can think of objections to the argument that I think are correct"

Expand full comment
Mar 25, 2023Liked by Michael Huemer

I've heard the argument "People don't have free will, but don't tell them that." That's about as strong a proclamation that free exist that I can think of, and I think this proof gets to that. I read your proof a while ago, and I had a little feeling that there's something wrong with it, but over time, I've come to realize that it's probably just because it's so simple that it makes me uneasy.

Expand full comment

I always believed in free will intuitively but this blew my mind; it’s brilliant!

Expand full comment

“Can" is ambiguous between two meanings. There are different senses of possibility. When disambiguating the different senses, at least one of the premises here seem false.

Let's say "can" = "logically possible". In that case, premise 3 is false. If determinism is true, that does not mean that all logically possible propositions are true.

Let's say "can" = "compatible with the laws of nature + the initial conditions of reality". In that case, premise 2 is false. The fact that we should do an action does not imply that that action is compatible with the laws of nature + the initial conditions of reality.

Expand full comment

“Determinism implies that good epistemic reasons play no role in explaining why one believes determinism itself because they would believe it regardless of whether it was true or not.”

My friend Connor asks: Couldn’t the determinist say “Having a good epistemic reason to believe something is part of what deterministically causes us to believe it?”

Expand full comment

To assert that free will is true, is to assert that determinism is false, and vice versa. Borrowing a mathematical term, free will iff not determinism. It's a form of tautology.

In essence, your proof assumes what it is proving. In premise #1 & #2, the "should" statement presupposes that there are choices you can pick between. In #1, the assumption is that you can choose to believe in the truth, or you can choose to believe in a falsehood, but you ought choose the truth. Your premise assumes free will.

The alternative statements, without that assumption would be something like,

1. We believe that we should believe only the truth.

2. If S believes they should do A, then S believes they can do A.

These statement presume neither free will nor determinism. Under determinism, we would simply be compelled to believe a falsehood (i.e. that "should"'s are even possible).

But once free will is removed from 1 & 2, your proof no longer works. For example, #4 no longer follows from 2 & 3.

Expand full comment

This argument proves too much: from (1) and (2) you can conclude that is it possible to perfectly distinguish truth from falsehood in all cases, that it's impossible for you to ever be mistaken about anything, ever, unless you choose to be. (Using, I suppose, your free will.)

That we should believe only the truth (that is, (1)) implies (via 2) that we can (if we want) believe only the truth—that we are capable of choosing to never believe falsehoods. This is only possible if we can infallibly distinguish falsehoods from truths, or, I suppose, if we never believe anything ever.

I therefore conclude that you, Michael Huemer, are either completely right about absolutely everything, or else you perniciously choose to intentionally err. That you are ever truly mistaken, about even the smallest point of fact, is a logical impossibility, by points (1) and (2) above.

Expand full comment
Jun 18, 2023·edited Jun 18, 2023

Imagine a counterfactual world where there is:

1) an extremely strong, irrefutable scientific proof of hard determinism;

2) a very clear scientific understanding of the causal mechanism that yields our thoughts, beliefs and attitudes;

3) a very clear scientific understanding of why an illusion of free will arises as a perceptual phenomenon;

But just like in the actual world, I am ignorant of these scientific theories and still maintain my belief that I have free will, as per premise 6.

In that world, is hard determinism still false (or not overall best justified by our evidence), despite overwhelming evidence to the contrary? This seems implausible.

Expand full comment

I think Professor Huemer is proving the self-described free will, but the will itself is not free.

As Schopenhauer said: 'A man can do what he wants, but not want what he wants

Expand full comment