A defence of libertarian (ish) free will

Metaphysics & Epistemology
1

The debates about free will are among the most confusing and unresolved in the whole of philosophy. This post explains what I believe free will really is, and why there is no good reason for not believing in it.

Firstly I’d like to exclude Compatibilism from this discussion, because it is an enormous waste of time. Compatibilists are simply determinists who don’t want accept the consequences of their own metaphysical beliefs, so invent a metaphysically-irrelevant meaning of “free will” – a meaning which serves no purpose but to generate confusion. I personally am not interested in discussing it further.

The position I really wish to contest is the idea that “everything must be determined or random”. This is typically the opening premise of arguments against metaphysical free will, and it assumes its conclusion in the premise: it begs the question against free will. IF everything is determined or random THEN free will is impossible. So the real question is how can something be neither determined nor random. “Determined” here means “fully determined by the laws of physics/nature”. “Random” means “happens for no reason at all – completely undetermined”.

So what kind of causality doesn’t fit into either category?

At this point we need to consider the interpretations of QM, because QM dictates what is possible. If the Many Worlds Interpretation is true then free will really is dead, because absolutely everything is determined (though it seems random from our branch-specific perspective). Similarly if there is some kind of physical collapse mechanism or purely deterministic metaphysical collapse mechanism then, again, determinism is absolutely true.

The only interpretations of QM which leave open conceptual space for free will are those which claim consciousness (somehow) is responsible for both wavefunction collapse and branch selection. I am aware of only two categories of these which clearly qualify. One is von Neumann’s interpretation, and two of its descendents (Stapp’s interpretation and my own Two-phase interpretation (2PC), which is a synthesis of MWI and VN). The other is any interpretation which links consciousness and collapse to non-computable value judgements, which includes Penrose’s Orch-OR and 2PC).

An argument might be made for including some versions of Bohm’s theory, Quantum Integrated Information Theory, and process-based theories (again includes 2PC).

I will now provide further details of exactly what is required under 2PC’s version.

Firstly there must be an agent. The agent of free will must be a unified structure (non-splittable) which models the external world, with itself in the model as a coherent entity persisting over time, is aware of multiple possible futures (so the model must span branches and be entangled with macroscopic superposed states in the outside world).

Secondly it must be able to assign non-computable valuations to different possible futures. They must be non-computable or the choice could itself be encoded in the wavefunction (which is computable).

Thirdly, these valuations must influence the selection of which branch manifests when the wavefunction collapses, especially with respect to those parts of external reality most closely entangled with the agent (most critically the rest of our own nervous system).

If these three criteria are met then we have a form of causality which cannot be random (because branch selection depends on non-computable conscious valuations), and cannot be fully determined by the laws of physics either (because the valuations are non-computable).

Can I prove this exists? No, of course not. But this does not matter, because a defence of free will only requires a refutation of the claim that everything is either determined by physical laws or random. Demonstrating that it is possible is enough.

And if it is possible, what justification is there for believing we don’t have it?

1 Like
2

Hi Geoff, I think this is a good attempt of defending libertarian free will.

But I think it will come down to whether the branch selection is reasoned or not. If it is reasoned, then it will be traceable to antecedents. If it is not, then it’s random.

This is why my position is not “there is no free will” – which is usually taken to be accepting some kind of fatalism and/or that our feeling of choice is an illusion. I don’t think free will even makes sense as a concept.

Choices are reasoned, so are necessarily the product of antecedents and randomness (and ideally, much more of the former than the latter). What do we even mean by an unreasoned, non-random choice and why should we desire it?

3 Likes
3

Think about what happens when we make a difficult decision. What makes it difficult is nearly always that reason on its own is not enough. We can reason all we like, and sometimes that is sufficient to arrive at a clear answer, in which case free will becomes little more than a rubber-stamping exercise. But sometimes it is not enough. This can be because the situation is too complicated for our brains to fully analyse, because there are critical unknown factors (including what other people might choose to do, or the outcome of a chaotic process like the weather), or because it involves a moral judgement which reason cannot resolve (and I am sure we could think of other examples). Either way, we nevertheless reach a point where a decision is no longer avoidable, because inaction itself equates to a conclusive decision. And at that last moment, we typically cannot even explain why we made the decision we did, for the same reason we can’t explain intuition or creativity. And yet we know it isn’t random – we don’t just give up and mentally “throw a dice” because we can’t decide. If we did that then we wouldn’t “own” the decision – we would have subcontracted it to randomness intentionally, which is the last thing we choose to do in those difficult moments. This is precisely because we really do feel like we “own” the decision, regardless of how difficult it was. And that is what makes it free will – the fact that we made the decision, and we have to own the consequences too.

Exactly what I described above. And we should desire it because it gives us an important level of ownership of our own fate. If somebody can’t explain how they wrote an exceptionally beautiful song, does that mean they shouldn’t be credited for writing it, because it must have been random?

1 Like
4

By “reasons” I am not claiming that it’s some concrete bullet point that can be verbally articulated.
I mean antecedents like your entire understanding of the world, and your predilections.
And it can be subconscious too. But ultimately, the rubber meets the road and a decision is made; and we can ask how something that isn’t an antecedent or randomness got involved in that process and how it works exactly?

Or I can try to put it another way.

Just now, why didn’t I jump over the moon? Because I am physically incapable of that.
Why didn’t I de-age myself 20 years? Because I don’t know how to do that.
Why didn’t I stab myself in the leg with a screwdriver? Because I am wired to find that deeply unpleasant.

No-one takes issues with statements like these, but my contention is that we could keep asking questions like these and ultimately funnel right down to the decision I *did* make. I did what I did based on my understanding of the world, and my predilections. That’s as free as free can be. But both “understanding” and “predilections” are antecedents.

1 Like
5

That is exactly why non-computable valuations are critically important. Penrose has spent decades defending the claim that consciousness is a process which must involve at least some elements of non-computability. He uses the example of mathematicians being able to “just know” that Godel sentences are true, even this cannot be computed. But it seems obvious to me that consciousness does this all the time – it is why we understand meaning, why we can make moral and aesthetic judgements, and why we do not suffer from the Frame Problem. This is in stark contrast to artificial intelligence, which shows no sign of being able to do any of these things, regardless of how powerful the computations are.

How does this work? Penrose doesn’t know. He says it must involve a deeper layer of physics (a non-computable physics). I think he’s strayed over the boundary into metaphysics here. But at the end of the day it doesn’t matter whether the mechanism involved is labelled physical or metaphysical – what matters is that the valuations are fundamentally non-computable. My own opinion is that this can only be possible if the agent involves both a superposed brain (which is nevertheless still a finite entity) and the ground of all being (aka “Brahman”, as an instantiated Atman), which is infinite.

So I think what this comes down to is the possibility that consciousness involves non-computable value judgements. If this is not possible, then neither is free will, and vice versa.

Just now, why didn’t I jump over the moon? Because I am physically incapable of that.
Why didn’t I de-age myself 20 years? Because I don’t know how to do that.
Why didn’t I stab myself in the leg with a screwdriver? Because I am wired to find that deeply unpleasant.

The first two examples aren’t relevant, because free will can only choose between physical possibilities, which those aren’t. The last one is much more interesting. There are some east Asian cultures where people actually pierce their own faces with knives or screwdrivers and run through the streets as a deeply spiritual act. They are proving they have sufficient free will to overcome the exact “wiring” you are talking about.

And surely that does indeed come down to whether or not consciousness is, or is not, a fully computable process.

1 Like
6

Don’t sugar-coat it, dude. Tell us how you really feel. :rofl: But I agree.

I think will is free from the physical laws. I’m a panpsychist, and I think consciousness is not physical. We are not nothing more than a bunch of physical events; a glorified, hugely complex, 3-D billiard game. Consciousness and the physical laws are both at play.

To the best of my understanding, I have often made choices that were a proverbial flip of the coin. Generally very small things, like which dessert to order at the restaurant, or which Bach piece to listen to. The idea that I could not have chosen other than the chocolate layer cake, or the chaconne for unaccompanied violin, seems preposterous. But I don’t know how meaningful that coin flip is. I don’t randomly pick between all the desserts. But when it comes down to the last two…? I’ve often ordered two desserts because I could not decide.

2 Likes
7

Let’s say you’re God. You can do whatever you want. Does this make no sense? Is there a contradiction?

8

In that case you do need to provide a possible mechanism, or the determinists will claim this is physically/logically impossible. Panpsychism also has many other problems, but they are tangential to this discussion. Stuart Hameroff’s position is pretty close to what you are describing (and Orch-OR is a mechanism).

From my perspective you’re expressing the view that physicalism can’t account for consciousness, and that free will is subjectively obvious, and it must have something to do with consciousness. This is a view held by a significant number of people, though agreement on any more details is elusive. Panpsychism, dualism and idealism are all well known places to land this intuition. Personally I reject all of those because I think brains are necessary for consciousness (my position is more like a dual-aspect neutral monism), and the specific model this leads me to gives me a specific defence of free will (including a mechanism).

In both cases, this is still a free will decision. You made a free choice to flip the coin, or to order two desserts.

9

Only if what you want to do is self-contradictory. No square circles. And no granting humans free will if you also claim perfect knowledge of the future…

2 Likes
10

Like God, a lack of positive evidence.

As it stands we have every reason to believe that all of our body’s movements have a deterministic physical cause; I raise my hand because the muscles in my shoulder and arm contract, these muscles contract because neurotransmitters were released by my brain, these neurotransmitters were released because of the neural activity in my brain, these neurons were active in these particular ways because of the prior behaviour of the neurons next to them, these neurons behaved in this way because of neurotransmitters released by the sense organs, and so on.

There’s simply no evidence that anything like a non-physical will interjects this causal chain. If such a thing were to happen then we could expect to see “energy from nowhere”.

1 Like
11

Lack of evidence is not evidence of lack. That would not be a justification.

1 Like
12

I didn’t say that lack of evidence is evidence of lack. But lack of evidence can, and does, justify believing that something isn’t the case. A lack of evidence that Zeus exists justifies my belief that Zeus doesn’t exist.

13

But I am not trying to prove free will exists. I am merely trying to show why it’s not impossible. The “positive evidence” is subjective (direct awareness of metaphysical freedom, which is exactly why most people find MWI-style mind-splitting to be preposterous).

And what if, at the core of all this, consciousness is a process which involves fundamentally non-computable value judgements? How can they have a deterministic physical cause? And again…the position I am defending doesn’t require that I prove this is true, only that it is physically possible.

Why would non-computable valuations require energy from nowhere? There is no reason to think they require energy at all.

14

Do you think we have “evidence” that consciousness exists?

Some people say no, because scientific evidence is impossible (we can’t even define consciousness scientifically, without rendering that word pointless). I’d say the majority think the idea that we have no evidence that consciousness exists is absurd – rather, it is the one thing we can be most certain of.

Free will is in exactly the same category. In my model of reality, it is the exact same thing.

15

These “non-computable valuations” presumably have a causal effect on the body; I “choose” to raise my hand and so my hand rises? Energy is required for me to raise my hand, and so if my hand is caused to rise, not by deterministic energy transfer between preceding physical causes, but by this “non-computable valuations”, then it would seem as if the energy that caused my hand to rise came from nowhere.

16

Maybe, but it doesn’t justify the assertion that Zeus doesn’t exist. It’s just an explanation for why you lean in a certain direction.

If this was a century ago and I told you there’s undetectable matter all around you, coursing through you, you would be justified in rejecting the idea, since it would be contrary to your worldview. But there actually is undetectable matter all around you. It’s called dark matter.

Same with free will. It may conflict with your worldview, but that only allows you to express your bias, because lack of evidence is not evidence of lack.

1 Like
17

Yes, but that is a subsequent effect, not the primary effect. The primary effect is the selection of one particular outcome over others, when the wavefunction collapses. This process does not require energy, and is not incorporated into any physical law. If energy was involved, we’d already know the physical mechanism for wavefunction collapse – it would be easy.

The energy required comes from exactly where science says it came from. This is more like the movement of a butterfly wing which causes a hurricane on the other side of the planet – your argument is equivalent to claiming that butterfly can’t cause the hurricane, because there’s nowhere for the enormous energy involved to come from.

18

Sure it does. My belief that Zeus doesn’t exist and my assertion that Zeus doesn’t exist are both justified by the lack of evidence that Zeus exists.

My belief and the assertion might still be false, but something doesn’t have to be true to be justified.

19

Lack of evidence for Zeus is not justification for asserting that there is no Zeus.

20

Yes, and the transfer of energy from one place to another follows deterministic physical laws. It doesn’t just “spontaneously” move from my fat cells to my arm because I (freely) willed it to.

The wave-function concerns such things as the spin or position of an electron, not the macroscopic movements of the human body.