Why I'm An Atheist Despite Psychophysical Harmony

Responding to the best argument for god

Explaining the argument

I am going to do something that might be a huge mistake, that makes me acutely aware of my own mortality; I am going to challenge the argument from psychophysical harmony.

For those who don’t know about the argument from psychophysical harmony — the following video provides a great explanation of it. I’d really highly recommend watching.

The original paper on it is also worth reading. The basic argument is the following. It turns out that there are harmonious relationships between the mental and the physical. For example, the mental perceives events as they occur in the real world. When you want your arm to go up, your arm does — there’s a harmonious relationship between our desires and how they’re manifested in the real world. You can experience value and then act to get more of it — this is a priori surprising.

C fibers firing cause pain. However, the intrinsic probability of this is super low. It’s just as intrinsically likely that C fibers firing would cause pleasure, or the taste of hummus. But the vast majority of possible psychophysical laws would involve disharmony between the mental and the physical.

To see this, there are two possibilities — either the mental causes physical changes or it doesn’t. Let’s consider each possibility.

If it doesn’t cause physical changes, then it’s incredibly unlikely that there would be a relationship between the physical and the mental. Let’s just take the example of pain. Pain happens when there are C fibers firing — but it could be the experience of hummus that comes when C fibers fire. There’s no reason why pain is more likely. Thus, on naturalism, the prior probability of a world exactly like our own except instead of pain accompanying the stub of a toe, or the feeling of a cut, or the sadness of loss, the taste of hummus accompanying it, is roughly the same as the odds that pain would accompany it. On this view, it’s an enormous coincidence that there’s a match between the physical and the mental. Why is it that our arm goes up when we want it to, if our brain doesn’t cause it to go up. If the mental doesn’t cause the physical, then we should expect total chaotic disharmony between the mental and the physical.

What if the mental does cause the physical? Suppose that C fibers firing cause pain and that pain causes the change of molecules in a way that causes an aversive reaction. Well, this will still be an enormous coincidence. The vast majority of possible psychophysical laws will involve pain not causing an aversive reaction. One equally intrinsically plausible psychophysical law, for example, will have the experience of Hummus be accompanied by aversive reactions. But it’s much more likely that aversion would be accompanied by hummus, or the experience of a warm bath, or the experience of being a lizard basking in the sun, or something else. Thus, any possible harmonious psychophysical law will be orders of magnitude less intrinsically probable than the multitude of possible psychophysical laws that produce disharmony. Given the multitude of psychophysical correlations, almost all of which are harmonious, the odds of this harmony, conditional on the psychophysical laws not being rigged, are incredibly low.

But I’m still an atheist

I think that the argument from psychophysical harmony is the best argument for theism by leaps and bounds. But I think theism starts out as really implausible. Thus, if there’s an alternative way to explain the data, even if it’s really implausible, I’m still probably going to find it more probable than theism. Psychophysical harmony is basically the only thing that moves me in favor of atheism.

Here, I’ll provide some reasons that this argument doesn’t make me a theist. Later, I’ll provide an alternative explanation of psychophysical harmony. This is, I think, my favorite solution.

Skeptical atheism

When confronted with the overwhelmingly powerful problem of evil, skeptical theists are those who say the following: Look, we don’t know why god allows evil, but god is very mysterious, and it’s likely that we wouldn’t know this even if there was some explanation. Thus, there’s some explanation for evil, even if we don’t know what it is.

This is basically my reaction to psychophysical harmony.

I think that it’s really hard to be super confident in the argument from psychophysical harmony given a few background facts.

One of them is that it’s such a new argument. Unlike the problem of evil, it came out around 2021. Thus, it hasn’t been subject to the myriad possible objections that philosophers should subject it to before we use it to conclude that theism is correct. The argument is just so new, but that it’s hard to be that confident in any of the premises.

Second, I have a really low prior in theism, largely because of the problems of evil and hiddenness. Given this, my prior in there being something wrong with the argument will be high, because I think it’s just so unlikely that it succeeds.

Third, consciousness is unbelievably mysterious. We have no idea why or how it exists. It opens up a huge number of mysteries. We don’t know whether it’s physical or non-physical, or even possessed by all atoms. This makes it more likely that we’ll have some dramatic paradigm shift; a dramatic paradigm shift being on the horizon makes it less likely that arguments built on assumptions about consciousness, pre-paradigm-shift, will be successful. It would be like making arguments about how theism solves the problem of motion in the 4th century.

Biting the bullet is fine; theism has a low prior

I think that, even if we have to bite the bullet and accept the full force of the argument, atheism is still plausibly going to be more likely than theism. This is because I think that atheism + psychophysically harmonious laws is more probable than theism.

While it’s true that there are a bunch of ways that the psychophysical laws could go awry, there are a bunch of ways theism could not happen. The unimaginably overwhelmingly large majority of possible existences don’t have god. Thus, theism starts as super implausible.

I also think that theism does have a revenge problem — it assumes, not explains, psychophysical harmony. After all, there will be a harmonious psychophysical relationship between god’s mind and his body — or, if he doesn’t have a body, his physical manifestation. When god wants any arm in the universe to go up, it does.

Additionally, god has several properties that have to all exist. If god were omniscient but lacked a physical manifestation, or was a causally inert spaghetti noodle, or were nearly any other possible omniscient thing, he wouldn’t explain the data. I’m basically unconvinced that these are less likely than theism — and because there are so many of them, theism will have a super low intrinsic probability.

Dustin thinks that this can be avoided by positing that God follows from pure perfection — thus, theism is just positing one property. But I don’t think that this works. Perfection sounds like a simple, pure property, but it really is a very complex thing. If you tried to write a computer program to simulate the idea of perfection, it would take a lot of bits. Thus, lots of information is contained in the idea of perfection. I don’t think that the idea of perfection simpliciter is simpler than just positing the fundamental features of perfection as brute — which means the revenge problem rears its head.

Apologetics squared — the person who made the video — argues that the revenge problem isn’t an issue because God’s psychophysical laws will be simple. But I think it’s plausible that the psychophysical laws that apply in the universe will also be simple — perhaps as simple as the ones that apply to god.

Defending this in detail would require an extraordinarily long article. I won’t do so here. But if you adopt this relatively controversial view, then the argument from psychophysical harmony won’t be very forceful.

Natural teleology explains it as do a few other things

Natural teleology is that idea that the universe is not purposeless, but instead it is inclined towards the good. It, despite lacking a mind, naturally promotes the good. It’s been defended by Nagel, among others. This explains psychophysical harmony without positing a god. If the laws of the universe are oriented towards the good, that explains why they promote goodness. Similarly, if there are fundamental laws that are oriented towards basically anything that is promoted by psychophysical harmony, they would also explain this. Buddhism also plausibly explains it, by having Karma explain everything. Axiarchism also says that things sometimes exist because it’s good if they exist — that would also explain harmony.

My favorite explanation

Suppose we accept the following few things.

1. There are huge numbers of universes.

2. Universes have their own psychophysical laws.

3. Beings in the universes can create new universes if they’re sufficiently motivated.

Plausibly this would create an anthropic selection effect for the sets of psychophysical laws that create beings that are more willing to create universes. But beings that are motivated to do good things would be more willing to create universes — after all, new universes are a good thing. Thus, most universes that would come to exist in the long run would have psychophysical harmony, because most of them would be created by good beings that want to create good psychophysical laws.

Additionally, if we expect that beings would evolve morality, they’d be motivated to do good. If they’re motivated to do good, they’d want to create sets of psychophysically harmonious laws.

Now maybe it would be the case that the optimal sets of psychophysical laws would be not ones that orient towards the good, but instead ones that orient towards the willingness to create lots of universes. But we can stipulate that they can’t create the universe exactly because universes are unpredictable — they can just create the harmonious laws. Thus, more finely tuned laws that orient beings to want to create lots of universes would plausibly die out, because they’re not precise enough to apply to the range of possible beings that would exist in the universe.