Psychophysical Harmony strongly argues AGAINST God

In 2024, philosophers Dustin Crummett and Brian Cutter took an idea that dates back at least to 1890 and sharpened the idea, formulating a genuinely novel argument for the likelihood of God’s existence—one that (popular Substacker and continued philosophical frenemy of this blog) ranks as an “S tier” argument. Considering the fact that monotheism has been hotly debated for at least a couple thousand years, this is impressive (I mean that genuinely); the idea is both clever and creative. It’s also unfortunately a backfire: Instead of proving the existence of God, the idea ultimately promotes the exact opposite position.

In this post, I’m not actually going to argue that God is dead, and if He is, I’m certainly not saying that Crummet and Cutter killed Him. The two things I’d like to establish are:

1. Where the “psychophysical harmony” argument goes astray.
2. How the psychophysical harmony argument actually distracts from the superior “psychophysical presence” argument for God.

I’m an atheist, but for that second part I’ll be essentially defending God (with one major caveat). Why? Because honest to God (please excuse the idiom), my goal in life isn’t the evangelization of atheism. My greatest desire is the killing of bad ideas, so people can spend more energy and time debating the better ideas. I’m happy for any reader who believes in God, and equally happy for any reader who doesn’t. I only feel aggrieved when that belief or disbelief is built atop backwards chains of illogic.

Like psychophysical harmony.

The basic idea behind psychophysical harmony is the observation that our actions and our emotions that seem to provoke our actions are apparently well-aligned, i.e., harmonious: Pain typically follows the things we would like to avoid (like touching fire), joy typically follows the things we would like more of (like hugs), and we don’t live in a strange and confusing world where touching fire brings us joy and hugs bring us agony, yet we somehow strive to avoid fire and maximize hugs despite this.

If this feels sort of dumb and self-explanatory—like, “of course we act to avoid pain. How would any other world even make sense?”—then one, welcome to philosophy, where we often question the basic assumptions that hardly even make sense to question. And two, don’t worry; I’m about to use this observation for fun effect and (combined with Crummet and Cutter’s logic) kill God. Then turn around and bring Him back again.

How to disprove God in 4 easy steps

1. Assume epiphenomenalism is false (we’ll revisit this assumption later).
2. Because of (1), we know that our conscious feelings influence our behaviors. Evolution would weed out disharmonious feelings as inefficient (a brain that has to constantly fight its own conflicting feelings would be less prone to survive), therefore:
   1. P(Harmony | Naturalism) = very high
3. God, however, has no such restriction against disharmony. A benevolent God is actually more likely to have created a disharmonious universe, because if disharmony is truly possible (as is the cornerstone of Crummet and Cutter’s argument), then all phenomenal states could simply be replaced with pleasurable ones. Therefore:
   1. P(Harmony | God) = low
4. By Bayes’ Theorem:
   1. P(Atheism | Harmony) >> P(God | Harmony).

Let’s break this down further.

For step (1), here is a nice explanation from Connor Jennings:

Epiphenomenalism is the view that consciousness is non-physical, but has no causal effect on the physical world. As I described in a previous article on Dualism, it’s as if consciousness is just a non-interfering passenger along for the ride.

If Epiphenomenalism is true, it would mean all the physical events would’ve played out the same way were no one conscious. If consciousness is causally impotent, then no matter what mental states you have, they’d have no effect on your actions. We could have white noise going on in here and we’d still all be walking around, going to work, and doing the worm at wedding parties without being asked (what, just me?).

For the rest of this “proof” to work, we need to assume that consciousness actually does something meaningful. I’ll discuss later what happens when we remove this assumption.

For step (2), we have a choice:

1. Either the conscious experience of pain intrinsically leads to avoidant behavior, and pleasure to attraction, and any deviation from this relationship requires non-zero extra mental processing to compute (even if only to flip the valence of a single neural bit)...
2. …or perfect masochism is a non-contradictory idea, and it’s possible for agents to maximize their own suffering without any extra steps nor loopholes in mental-programming.

If we accept the first, then that extra marginal bit of mental processing (to take literally any action) would be a maladaptive evolutionary trait. After a few billion years, psychophysical disharmony would be mostly impossible outside of the rare edge case or intermittent backfire mutation.

If we accept the second, then we can imagine harmony as a spectrum, with infinitely pleasurable/painful “appropriate” feelings on one end (pleasurable attractor states and painful avoidant states), and infinitely “perverse” feelings on the other end. With no impact on evolution, we’d expect P(Harmony | Naturalism) = 1/2, which leaves us far from a proof, but still makes naturalism more favorable than theism’s chances defined in step 3.

For step (3), I must give attribution to Crummet and Cutter. I myself would have assumed something like the following:

There is no such thing as psychophysical harmony, because no particular brain states are a normatively appropriate match for any particular phenomenal states. It might be an interesting metaphysical fact that a certain arrangement of grey matter serves up one experience, while another arrangement serves up something different. But it’s not as if you’d ever see a full description of some brain state and then think to yourself, “Man, that’s a perfect fit for what tomatoes taste like!” or “Gosh, I hope that’s not the experience of stubbing your toe.” Ultimately, it’s all just brain goop, and no specific style of brain goop is ever more or less “fitting” for one experience versus another.

(Thank you , post here.)

If we take for granted that actually, no, phenomenal states are more than just arbitrary metaphysical facts about certain arrangements of experience; they are properties independent of the physical facts of a brain’s processing, and therefore can be paired in any conceivable way, and adding onto this the fact that God is by definition omnipotent—then we’re forced to face the fact that God could have chosen to make us feel pleasure for both positive and negative stimuli, and chose not to. He could have at least minimized the magnitude, frequency, and maximum possible amounts of phenomenal pain, but chose not to. God could and would have done this were He benevolent; God is defined as benevolent; ergo God DNE.

It’s the Problem of Evil all over again, except this time eliminating some of the usual objections to the Problem of Evil—like the need for suffering to make our choices meaningful. When Crummet and Cutter sever phenomenal states from behavior, then they also sever phenomenal states from our free will: Even if I felt pain when I should feel pleasure, and vice versa, I’d be unwise to assume that anybody else is the same; safer to assume I’m a freak mutation (or that I’ve been cursed by the devil), and working to maximize other people’s joy will still do so in truth (on average). Therefore my choices would absolutely still be morally meaningful, and in short:

The possibility of psychophysical disharmony negates theists’ free will counter to the Problem of Evil.

If you agree, maybe consider subscribing - RA is a blog supported by free subscribers. Or if you disagree, please drop a comment, let me know why!

Reviving God in 3 easy steps

Crummet and Cutter’s paper makes the critical error of not imagining how much positive utility God could reap from disharmony. Focusing only on possible naturalistic disharmonies paints a more persuasive picture for their purposes, and I mean that sincerely—even if we accept as afore-argued that evolution would disfavor pain-for-positive-stimuli and vice versa, that still leaves open other plausibilities:

• Under naturalism, we could have lived in a hollow world completely lacking in consciousness or any sort of phenomenal experience.
• Under naturalism, we could have had a uniform experience of constant, bland white noise.
• Under naturalism, we could have had far more extreme degrees of pain and pleasure.

Out of all these possibilities, why would naturalism have led to our particular configuration? To me, this is the kernel of insight at the heart of Crummet and Cutter’s paper. And their argument gets stronger when we simplify. Cut away all the confusing harmony stuff; instead we could just ask about psychophysical presence: Why do we feel anything at all?

Here’s a weak proof of God’s existence:

1. Why do we feel pain?
   1. We might say: That’s how we learn from our mistakes. It builds character. But it seems to me that learning can happen in the absence of pain. For instance, I don’t think anybody believes the YouTube algorithm feels pain—yet it’s certainly capable of learning that I enjoy short-form content around archery, linguistics, and Taskmaster.
   2. We might also say: Pain acts as a strong signal we can’t ignore. We wouldn’t be so motivated to address the roots of our problems if pain wasn’t so painful, right? Except I don’t see any reason why that should be true either. The YouTube algorithm presumably wouldn’t work any better if we simulated physical agony any time it made a poor recommendation.
2. There doesn’t seem to be any good naturalistic answer to (1). Phenomenal states seem extraneous to the needs of evolution.
3. If there isn’t any alternative to naturalism other than theism, then… there’s got to be a God, right? By process of elimination?

Alone, this isn’t enough to kill naturalism. If phenomenal states are indeed extraneous to the needs of evolution, and in fact extraneous to the physical world entirely—if epiphenomenalism is true, unlike my proof’s earlier assumption—then we might defend naturalism by saying: Maybe phenomenal experiences emerge from the physical, like some sort of opposite of a shadow world. Maybe pain is just the manifestation of complex dispreferences. Questioning the phenomenal is akin to questioning why someone sees the color blue the exact way that they do. The color had to look like something, just like dispreferences have to be stimulated in some way.

Except I can’t accept this line of reasoning either, because I know for a fact that epiphenomenalism is false.

A world of hollow dispreferences

Imagine a hollow world in which consciousness is absent, or white noise, in which agents still behave the same way as us. Babies still cry loudly to express their dispreference at hunger or dirty diapers. People still say, “THAT HURTS!” in response to bodily or relational dispreferences (synonymous to shouting “I DISPREFER THAT!”). Philosophers still argue about hedonism versus eudaimonism versus other theories about utility, suffering, and virtue.

In a hollow world, each of the following questions still make sense to ask:

• Why do I hold such strong (bodily/other) dispreferences?
• Why couldn’t God have made a world that always satisfies my preferences?

But in a hollow world, there’d be no equivalent to the post you’re reading right now. They wouldn’t have any provocation to ask:

• Why must my strong dispreference feel this way, instead of being something hollow?

I know that my readers will understand what I’m referring to when I refer to phenomenal states, but a hollow person reading this post wouldn’t understand this discussion anymore than someone born colorblind could imagine the color blue.

It sounds circular, but engaging in discussion about the mystery of consciousness proves that there does indeed exist a mystery of consciousness—which means we’re back to having something naturalism can’t explain.

The Problem of Evil, combined with psychophysical harmony, pretty convincingly shows that we can’t have a God who’s both omnipotent and omni-benevolent. If naturalism is out too, maybe that leaves only the possibility of some theistic entity with an imperfect amount of power or not-quite benevolence—a God that might be worthy of your faith, but not Yahweh nor Allah.

Maybe.

The Principles of Psychology, William James

I tend to agree with his takes on shrimp, insects, politics and AI, but I’ve oft strived to correct his takes on anthropics and theism.

Or there exists a mental layer that stimulates both the consciousness and the behavior, and the nature of that stimulation works the way you’d intuitively assume.

This can also be shown a second way: The experience of meaningfulness is itself another phenomenal state which God could have granted us, no matter the circumstance of our behaviors.

One extremely speculative idea I have that might explain evolution’s need for our phenomenal states, inspired by the layered sort of logic of the Blue Eyes logic problem: You’ll be considered more reproductively fit if your potential mates know that their preferences will be taken seriously, which can be evidenced by your taking your own preferences seriously, but which can be better evinced by melodramatic lamentations, lamentations which themselves can only be taken seriously if we dig down layer after layer and find a cornerstone that’s not hollow, in a way hollow-men would never perfectly imitate; as I pointed out, a hollow-man would never be able to understand and certainly would never have written the post I’m writing now, and everything else builds up from there.

Me, I don’t know. My atheism is self-professedly irrational; rationally, I’m an agnostic.