Reflections on my Conversation With James Fodor

Who was right about psychophysical harmony?

1 Introduction

I recently had a chat about psychophysical harmony with James Fodor. Fodor’s always interesting to talk to—he’s really smart, yet quite critical of basically every theistic argument. I’m particularly glad that James hasn’t bizarrely turned on me the way his compadre Digital Gnosis has—after previously friendly interactions!

(The funny thing: I wasn’t even smiling until about 30 seconds after the linked time stamp. Apparently the best explanation of people smiling is that they’ve been caught, pinned down, and trapped, despite their attempts to be like jello, rather than, say, that they found something amusing).

I find the argument from psychophysical harmony to be a powerful piece of evidence for theism, something that makes theism much more likely than it would otherwise be. Ultimately, I’m still agnostic because I find the problem of evil to be such a powerful anti-theistic consideration.

James had three broad concerns: first, he isn’t sure that psychophysical harmony is a real thing; second, he isn’t sure that naturalism lacks an explanation; third, he isn’t sure that theism has a better explanation. I disagree on all three counts. I’ll be writing this from memory, so it’s possible I’ll forget a point or two that he made, but I should get the broad outline.

2 What is psychophysical harmony?

Psychophysical harmony denotes the fortunate pairing between the mental and the physical. It’s easiest to see with examples like the following:

1. When I have the conscious thought “I’d like my arm to go up,” my arm goes up. My thought is connected, in a fortunate way, with what happens.

2. My mental experience roughly represents the way reality actually is. For example, I see a laptop in front of me, and there in fact is a laptop in front of me.

3. I respond in rationally appropriate ways to stimuli. For example, when I feel pain, I act to avoid it.

This is puzzling given that it could have—in at least an epistemic sense—been otherwise. For example, C fibers firing causes pain and also aversive behavior. But it could have been otherwise—C fibers firing could have caused pleasure instead, while bringing about the same behavior.

Every time you have a conscious thought, there will be a physical state—call it A—which gives rise to a mental state, B, which gives rise to another physical state, C. For example, there’s some brain state (A) that makes my nose itchy (B) and gives me a desire to scratch it, which results in me scratching it (C). But it could have been otherwise—B could be replaced with D, or E, or F, or G, or an infinite number of other mental states. Given that these could be swapped out while keeping the physical states the same, evolution can’t explain it—for if B was replaced by D, the world would be physically the same, so there would be selection for disharmony.

So, for example, a much simpler pairing would be if every single possible mental state produced by any brain was of, for instance, eating cheetos. Or if, rather than prior to raising my arm willing my arm to be raised, I had the experience of blinding light. James says he’s not sure what would be disharmonious or bad about this and isn’t sure what this supposed harmony is supposed to consist in.

It’s notoriously tricky to define words like harmony, but I’ll give it a try. Psychophysical harmony arises when there’s a fortunate pairing between the mental and the physical. The content of the mental resembles, in some way, the content of the physical. So it’s fortunate and harmonious that when I want to move my arm, it moves in the way I wanted to move it, because the content of my desire resembles what happens. Same with accurate perception.

I think there are four things that are especially valuable about harmony, of the type that we have.

First, it enables us to have accurate beliefs. If, rather than having perception that matches reality, our experiences were constantly flashes of light or the experience of eating fish, we would never come to have knowledge. That seems bad!

Second, it enables us to be agents acting in the world. If our conscious experience was just of seeing a red wall, then we wouldn’t be agents in any meaningful way. If we never had a will, or had a will that didn’t correlate at all with what we did physically, then our sense of agency would be lost. Being agents acting in the world is a requirement for any sort of valuable relationship.

James claimed, in response to this, that we could still be agents if prior to raising our arm, there was the qualitative experience of a flash of yellow light, rather than of us willing to raise our arm. But this seems to strip us of agency! If we don’t choose to raise our arm, but instead have a random experience followed by some action, then there’s no meaningful sense in which we’re choosing to act. The thing that’s especially harmonious about the thought preceding limb movement, for example, is that it involves making a choice, rather than just having some experience that results in movement. That’s why there’s an important sense in which you control your limb movements but don’t control your heartbeat.

Third, even putting aside questions of harmony and fittingness, we have complex experiences. The simplest psychophysical laws would produce very simple consciousness—say, of just seeing a red wall. So if you think that there’s something about our experience that is better than just having a constant experience that doesn’t change, probably having no pleasure, then that calls out for explanation.

Finally, psychophysical harmony enables us to have conscious experiences that are about other things. The simplest psychophysical laws don’t enable us to have any rich knowledge, love, or deep pleasures, for they wouldn’t represent anything in the world. But if knowledge, love, and deep pleasure is good, then so too must psychophysical harmony be.

3 Why does naturalism have trouble explaining it?

James is a physicalist. He thinks that consciousness is just a physical thing, and that through empirical investigation, we discover which physical things are equivalent to which mental states. For instance, by looking at brains, we learn that C fibers firing and pain are the same thing. I find physicalism implausible, but think that the version of physicalism that James endorses is just as vulnerable to the psychophysical harmony argument as any other view, so we didn’t need to argue about physicalism (except the roundabout way that psychophysical harmony gives one a reason to abandon physicalism).

James’s argument seemed to be that on physicalism, the connection between the mental and the physical is necessary. Just as water couldn’t be anything other than H20, pain couldn’t be anything other than C-fibers firing. So then the fact that pain is C-fibers firing doesn’t call out for an explanation.

This, however, does not succeed. Pointing out that some fact is necessary is not an adequate explanation of why it is so. Imagine someone is a necessitarian—they think that everything that happens necessarily happens. If they got 100 royal flushes in a row in poker, it wouldn’t do to explain it by saying “it’s just necessary!”

James responded to this example by saying that this still wouldn’t explain the process by which it was necessary. How did they get so many royal flushes? But there is a perfectly good necessitarian explanation—it’s necessary that the person would, by chance, keep getting royal flushes! It’s not that this theory can’t explain the data, it’s that the theory isn’t likely.

Let’s investigate this example a bit more, because I think it’s crucial for understanding where James goes astray. The problem with this explanation is that it has a low prior probability. The prior probability of a theory is the odds that it’s true prior to looking at the evidence. The theory that a fairie rigged a deck of cards to get some particular sequence explains why there’s that sequence, but it’s a bad theory because it has a very low prior probability.

The necessitarian explanation of the royal flush is not a good one because it has such a low prior. Suppose all you knew was that necessitarianism was true—you’d still think it very unlikely that the person would get so many royal flushes. So the theory that necessitarianism is true and happens to be necessary that a person gets a bunch of royal flushes has an absurdly low prior probability.

Thus, you should think the hypothesis that they’re cheating is a better explanation than the necessitarian explanation because it has a higher prior probability.

James noted that, through empirical investigation, we can learn what things necessarily are. For example, through empirical investigation, we know that water is H20—necessarily! Thus, it isn’t epistemically possible that it be anything else.

This is right as far as it goes. If we grant that physicalism of the type James endorses is true, then there’s nothing to be explained. But psychophysical harmony gives a reason to reject type B physicalism, because conditional on type B physicalism, it’s unlikely that the psychophysical laws would necessarily be harmonious.

This is because P(psychophysical harmony)|theism+non-physicalism>|~theism+physicalism. If all you knew was that physicalism was true, you wouldn’t expect the necessary psychophysical laws to be harmonious.

This is a pretty abstruse point so let me give two examples. Imagine we discovered that the laws of physics were such as to produce, on every atom, in tiny font, the words “in the beginning was the word, and the word was with God, and the word was God.” Then, we’re considering between two hypotheses: theism and atheistic necessitarianism.

This would, I claim, be extremely strong evidence for theism. This is because the odds of this occurring if theism is true are much higher than the odds of it occurring on atheistic necessitarianism. But the atheistic necessitarian could reason like James, saying “we’ve done empirical investigation and discovered that the laws are such as to necessarily produce these letters, so therefore necessitarianism predicts this.” This would be silly reasoning for, I claim, the same reason James’s reasoning goes awry.

It’s true that if atheistic necessitarianism were true, then those laws would have to be necessary. But the odds of the laws spelling out “in the beginning was the word, and the word was with God, and the word was God,” conditional on atheistic necessitarianism, are super low. The more specific hypothesis, that atheistic necessitarianism is true + the laws spell out “in the beginning was the word, and the word was with God, and the word was God,” explains the data, but it has an absurdly low prior probability—why’d the laws spell out that in particular.

Here’s another analogy: imagine that Pedro shows you a particularly weird polygon. It exhibits some strange mathematical property, wherein it produces some string of numbers if some procedure is done to it. You’re not sure if this is a necessary property of polygons or if it is specific to this polygon made by Pedro. So you calculate the string of numbers, convert it to English, by pairing 1 with A, 2 with B, etc, and it spells out “this polygon was made to have this property by me. Much love, Pedro.”

Upon finding this out, you get very strong evidence that this is not a generic property of polygons. This is because the hypothesis that Pedro would make a polygon with this property has a much higher prior probability than the hypothesis that the string of numbers would have the property of, by chance, spelling out, if converted into English, “this polygon was made to have this property by me. Much love, Pedro.” Both hypotheses explain the data, but one has a much higher prior probability.

It’s true that, through empirical investigation, you’ve discovered what the sequence spells out. But you haven’t discovered why it spells it out. Given that the odds of it spelling out that sequence are higher conditional on the hypothesis of design than on the hypothesis of it being a necessary mathematical fact, you get strong evidence that Pedro made it specifically to have that property.

James rightly notes that we can scientifically establish that A reduces to B, and then there’s nothing more calling for an explanation. This is true—if we know that A reduces to B, then this doesn’t need an explanation. But if we’re not sure if A reduces to B, and there’s another hypothesis that better explains the properties of A, then that theory is thereby better. The reason we don’t infer that there’s a God who made the specific laws of physiology, rather than having them reduce to chemistry, is the reductionist hypothesis is a better explanation of features of physiology.

4 Does God explain it?

I think theism is a pretty good explanation of psychophysical harmony. This is because harmony is both improbable and valuable. James said he wasn’t sure why God would make a psychophysical harmonious world rather than, say, a world without harmony where everyone is in constant unending ecstasy.

I agree that the ecstasy world is pretty likely on theism. But I think that it’s also reasonably likely that God would make a world with psychophysical harmony, so the odds of harmony conditional on theism are much higher than the odds of it conditional on atheism. Psychophysical harmony is, as I’ve argued before, needed for any valuable relationships—and God would be likely to make valuable relationships.

James disputed that a psychophysically harmonious world would be more valuable than one without harmony. We could still have interactions with people if they were disharmonious. I disagree, however—those experiences wouldn’t be valuable in the same way. If objective list theory is true—which theists should think is, as most philosophers do, given that absent objective list theory the problem of evil is basically unsolvable—then it would be bad for you if your loved ones never had thoughts about you, but were constantly experiencing drug-trip-like ecstasy. This is similar to an example Richard has given.

Next, James claimed that he wasn’t sure why the theistic explanation was any more legitimate than the naturalistic explanation. Both theories can explain it—theism by saying that God exists and wants psychophysical harmony, naturalism by saying that harmony is necessary.

It’s true that both views can explain harmony. But theism is a much better explanation! Conditional on there being a perfect being, psychophysical harmony is pretty likely—maybe 60% likely. In contrast, given the infinite number of ways disharmony could be necessary (I use could be in an epistemic sense—prior to finding out the laws, we don’t know which way it would be), the odds are very near zero conditional on naturalism. Therefore, if one thinks, prior to considering psychophysical harmony, that the odds of theism are, say, 1%, psychophysical harmony should make them think that theism is almost definitely true.

Let’s go back to the earlier example of the laws of physics spelling out “in the beginning was the word and the word was with God and the word was God,” on every atom. That’s best explained by theism, because even though the odds of that are low on theism—perhaps 1 in 100,000—the odds are much lower on atheism. Psychophysical harmony is similar—it’s so improbable on naturalism that it’s strong evidence for theism.

Now, if you think that perfect being theism is ridiculously unlikely, then you shouldn’t be moved by this. The presence of psychophysical harmony is strong evidence for a pixie that just creates psychophysical harmony, but that’s still a ridiculous view to have because of its low prior probability. But I never claimed that psychophysical harmony gives a person who starts out believing theism is almost guaranteed to be false a reason to be a theist all things considered. I merely claimed it’s strong evidence for theism.

All in all, I really enjoyed the conversation with James and hope to have more in the future! However, the conversation did nothing to shake my conviction that psychophysical harmony is very strong evidence for theism.