Dualism is an attractive philosophy with an Achilles’ heel. Mind and body seem to be fundamentally separate things, yet dualists since Descartes have never been able to solve the famous problem of interaction. If mind and body are in different ontological categories, then how could they possibly interact with each other, even in principle?
Descartes didn’t give a good answer, nor has any other dualist I’ve ever encountered. They tend to respond, “Well, we don’t know how mind and body interact, but that doesn’t mean it’s impossible!” This is not a satisfying answer, even though I agree with them, and lots of philosophers don’t find it compelling at all. In fact, many have claimed the problem of interaction is so severe, it’s a refutation of dualism. They think the problem can’t be solved even in theory.
I disagree. I’ve been trying to figure out a plausible mechanism of interaction for many years, and finally, I’ve got one. I have a working resolution to the mind-body problem that solves the problem of interaction. It not only supports substance dualism, but substance pluralism, which doesn’t restrict the amount of ontological categories to only two. If you’re sympathetic to dualism, or are familiar with its history, this is the type of theory we wanted Descartes to figure out a few centuries ago.
I call it a theory of indirect interaction. Mind and body do not directly interact, but they effectively interact – i.e. the state of one affects the state of the other. This theory has many favorable properties:
1) It gives a plausible mechanism for interaction between objects in any ontological category – not just mental and physical. Even if there are a hundred more categories, the mechanism could still work.
2) It allows for two-way causality. Physical states can affect mental states; mental states can affect physical states.
3) It is free-will-agnostic. There is a clear opening for the role of free will, but the system works perfectly fine without it.
4) It doesn’t break the laws of physics – or perhaps more precisely, it doesn’t break the laws of causality. It might simply broaden the scope of the laws of physics.
The purpose of this article is not to claim that “This is the way things actually work in the world!” Rather, it’s to demonstrate that in principle there could be a mechanism for things in different ontological categories to effectively interact with each other. Whether or not I’ve discovered the “real” mechanism is a separate question.
Examining Causality through Billiards
The best way to illustrate the theory is by first taking the mind out of the equation and analyzing purely physical phenomena. By breaking down physical phenomena into their most fundamental form, it will elicit the concepts necessary to understand indirect interaction.
My favorite example of “purely physical phenomena” is the motion of balls on a billiard table. Let’s take a simple scenario. Imagine that there are only two balls left on the table – the white cue ball and the black 8-ball. Imagine the cue stick strikes the white ball, the white ball rolls forward and hits the 8-ball, then the 8-ball rolls into a pocket.
Let’s break this scenario down as thoroughly as possible. What are we really talking about when we reference “pool balls”? What exactly are such objects? What are we describing when we say, “The 8-ball rolled into a pocket”?
Rather than give an extended analysis of the metaphysical status of pool balls (which you can read about in my article “No, Chairs Do Not Exist”), let me give you one plausible position. What a “pool ball” really is is “units of matter arranged in a particular way in a particular part of space.” Let’s call those units “atoms.” Some philosophers might put it, “A pool ball is really just atoms arranged ball-wise.”
Let’s consider this phrase: “The 8-ball rolled into a pocket.” What exactly are we describing? What is this “rolling” phenomenon? If the pool ball is just atoms, then we can rephrase it this way: “Some particular atoms changed their positions in space.”
Another way to understand it would be to say, “At Time 1, atoms were in Position 1. At Time 2, atoms were in Position 2.” In fact, that’s a pretty good description of motion in general.
That’s an abstract way to understand pool balls rolling on a table. Now let’s ask two interesting, yet difficult questions:
1) When the white ball hits the 8-ball, why does the 8-ball start moving?
2) Why does the 8-ball move on its particular path rather than some other path?
Notice that when we reduced the phenomenon of motion to “atoms changing position,” it doesn’t actually communicate an extremely important piece of information:
The changes in position are not random.
We didn’t say, “At Time 1, the white ball struck the 8-ball, and at Time 2, the 8-ball started orbiting Jupiter.” No. There is a pattern to the motion. A predictable, observable pattern.
Why isn’t the motion of the 8-ball completely random? Why should the motion be predictable at all? Hell, why doesn’t the universe just spontaneously fall apart when the balls collide? What holds all of these objects into the same coherent, predictable system?
One plausible answer is this: There are laws of the universe. Physical laws keep the whole thing together. They make motion predictable. The reason that the 8-ball rolls into its pocket instead of orbiting Jupiter is because there are laws of physics which govern the behavior of objects. These laws have a real existence.
That’s a nice-sounding answer – and physicists might like it – but it provokes many more questions. For example:
Are the laws of physics physical themselves? Do the laws of physics take up space or weigh anything?
What is the relationship between the laws of physics and the objects governed by them?
What is the mechanism for the laws of physics?
In other words, how do laws keep objects in order?
If our explanation for physical phenomena appeals to laws, then we’ve posited the existence of two radically different types of things: physical phenomena and the laws which govern them. Atoms in space, by themselves, are not sufficient to explain why they move in predictable ways. There must be underlying principles, or laws, which determine their behavior. By thinking about “purely physical phenomena” this way, it gets us one step closer to solving the mind-body problem.
Inputs and Outputs According to Laws
Let’s break down our billiard example even further. Instead of only identifying what we see, we need to identify exactly what we don’t see.
Treat the table and balls as a whole system. We see changes in the position of the balls, which means we see the system in different states at different times. But we don’t see the glue between the states. We don’t actually see the laws that we’re appealing to in order to explain the phenomena. We’re simply inferring the existence of laws and causality to explain the patterns in our observation, but we don’t see the laws themselves.
It’s helpful to keep rephrasing and condensing our language. Instead of saying, “At Time 1, the object was in Position 1, and at Time 2, the object was in Position 2,” we can simply talk about “states.” We can say, “State 1 was followed by State 2.”
So another abstract way to understand physical phenomena is to say, “There is a series of states. Each state contains a particular arrangement of atoms in space. The changes between states are non-random and happen in accordance with laws.”
This allows us to re-ask the previous questions:
1) Why is State 2 the way that it is and not some other way?
2) Why doesn’t State 2 include the 8-ball orbiting Jupiter?
Any given state is determined by its previous state. Since State 1 was a particular way, the laws of physics determine that State 2 must follow State 1.
Or, to put it more succinctly: Preceding states determine future states.
Let’s consider these states of the universe from another perspective: As “inputs” and “outputs.” Inputs yield outputs. So if we call State 2 an “output,” we could say that State 1 was its “input.”
What determines that State 2 is an output of State 1 in particular? The laws of physics. If we think about states of the universe as being inputs and outputs, we can understand the laws of physics as a kind of mathematical function – they take inputs and turn them into specific outputs!
This is a theoretical picture in which the universe is like a gigantic computer that keeps churning out new output states. The outputs are determined by their inputs. Then, those outputs are used as inputs for the next state. The laws of physics are the specific code that determines exactly how inputs relate to outputs.
So, we can reduce the physical universe to a very abstract formula:
Input state + laws of physics -> output state.
Then, that output state is treated as the next input, and the universe churns out a new state.
Information and State
We’ve posited the existence of two radically different types of things to explain physical phenomena – spatially-extended atoms in space, and non-spatially extended laws of physics which govern their behavior. Whether or not it’s necessary to give the laws of physics a real existence is an interesting question (and it turns out that it’s awfully hard to explain the regularity of physical phenomena without them!). Regardless, this metaphysical picture allows us to understand how objects in different ontological categories might be able to interact with each other. However, we must go deeper.
Reduce the physical universe to “atoms in their position in space at any given time.” Those atoms themselves are not enough to determine the future state of the universe. There must also be laws. But that brings up several more difficult questions:
1) What connects the physical states to the laws?
Why aren’t the physical states completely separated from the laws? What’s the glue between the laws and the physical states?
2) How do the laws of physics “know” the state of the universe? Why doesn’t the universe “get it wrong” when determining future states?
3) How are states treated as inputs? What’s the format?
All of these questions can be answered by the final piece of the puzzle: Information. The universal mathematical function that takes inputs and turns them into outputs has information about the physical state. This information is itself non-physical. The information is the glue between the laws of physics and the physical states themselves.
So, what actually gets used as the “input” is information about the physical state, rather than the physical state itself. The subsequent output is another purely physical state, then information about that output is used as the new input state!
This is an abstract way to understand the mechanics of a physical system. Crucially, it allows for real ontological differences between the physical state, the information about the physical state, and the laws of physics which take that information and generate new output states.
Think about the relationship between ordinary objects and your knowledge about them. Take your information about chairs. There’s a categorical difference between physical chairs and your information about physical chairs. Chairs take up space, while your knowledge about chairs does not take up space. The concept of a chair is not somehow embedded inside of chairs. Information is not the same thing as what the information is about. Information about physical states does not need to be embedded within physical states.
In this theory, the physical states are entirely concrete, not abstract. They are reducible to “atoms in space.” Yet, there can be information about those physical states which is abstract and not reducible to atoms in space.
There’s an interesting question about the metaphysical status of information. It’s “abstract,” but what exactly are abstract things? Are they mental? Platonic? This theory doesn’t require a particular answer, but it should be clarified that it doesn’t necessarily imply consciousness. Your knowledge of chairs is within your mind; you can have a kind of conscious experience of it. The universe doesn’t need to have any internal experience of knowing information about physical states, just like your computer doesn’t have to have an internal experience of “reading and knowing” the state of your hard drive. Information is processed in your CPU without consciousness.
So, to revise our picture of a physical system one more time:
We start with atoms in space. The universe has information about the position of the atoms in space. That information is used as an input into a function that we call “the laws of physics.” It then generates a new output state – i.e. atoms change position. The universe has information about this new state, which then gets put back into the function to generate subsequent output states. The universe progresses.
If this theory works, then we’ve just solved the mind-body problem and the problem of interaction. All we’ve got to do is add mind.
Mind and Brain
The picture I’ve just painted includes effective interaction between at least two ontological categories – the laws of physics and the spatially-extended objects that are governed by them. Now, it doesn’t matter how many ontological categories we posit; the same mechanism can still work. Instead of restricting output states to only spatially-extended physical stuff, we can expand the category of output states to include mental stuff as well – feelings, experiences, qualia, etc.
For example, take the conscious experience of seeing red. It’s a particular kind of mental state. In this system, it’s simply another output that will get generated with the correct input. Whenever the physical universe is arranged in a particular way, the output state of “experiencing redness” is generated. That output state does not need to be physical. It can be in an entirely different ontological category!
This allows us to expand the laws of physics to include laws of mental representation. Just like particular physical inputs yield particular physical outputs according to laws, the universe can also generate particular mental outputs with the right input. In other words, the universal function includes the informational criteria for generating both physical outputs and mental outputs.
This theory accords perfectly with the physical mechanics of sight. When physicists talk about “light rays entering the eye, stimulating particular nerves, etc.” they’re simply talking about changes in physical states. As these physical states change, the information going into the universal function also changes, and at some point, when the correct physical state has been reached, mental states start getting generated.
Notice: it’s not the physical state itself that’s generating mental phenomena. It’s not some mechanism in the brain. It’s information about the physical state which gets used as an input to generate a mental state in a different ontological realm.
In this theory, brains are not some unique object that “secretes consciousness,” as some philosophers have suggested. Consciousness is not to be found within a skull. There’s nothing intrinsically special about the atoms that compose a brain. What’s important is their arrangement and the corresponding information about them. If patterns and information about the brain are indeed what generates consciousness, then we also have no need to posit panpsychism, which suggests each atom might be “a little bit conscious” itself.
The reason that the brain is so closely correlated with conscious states is because it’s precisely the information about the atoms in space that we call a “brain” that yields consciousness. The brain state itself is not enough; it requires brain states plus the laws of physics/mental representation. So it shouldn’t be surprising that when people get brain damage, their conscious experience changes. This isn’t because the brain loses the ability to create consciousness. It never had that ability. It’s because when the brain state changes, information about the brain state changes, which then changes the input and subsequent output of the universal function.
This is why I call the theory a mechanism of “indirect interaction.” The brain isn’t directly generating consciousness. Instead, it’s the pattern of information corresponding to the physical state of the brain that generates consciousness. The effect is essentially the same. The state of the body affects the state of the mind, but it’s via an abstract mechanism instead of a purely physical or mental one.
Our experience of the world suggests that physical states can affect mental states and that mental states can affect physical states. For example, experienced meditators can regulate their body temperatures through deliberate mental focus. Even regular people can make their mouth water simply by envisioning a juicy steak when they’re hungry. Or, take one of the most significant examples of mental states appearing to affect physical states: The placebo effect. How is it possible?
Well, just like the outputs of the universal function can be mental or physical, so can the inputs! The universe can have information about physical and mental states. So information about mental states might also be used as inputs to generate outputs.
Let’s take the placebo effect as an example. Simply taking a sugar pill is not enough to generate improvement in one’s symptoms. It also requires belief that the pill will help you. So, in order to generate the desired result, the universal function requires an informational input from both physical and mental states. Having only the correct physical state or mental states is not enough. Both must be in the correct state.
Two-way causality accords with our experiences, and contrary to the claims of some philosophers, it doesn’t need to break the laws of physics. The laws of physics can simply be expanded to include mental states as well. Instead of calling them “the laws of physics”, perhaps it would be better to call them “the laws of the universe” to include governance over all kinds of phenomena.
Another benefit of the theory is that it allows for the existence of free will in a rather straightforward way. If mental states are used as inputs into the universal function, then what if some mental states are volitional? If not all mental states are determined by previous states of the universe, it could allow for volitionally-determined mental states. Those volitional states would then be used as an input to generate a particular output.
For example, whether or not you eat dinner at 6pm or 7pm might not be a predetermined fact. The universe could require a volitional state in order to determine which output gets generated. In other words, information about your choice, whether 6pm or 7pm, will determine what happens. Without your choice, you might not have dinner at all.
Now, I don’t currently have an answer to the question of free will, but I think it’s a strong benefit of this theory that it can seamlessly allow for its existence. The mechanics of indirect interaction gives us a concrete mechanism for minds to affect the world, whether that mind is controlling its mental states or is merely a predetermined output of the universal function.
The theory I’ve just explained is extremely flexible. It allows for the existence of arbitrarily many ontological categories. If you think the world is constituted by only physical and mental stuff, it can work. If you think Platonic objects also exist, that’s fine too. If you think there are 100 other categories, all of which interact with each other, that’s fine as well. The ontological categories can be completely separated, so long as there’s a simple fact about them: The universal function has information about their state. That isn’t difficult to imagine, since in this theory, the universal function is the thing outputting the different states into their various ontological categories in the first place!
Indirect interaction also allows for a plausible story of emergence. If might be the case that the universe began with only physical phenomena and laws. Then, over time, as matter rearranged itself, a pattern of information yielded the very first conscious output. If this actually happened, then other types of emergence might also be waiting to come into existence with the correct informational input.
The theory works whether the interaction is causally one-directional or two-directional.
It also allows for the existence of free will.
It’s also consistent with the modern conception of the relationship between body and mind that views the body/brain as fundamental. It might be that the physical state of the brain entirely determines mental states. Mental phenomena can be purely epiphenomenal. Indirect interaction simply gives a causal mechanism for brain to affect mind. So if stimulating one area of the brain causes changes in mental phenomena, it’s not because some particular gland starts secreting consciousness a bit differently. It’s because the underlying physical structure of the brain changes, which changes the information going into the universal function.
This mechanism is even consistent with idealism. Even if one rejects the existence of physical stuff completely, the regularity of mental phenomena still requires explanation. If the laws of the universe govern only mental phenomena, because that’s all that exists, it might be that the underlying mechanics are the same: state + information + laws -> output.
It also explains why the interaction problem has lasted so long. People keep looking in the wrong places. While there’s a tight correlation between brain states and mental states, you’ll never find consciousness within the brain. You’ll only find correlating physical states. The mechanism is not within the skull, because consciousness is simply not a physical phenomenon. You can’t see its generation from the outside. There are no levers, pulleys, glands, or fluids that contain it. That’s because the relationship between brain and mind is abstract. Information about the physical state is not to be found within the physical state.
There are many parts of this theory that one can object to. Perhaps you think the laws of physics aren’t real, for example. Or perhaps you think the continuity of time makes this story less plausible. The details don’t matter. The point is to paint a picture of at least one conceivable mechanism for objects in different ontological categories to effectively interact with each other. If such a picture exists, then the interaction problem is not a refutation of substance dualism or pluralism.