Interacting With the Interaction Problem

Or: What To Expect When You're Introspecting

Pictured: Me reassuring physicalists

Earlier this week, noted Youtuber, podcaster, and undercover Mormon apologist Emerson Green posted a new video with Ralph Stefan Weir titled “The Myth of the Interaction Problem,” where they looked over a few classic objections to interactionist dualism and discuss how they (supposedly) go astray. Emerson is one of my favorite people in the online philosophy space, so I was excited to check it out. Overall, I thought the exchange was really interesting, and I especially enjoyed their historical look at how the concept of causal closure and the arguments related to it have been reformulated since the days of Descartes. But I was also admittedly ENRAGED intrigued by the pervasive sense of incredulity they expressed towards advocates of physicalism and proponents of interaction-based arguments for it. Now, to be clear, this isn’t a criticism of Emerson and Ralph Weir specifically - in my experience, most substance dualists are legitimately, earnestly baffled when it comes to why anyone would prefer a physicalist theory of mind (and especially my particular brand of hardcore reductionism). So I thought it would be a good time to lay out a quick worldview-level case for why I personally find the idea of non-physical causation to be so unappealing, and why the alternative account of consciousness isn’t as crazy as some of its opponents seem to think.

I’ll start with a point of agreement: I don’t think it’s a good idea for physicalists to just assert, a priori, that different sorts of substances can never interact. That’s a fine intuition to have - it certainly strikes me as the right default assumption - but actually showing there’s any logical or metaphysical contradiction involved is a lot tougher, and we should at least be open to considering that our commonsense model of how a mental substance would work is misguided. After all, as physicalists, we’re already committed to thinking physical substances can do some pretty crazy stuff you wouldn’t expect, like constituting consciousness. So no matter which “side” you take, there’s going to be some point down the line where you’ll have to dismiss an intuition. For this reason, I think it’s probably best for everyone to call a truce, put aside the sweeping a priori statements, and start looking at things from an evidential perspective.

But what sort of evidence would even be relevant here, and how does it shake out? That’s where the disagreement starts. I think everyone will accept that, short of computationally modeling the entire universe at the level of quarks and gluons, there’s no possible method for experimentally confirming the causal closure of the entire physical universe all in one go. At the same time, I think everyone will also accept that, apart from consciousness, there’s not much debate about the isolated causal closure of any other major natural phenomena like digestion or respiration. So the question really comes down to how you weigh the uniqueness of consciousness on one hand, versus the widespread absence of non-physical influence in other areas of the natural world and the general success of reductive physicalist explanations on the other.

In other words, if there’s one single question to ask, it’s this: Are the features of consciousness so radically distinct that they justify rejecting the inductive assumption of closure that we generally apply to other features of the natural world that we may not fully understand? To explain why I answer this question by coming down hard on the side of causal closure “across the board,” I want to focus on three basic claims:

1. Consciousness is a rare, localized natural phenomenon that only arises under extremely specific conditions.

2. Previous historical claims about the necessity of some non-physical cause for natural phenomena have universally turned out to be false.

3. Evolutionary biology gives us good reason to believe our introspective abilities would not present the metaphysical status of our own cognition accurately.

My contention is that anyone who accepts (or is even somewhat open to) these three claims should reject the theory of interactionist dualism and embrace physicalism, at least until further scientific and philosophical exploration reveals decisive reasons to abandon it. Let’s cover each one in order.

The Place of Consciousness in Our Picture of Reality

Consciousness matters a lot to human beings. Arguably, it’s the thing that matters to us most, at least in terms of how central it is to living a life of value. So it makes sense that human beings might be tempted to put consciousness in a privileged position when we build our ontologies. But a very different story comes out when you step back and look at things from a cosmic perspective. As best we can tell, consciousness is something that only arises in exactly one sort of biological structure found inside a small number of organisms on one particular planet out of untold trillions and trillions. (Even if you assume, as I do, that consciousness exists elsewhere in the universe, there’s still an unfathomably large ratio between planets with consciousness and planets without.) In other words, it’s much, much rarer than things like rain, tides, flight, or diamonds; vast stretches of space, light-years across, will never, ever see it, and in terms of the grand timeline of all existence, the window where it will be found anywhere at all is infinitesimally narrow.

I’m not bringing this up just to be a bummer. I’m bringing it up because it’s easy to forget that our conventional divide between mental and physical isn’t the clean split down reality’s middle that dualists often make it out to be. Instead, it’s a dramatically lopsided distribution, with one substance comprising 99.999999999999% of everything and another plugging the one particular, contingent, human-centric hole that’s left. Without some other extravagant commitments, consciousness considered from the outside has none of the features that we would expect from a fundamental facet of reality: It isn’t widespread, or resilient, or present at all times, and it seems to rely on complex arrangements of other, more basic constituents to even arise in the first place. There’s nothing a priori impossible about a truly irreducible primary substance having (or appearing to have) these features, but it still does seem suspicious, and our presumption against it should be strong - just like it would be if someone invoked a novel non-physical substance to explain the process of protein folding, or the photic reflex that makes some people sneeze when they look at bright lights.

To be clear, this little preamble isn’t a point against interactionism in particular. But it’s absolutely something that should structure the prior probability we assign to dualism more broadly; locking down an explanation for one fragile, sporadic, and high-level phenomenon at the cost of doubling our most basic ontological commitments just seems like an obviously bad deal even before you get to looking at specifics. The overall calculus would shift, of course, if we were constantly coming up against natural phenomena we couldn’t explain in physical terms, or if the evidence for the irreducibility of conscious states specifically was so overwhelming it forced our theoretical hand. Unfortunately, in both cases, that’s not how things turn out.

The Success of Reductionism and the Failure of Intuition

If you take a broad look at history, it’s pretty clear that human beings are not physicalists by default. In fact, pretty much every human culture to ever exist - from hunter-gatherer bands in the neolithic up on through to the very recent past - has relied on an eclectic mix of curses, karma, divine favor, élan vital, intelligent design, and more to explain how the world around them works. But since the dawn of the scientific age and the advent of methodological naturalism, we’ve been able to learn beyond a shadow of a doubt that human beings were largely wrong about all these things. I know our collective trauma from the New Atheist era makes it feel tacky to drive this point home, but it really does matter: If you line up every single non-physical causal explanation for observed natural phenomena except for substance dualism and then go one-by-one asking whether those explanations turned out to be good ones, the answer is universally no.

At the risk of overusing some italics, I’m going to repeat that: There are, so far as I can tell, literally no other instances at all of an explicitly non-physical explanation for a particular natural phenomenon triumphing over a competing reductive physicalist account. Of course, there might be some wiggle room depending on how you define “natural phenomenon.” If you broaden the term so dramatically that it includes the very existence of anything at all, then I guess you could say creation ex nihilo by God is a theory that’s still out there (although, of course, I still think it’s a terrible one). But if we’re talking about the sort of stuff we see rivers and mountains and plants and animals doing - the category consciousness should broadly be placed in - then I really do think it’s a solid 0% success rate for any other non-physicalist intuitions.¹

And it’s not just that human beings are really, really bad at getting suckered by these intuitions. It’s also that they’re really, really bad at giving them up when challenged. Pick any candidate from the long list of non-physical frameworks supplanted by physical ones and you’ll find devotees earnestly insisting the reductive physicalist explanation just couldn’t possibly be true, almost always on the basis of pure incredulity. The later vitalists are a great example of this dynamic; even as developments in modern biology began to decisively undercut their claims, many still spent decades stubbornly insisting there just had to be something about life that mechanistic theories were missing because, you know, just look at living things and non-living things and how obviously different they are! But of course they were wrong, and eventually their “hard problem of vitality” was chiseled down to nothing. Even if you don’t think substance dualism will face the same ultimate fate, being intellectually honest still requires admitting we’ve got a truly terrible track record when it comes to certainty on this stuff.

Now look, I get it: Any time a dualist hears someone start talking about What The Science Says, they start rolling their eyes - and sometimes for good reason! So let me be clear about what my argument isn’t. I’m not saying the neurological correlates discovered by modern science are incompatible with dualism, or that they decisively prove all causation is physical, or that we’re allowed to dismiss the concept of non-physical causation forever just because some people were wrong about whether demons made it rain. None of this is a deductive, 100% certainty sort of thing. Rather, it’s about building an inductive case for skepticism on the basis of sociohistorically informed priors. So it’s not, as Emerson suggests in his video, like checking the first few train cars in a railyard, finding no one inside, and deciding that means all the rest must be empty too. Instead, it’s more like finding hundreds of train cars empty, and realizing the hunch you had about the last couple dozen kept on steering you wrong.

Does that absolutely guarantee the last locked car is empty too? Of course not. But it definitely makes that a great working theory, and you’d be well within your rights to demand some seriously novel evidence before you take another look. It’s the same for me with physicalism: Before we get to considering the specific evidence involved, we’re totally justified in taking causal closure as the default assumption on the basis of its wildly successful historical application and the total lack of identified (or even suspected) non-physical causation popping up anywhere else. In fact, Emerson’s guest pretty much admits this about sixteen minutes in, when he says we’d have good inductive reason to reject interactionism in the absence of our first-person conscious experience. So let’s turn to that first-person data now and ask whether it’s really powerful enough to overcome such a dismal prior probability. (I think you can guess what my answer will be.)

What Introspection Does and Doesn’t Get You

I want to start off this section by saying I’m a huge fan of introspection. Can’t get enough of it. I love being able to thoughtfully consider my own desires, values, and beliefs. I love reflecting critically on the things I perceive and how they impact my behavior. I love the process of rational deliberation and the experience of developing second-level perspectives on my own psychological states, all with the goal of better understanding my own mind and directing it towards productive ends. Introspection is great for doing all this stuff! And that makes sense, because these are exactly the sorts of things we need to do if we want to survive and flourish. Our introspective abilities are perfectly tuned to help us navigate our complex mental lives as they interact with a complex external world and for that, we should be grateful.

But, with that said: One thing introspection is very clearly not tuned towards is giving us veridical reports of the metaphysical nature of what it is we’re introspecting! And why would it? Before those pesky philosophers popped up around 500 BC, no living creature in existence had any reason whatsoever to care about whether their mental states were constituted by physical or non-physical substances. That sort of knowledge doesn’t help you find food, or stay warm, or produce more offspring, and it wouldn’t even make the first-level process of introspection any more productive. In fact, viewing one’s own mental life in purely reductive physicalist terms, rather than through the lens of our intuitive folk-psychological categories, would almost certainly be a downside evolutionarily. It’s precisely the fact that we can sort through an unbelievably complex and interconnected jumble of neurobiological activity by engaging with simplified, essentialized mental placeholders instead that makes introspection such a powerful tool in the first place. So the idea that “looking inward” at our own conscious experience would necessarily give us a correct understanding of the substance that grounds it isn’t just unwarranted - it’s actively implausible.

And we don’t have to rely on evolutionary just-so stories to make this point. We can also look at how we, as humans, design interactive systems. The computer you’re reading this on right now is nothing but a unfathomably complex system of electric signals moving through silicon and copper. But is that what the user interface has been programmed to communicate? Of course not - if that was the designer’s approach, it would be unusable for any of the tasks it’s meant to help with. Instead, the system has been designed to take that incomprehensible mass of physical data, represent it as a series of simplified, essentialized objects, and then present those objects in a way that’s maximally productive for the first-level tasks valued by the user but totally misleading in terms of capturing the “true nature” of the computation. A prescientific time traveler with no knowledge of computer science could scroll through your phone or work at your desktop for a hundred years and never come to even imagine it was anything other than a box that was really full of folders and documents and little buttons to press; if you finally popped off the case and showed them it was just a bunch of green wafers, they’d be absolutely incredulous, in the same way substance dualists are now when you suggest a similar story about minds and brains.

Of course, no one intentionally designed human brains the way we’ve intentionally designed computers. But whether artificial or natural, systems that require the ability to sort through and engage with huge amounts of physical data are generally incentivized to obscure that data’s true nature in favor of simplified representations that allow for easy manipulation. So while we should absolutely value our evolved introspective capabilities for all the good they do when it comes to effectively engaging with our own thoughts and feelings, we have strong reasons for believing that same process of introspection will be systematically unreliable when it comes to answering deeper metaphysical questions about their ontological nature. And to be very clear, nothing about this is a denial of consciousness itself, or a denial of any intuitive appearance it might have. Personally, I do think it’s pretty clear there are certain culturally bound, contingent linguistic and philosophical traps that lead people to weigh down their intuitive sense of the mind with more metaphysical commitments than it really needs to have, and there’s some good evidence that cultures without exposure to classic Cartesian memes conceptualize their own mental lives in very different ways. But even accepting those complications, it’s perfectly reasonable to say that consciousness does inherently represent our thoughts and feelings has having properties that physicalism can’t account for - something dualists are very, very happy to remind us of.

The problem for dualists is that physicalism doesn’t need to account for those properties’ actual existence. It just needs to account for how it is they appear to exist, and there’s no reason whatsoever at this stage of investigation to assume that’s a lost cause. So when dualists talk about the evidence provided by first-person conscious experience, it just falls totally flat for me. What we actually have evidence of is how consciousness appears to us, which physicalism doesn’t deny - what we really need to salvage dualism would be evidence those appearances accurately represent what’s really going on. And so long as we have no positive reason to expect our introspection to be veridical in this way, as well as some reason to think it wouldn’t, then it’s hard to see how dualists could even start to challenge the presumption of physicalism given by my first two points.

Case (Causally) Closed

Putting these three points together isn’t hard: Consciousness is an obscure and poorly understood natural phenomenon that we shouldn’t expect to have reliable insight into, and we’ve got a a persistent habit of writing off reductive explanations for exactly that sort of thing before being decisively proven wrong. For me, that’s enough to make non-physical causal explanations a total nonstarter. Dualists often frame physicalists as overly confident, or even arrogant, when we assume that the natural sciences will one day provide a reductive account of our entire mental lives. But ironically, I see physicalism as a profoundly humble and conservative approach, one that recognizes our tendency to be seduced by easy answers and asks instead that we stick with what works until it’s absolutely necessary to abandon it. Maybe, in ten or fifty or five hundred years, we’ll have a perfectly detailed model of every little synapse and still find ourselves coming up short. Then I’d be happy to give dualism another look. But until then, I’m happy to bet on the historical success of reductive explanations and the historical failure of anything else, even if it means acknowledging that my own mind might not be an expert on its own fundamental nature - as much as I love the guy, I just don’t think that’s his area of expertise.

Obviously, I don’t expect this piece by itself to convince anyone firmly entrenched on the other side. But I do hope it can at least help explain to a dualist why it is that physicalists are so comfortable with a position that might strike others as crazy. There are certainly bad arguments in both camps - naive scientism for me, naked appeals to incredulity for others - but ultimately it all comes down to fundamental differences in how we weight our intuitions against external confirmation and what we take to even count as meaningful evidence in the first place; if I believed everything Emerson believes about the nature of consciousness and its place in the universe, then embracing interactionism would be no problem at all, and if dualists had the same worldview-level commitments I do, they wouldn’t be dualists. For this reason, I’m pretty sure the debate will continue on as usual for quite some time, until actual achievements in neuroscience shrink the sphere of mystery tight enough to squeeze most dualists out. But until then, while it’s still a matter of dueling theoretical frameworks, I hope I’ve at least made a reasonable case.

1. This is true, by the way, whether you think about things in terms of causal closure or in terms of the conservation of energy. I don’t want to get bogged down in the technical details here, but suffice to say - again, putting aside consciousness - there’s no example in the whole of existence we can point to where additional energy enters a closed physical system from an immaterial source.