Dualism and Dual Standards

Define physicalism as the thesis that the things described by physics — whatever those may turn out to be — are the only things which ‘really’ exist. Define materialism as the additional thesis that those ‘things’ happen to be the material substances described by Newtonian physics: the idea of the world as a “great machine.” Under this definition, there can hypothetically be a non–materialist physicalism, and I will describe a possible one shortly (though I will assume that idealistic interpretations of quantum physics, even if accurately representing the cutting edge picture of the world revealed by the most modern physics, should not be characterized as ‘physicalist’). But now, consider materialism. What would have to follow about the nature of consciousness, were the materialist thesis as defined here true?

First, let’s be abundantly clear as to what we mean here by “consciousness.” We are not referring merely to a person’s ability to do the kinds of things they are able to do when in a waking state as opposed to a sleeping state—the ability to react to stimuli, give verbal reports, and so forth. Rather, we are referring very specifically to phenomenal conscious experience per se: the ‘redness’ of red, the ‘painfulness’ of pain, the ‘happiness’ of happiness, and so forth, whenever any of these are consciously experienced. By phenomenal, we mean that there is a character to these experiences that simply cannot be explained—when I talk about the ‘painfulness’ of pain, you know what it is that I’m referring to; but only because you have experienced the phenomena yourself. This is simple, but it is the primary data which philosophy of mind is concerned with, and discussions in philosophy of mind routinely get bogged down in confusion over just which sense of the term “consciousness” someone is using at any given time.

Now, if materialism is true, then one thing must necessarily follow: if material events are all that really exist, then either phenomenal conscious experience must literally be a material event, or else it simply cannot exist at all—and as will soon become clear, these options aren’t actually alternatives to each other. But it is indisputably clear that conscious experience does exist. It would be meaningless to try to call it an “illusion,” because this merely leads to a regress: is that “illusion” being experienced, or not? If it is, then it really exists, and it isn’t an illusion; and if it isn’t, then it isn’t an illusion here either, as illusions have at least a certain form of ‘existence.’ Thus, it may make sense to call certain aspects of the way consciousness appears illusory, but it would make no sense to call consciousness itself an illusion: either it exists, or it doesn’t; and as is indisputably evidenced by the fact that it feels like something for you to be sitting where you are, reading this at this very moment, it incontrovertibly does. This paragraph may sound trite, but a small subset of philosophers who recognize the failure of other forms of physicalism yet remain dogmatically attached to the materialist thesis anyway call themselves “eliminative materialists” and happily deny that conscious experience really exists at all—preposterously, they are more convinced of an indirect inference they have consciously made (that is, the one from scientific observations to the real existence of ‘material substance’) than of the very existence of their conscious experience itself.

So, might trying to identify conscious experience with a material event rescue materialism? Not any more than eliminating it could, because this is in effect just what “identification” does: recall that we described the definition of consciousness we are using here by using as example the ‘redness’ of the color red, that is, the ineffable phenomenal experience per se. When we give a description of the process of this perception in strictly material terms, it comes out something like this: lightwaves vibrating at 650 nanometers strike electrical synapses wrapped inside balls of vitreous, are converted into electric impulses which excite neurons in section V4 inside a grey organ inside the skull. Then a reaction is produced: blood flow to the section of the grey organ known as the amygdala increases, and the signal from neurons in V4 travels to the nervous system, causing the body to move.

Is there a problem with this description? Obviously, it totally omits any mention of experience. Not only is there no possible way that a ‘material’ account can hint that the excited neurons in section V4 of the brain are perceived as the redness of red in particular — it can’t even hint that any experience accompanies those neurons at all. There is no hint here — and on materialist terms, no possible way to hint — that the increase of blood flow to the amygdala feels like what we know as fear; or that it feels like anything at all — or that the transfer of the signal from photoreceptor neurons to the nervous system feels like the intention to move (or anything at all), and so on. But if two things are identical, then a full description of each should reveal that identity: if it turns out that my next door neighbor Jim and the robber on fourth street happen to be the same person, then a full description of both should make this obvious. Yet there is no way in principle that experience–per–se can ever be put in ‘material’ terms, and no way in principle that any material description of anything can ever hint at whether or not there is any experience accompanying that process at all (nevermind its particular character and quality). So “identifying” conscious experience with something other than conscious experience necessarily can only ‘work’ by wholly eliminating it.

Another way of talking about identism is with the term ‘supervenience.’ But it is clear that this term introduces nothing new to the conversation thus far: it collapses into either identism or eliminativism. Consider the analogy that is usually made—one example of a ‘supervenient’ relationship in nature is in the relationship between ‘wetness,’ a property of water, and the properties of molecular H2O behavior: wetness is said to ‘supervene’ on this micromolecular activity. What does this really mean? In essence, what it actually means is that ‘wetness’ is a shorthand term we use to describe the behavior of H2O molecules when a detailed description of the latter is unnecessary (i.e., all practical everyday intents and purposes). So the disanalogy in the relationship between wetness and H2O molecules on the one hand, and conscious experience and the brain on the other is actually instructive as to why identism fails: when we understand the behavior of H2O molecules, there is nothing left about ‘wetness’ to understand! Yet there is clearly something crucial left to understand about consciousness even after we have explained the physical processes of the brain.

Furthermore, the phenomena we refer to as ‘wetness’—precisely because it is identical with the micromolecular behavior of H2O—simply follows logically from that very behavior: there is no way that something could exhibit that behavior at the microphysical behavior, and not be “wet.” But something could easily perform the physical processes of the human brain, and not be conscious. When we understand how molecules of H2O roll across each other because of their loose connections, and when we understand how another object composed of more tightly connected molecules would fall through the gaps between those loosely collected molecular rings of H2O (that is, sink), then “wetness” follows as a straightforward logical consequence and there is nothing left about it to understand. But when we understand how neurons connect and relate in material terms, however, we do not understand conscious experience; it does not follow as a straightforward logical consequence. So the relationship is not one of ‘supervenience’ and consciousness is not identical with any material substance. So, the materialist version of physicalism (as I have defined these terms) necessarily fails; it doesn’t even exit the gate to become a live option, as it could only work by either eliminating conscious experience or ‘identifying’ it with something that is not conscious experience (which amounts to eliminating it). With materialism done away with, physicalism isn’t necessarily dead. There is at least one form of physicalism which is not materialist: nonreductive emergentism. I will assume here that it is the only form of ‘physicalism’ worthy of the name and thus conclude that between eliminativism, identism, and emergentism, ‘physicalism’ is exhausted.

Nonreductive emergentism admits that consciousness is a real phenomena in its own right, and contends that it is in some sense ‘produced’ by the physical processes of the brain. The first problem is already clear: how? The interaction problem — “how is it that mind and brain, conceived as categorically different types of things, interact?” — is often taken to be a conclusive argument against accepting substance dualism. But it seems that any nonreductive physicalism must necessarily put itself in precisely the same boat, unable to give any explanation of how it is brains produce minds — which is, clearly, a kind of interaction. Furthermore, a nonreductive emergent view could either proclaim consciousness an epiphenomena, or grant that conscious experience has genuine causal efficacy, and either way, there is a definitive problem. Arguably, emergent physicalism has to proclaim consciousness an epiphenomena, since its claim by definition is that consciousness is produced by physical processes. If so, the problem is that this qualifies as a modus tollens conclusively refuting emergentism, because epiphenomenalism is demonstrably, irredeemably false.

The contention of epiphenomenalism is that conscious experience has no causal efficacy of any kind. That is, the relationship of the physical processes of your brain to your conscious experiences is like the relationship of your actual body to your reflection in a mirror: your body at time slice (1) makes movement (a), which causes mirror movement (A) in the reflection. At time slice (2), your bodily movement (a) continues into movement (b), which causes mirror movement (B) in the reflection. The actual causality, on this view, is running from (a) to both (A) and (b), and then from (b) to both (B) and (c). If it appears that (A) causes (B), then this is only an illusion, and (A) certainly has no causal impact on (b). Alternatively, Thomas Huxley in the nineteenth century famously described it by stating that mental events are like “the steam whistle which signals but doesn't cause the starting of the locomotive.” Now, philosophies should not be dismissed out of hand simply because they seem unintuitive. But there is a very simple reason why epiphenomenalism is in a pathetic degree of error about the nature of consciousness: we are talking about it right now. You are reading these words, which refer to conscious experience, and you understand what they mean. If conscious experience per se were purely a byproduct of physical processes with no causal efficacy of its own, this would simply not occur, because the supposedly ‘physical’ processes producing your supposedly epiphenomenal conscious experience of thought would have no concept of conscious experience in order to be able to think about it as you are at this very moment: conscious experience would not exist in the causal order such that your physical brain could ever produce any thoughts directly about it. Yet, we demonstrably can communicate ideas about conscious experience per se through (physical) speech—and on that one consideration alone, epiphenomenalism is conclusively refuted.

So might nonreductive emergentism contain room to acknowledge the reality of causally efficacious experience? Probably not. Emergentism is a very similar notion to the idea of supervenience which we discussed and dismissed above: all the natural examples involve cases where the supposedly ‘emergent’ (or supervenient) phenomena are not truly new phenomena over and above that which they ‘emerge’ from (or supervene upon). But suppose for the sake of argument that it truly could be possible: how, then, would this option be any better than substance dualism at all?!

Not only are we left with no possible way to explain how conscious experience emerges from material substance, we are also left with an exact analogue of the interaction problem again: how does conscious experience effect those material substances? In fact, nonreductive physicalism faces an even worse version of the problem than dualism does, and it very likely faces it in a way that is—unlike for dualism—insoluble because it is simply flat out conceptually incoherent: how does conscious experience exert causal power over the very microphysical processes which it emerges from? At worst, this is equally as incoherent as supposing that the phenomena of ‘wetness’ could somehow exert causal power over the very molecules of H2O which it supervenes upon. At best, this leaves nonepiphenomenal emergentism facing the same problem which supposedly refute dualism and left physicalism as a given by default … even more severely than dualism itself. So if the interaction problem was our motivation for abandoning dualism, then it should be a far greater motivation for abandoning emergentism.

But there is a second, and largely epistemological, problem with both classes of emergentism too—take my definition of it: “consciousness is a real phenomena in its own right … that it is in some sense ‘produced’ by the physical processes of the brain….” Dualism admits that consciousness is a real phenomena. So what does the step towards emergentism do, besides add the additional premise that consciousness must be produced by physical processes in the brain? Nothing—it does nothing else at all. But how could this step be justified? What kind of observation could possibly establish once and for all that consciousness is ‘produced’ in this way (besides, perhaps, the cessation of consciousness at death?) Observing that reported states of consciousness consistently correlate with particular physical brain states could just never accomplish this in principle, since the mere observation of correlation is (again!) not proof of any particular direction of causation. So could anything possibly falsify emergentism? It seems not. There is a further, important confounding variable: we never observe others’ states of consciousness directly. Rather, we are relying on verbal reports—and those reports necessarily have to be spoken, by mouths attached to brains. Thus, relying on such verbal reports to try to determine whether consciousness can exist without the brain places us in a position like that of primitives who have discovered a television set and are attempting to determine whether the images flashing across its screen are produced by it, or whether the television is merely transmitting a signal of some sort from elsewhere: what kind of test could we possibly devise; what kind of empirical finding could possibly establish once and for all that the television is indeed producing the images individually? Observing that the images stop when a button is pressed would clearly not suffice. So even if the problems above did not conclusively refute emergentism, is there anything that could possibly warrant our belief in it?

And thus, the “physicalist” options are exhausted. Eliminativism is preposterous. Identicism is incoherent. Epiphenomenalism cannot account for the fact that we are discussing consciousness here at all. And emergentism, the most plausible option of the bunch, not only faces exactly the same problem that dualism faces—twice, and even more severely both times (on the upwards direction, we have not only the problematic question of how material processes could provide inputs into conscious experience but the further problematic question of how material processes could produce consciousness itself; in the downwards direction, we have not only the problematic question of how consciousness can influence material processes but the further problematic question of how it could influence the very same material processes which it is emerging from). Perhaps barring idealistic interpretations of quantum physics which I have excluded by definition, there are no ‘physicalist’ options left.

Now that the options it is being compared to are clearer, let us revisit the interaction problem for substance dualism. Just how destructive is this problem for dualism by comparison? We have outlined how emergentism faces exact analogues of the interaction problem which are even more severe than those facing dualism, and summarized those in the preceeding paragraph. But on emergent physicalism, it is clear that there must be intermediate mechanisms between microphysical processes and consciousness: there is clearly something to explain about how the microphysical could ‘produce’ conscious experience. But the same is not obvious on dualism! Why is this? Because it is not clear that, on dualism, there would be any intermediate mechanism in need of explanation.

Consider: we can ask why it is that turning the ignition on a car causes the engine to start, because there is something in between the two that can be given an explanation. But it would make no sense at all to ask how pressing the gas pedal causes it to move, since this is so basic that there is nothing inbetween the two that could possibly be elaborated on. On dualism, the interaction between mind and brain is plausibly more like the relationship between pressing the gas pedal and causing it to move than it is like turning the keys and wondering why it should follow that the engine then starts. But even supposing there were intermediate mechanisms mediating the dualistic interaction, it would still follow that the burden of explaining them would not be so heavy for dualism: for, on dualism, it probably follows as a consequence that we should have no way of understanding them! To understand this point, consider a case in which a sort of dualism would hold: suppose we are all, at this very moment, plugged up into virtual reality machines.

Our memories of our outside selves have been erased, so that we assume we are identical to our virtual avatars. In this world, there is a dualism between the real self and the virtual self. From within that virtual world, we would never — in principle — be able to locate or identify our ‘real self’ (except at best perhaps by inner reflection). No matter how long we probed our virtual brains, we would never locate our actual real minds inside of them (since our real minds would not, in fact, be inside of them). And for just the same reason, it should follow that we would never — from the limited perspective we are capable of within that virtual world — know anything about whatever mechanisms intermediate between our real, and virtual, selves to conduct the interaction. Therefore, on dualism, either there are no intermediate mechanisms and thus nothing about interaction in need of explanation, or else there are intermediate mechanisms which dualism itself predicts we should have no capacity to understand. By contrast, emergent physicalism (the only kind of physicalism potentially capable of avoiding the eliminative–epiphenomenal trap) in the best case faces that problem twice as many times as dualism would have, were it truly even a serious difficulty for dualism at all.Dualism therefore remains a live option which is neither trivially nor obviously false. As the renouned materialist philosopher William Lycan, admits in Giving Dualism its Due, “the standard objections to dualism are not very convincing; if one really manages to be a dualist in the first place, one should not be much impressed by them.”

He continues, “Being a philosopher, of course I would like to think that my stance is rational, held not just instinctively and scientistically and in the mainstream but because the arguments do indeed favor materialism over dualism. But I do not think that, though I used to. My position may be rational, broadly speaking, but not because the arguments favor it … My purpose in this paper is to hold my own feet to the fire and admit that I do not proportion my belief to the evidence.” So, in conclusion: depending on whether an emergent physicalism which does not imply epiphenomenalism is possible (which in turn depends on whether it is even conceptually coherent to suppose an ‘emergent’ phenomena could have causal efficacy over the very processes it ‘emerges’ from), materialist physicalism is at worst not even a live option, and at best an even larger ‘leap of faith’ than dualism.

“Saying ‘science may someday find a way to reduce consciousness (or reference, or whatever) to physics’ is, here and now, saying that science may someday do we–know–not–what, we–know–not–how. And from the fact that those words may in the future come to have a sense we will understand, … (it does not follow that they mean anything now)….” (Putnam, 1999, p. 173).