Materialist Conceptions of Mind, Part 1: Identificationism

In the next few posts I want to (a) consider what I take to be the main ways that a materialist can account for consciousness, and (b) try to identify the main hurdles that these accounts must overcome in order to be generally convincing.


For my purposes here a “materialist” will refer to someone who holds that phenomena in general and mental phenomena in particular (that is, conscious states and subjective experiences such as thoughts, sensations, emotions, and desires) are (i) purely material or physical, and (ii) wholly and adequately explained by the sorts of physical mechanisms and processes that, in the sciences, are routinely posited as explanations for empirical phenomena (even if those explanations might not be accessible to us right now). By the “material” or “physical” I mean, at the most basic level, entities characterizable in spatio-temporal terms that interact and affect one another in accord with causal laws whose consistent operation can be tested through scientific means—in short, matter and energy.

So, (i) is an ontological thesis about what mental phenomena are. To say that they are purely physical is not, however, to say that they must be entities composed of matter an energy, like tomatoes and rocks. While that is one way in which something can be purely physical, there are other ways. Mental phenomena might, for example, be physical process, or a certain class of properties possessed by these processes, or the way that certain physical processes “look” from a peculiar perspective.

The related thesis, (ii), is more of an epistemological one about what methods we should employ for coming to an adequate understanding of mental phenomena. The methods are scientific. That is, we will come to adequately understand mental phenomena in just the way that we have come to understand such phenomena as illness, lightning, or the cycle of the seasons: by uncovering the physical building blocks, structures, and mechanisms that (in accord with physical laws) explain the phenomena in question.

The materialist, thus understood, has (I think) three main options for dealing with consciousness. These options are:

1) “Identificationism,” by which I mean a direct identification of mental phenomena with brain processes—what we call mental phenomena just are brain events in much the way that water is H20. (Note: while some equate this view with what is called “eliminative materialism,” I reserve “eliminative materialism” for the more radical view, seemingly embraced by Feyerabend and a few others, that mental phenomena don’t really exist at all).

2) “Emergentism,” by which I mean the view (most famously advocated by Searle) that consciousness is an emergent property of an active brain. What we know as our consciousness is a feature of the brain in something like the way that brittleness is a feature of glass—it is a property of the whole that arises by virtue of the activity of the constituent elements organized in the distinctive way that they are, and can thus be adequately explained by reference to those elements.

3) “Perspectivalism,” by which I mean the view that consciousness is the way that the complex processes of the brain “look” from a distinctive perspective—specifically, from the “internal” perspective that arises when the brain itself engages in self-monitoring/self-representing activity. This is the kind of view advocated by Douglas Hofstadter in I Am a Strange Loop.

In this first post, I would like to look at the first of these three options to explain why materialists would do well to look elsewhere. The view gets what plausibility it has from the growing body of neuroscientific research that has succeeded in correlating specific “inner” experiences (the sensation of bright light, pain, etc.) with (partially understood/described) physical processes in the brain. The idea here is to simply identify the former with the latter—to say, in effect, that what neuroscientists are slowly describing just is the conscious experience with which it has been correlated.

There are, in effect, two substantial difficulties with making this kind of direct identification. The first has to do with the failure of brain states and mental states to meet ordinary standards of identification. The second has to do with the prospects for “multiple realization” of mental phenomena.

Let me begin with the ordinary standards of identification. In any usual understanding of identity, in order for A and B to be identical, what is true of A must be true of B and vice versa. If A has properties that B lacks (or vice versa), then the relation between A and B must be more complex than one of simple identity (perhaps A names X-as-experienced-from-perspective-1 while B names X-as-experienced-from-perspective-2).

But mental phenomena—thoughts, sensations, desires, emotions, etc.—clearly have properties that the corresponding brain processes lack, and vice versa. To see this better, it is helpful to introduce the notion of “qualia” (“quale” in the singular), by which is meant the subjective contents of experience, or, perhaps, the “way it feels” or “what it’s like” to undergo this or that mental experience (for example, what the peculiarly bitter sunflower seed I just ate tasted like as I chewed it up). This “what-it’s-like-ness” of mental phenomena isn’t some incidental feature of these phenomena, but rather their chief constitutive element. Put another way, what I mean by the conscious experience of pain just is the associated quale.

And, as Thomas Nagel and Frank Jackson have both famously argued in different ways, these qualia are inaccessible to scientific study in a way that the details of brain processes are not. Nagel, in “What Is It Like to Be a Bat?”, points out that we know a great deal about the neurophysiology of bats, and we know a great deal about how their acoustic radar works. But none of this tells us what it’s like to experience the world through such a radar in the way a bat does. The qualia of a bat’s inner life seem to be inaccessible to science in a way that the brain processes of the bat are not.

Jackson offers a different thought experiment, in terms of a hypothetical scientist who knows everything there is to know, scientifically, about the physiology of color perception, including an exhaustive understanding of the brain processes that correspond with the experience of color…but she has never herself experienced color. And then, one day, she sees a red rose for the first time. Jackson asks: has she learned something new? He thinks the answer is obviously “yes,” which means that there is something about the experience of color that is “above and beyond” the neurological processes science can study.

Here’s a way to put their point: brain processes are third-party accessible objects of study. But there is something related to these brain processes that is inaccessible to such third party study, being instead only available through first-person introspection. This something else is what has come to be called the “qualia” of consciousness. But the fact is that we might as well say that it is consciousness itself—because what we are referring to when we speak of our conscious life just ARE these qualia.

These facts make a simple identification between mental phenomena and brain states at best highly problematic. Mental phenomena are something that each of us has immediate first-person familiarity with; but these phenomena seem inaccessible to neuroscientists engaged in third-person investigation of our brains. And we can claim familiarity with what this or that conscious experience is like—and so honestly claim familiarity with mental states—even if we know absolutely nothing about the brain states with which they correspond. As such, it seems as if mental states have relational properties which brain states lack, and vice versa.

Some defenders of identificationism try to escape this conclusion by pointing to the fact that we can identify water with H2O even though, for a long time, people were entirely unfamiliar with the chemical composition of water (and many remain so). Doesn’t that show that someone’s being familiar/acquainted with A but not with B doesn’t preclude the identity of A and B?

The problem with this rejoinder is that it fails to make a crucial distinction. “Water” denotes a certain entity out there in the world—one we are acquainted with primarily through a range of properties such as liquidity, clarity, capacity to quench thirst, etc. “H2O” denotes the same entity out there in the world understood from a scientific perspective in terms of chemical composition. It turns out that the entity characterized by the former set of properties is the same one that has the given chemical composition.

That is, water and H2O are identical because they refer to the same thing. What is referenced by "water" is not a set of properties but a certain kind of thing (a "natural kind," some might say) that exists in the world but which we are acquainted with through the given set of properties. What we are referring to when we use "water" is the underlying thing which has the properties. But when we speak of pain, we are referring to the quale itself--the "what it feels like" of pain, as opposed to any underlying physical mechanism which might or might not be able to explain the emergence of this sensation.

To better see the problem here, let's consider a different example: the statement "Clark Kent is Superman." Now, consider two possible ways to look at this statement, related to two distinct ways of understanding “Clark Kent” and “Superman.” In the first case, the statement comes out true while in the second it comes out false (at least if we set aside for the moment the unique challenges associated with fictional individuals).

Consider Case 1. In this case, by “Superman” we mean “the individual who embodies a range of traits, including super strength, ability to fly, a do-gooder personality, etc.”; and by “Clark Kent” we mean “the individual who embodies a range of traits, including being mild-mannered, dorky, a reporter for the Daily Planet, etc.”. In this case the identification holds because the individual who possesses the former traits is the same individual who possesses the latter. It isn’t that Superman and Clark Kent possess different properties not possessed by the other. Rather, it is that we know the individual under one name in terms of some of that individual’s properties, and we know the individual under the other name in terms of a different set of that same individual’s properties. But given that the two names refer to the same individual, every property possessed by Clark Kent is also a property possessed by Superman, and vice versa. Hence, a simple identification is possible.

Now consider Case 2. In this case, by “Clark Kent” we mean “the identity taken on by Kryptonian Kal-El in which he masquerades as a normal Earthling”; and by “Superman” we mean “the identity taken on by Kal-El in which he exercises his unique powers in public view.” In other words, under this usage each term refers, not to an individual person, but to an “identity” or “persona” that an individual person has adopted. Well, given THOSE meanings of the respective terms, it would be a mistake to say that Clark Kent is Superman. After all, Kal-El’s “secret identity” is very different from his “superhero identity”, and so they cannot be the same identity—even if it is the same individual who adopts them both.

But given the way in which we use the language of mental states and the way that we use brain-state terms, the two are related in something more like Case 2 than Case 1. And this means that a simple identification of the kind “Clark Kent is Superman” is false, even if it might be true that there is a common underlying entity which is the bearer of both the “mental state” and “brain state” identities. This means that the materialist is better served by the materialist version of what has come to be called “property dualism”: the view that brain states have “mental properties” that are distinct from the physical properties studied by scientists. And while there are different ways to explicate this sort of property dualism, the one that strikes me as being the most plausible is emergentism—the idea (energetically defended by John Searle) that consciousness is an emergent property that supervenes on an active neurological system like the brain taken as a whole.

More quickly and briefly, let me stress one additional difficulty with identificationism. The problem lies in the fact that, by all appearances, on the assumption of materialism we should suppose consciousness admits of “multiple realizations.” The idea of multiple realizations is this: the same property or process could be produced by very different physical building blocks and arrangements. If several engineering firms are given the same task—to design a device that dispenses Starbucks-quality lattes in response to the voice command, “Overpriced coffee! Now!”—they might realize their objective in very different ways. The materials might be different. How those materials are physically arranged might be different. The internal processes of the resultant machine might be quite different in many ways—and yet, both machines might have the very same “emergent” property of dispensing Starbucks-quality lattes in response to appropriate voice commands. This property thus cannot be identical with any particular set of physical components organized and operating in a particular way, because the same property can supervene on very different physical building-blocks.

Now I suppose it is possible that only organic materials exactly like those in terrestrial brains, organized in only the ways that more highly developed terrestrial brains are organized, and operating only in terms of those brains’ distinctive “programming parameters,” can give rise to what we know as consciousness. But this seems implausible to me. Maybe that’s just because I watched one too many episodes of Star Trek--but I think it's more than that. Among other things, rejecting multiple realizations makes the emergence of consciousness through natural selection seem to be an even bigger miracle than it already is (something I doubt a materialist would find congenial).

So, I’m inclined to think that if consciousness is the kind of thing that can emerge from physical systems at all, it will be the kind of thing that will admit of multiple realizations. If so, then one kind of mental phenomenon (say, the experience of pain) cannot be identified with the kind of neurological activity it happens to be correlated with in human brains. And so, if you are a materialist, you will need a different account than identificationism. And since the contender which seems to arise most naturally out of the ashes of identificationism is emergentism, this is what I will take up in my next post.