Consciousness, Atomism, and the Ancient Greeks

Atomism about matter was not always the entrenched and incontestable theory it is today. Originally, it was a bold and virtually groundless speculation—a sheer stab in the dark. Only much later did it receive serious confirmation (and elaboration). The basic idea of atomism is that the physical substances that we observe, in all their immense variety, might be composed of imperceptible units, relatively few in number, whose various combinations give rise to observed reality. The properties of these hidden units explain the behavior and interactions of the manifest substances, so that we may derive descriptions of the macroscopic world from information about the microscopic world. Since the atomic units themselves are not observed, they are initially introduced as theoretical posits, justified by their explanatory power. By definition, they are simple—that is, not composed of more elementary parts—and hence have no further explanation in terms of composition. They exist in some sort of medium—space or “the void”—and they are subject to certain sorts of adhesive force—that which holds them together into the stable forms that we observe. Thus atomism is a reductive theory, in that it seeks to account for apparent natural variety in terms of relatively homogeneous elements; the world is more parsimonious than it appears, more uniform. Atomism views the manifest world as a series of variations on a small number of themes—with the themes not apparent to the naked eye. It decomposes the variety of the world into simpler unobserved units.

In this paper I want to explore the possibility that atomism might also be true of the mind, specifically of consciousness. I do not mean by this the idea that the mind might be composed of the same atoms that matter is composed of; my question is not whether materialistic atomism might be true. My question is more abstract: might some version of atomism be true of consciousness? That is: might the mind be composed of unobservable units, relatively few in number, that combine to produce the richly varied phenomenology that we find in mental life? Might there be a reduction of the varieties of consciousness to the properties of these atoms, where these properties would permit a derivation of the manifest properties of consciousness—an explanatory theory analogous to our atomic theory of matter? Whether these atoms of mind are “physical” or “mental” (or whether these are useful or meaningful categories) is not my chief concern; I am concerned with the structure of the theory—whether it fulfils the abstract description of atomism so far presented. The question is whether consciousness might have an analysis (in roughly the chemist’s sense) into atomic elements that combine to generate its manifest qualities—whatever the nature of these elements turns out to be. In other words, is anything analogous to the atomic theory of matter true of the mind?

In order to pursue this question, I propose that we revisit the ancient Greeks. My reason for proposing this is that I believe that the current state of understanding of the mind resembles pre-Socratic understanding (or the lack of it) of the material world, so that we can derive some valuable lessons by comparing our own state of knowledge to theirs. Of course, they had no appreciation of how primitive their theories were—those theories no doubt impressed their proponents as the height of sophistication; and I believe that we too have an overly rosy perspective on the adequacy of our present understanding of consciousness. Indeed, there are striking parallels between the theories contemplated then—and no doubt maintained with great seriousness and vehemence—and our contemporary theories of the mind. This, then, is to be a lesson in humility—or historical perspective (which may be the same thing). Atomism, I shall suggest, is the best bet in the current state of knowledge, i.e. ignorance. It is not that we can know atomism to be true, but as stabs in the dark go it is a pretty good one—about as good as the stabs of the ancient atomists.

When Democritus first hit upon the atomic theory it was totally speculative, with almost no empirical support or explanatory success; it was a theory schema, pulled from thin air, an overreaching hunch. It is said that he noticed motes of dust in the air and conceived the idea that all of matter might be composed of such tiny parts, only tinier. The level of explanation was along the lines of postulating sharp spikes on the atoms composing certain foods to explain their bitter taste—that is, wholly false. But Democritus did, crucially, have the idea that there ought to be some explanation of this form—some account of the macro in terms of the micro. He sensed that the observable world might conceal its own workings—the right level of explanation might go beyond anything that we can observe. This was his central insight—an idea about the form of the correct theory. The theory was, of course, widely dismissed in the ancient world, even ridiculed; but as we all know it was spectacularly confirmed some two millennia later, though in a form that Democritus could have hardly have envisaged. It was the best theory in the ancient world, though it didn’t look like it then; it was just too bold, too distant from common observation. Less ambitious theories looked better to ancient thinkers. My conjecture is that the same is true of atomism about consciousness now: it looks wild, unfounded, counterintuitive, but it will turn out to be the best theory—better than its contemporary rivals, for all their seeming sanity. Let me then recall those ancient theories and indicate their contemporary counterparts in theories of the mind. I hope this will not appear frivolous, though the humor is sometimes hard to resist; there is a serious point to it.

The pre-Socratics were committed naturalists--no appeal to supernatural entities and forces. Not surprisingly, their theories tended to be grossly reductionist. Thales is the paradigm, with his bracingly simple proposal that “everything is water”. Water comes in several forms—ice, liquid, steam—and is essential to life, so maybe we can reduce everything to water. It seems as if there is a greater richness to nature than this—more types of substance—but maybe variations on the water theme can account for everything. Fire? A very agitated form of water. Earth? Coagulated water. Air? Exceedingly light thin water. Soul? Well…we do need water in order to keep the soul alive. So it’s not as if Thales had nothing to say about these natural questions, though indeed the theory looks remarkably far-fetched. It is the same with Anaximines’ theory that all is air. The theory certainly has an economical basis—Occam would have approved—and air is obviously implicated in life through the process of breathing, and maybe air can take solid forms (steam and air are similar, and steam can turn to water), and fire might be very hot air. Like the water theory, the air theory strikes us as too sparing in its allowable theoretical primitives, but it is not as if these theories are just mad ravings or pieces of poetry—they have their rational motivations.

Now what do these two theories remind you of? They remind me of various brands of materialistic reduction that have gripped more recent thinkers—particularly, behaviorism and “central state” materialism. Faced with the seeming heterogeneity that separates mind from matter, consciousness from the brain and body, and also with the heterogeneity within the mind, materialists assure us that this is all illusory: really “all is body”. In the central state version, we are told that neurons and their firings are a sufficient basis for every aspect of consciousness: consciousness itself is “really” just the firing of neurons, appearances notwithstanding, and the distinctions between conscious states—say, between seeing red and feeling pain—are really nothing but variations on the neural firing theme (faster or slower firings, or where in the brain the firing takes place). There are neural states “corresponding” to conscious states, so why not invoke Occam’s razor and simply identify the two? Granted, it seems hard to believe—mental states are not on their face neural states—but then science has often surprised us. Thus the materialist reconciles himself to the less palatable aspects of his view; and he can also ask what else you have to offer if you don’t like what he is peddling.

Now I am not here attempting to argue that materialism is false; I am simply making a tendentious comparison to certain aspects of the debate about materialism and what occurs to anyone thinking about the theories of Thales and Anaximines. And I do think there is a striking similarity here: water and air are taken as given, and then the rest of nature is squeezed into these boxes; similarly, matter is taken as given, and then we have to squeeze the mind into that category in the form of firing neurons—as a minor variation on the theme of matter. Might materialism actually be as crude a theory as those old theories? Did the theories of Thales and Anaximines seem crude to them? Surely they seemed like the cutting edge of human thought, and anyway no one around at the time had a better theory. Their manifest defects could be put down to their novelty and audacity. Materialism might strike you as counterintuitive, even absurd, when you first encounter it, but you can school yourself to accept its leveling consequences—and there are always “moves” that can be made to ward off the doubters.

Or consider Pythagoras: he held, in effect, that “all is number”. He had the curious view that physical objects were composed of numbers—that they were abstract and mathematical. They don’t look like they are composed of numbers, but actually they are; they can be mathematically described, and this is because they are themselves mathematical. Remind you of anything? What about contemporary computational functionalism? Mental processes are computational procedures, abstract properties of brains, which can be modeled by computer programs. Granted they don’t seem this way to introspection—they seem somehow “subjective”—but in reality they are abstract functions, mathematically describable operations. A computer program is an abstract algorithm, which is “realized” in the hardware of the computer; similarly, a mental state is just a state of such an algorithm, “realized” in the brain. This is not a materialist view, since mathematical entities are not reducible to physical entities—hence the possibility of “multiple realization”—but it is reductive nonetheless. In the same way, Pythagoras’s theory of physical objects is not “materialist”, since he explains such objects in terms of abstract numbers; and, yes, maybe the same collection of numbers might crop up in (be “realized by”) distinct objects. Pythagoras held that objects are really “software”, just as the contemporary functionalist holds that mental states are just software. After all, we get such a nice theory if we adopt these mathematizing reductions: both physics and psychology turn out to be branches of mathematics.

What abut contemporary eliminativism--does it have an ancient counterpart? Parmenides was puzzled about change; he couldn’t see how it was possible. So he declared it illusory, along with the entire observable world. In his view, “all is one”, and diversity is an illusion too. The world of changing diverse objects is declared unreal, a prejudice of common sense that cannot withstand rational scrutiny. Zeno, influenced by Parmenides, famously declared motion unreal, because of his paradoxes; he “eliminated” motion from his world-view. These were acknowledged to be extreme moves, radically counter to common sense, but they seemed by their proponents to be backed by solid skeptical arguments. Here I am reminded of the whole eliminativist tradition from Quine and Rorty, to the Churchlands, to Kripke’s “skeptical paradox”. Quine can make no sense of reference (as Parmenides could make no sense of change), so he eliminates it; the Churchlands can’t fit folk psychology into their pre-conceived scientific picture of human beings, so it is jettisoned; Kripke’s skeptic (just like Zeno) claims to find a real paradox at the heart of the notion of meaning—it just doesn’t seem possible—and so proposes to abandon the idea that meaning is a fact. Each of these thinkers argues himself into the denial that apparently self-evident facts are real—no change, no motion, no beliefs, no meanings, no mind. It turns out there is a lot less to the world than we thought; desert landscapes, etc. The more robustly commonsensical among us shake our heads at such sweepingly eliminative recommendations.

Panpsychism has a beautiful precursor in the shape of Anaxogoras. Anaxagoras was deeply puzzled about how hair grows on the head: hair luxuriantly cascades about the scalp, but where does it come from—after all, there is no hair inside the head from which it might proceed? Come to think of it, how does anything turn into anything else? How does bread become flesh and bone? His answer was radical: it’s because every substance contains a little bit of every other substance (“all is all”, as one might say). Hair grows from the scalp because the scalp itself contains tiny hair-like elements; indeed, hair comes from bread, through digestion, because bread has slivers of hair in it too—though these hairy portions are too small to see. And there must be hairy elements in whatever bread comes from. Pan-hairism! Anaxogoras has a real explanatory problem--how to explain the “emergence” of hair from heads—and he solves it with a bold stroke—he postulates hair in the head already, but in unobservable form. We might call these theoretical posits “proto-hairs”, since they are not quite full-blown hairs, but they have hair-ish properties. Compare: how does consciousness emerge from the brain, given that the brain seems so different from consciousness? Answer: the brain is composed of material elements that themselves have mental properties, and indeed such properties are found in matter generally. Panpsychism! These properties might be called “proto-mental”, since they are not full-blown conscious states, but they do have the seeds of consciousness within them—they are conscious-ish. This, too, is a response to a genuine explanatory problem, and the logic of the response seems exactly analogous to Anaxagoras’s response to his problem. The resulting doctrines are admittedly startling, perhaps offensive to common sense, but they are the only thing that could explain what needs to be explained. True, there isn’t any direct evidence for hairs in every bit of matter, or for mental states in every molecule, but there have to be such things or else the world makes no sense.

What about the more moderate thinkers of antiquity? Empedocles suggested, non-reductively enough, that the natural world is composed of four elements—air, earth, fire and water. Each is primitive; none can be reduced to the others. For him, Thales was a crude dogmatic reductionist; there was more to heaven and earth than forms of water. Empedocles has his counterpart in the non-reductive theorists of today—everyone from Cartesian dualists to non-reductive materialists like Davidson. Water and fire are distinct, sui generis and “incommensurable”; there are no (strict) water/fire laws; yet there are causal interactions between the two (water puts fires out and fires heat water). Empedocles had four fundamental constituents; nowadays we have two (mind and matter). There is no true completion of “all is…”. Such theorists make no attempt to explain how mind and matter are related—how the former “emerges” from the latter—just as Empedocles made no attempt to explain how one of his elements might come from the others or specify what fundamental properties they might share. He was content simply to declare a set of basic divisions within reality. After all, look into what absurdities the reductionist drive had led others!

Then there is good old Socrates, a relative latecomer to these debates. Socrates, it is reported, tired of these ancient debates, after a youthful infatuation with them, viewing them as so much groundless speculation—they were not fit subjects to deliver genuine knowledge. Thereafter, he devoted himself to ethics, which he felt offered a greater chance of knowledge. His position on questions of natural philosophy was one of resolute agnosticism. This, I think, confirms Socrates’s reputation for wisdom: he saw that all this speculation was precisely that and didn’t want to waste his life in fruitless wrangling over unknowable matters. As it turned out, this wrangling eventually metamorphosed into genuine science, but in the ancient world this was very far being from the case. Socrates was a pessimist about finding out the truths of natural philosophy, so he turned to ethical questions. His emphasis on human ignorance no doubt reflects his youthful frustrations over the science of nature. His contemporary counterpart is the skeptic who questions all current theories of the nature of consciousness (Nagel would be a good example). I too sense a deep well of ignorance here, with the various theories on offer merely crude attempts to come to grips with the phenomena. But in this paper I am interested in trying out the atomist vision; not because I think we will soon (or ever) be able to convert this vision into an actual theory—rather, as a speculation about what the form of a correct theory might be like. And I expect my speculations to be at least as crude and groundless as Democritus’s were; that, indeed, is part of my point.

But let me first make a comment about materialistic atomism—the theory that the mind consists of material atoms, the same ones that compose brains. This is the thesis that the varieties of conscious phenomena are simply aspects of the atomic constitution of those phenomena: chemical kinds are aspects of atomic constitution, and so are phenomenal kinds. Properties of consciousness are therefore explicable in terms of the properties of physical atoms, just as properties of chemicals are. This is clearly a highly reductive theory, and the reduction proceeds upon the basis of an atomistic conception of the nature of mental phenomena. The point I want to make about this is that this is a responsible and respectable materialistic theory of the mind—though one that we have no good reason to believe and a lot of reason to disbelieve. But it has the form of a decent theory, because of the explanatory obligations it acknowledges; it accepts the same obligations as an atomistic theory of chemical phenomena does. It is no mere “identity theory”, according to which it is more or less stipulated that mental states are physical states like neural firings. Atomism is a claim of explanatory power, and it earns its right to be taken seriously by making good on its explanatory obligations. The simple identity theory does not do this. Thus the difference between, say, seeing red and feeling pain must be reductively explained in terms of the differences in the properties of the underlying physical atoms, since materialism of this kind must claim that every distinction is a physical distinction, and physical distinctions must always come down to differences in the properties of atoms—in particular, the nature of the particles that make up the atoms (electrons, protons, etc). I think it is obvious that no such explanation of the phenomenal has ever been given, and indeed it is hard to see how it could be given. So there can be no atomistic reduction of the mental in the way there is an atomistic reduction of the chemical. The broader kind of atomistic reduction I am contemplating in this paper will not then take the form of reducing the mind to physical atoms (of the kind now known about). Mental kinds are not this type of atomic kind. I am envisaging a theory that does for phenomenal kinds what our physical atomic theory does for chemical kinds—and cannot do for phenomenal kinds.

The kind of atomism I have in mind is not committed to any such materialism. It says simply that conscious states and processes consist of underlying states and processes that are not observed but which combine to produce what we do observe. We can assume that these underlying entities are not elsewhere observable; they are not part of what we ordinarily take the world to contain—any more than the atoms of Democritus or modern atoms are. So I am not taking them to be phenomenal atoms—that is, elements of consciousness derivable by phenomenological analysis. They are nothing like the constituents of sense-data that some philosophers have posited. They are essentially hidden, and presumably conceptually alien—just like physical atoms. Of course, I have no idea of what they are like intrinsically; all I am suggesting is that they exist and compose the mind according to some principles or laws of combination. When you put the atoms together in certain ways you get a sensation of red, say; and this result is wholly explicable in terms of the properties of the atoms—just as the properties of a chemical are so explicable in terms of physical atoms. The appropriate theory is therefore fully reductive. We can also suppose, following the physical model, that the types of consciousness atoms are relatively few in number, so that we can derive the richness of mental life from a smaller set of primitives; an underlying uniformity is masked by the surface variety of the mind. Physical atoms consist of a nucleus and revolving charged particles; from this we can derive the properties of the composed substances. In the case of the atoms of consciousness we can expect a similar simplicity of underlying reality; so the mind is simpler than it seems—just as the physical world is. Physical atoms also exist in space and enjoy spatial relations to each other. It is doubtful that mental atoms can be conceived in this way, at least as space is now understood, because the mind does not appear to be spatially describable (in current terms); but presumably they too must exist in some sort of medium, within which they can separate and combine. Physical atoms can also exist in a detached form, not as part of a macroscopic object; at any rate, they can be conceived apart from other atoms. So we might expect that atoms of the mind can (be conceived to) exist in a detached form, pre- or post-consciousness. Where these atoms might exist, independently of constituting a full-blown mind, we cannot say; but we cannot rule out their existence in other places (or “places”).

Is this panpsychism by another name? Well, it doesn’t say that mental atoms are everywhere, nor that physical atoms have conscious (or proto-conscious) properties, nor that we have any idea what the intrinsic nature of the atoms of mind is. It shares with panpsychism only the idea that there is a hidden theoretical level of specifically mind-generating properties—but unlike panpsychism, properly so called, it does not identify these properties as mental, in the sense that the concepts that express them are anything like our concepts of the mind. Also, panpsychism isn’t generally an explicitly atomistic doctrine, claiming that conscious states have atomic structure in their own right. For all we have said, the atoms of mind might compose a Cartesian immaterial substance; so there is nothing in the theory as such to entail the distinctive doctrines of panpsychism. Just as Anaxagoras’s doctrine includes some atomistic ingredients, because the bits of hair in everything are microscopic, so panpsychism contains a hint of atomism; but Anaxagoras’s doctrine is not the same as Democritus’s--as panpsychism is not the same as atomism about the mind. In fact, atomism about the mind is not itself a theory of emergence at all; it is a theory about the intrinsic nature of mental phenomena. It doesn’t attempt to explain how consciousness arises from brain cells; it says what kind of constitution conscious minds have.

Is there any reason to believe the atomistic theory? Very little, so far as I can see—just as there was very little reason to believe the atomic theory at the time it was first enunciated. I am describing the theory, not endorsing it—as anything other than a hunch. But there are a couple of points that might seem suggestive. First, it does appear that consciousness is ontologically dependent on the brain (not just causally), and the brain is a physical substance composed of physical atoms--so it would make sense to suppose that consciousness has an atomic structure too. It would be odd if it were a continuous thing, given that it arises from a non-continuous substance. And it would be odd if its atomic nature were not hidden, given that the atomic constitution of its basis (the brain) is hidden. How can brain and mind mesh if one is atomic and the other is not? (I mean this to sound like one of those suspicious but ingenious arguments that the ancient Greeks were so fond of.) Secondly, what model of change do we have if we cannot base it on the idea of atomic structure? When physical objects change their atomic parts get re-arranged; there is no macro-change without a micro-change. The mind changes too, of course, as mental states come and go; it seems reasonable to suppose that this also is a matter of the re-arrangement of its parts. The changes we observe are the reflection of changes at the underlying level, as the atoms come and go and connect to other atoms. Changes in the stream of consciousness are the visible upshot of a vast sea of changes at the atomic level. Of course, none of this is probative, to say the least, but at least the idea makes some sort of theoretical sense—it is not just groundless assertion.

Let me emphasize that I don’t want this theory to be provable, on empirical or a priori grounds. My view is that our ignorance is far too deep to allow any such knowledge. I mean to be occupying a position analogous to that of Democritus so long ago: this is a theory to which I have taken a fancy, that has come to me from thin air, that I really have no right to hold. But, as Democritus’ theory was eventually so stunningly vindicated, so this theory might be (or might not); in any case, it has been stated. At least I have the success story of Democritus to encourage me; he had nothing but an obscure intuition with no antecendents. To tell the truth, I have a stubborn conviction that the atomic theory of consciousness has to be true--that nothing else is on the cards; but, as I say, I have no right to this view, or to my reader’s assent to it. Have I even had anything like Democritus’ experience of the dust motes in the light? Not really, except perhaps seeing those little points of iridescence that inhabit the visual field as one tries to go to sleep at night; and maybe an obscure intimation that beneath the surface of my consciousness there throbs a universe of discrete and buzzing nodes of sensitivity--little blobs of mind dust. But here I stray, lamentably, into poetry, worthy perhaps of a bemused and probing ancient Greek. I know not whereof I speak, but the words come to me anyway.

Democritus felt that there had to be something more to reality than meets the eye; the observable world cannot explain itself. One part of it cannot supply the true nature of another part—as Thales in effect believed. There has to be an underlying reality, relatively simple, subject to rigid law, highly structured, that accounts for what we observe. The surface needs support from something else. I have the same conviction about the mental: what I introspect of my consciousness cannot be the whole story—there has to be another level of reality that brings it all together. I don’t believe the so-called physical world can do this, at least as currently conceived; so I am compelled to posit a hidden world of mental atoms and their interactions, where these atoms may be far removed from anything we observe of consciousness (though of course they constitute it). The atoms of consciousness might have properties that are not at all predictable given the way consciousness appears to us. Will they take the form of a nucleus surrounded by a shell of particles? That sounds like too slavish an adherence to the physical model, but they might have a complex structure, and hence not be strictly atoms, i.e. indivisibles. Will these atoms combine to form molecular compounds? That sounds more likely, since there must be a hierarchical process of construction that takes us from the most primitive level to the level of felt experience: atoms, then molecules, then manifest kinds. Physical contiguity is not likely to be the preferred mode of combination—adhesion to neighboring atoms—but there will have to be something analogous to it—some principle of concatenation. Electro-magnetism won’t be the force that binds the atoms into bigger wholes, but again something analogous to it needs to be assumed; for something must cause the atoms of mind to come together into stable, if shifting, forms. We know nothing about any of this, except perhaps that something playing these roles must be postulated, given the very structure of an atomic theory. We know what the atoms of mind have to do, but we have no idea what they are.

An interesting question to consider is whether a distinction between primary and secondary qualities will be a part of the atomic theory of consciousness. Galileo and others distinguished the intrinsic properties of objects from properties that depend upon the observer’s response—shape from color, for instance. The atomic theory was intended to apply only to the former properties; the latter properties were not held to be subject to atomism. This distinction between primary and secondary qualities is essential to the success of physical atomism, given that some of the properties of objects—the mind-dependent ones—are not capable of analysis in terms of physical atoms. Thus atomism cannot account for the color of objects, according to Galileo and company, but this is no defeat for an atomic theory of the mind-independent intrinsic properties of objects. My question is whether such a distinction would arise for the case of the mind itself: do we need to distinguish intrinsic features of consciousness from features it has in virtue of its appearance? The answer, I think, is that we might make such a distinction but that it would not have the result of exempting the secondary qualities of the mind from the reach of atomism. The reason is obvious: those appearance-dependent qualities would themselves be aspects of consciousness, and so a general atomism about the mind has to account for them. Put differently, atomism about the mind has an obligation to account for the appearance-dependent qualities of the mind as much as for the intrinsic qualities; there can be no shifting of these qualities elsewhere, as with physical atomism.

Suppose that being in pain is a secondary quality of mind, with some primary quality basis—we know not what: we would then need to produce an atomistic account of pain, and not merely of its basis in the primary qualities of consciousness. So the significance of any distinction between the primary and secondary qualities of consciousness would be quite different in the environment of an atomic theory of the mind. The atomism needs to go all the way up, so to speak. And, clearly, any attempt to re-introduce a primary/secondary distinction at the level of an appearance-dependent property would run afoul of an infinite regress, since there will always be an appearance-dependent residue.

Would the production of such an atomic theory solve the mind-body problem? Not necessarily, though it might contribute to such a solution. For there is no guarantee that the atoms discovered to constitute the mind will relate intelligibly to the atoms that constitute the brain; there might still be an explanatory gap (indeed, as I have noted, this kind of atomism is in itself consistent with dualism). A yet deeper analysis of the world might be necessary to bring the two schemes together—to make the two sorts of atom interlock (which I assume they must eventually do). We already know, I think, that physical atoms cannot do the job of constituting the mind, so there will certainly be no identity between the atoms of the mind and the atoms of the brain—the one atomic theory will not reduce to the other. On the other hand, surely the atomic theory of mind would provide a deeper understanding of what the mind is, and hence illuminate how it might fit into the material world. It might, for example, help us understand psychophysical causation: how the re-arrangement of one sort of particle brings about the re-arrangement of the other sort. But, again, we are flailing in the dark here, in view of our pre-Socratic predicament. Conceivably, the correct atomic theory of consciousness might make the psychophysical divide look even more unbridgeable, and some quite new conceptual framework be needed in order to get a unified picture—say, something that revealed both sorts of atom as expressions of a deeper reality (Neutral Monist String Theory?).

Finally, I want to say a few (amateur) words about chemistry. Modern chemistry is the area in which the atomic theory of matter receives its fullest confirmation. Bohr’s theory of the atom makes sense of the periodic table of elements, in terms of atomic numbers and weights, and predicts the behavior of chemical substances. It is worth reflecting briefly on the structure of this understanding. Chemists used various techniques, such as spectroscopic analysis and crystallization, in order to arrive at a table of the elements, the basic building blocks of the substances we observe; only later did the atomic theory, involving orbiting electrons, come in to give this mass of observation a solid theoretical foundation. So the order of discovery was, initially, taxonomic and macroscopic, and only subsequently explanatory and microscopic. Question: could there be anything analogous to the periodic table for the mind? This would not be a purely phenomenological classification, but would have to involve some techniques of experimentation and analysis; it would have to isolate the underlying building blocks—the ingredients of mental compounds and mixtures. What sort of experimentation and analysis? I have no idea—I am but a limping Democritus in all this. Perhaps dissociations caused by brain injury might play a part in this—basic mental modules might approximate to chemical elements. Perhaps these could be ordered in some revealing way. The stage would then be set for a Bohr of the mind to come along and explain how these elements derive from a few basic atomic structures—the analogues of the electrons orbiting the nucleus. In other words, we don’t go for the atomic theory straight off; we wait till we develop the mental chemistry far enough. The mental chemistry—exemplified by the periodic table—is the intermediate level of theory, needed to bring together consciousness as it appears and whatever atomic structures might underlie it. We do the chemistry of consciousness before we attempt the physics. Looking at the history of chemistry from the time of Democritus, this sounds like the best strategy for working towards an atomic theory of mind: the chemistry took shape before its atomic rationale ever became apparent. But, as I say, I haven’t the slightest idea how we might proceed here; I am merely drawing an analogy that might be worth pondering.

My aim in this paper has not been to defend a theory of consciousness. My aim has been to sketch the form that such a theory would have if it were ever discovered. My reasons for putting forward such a sketch are not that I have any real arguments for it, only that it seems like a theory-form that might actually be true. It is a pure conjecture on my part. Speaking as one of the ancient Americans—seen from a vantage point in the distant intellectual future—I am giving my pre-Socratic self a voice. Democritus had it right, after all, by some combination of chance and strange divination; so maybe an atomic theory of consciousness is also the truth, even though obscurely glimpsed now (even this is putting it too strongly). And who can resist an occasional pre-Socratic spasm when it comes to the topic of consciousness? It is hard to play the wise agnostic Socrates all the time, declining to speculate and devoting oneself to ethics.

Colin McGinn