Consciousness, Paralysis, and Functionalism

Paralysis has been regularly used to defeat behaviorism and its descendant functionalism. How can the mind consist of behavior if paralysis is consistent with having a mind? The objection is clear and strong: paralysis shows that behavior is not required for consciousness to exist. To have a conscious state is not to move in a certain way. It might even be that a conscious being has never moved and never will and has no disposition to move. What I want to point out is that much the same objection applies to attempts to define mental states in terms of their mental effects: what a mental state does internally is equally incapable of defining it, because of the possibility of mental paralysis. I will call this the “paralysis argument”. The argument is not complicated; in fact, it is embarrassingly simple. Consider a pain occurring at time t: it may have many mental effects—forming the belief that you are in pain, desiring to escape the pain, comparing it to other pains you have had, laying down a memory of the pain, thinking you will later write about the pain, deciding not to do that again, and so on. All these occur after t, radiating out from the pain, eventuating in overt action. Now suppose that some neural catastrophe were to afflict you, interrupting the causal connections: you don’t and can’t have those mental effects of the pain; you have the pain but the causal sequence is abolished. The neurons fail to send the signals to other areas of the brain, shutting down the mental effects of the pain. You are mentally paralyzed. Clearly, this is consistent with having the sensation of pain, since the effects occur after the pain. But then, such effects are not a necessary condition of feeling the pain, and so cannot constitute having it. The situation is exactly like bodily paralysis except that the causal sequence is interrupted at an earlier point. The effects could extend further out than the body into the world at large and could be interrupted at any point; the situation is not different going in the opposite direction. There can be behavioral paralysis and mental paralysis—as well as “paralysis” with respect to environmental effects. But then, we can’t define mental states in terms of their effects, physical or mental, which means that functionalism won’t work. What a mental state is can’t be reduced to what it does, mentally or physically. To be a pain is not to have a certain functional role, since the pain can exist without the functional role existing. People have argued that functional role is not sufficient for a mental state, but it is also not necessary, by the paralysis argument. We don’t see this kind of paralysis often, if at all, but that is because the relevant parts of the brain are not as exposed to injury or interference as much as the spinal cord; but as a matter of principle the cases are not different. The fundamental problem is that pain and its effects are distinct existences (as Hume would say), so that one can’t be the other. Once this point is appreciated (it isn’t a difficult point to grasp) we can develop two further anti-functionalist arguments: the “time-lapse argument” and the “ignorance argument”. The time-lapse argument points out that the effects of pain post-date the pain, but the pain doesn’t post-date itself; so there can be no identity between them. The ignorance argument points out that you could know you are in pain without knowing its effects, since these are different things; so there can be no analytic or a priori connection between them. You might simply become mentally blind following the occurrence of the pain, so that you are ignorant of its mental effects, or those effects don’t occur because of mental paralysis. Your knowledge that you are in pain cannot then consist in the knowledge that you are in a state with such and such mental effects. Unlike materialism, functionalism has problems with time: materialism invokes a contemporaneous brain state, but functionalism invokes a temporally extended causal sequence, thus incurring the problems outlined. As I said, pretty simple stuff.[1]

[1] This isn’t to say that functional role could not be employed less ambitiously in an account of the mental, as merely an aspect of the mental state; but then, it isn’t identical to the mental state. This would be a kind of weak functionalism analogous to various kinds of weak materialism. Classical functionalism, like classical materialism, is a case of theoretical exaggeration, taking an essentially sound point and overdoing it.

A New Metaphysics of Necessity

There is a tacit recognition in the history of philosophy that in order to account for necessity we need to introduce a split in what we regard as overall reality. Thus, we have the idea that necessity resides in meaning, conceived as separate from the ordinary reality of things and properties, and also the idea that necessity resides in a totality of possible worlds, conceived as separate from the actual world. We need to adopt a type of dualism (doubling of worlds): of world and meaning, or of actual world and possible worlds. The latter worlds are conceived as shadowy (if not shady) in relation to the primary reality of the non-semantic and actual worlds. Necessity belongs in a separate sphere of reality, specially designed for it. These are both good ideas, in their way, because they accept that reality needs to be carved into two parts in order to do justice to necessity. A metaphysics of necessity needs to be ontologically expansive (non-local) in some way. Necessity can’t exist in a world consisting only of ordinary objects with ordinary properties (what we might call the-cat-sat-on-the-mat metaphysics). We might even have to countenance something relatively exotic. Along these lines, I will defend a split-reality metaphysics of necessity, specifically a two-worlds approach. I shall say that necessity is truth in the deep world and contingency is truth in the surface world. So, I believe in two worlds existing side by side, one deeper than the other. I am a “world dualist”. But my dualism is not the same as other dualisms that dot the philosophical landscape: mind and matter, appearance and reality, concrete and abstract, fact and value, particular and universal, noumenal and phenomenal. I have in mind a different sort of distinction that has no accepted name (or even being), which I will call the world of essences and the world of facts. The intuitive idea is that some truths correspond to things that have been brought together, assembled, combined, while other truths do not correspond to such joining, but concern inseparables, indivisible wholes. For example, my table may have a cup on it, which has been brought to the table from elsewhere and can be disjoined from it; it may also have the property of being brown or shiny. Such circumstances may be described as combinatorial in somewhat the manner in which Wittgenstein spoke of facts as combinations of objects (with properties and relations included under “objects”). These arrangements are contingent, accidental, mutable, conceivably otherwise. But my table might have other properties which are not like this, e.g., being made of wood (or a particular piece of wood). Here we cannot separate the object from the property: the wood was never brought to the table, combined with it, placed next to or on it. The wooden table is not the result of a combination of objects with a prior and independent existence: the table has no being without the wood that composes it. The truth that the table is made of wood (this wood) is a truth of essence not a truth of fact in the aforementioned sense; it isn’t combinatorial. It is a truth about the nature of the thing not about what accidentally befalls it. And natures are not combinations of independent existences, since without a nature a thing cannot exist at all. So, we have two levels or layers of being: a world of essences and a world of combinations (facts in roughly the sense Wittgenstein had in mind). The former give rise to necessities and are eternal and immutable; the latter give rise to contingencies and are (typically) temporary and mutable. We can say that necessities exist in the former world and contingencies exist in the latter world. These are different kinds of world (compare the world of meaning and the non-linguistic world) in that they have a different kind of “architecture”—one being combinatorial and the other being non-combinatorial. If the world is the totality of facts in Wittgenstein’s sense (complexes of objects), then that world does not include necessities (essences); and the world of essences does not include the world of facts, since it doesn’t admit of combination and recombination. (Note that Wittgenstein does not locate necessity in the world of facts, since for him it consists solely in tautologies). We have two different modes of being here—two different ways of constructing reality. And the point generalizes: numbers have no existence without their essential properties (there is no such thing as bringing evenness to the number 2), and bachelors are not combined with the state of being unmarried (they are unmarried by definition). In other words, de re necessities (and de dicto) are differently formed from contingencies—they are found not made, constitutive not compositional. You can’t arrange for the table (thattable) to be made of wood; it just is, essentially, ineluctably. The point is that talk of two worlds is justified by the inner architecture of the realities in question (I won’t say facts because I am using “fact” here in the narrow Wittgensteinian sense). But why do I say “deep world” and “surface world”? First, because we don’t see the world of essences; it is hidden from view. We know it but we don’t perceive it: we never observe the coming to be of an essential nature on the part of a pre-existing object, since there was no table before the piece of wood came into its life. But we do see the combining operation that forms contingent facts (we see the table being painted, for example). Second, the surface world presupposes the deep world in that there could be no combination of objects without antecedent objects possessing a determinate nature: essence precedes existence. Facts need constituents and constituents need natures. But the world of essences doesn’t need the world of facts (compare Plato on particulars and universals): things can have natures in logical independence of combining with other things. Third, the essence of an object is a deep truth about it in contrast to its contingent and fleeting properties. So, the world of essences is a deeper world than the world of facts; and this means that necessities are deeper than contingencies—more fundamental, more formative. Necessary truths are therefore metaphysically deeper than contingent truths. They call for a further world and this world is deeper than the world of contingency. We don’t need meanings and a plurality of worlds to capture the nature of necessity (in my judgment anyway[1]) but we do need an extra layer of reality; we need that ontological split. Metaphysically speaking, necessity has to reside in a different place from contingency; it needs its own dimension. No doubt this is why it has always seemed metaphysically threatening (it was the biggest threat to logical positivism). Linguistic theories seek to confine it to the realm of meaning, while possible worlds theories multiply non-actual worlds for it to inhabit; according to the deep world theory, we need to recognize two layers to our world, with necessity existing in the deeper layer. Necessary truth is defined as truth in virtue of the deep world. Let me add that this theory, like the others, brings with it a vision of reality that can alter your way of seeing the world: you see it as having an outward perceptible form consisting of myriad combinations of objects and you see it as harboring an underlying reality of unchanging essences that are not combinatorial at all. There is no juxtaposing at this deeper level, no moving of pre-existing pieces around (like a chess game); rather, there are fixed and immutable indivisible realities—as that the table is essentially made of wood. In particular, the table and its constitutive chunk of wood were never at distinct points of space from which they moved in order to live happily together; rather, the table has no reality other than being made of that piece of wood. When you look at the table you are gazing at the intersection of two worlds, a world of essence and a world of accident (e.g., the cup sitting on the table for a while). Both these worlds occupy your visual field. You see the table as a metaphysical nexus consisting of a surface structure and a deep structure (to coin a phrase). The (whole) world is the totality of facts and essences. This is what you feast your eyes on every day. The universe comes to seem like a happy superimposition of one world on another, with necessity lurking just beneath the surface.[2]

[1]There is no inconsistency between the deep world theory and the other two theories, though I would object to them on other grounds. My point is that we all agree that necessity requires more than the ordinary world of perceived particulars; it needs its own special place—the question is where. An advantage of my proposal is that the necessity lies exactly where it is supposed to lie, i.e., in the object of predication, not in language or other worlds. It exists in the world of essences that underlies the world of facts, which is not somewhere else entirely.

[2] There is a question what to say about the necessity of origin, given that parents are spatially separate from their offspring. Here we should say that the child doesn’t have the property of being the child of these particular parents in an external manner, because without having that property it would not be the individual that it is. The case is like natural kinds: I don’t combine with the property of being human as I combine with the property of being a philosopher, because there is no conceiving of me as existing without being human. You can’t be said to combine with something that is a necessary condition of your existence. Combination requires logical independence.

And and Not

Sharp thinkers (Frege, Russell, Wittgenstein, among others) have felt that there is something special about the classical logical connectives, and, or, not, and if. I will list the features commonly attributed to these concepts. They are truth-functional and referentially transparent. They are disquotational in the manner formulated by Tarski’s so-called recursion clauses (“p and q” is true if and only if “p” is true and “q” is true). They are inter-definable and permit reduction to just two primitive concepts (and and not are the most intuitive—I set aside the Scheffer stroke). They can be iterated indefinitely. They are logically central in that they form the structure of logical arguments. They are topic-neutral. They can be used to define the quantifiers all and some (more or less). They are constructive in that they build new propositions from old. They are universal in belonging to any language or system of thought worthy of the name. They have been thought to be unique in that no other linguistic constructions have the properties just listed, despite some superficial syntactic resemblance (“because”, “necessarily”, “believes”); and it is true that other locutions don’t have the full range of properties that mark the classical connectives. They do appear special. So, how are we to understand their special role in our thought and language—what do they do for us? I think the answer is as follows. The connective and (i.e., conjunction) operates so as to create totalities: these can be propositions or facts or objects (Bob and Ted and Carol and Alice). It is accumulative (additive, aggregative). This is why it is so closely connected to all, which also collects things together in thought. That is, and is the basis of set-theoretic thinking, which is the basis of mathematics and much of science. It is the foundation of classes and categories. With and we go from the particular to the general. We can even formulate ideas such as that of the whole world (the totality of facts—this fact and that fact and the other fact, etc.). Conjunction is a fecund operator. In the case of not we get the possibility of unreality—of not being. If you negate a proposition, you create a representation of something (purportedly) unreal—something that is not the case. Snow is white, but it might not be white. Thus, possibility enters our thoughts, our understanding of reality: reality contrasts with unreality, truth with error, presence with absence, life with death (not-aliveness). The existentialists were right to see in negation an expression of the human condition—our awareness that reality is suspended over an abyss of unreality. The false is as real as the true in the sense that error really exists. It is doubtful that other animals grasp this contrast; they live in the not-not, i.e., the is. So, and gives us the idea that reality forms collections and not gives us the idea that reality has alternatives: it isn’t that reality consists of nothing but unrelated particulars and what you can see with your eyes. There are totalities and non-actualities, wholes and absences. Logic is built around both ideas. So, andand not have a claim to being conceptually basic and indispensable to reason—hence as the essence of logic. Those sharp thinkers were onto something, even if they couldn’t quite say what it was. They gazed in wonder at conjunction and negation, and we can join them in that. As Frege might say, without and and not thought would be crippled, but with them it soars. They form the backbone of logic, which is the backbone of reason, which is the mark of humanity: and and not make us special.[1]

[1] I don’t mean they are projections of human nature—that would be psychologism. I mean our grasp of these objective operations is what sets us apart from the rest of nature, though I wouldn’t rule out some primitive grasp of them in other species, especially primates.  

What is a Mental State?

The first-person viewpoint is apt to skew our conception of what a mental state essentially involves. We introspect the state and think we have attained a pretty comprehensive picture of what it is, intrinsically, constitutively. Fundamentally, this is a confusion (conflation) of epistemology and metaphysics—privileging one mode of knowing over other modes. The nature of a thing is unlikely to reveal itself fully to a single viewpoint. The introspective viewpoint is so familiar, so natural, and so automatic that we easily fall into overestimating its revelatory power. Phenomenology becomes comprehensive insight, not just one perspective among others. Enormous amounts of philosophy have flowed from this movement of thought (we might even call it “the introspective fallacy”). But there are two other viewpoints that need to be taken into consideration, which have no natural names: I mean the viewpoint of an observer of the behavior of another, and the viewpoint of the brain scientist (including brain surgeon). Both are third-person, but they differ markedly—one being natural and shared by all, the other esoteric and specialized. They should be sharply distinguished. I will characterize the former in a variety of ways, as befits its complex nature: it is functional, action-oriented, teleological, biological, dynamic, and behavioral (the catch-all term). It is the mind in action, engaged, goal-directed; it is where the biologically adaptive nature of the mind shows itself (not so much from the introspective viewpoint). It is the reason the mind exists at all—its evolutionary sine qua non. It is also where the social dimension of our psychological discourse belongs: how we think of each other (animals too). It is not to be forgotten or downplayed. Mental states have causal roles, purposes, interactions, bodily expressions. It is quite wrong to think of them as purely “inner”; they have an outer nature too. Behaviorism, in its many varieties, is by no means wildly off-target: it stems from a deep truth about the nature of the mental. The mental, like all biological phenomena, is an active functional reality; the deed is woven into it. But this nature is not immediately available from the introspective standpoint, which confines itself to feeling, subjective appearance, what-it’s-likeness. That standpoint has to be prescinded from in order to grasp the functional-teleological aspect of mental states. In addition to this conception of the mind we have the unfamiliar and artificial standpoint employed by the student of the brain. For most of us this is obtained only from books: we don’t observe brains every day, with the naked eye or under a microscope (etc.). Truth to tell, this cerebral (cortical, neural) standpoint is quite alien, even alienating: it taxes our normal ways of understanding people and animals. Yet the involvement of the brain in the mind is undeniable, obscure as it may be: somehow the mind is rooted in the brain, dependent on it, impossible without it. The brain is as essential to the mind as its phenomenology is, though this is not apparent to us in ordinary life. Mental states must have a cerebral nature; they don’t exist outside the brain, as if owing nothing to it. In the vernacular, the brain is up in the mind. Thus, materialism is also not a wild and gratuitous imposition on reality; it is based on the sound perception that the mind is deeply indebted to the brain and must indeed be a brain state (of some sort). Mental states have a nature that is revealed by adopting the viewpoint of the brain scientist (or surgeon). So, they have a triple nature—a tripartite architecture. A triple aspect ontology is what they demand. But these three aspects don’t easily slot together: they don’t entail each other, not by a long chalk. Indeed, they are fundamentally conceptually distinct, seemingly jammed together, barely on speaking terms. Integrating them takes serious work (and may not be possible for us). But they are not inconsistent with each other, just heterogeneous. Mental states are therefore not like regular natural kinds: water is just H2O and heat is just molecular motion, neither more nor less, but pain is a feeling plus a functional kind plus a brain state. It is a composite being, existing on three plains, as it were. It is a phenomenological-teleological-corporeal kind, part feeling, part purpose, part neurology. The mind-body problem is really the problem of integrating these three aspects without denying or shortchanging any of them. For how can the same thing be all three, and how are they connected? Once we see the triple nature of mental states, we can make sense of the various positions that have been adopted towards them. Dualism, behaviorism, and materialism are all exaggerated responses to the threefold being of the mental; they select one aspect and make too much of it, neglecting the other aspects. Thus, the history of psychology goes from the introspectionist (phenomenological) school to the behaviorist school to the neurological school (what is now called neuroscience), each declaring that it alone is true to the real essence of the mental. But the mind has three aspects, intrinsically, essentially, irreducibly. We also see how various thought monsters insinuate themselves: the disembodied mind, the mindless zombie, the free-floating mental stuff supposed to constitute physical reality (purposeless, non-biological). These all involve detaching one aspect of the mind from the other aspects and declaring it autonomous. The truth, however, is that all three are equally integral to the nature of a mental state: the phenomenal, the causal, the physical. Disciplinary boundaries don’t correspond to real ontological divisions. The mind needs three disciplines to study its complex nature (plus a fourth to integrate them). Above all, we must not let epistemology dictate metaphysics: just because we have three different epistemic perspectives on the mind doesn’t entail that the mind itself must fall into three separate non-communicating compartments.[1] Mental states are not as cleanly divisible objectively as the viewpoints we take to them. We could in principle have one of these viewpoints without having the others, but the mind itself couldn’t exist in such a divided state: it necessarily has each of them. It has a physics, a biology, and a phenomenology—each as essential as the other two. It can’t therefore be modeled on cases in which only a single level exists (water, heat, light, etc.). Idealism is thus impossible, as is reductive (eliminative) materialism: mental states can’t exist without a biological function and a cerebral realization, and they can’t exist without an introspectively known phenomenology (the brain can’t swallow up the mind). Nor can behaviorism, however relaxed, provide a complete account of the mind. Each is part of the truth, but not the whole truth. Biological traits have two aspects, physical and teleological, while physics and chemistry have just one, but psychology has three. The concept of mind, then, can be analyzed into three parts, each with its own proprietary conceptual apparatus, somewhat as the concept of knowledge can be analyzed into three parts (roughly). No one scheme of concepts will suffice, and no one is more central than the others. In particular, concepts of the brain are essential in giving a full accounting of the mind, despite their general invisibility. Much of the mind is hidden, though some is out in the open. That’s just the way the mind is: the hidden part is as vital as the open part. In other terminology, the mind is subjective and objective.[2]

[1] For some reason it feels natural to describe the introspective perspective as the view from above, the cerebral perspective as the view from below, and the behavioral perspective as the view from the side. Or is it just me?

[2] This paper follows on from my earlier paper “A Triple Aspect Theory” with some changes of emphasis. I now see the functional aspect as more indispensable than I did.