From "Philosophy of Mind: Classical and Contemporary Readings" ed. David Chalmers (Oxford: Oxford University Press, 2002)
There is a family of seven arguments advanced by John Searle urging the ontologically distinct and physically irreducible nature of conscious phenomena. These are joined by three arguments from Frank Jackson and David Chalmers which tend to the same conclusion. My aim in what follows is to construct systematic and unitary analogs of all ten arguments, analogs that support a parallel family of antireductive conclusions about the nature of light. Since those analogous conclusions are already known to be false in the case of light (its physicalist reduction is one of the many triumphs of electromagnetic theory), it becomes problematic whether the integrity of the original family of antireductionist arguments is any greater than the purely specious integrity of their deliberately constructed analogs.
I. A Searle-like Family of Arguments Concerning the Nature of Light
(A) A fundamental distinction:
original (intrinsic) visibility versus derivative (secondary) visibility
Only light itself has original visibility; for light alone is visible, when directed into the eyes, without the causal intervention of any mediating agent. By contrast, any physical object, physical configuration, or physical event is visible only when and only because light is somehow reflected from or emitted by that object, configuration, or event. Such physical items have at most derivative visibility, because they are utterly and forever invisible, save as they interact appropriately with the one thing that has original visibility, namely, light itself.
These conclusions reflect the obvious fact that, if the universe contained no light at all, then absolutely nothing would be visible, neither intrinsically nor derivatively.1
(B) The original visibility of light marks it off as belonging to a unique ontological category, distinct in its essential nature from the essential nature of any physical phenomenon, which must always lack original visibility. In other words, for any physical object, configuration, or event, it is always a contingent matter whether or not it happens to be visible on this occasion (it is a matter of whether or not it happens somehow to be illuminated). By contrast, light itself is always and essentially visible. The ontology of light is an ontology of things and features that are uniquely accessible from the visual point of view.
This means that the phenomenon of light must be irreducible to any complex of purely physical or not-essentially-visible phenomena. You simply cannot get original visibility from things that have, at most, derivative visibility.2
(C) The consequence just reached is denied by a celebrated research program called Strong EM. This program claims not only that light can be "instructively simulated" by the behavior of interacting electric and magnetic fields (to which all may agree); it makes the stronger claim that light is actually identical with electromagnetic (EM) waves. The folly of Strong EM can be seen in the following obviously sound argument.
1. Electricity and magnetism are physical forces.
2. The essential nature of light is original visibility.
3. Physical forces, no matter how they are deployed, are neither identical with, nor sufficient for, original visibility.
Therefore,
4. Electricity and magnetism are neither identical with, nor sufficient for, light.
Premises (1) and (2) are obvious. That premise (3) is obvious can be seen by the following thought experiment. According to EM theory, an oscillating magnet or charged particle will generate an expanding sphere of oscillating EM fields: an EM wavefront. And by the same theory, this is strictly sufficient for the existence of light. But imagine a man in a pitch-black room who begins to pump a bar magnet back and forth. Clearly, it will do nothing to illuminate the room. The room will remain wholly devoid of light.3
(D) The ontologically distinct nature of light is further reflected in the fact that the distinction between (visual) appearance and reality, which holds for any broadly physical phenomenon, cannot be drawn in the case of light itself. It there disappears. For while light is an agent that typically represents the physical objects, configurations, or events from which it has been differentially reflected or emitted, light does not represent itself. It is neither reflected nor emitted from itself. It thus cannot possibly misrepresent itself, as it may occasionally misrepresent things other than itself from which it has been reflected or emitted. Accordingly, where the reality at issue is light itself (as opposed to any and all physical phenomena), the appearance just is the reality.4
(E) The irreducibility here claimed can be further seen as follows. Suppose we tried to say that the redness or blueness of light was nothing but a specific wavelength of EM waves. Well, if we tried such an ontological reduction, the essential features of the light would be left out. No description of the extrinsic wavelengths of EM waves could possibly convey the intrinsic character of (objective) visible redness and visible blueness, for the simple reason that the visible properties of light are distinct from the physical properties of EM waves. This argument is ludicrously simple and quite decisive.5
(F) Light is always and necessarily visible: there can be no such thing as invisible light. Granted, not all light is visible at any given time or place: light can be "shallowly" invisible to me simply because its path does not lead into my eyes. But if light exists at all, then there is some perspective from which it will be directly visible. Let us call this the connection principle, since it unites (i) being light and (ii) being accessible-from-the-visual-point-of-view.6
(G) Considerations (A)-(F) indicate that light is a phenomenon that is ontologically distinct from and irreducible to any purely physical phenomena. And yet, while nonphysical in itself, light is plainly caused by certain special physical phenomena, such as very high temperatures or the electrical stimulation of gases. Let us call our position here nonreductive physical naturalism: it holds that light is a natural (but irreducible) phenomenon caused to occur within certain special kinds of physical systemsspecifically, within self-luminous objects, such as the sun, fires, and incandescent filaments. The aim of a scientific account of light should be to explain how such a nonphysical phenomenon is caused to occur within such highly special physical systems as stars and light bulbs.7
II. Three Jackson/Chalmers-like Arguments Concerning the Nature of Light
(H) In the study of the nature of light, there is a distinction to be drawn between the "easy" problems and "the hard problem." The first class concerns such problems as the emission, propagation, and absorption of light, its reflection and refraction, its velocity, its carrying energy, its self-interference, and so forth. These are all causal, relational, functional, and in general extrinsic features of light, features variously accessible by a wide variety of physical instruments and techniques; and it may well be that someday they will all be satisfactorily explained in terms of, for example, the propagation and interactions of EM fields.
But there remains a highly special intrinsic feature of light whose explanation must be found along some other path. This intrinsic feature is luminance, and it is what is responsible for the "original visibility" that is unique to light. Unlike all of the extrinsic (that is, physical) features of light listed above, luminance is unique in being epistemically accessible only from "the visual point of view."8
(1) We can illustrate and reinforce the contrast just drawn with a thought experiment about a physicist named Mary who is completely blind, but comes to know everything physical there is to know about EM waves, about their internal structure and their causal behavior. And yet, because she is blind and thus has no access at all to "the visual point of view," she cannot know about, she must remain ignorant of, the special intrinsic feature of light-luminance-which is accessible from that point of view alone. Evidently, even complete knowledge of the physical facts must still leave her ignorant of the nature of luminance. Luminance must therefore be, in some way, nonphysical.9
(J) As just illustrated, any possible physicalist story about the structure and causal functions of EM waves must still leave open an "explanatory gap" between the physical processes and luminance. In particular, it leaves unanswered the following question: Why should mutually-inducing electric and magnetic fields (for example) oscillating at a million billion Hertz and propagating at 300,000 km/sec ever give rise to the intrinsic feature of luminance? After all, we can easily imagine a universe that is filled with oscillating EM fields propagating back and forth all over the place, a universe that is nonetheless utterly dark, because it is devoid of the additional feature of luminance. We need to know how, when, and why oscillating EM fields cause the ontologically distinct feature of intrinsic luminance. Until we understand that mysterious causal relation, we shall never understand the ground and real nature of light.10
III. Critical Commentary
Concerning (A). As an exercise in term introduction ("original" visibility, and so on), this is strictly harmless, perhaps. But it falsely elevates an extremely peripheral feature of light-namely, its capacity to stimulate the idiosyncratic rods and cones of terrestrial animals-into a deep and presumptively defining feature of light. This is thrice problematic. First, it is arbitrarily selective. Second, it is strictly false that only light will stimulate rods and cones (charged particles of suitable energy will also do it, though at some cost to the retina). And third, infrared and ultraviolet light is quite invisible to terrestrial eyes. Our eyes evolved to exploit a narrow window of EM transparency in the earth's idiosyncratic atmosphere and oceans. Nothing of ontological importance need correspond to what makes our rods and cones sing.
Concerning (B). The dubious distinction legislated in (A) is here deployed to consign all physical phenomena to a class (things with merely derivative visibility) that excludes the phenomenon of light. This division certainly appeals to our default stereotype of a physical object (a tree, or a stone, has merely derivative visibility), but it begs the question against the research program of physicalism, because some unfamiliar physical things may indeed have original visibility, our common-sense expectations notwithstanding. As it turns out, EM waves with a wavelength between .4 and .7 gm are capable of stimulating the retina all by themselves, and thus have original visibility as defined in (A). The argument of (B) is thus a question-begging exploitation of superficial stereotypes and EM ignorance.
Concerning (C). The crucial premise of this argument (premise 3) may seem highly plausible to those who have a common-sense prototype of forces and who are ignorant of the details of EM theory, but it plainly begs the central question against physicalism. (Premise 3 is the direct denial of the basic physicalist claim.) Moreover, it is false. As mentioned in the preceding paragraph, EM waves of suitable wavelength are sufficient for original visibility. The "Luminous Room" thought experiment, concerning the oscillating bar magnet in the pitchblack parlor, is designed specifically to make premise 3 plausible, but that prejudicial story illegitimately exploits the fact that some forms of EM radiation have wavelengths that are simply too long to interact effectively with the rods and cones of terrestrial retinas. The darkened parlor may look to be devoid of light, but, thanks to the oscillating magnet, a very weak form of light is there regardless.
Concerning (D). While superficially plausible, perhaps, this argument refuses to take into account the many ways in which we can be mistaken or misled about the character of the light entering our eyes (for example, the light from a cinema screen appears continuous, but is really discontinuous at 36 frames/sec; the light of an incandescent automobile headlight, while really yellowish, looks white at night; and so forth). Its brief plausibility is a reflection of nothing more than our unfamiliarity with how light is perceptually apprehended and with how that intricate process can occasionally produce false perceptual beliefs. It is a reflection of our own ignorance, rather than of any unique ontological status had by light.
Concerning (E). This argument is sheer question-begging assertion rather than instructive argument. Whether objective properties of light such as spectral redness or spectral blueness are identical with, or distinct from, specific wavelengths of EM radiation is precisely what is at issue. And in this case, it has been plain for a century that these properties are identical. It is also plain that spectral redness, spectral blueness, and their various causal properties-their refractive and absorptive behavior, their velocity and interference effects-are positively explained, rather than impotently "left out," by their smooth reduction to EM features.
The point about what an EM vocabulary can or cannot "convey" about certain perceptual properties is a distinct point (and a red herring) to be dealt with below in "Concerning (I)."
Concerning (F). This argument also would be found plausible by someone still imprisoned by prescientific prototypes of light. Invisible light may well be a conceptual impossibility against the assumptions of the story just told, but we now know better. Indeed, we have learned that most light is invisible-and not just "shallowly" invisible, but permanently beyond human visual apprehension. Once again, we find ignorance being paraded as positive knowledge.
Concerning (G). This summary attempts to find a proper place in nature for the phenomenon touted as ontologically distinct and physically irreducible in arguments (A)-(F). The place suggested is that of a nonphysical causal consequence of certain special but purely physical events.
Such a move threatens to violate well-established laws concerning the conservation of both energy and momentum, at least if light is presumed to have any causal powers of its own. But we need not enter into these matters here, for as the critical commentary to this point shows, there is no significant motivation for any such antireductionist research program in the first place. And in the second place, the proper place in nature of light has already been made clear: it has been smoothly and systematically reduced to EM waves.
Concerning (H). Light is here conceded to have a wide variety of physical features-its so-called "extrinsic" or "structural/functional" features-to which some sort of physical explanation is deemed appropriate. But light is also assigned an allegedly special or "intrinsic" feature, a feature that is epistemically accessible through vision, but not through the "structural/function-al" stories to which current physical science (alas) is limited.
Once again, our prescientific noninferential epistemic access (namely, vision) to certain entirely physical properties is portrayed as a unique window onto an ontologically special domain. And to compound the felony further, the potential reach of physical explanation is restricted, by arbitrary fiat at the outset (rather than by any empirical failures revealed during the course of ongoing research), so as inevitably to fall short of the so-called "intrinsic" features within the "special" domain at issue.
The "hard problem" is thus made transcendently hard at the outset by presumptive and question-begging fiat, rather than by any substantive considerations. As EM theory has taught us, there is no "hard problem" here at all, and no defensible ontological distinction between intrinsic and extrinsic features. "Luminance"-if we concede the integrity of this notion at all-is just the normal and entirely physical capacity of EM waves to excite our own rods and cones (and to induce chemical changes in photographic film, to free electrons in a television camera, and so forth).
Concerning (I). This "knowledge" argument equivocates on `knows about.' It elevates two distinct modes of epistemic access to light into a false dichotomy of distinct phenomena thereby accessed-physical features by scientific description, and a special range of nonphysical features by normal human vision. But for light, at least, we know perfectly well that there is only one thing here rather than two, only one class of objective features rather than two.
What Blind Mary is missing is one common form of knowledge about light: she lacks perceptual/discriminative knowledge of light. And yet, people who have such knowledge are accessing the very same features of reality that she is obliged to access in other ways. The difference lies in the manner of the knowing, not in the nature of the thing(s) known. It is true that no amount of propositional knowledge of light will ever constitute the visual apprehension of light, but that is entirely to be expected. They are different forms of knowledge; they operate with different representational "palettes" inside Mary's brain. But they both represent, each in their own distinct way, one and the same entirely physical thing: light.
Our contemporary scientific knowledge about light aside, one can see immediately that the crucial divergence here is merely epistemic rather than ontological (as the argument pretends). For while it is indeed true that Blind Mary does not know what it is like to see spectral-red light, it is equally true, and for exactly the same reasons, that she does not know what it is like to see EM waves at .65 um. The deficit here evidently lies with Mary and her epistemic failings, not with EM waves and their ontological shortcomings visa-vis light. For Mary would continue to have her deficit even if light were (as it is) identical with EM waves. Her deficit, therefore, can hardly weigh against that identity.
Concerning (J). This "open question" argument begs the question in favor of the ontological distinctness of "luminance," and then insists on our providing a causal account of how EM waves might produce it. This gets everything backward. We no longer have need for an account of how EM waves might "cause" the various phenomena associated with light, because the systematic reconstruction of optical phenomena within EM theory leads us to believe that light is simply identical with EM waves, and that the assembled properties of light are identical with, rather than caused by, the corresponding properties of EM waves.
The conceivability of a dark universe filled with EM waves shows only that the various cross-theoretic identities motivated by the EM reduction are, as they should be, contingent rather than necessary identities. It should also be pointed out that such an "open question" argument will be maximally appealing to one who is minimally instructed in EM theory. This is because the more one learns about EM waves, about their effects on matter in general and on our eyes in particular, the harder it becomes to imagine a consistent scenario in which a universe abuzz with EM waves of all wavelengths remains dark even so. Here, as in so many of the earlier arguments, the audience's presumed ignorance is once more a lubricant that smooths the path of a worthless argument.
This concludes my attempt to construct, and to deflate, a systematic analog for the family of arguments currently so influential in the philosophy of mind. My point, of course, is that the family of arguments on which they are modeled is just as empty of real virtue.
IV. A Final Nagel/Searle Argument for Irreducibility
A question will inevitably arise over the fairness of the global analogy deployed above. In particular, it will be complained that the global analogy is faulty in placing the objective properties of "original visibility" and "luminance" in the role played by the subjective properties of original intentionality and inner qualia in the arguments under attack.
The analogy deployed does indeed proceed in precisely this fashion, but this assimilation is the central point of the exercise. It should at least give us pause that the original family of arguments can be collectively and successfully mirrored in a ten-dimensional analogy that deliberately and self-consciously concerns "objective" features. After all, if the analog arguments are at all compelling-and to the electromagnetically uninformed, they will be-then the essential appeal of both families of arguments presumably derives from something other than the unique status of the "subjective."
Second, there is no mystery about what drives the plausibility of the analog arguments. It is the ignorance-fueled appeal of the idea that the epistemic modality of vision is or might be a unique window onto an ontologically distinct class of properties. But in the case of light it is also plain, at least in retrospect, that nothing substantive motivates that repeated insistence. We have to wonder if the same failure might be true of the original family of arguments. After all, and whatever else it might be, introspection is an epistemic modality, or perhaps a family of them. And while it may have its own quirks and distinguishing profile, it is entirely unclear whether it, alone among all of our epistemic modalities, constitutes a window onto a unique ontological domain of nonphysical properties. None of our other epistemic modalities has any such distinction: they all access some aspect or other of the purely physical world. Why should introspection be any different?
Searle has a further argument, unaddressed to this point, whose burden is to illustrate the ontological cleft he sees between the domain of "outer sense" and the domain of "inner sense," as Immanuel Kant called them. Searle's argument here appeals, uncharacteristically, to the history of science. The argument originally appeared, very briefly, in Thomas Nagel,11 but Searle has more recently developed it in detail.
Premise (1). We must draw a distinction between the real and objective properties of objects and the contingent subjective effects those properties happen to have on the conscious processes of humans. For example, objective heat (molecular KE) is one thing; the subjective feeling of warmth in humans produced by objective heat is quite another.
Premise (2). The scientific reduction of observable phenomena typically ignores or "carves off' their contingent subjective effects on the conscious processes of humans, and reduces only the nonsubjective aspects of the phenomena. (For example, kinetic theory successfully reduces objective heat to molecular KE, but leaves its subjective conscious effects on humans aside. EM theory successfully reduces objective spectral colors to different wavelengths of EM radiation, but leaves their subjective conscious effects on humans aside. And so forth.)
Premise (3). When we attempt to provide a physicalistic reduction of those subjective conscious effects themselves, we must realize that here we cannot "carve off' their subjectiveeffects-on-us from their objective properties, and reduce only the latter, because it is precisely those subjective-effects-on-us which we wish to understand. Here, inside the mind, there is no longer any meaningful or defensible distinction between the "objective" and the "subjective" which would allow us to repeat the pattern of reduction described above. The subjective phenomena are exclusively and essentially subjective. Any alleged "reduction" would simply leave out what is essential to their nature.
Therefore, mental phenomena are irreducible to physical phenomena. The proper pattern of a physicalist reduction (an "objective"-to-"objective" mapping) uniquely precludes any reduction of the subjective.12
What is going on here? Simply this. The Nagel/Searle argument treats a contingent, minor, and remediable feature (of a handful of historical examples of reductions) as if it were a necessary, central, and permanent feature of any possible physicalistic reduction. Specifically, the merely contingent feature that is paraded as essential is the feature: leaves aside the-effectson-human-consciousness (the "C-effects," for short). The argument then points out that this "essential" feature of physicalistic reduction precludes any such reduction in the unique case of C-effects themselves, since "leaving the C-effects aside" is here not an option.
It is indeed true that historical property reductions pay little or no attention to, or provide us with little or no insight into, the C-effects of the various phenomena being reduced. Searle and Nagel seem antecedently convinced that this historical fact is the inevitable reflection of an ontological gulf already fixed between "objective" phenomena and "subjective" phenomena.
That is one (distant) possibility. But there is an obvious alternative explanation of why physicalistic reductions so regularly leave out any account of the human C-effects of the phenomena being reduced, as the historical reduction of heat to molecular energy made no attempt to account for the subjective sensation of warmth, or as the historical reduction of light to EM waves made no attempt to account for the subjective sensation of redness.
The obvious alternative explanation is that such C-effects are the proper province of a distinct science, a science such as cognitive neurobiology or computational neuroscience. Searle is wrongly demanding that the kinetic theory of heat do, all by itself, something that clearly requires, in addition, an adequate theory of the brain. The fact is, during the late nineteenth century, we were too ignorant about neurobiology for the kinetic theory to suggest any worthwhile hypotheses about the human C-effects of molecular energy. It is no surprise, then, that physicists simply walked past that arcane problem, if it ever occurred to them to address it in the first place. The same is true for the EM theory of light and the problem of our subjective sensations of redness.
Accordingly, this incidental "leaving-aside" need have no metaphysical or ontological significance. This deflationary view is further encouraged by the fact that physicalistic reductions such as the kinetic theory also "leave aside" any explanatory account of millions of other phenomena, so there is no automatic reason to find any special significance in its ignoring of human C-effects in particular. If I may give several examples, historical reductions of heat typically leave aside any attempt to account for:
heat's effect on Antarctic anchovy production heat's effect on bluebird-egg cholesterol levels
heat's effect on pneumonial infections
heat's effect on the Gross National Product of Peru
heat's effect on the rotting of vegetable matter
heat's effect on the conscious states of humans
(this list is extendible indefinitely)
The great reductions of classical and modern physics typically leave out any account of heat's (or light's, or sound's) effect on all of these things, and of millions more, because no reduction all by itself can presume to account for the ever-more-distant causal effects of its proprietary phenomena as they are progressively articulated into all possible causal domains. There are far too many domains, and causal understanding of the phenomena within those other domains will typically require the resources of further theories in addition to the theory that achieves the local reduction at issue.
It is in no way noteworthy or ontologically significant, then, that the kinetic theory of heat, all by itself, provides no account of any of the arcane phenomena listed above, nor of millions of others as well. In particular, it is neither noteworthy nor ontologically significant that the kinetic theory of heat provides no account of the human conscious response to heat. This marginal and idiosyncratic phenomenon has no more ontological significance than any of the other arcane phenomena just listed. And they all require the resources of theories beyond the kinetic theory of heat to address them adequately. Specifically, heat's effect on
anchovy production needs ecology egg cholesterol levels needs metabolic chemistry
pneumonia needs immunology and bacteriology
Peru's Gross National Product needs biology and economics
vegetable rotting needs bacteriology and cell chemistry
human conscious experience needs cognitive neurobiology
My counterclaim, then, against Nagel and Searle, is that it is not an essential feature of physicalistic reductions that they always "leave aside" human C-effects, or any of the many other effects cited. It is a merely contingent and wholly explicable fact that historical reductions have so far done so. It is not an essential pattern that all physicalistic reductions are doomed-bynature to follow; nor is it a self-imposed definitional stipulation on what counts as a reduction, as Searle13 at one point inexplicably suggests. Once we begin to address human C-effects with some appropriately focused science-as neuronal vector-coding theories are already doing, with striking success14-then that earlier "pattern" will be well and truly broken. For that pattern reflected only our own scientific ignorance, not some ontological division in nature.
In sum, human conscious experience has no quicksilver history of darting off to one side each time our reductive scientific thumb has tried to pin it down. There have been no significant reductive attempts at that target, not, at least, within the grand historical reductions of physics and chemistry. Instead, the phenomena of human conscious experience have quite properly been waiting, patiently and at the sidelines, for the maturation of the only theory that has any realistic hope of providing such a reductive account, namely, an adequate theory of the brain. If and when that approach has been fully tried, and proves a failure, then, perhaps, it will be time to insist on nonphysical approaches.
Both the appeal to ignorance and the question-begging nature of the Nagel-Searle argument become finally vivid if one plays at constructing a series of parallel arguments to "establish" the physicalistic irreducibility of whatever arcane, complex, and puzzling phenomenon one might choose to consider (something from the preceding list, for example). Simply note that historical reductions of various important phenomena have invariably left that particular phenomenon aside as an unaddressed mystery; pretend that this is an essential pattern, a reflection of an antecedent metaphysical division, or the result of some appropriately exclusive definition of "reduction"; note that said leave-aside pattern (surprise!) precludes any similar reduction of exactly the phenomenon at issue; and you are home free. You will then have performed for us the same empty service that Nagel and Searle have performed.
V Some Diagnostic Remarks on Qualia
There is a chronic temptation among philosophers to assign a special epistemological, semantical, or ontological status to those features or properties which form the "discriminational simples" within each of our several sensory or epistemic modalities, such as brightness and colors in the case of vision, sweetness and sourness in the case of taste, and so on. These are the features of the world where one is unable to say how it is that one discriminates one such feature from another; one simply can. As well, one is unable to say how the meaning of `red' differs from the meaning of `green'; one simply has to point to appropriate exemplars.
Such discriminational simples are typically contrasted with properties, such as "being a horse," where one can usually articulate the more elemental constituting features that make up the type in question: size, shapes, configuration, color, texture, and so forth, which more elemental features lead us stepwise back toward the discriminational simples.
Too much has been made of these "simples," for the existence of such discriminable but inarticulable features is entirely inevitable. Such features must exist, if only to prevent an infinite regress of features discriminated by constituting subfeatures discriminated by constituting subsubfeatures, and so on.15 And their existence is inevitable even on wholly physicalist conceptions of cognition. It simply cannot be the case that all conscious feature discriminations are made on the basis of distinct conscious (sub) feature discriminations. Given any person at any time, there must be some set of features whose spontaneous or noninferential discrimination is currently basic for that person, a set of features whose discrimination does not depend on the conscious discrimination of any more elemental perceptual features. In short, there must be something that counts, for that person, as a set of inarticulable qualia.
Accordingly, we should not be tempted to find anything physically irreducible or ontologically special about such inarticulable features. They need reflect nothing more than the current and perhaps changeable limits of the person's capacity for epistemic and semantic articulation, the current limits, that is, of the person's knowledge of the world's fine structure and his own epistemic access to it. Most importantly, there is no reason to expect that the current limits of the typical person's knowledge must mark the boundary of a distinct ontological domain. This is just as true, note, for the epistemic modalities that underwrite (what we loosely call) "introspection" as it is for the epistemic modalities of vision, taste, and audition.
And yet, philosophers have regularly been tempted here, some beyond redemption. Bishop Berkeley rejected the identification of sound with atmospheric compression waves; William Blake and Johann Wolfgang Goethe rejected the identification of light with Isaac Newton's ballistic particles; and Nagel, Jackson, Searle, and Chalmers reject the proposed reduction of inner qualia to physical states of the brain.
There is an important factor here that may help to explain why such features have so frequently been held to be beyond the reach of any physicalist reduction. Specifically, any reduction succeeds by reconstructing, within the resources of the new theory, the antecedently known nature, structure, and causal properties of the target phenomena. That is what intertheoretic reduction is. But if the target phenomena, such as sensory qualia, are features whose internal structure (if any) we are currently unable to articulate, and whose causal properties (if any) are largely unknown to us, then the target phenomena will inevitably seem to offer the minimum purchase possible for any aspirant reducing theory. They will display no structure worth reconstructing. They will present themselves as smooth-walled mystery. They will appear to be irreducible to any "structural/functional" theory from conventional science.
But the appearance of seamless simplicity need reflect nothing more than our own ignorance, an ignorance, we should note, that already holds promise of repair. In sum, we should not be too quickly impressed by qualia, whether outer or inner. If cognitive creatures exist at all, then the existence of inarticulable qualia is inevitable, even in a purely physical universe.
If ultimately they are physical, then inner qualia ought to be epistemically accessible from more than just the first-person or "subjective" point of view; they ought to be accessible as well from one or more "objective" points of view, via some appropriate instruments that scan brain activity, for example. Some will continue to find this implausible on its face. That is mainly because the terms ,objective' and `subjective' are commonly used in mutually exclusive contrast. But the default implication of mutual exclusivity may well be inappropriate in precisely the case at issue. After all, we know that the two epistemic modalities of vision and touch, for example, are not mutually exclusive in the phenomena that they access-one can both see and feel the shape of an object, see and feel that the sun is out, see and feel that rain is falling, and so forth. Why should it be impossible a priori that the epistemic modality we call "introspection" have some similar overlap with one or more of our other epistemic modalities?
Indeed, such overlap appears actual, even by the standards of common sense. One can tell by introspection that one's own bladder is full, but an ultrasound image will tell anyone the same thing. One can tell by introspection that and where one's retinal cells are photo-fatigued (we call it an "after image"), but that too is accessible by nonsubjective means. One can tell by introspection that the cochlear cells of one's inner ear are firing randomly (the condition is called "tinnitus"), but others can access their behavior instrumentally. There are, of course, thousands more such examples.
It would seem, then, that the "subjective" and the "objective" are not mutually exclusive after all. In at least some cases, one and the same (physical) state can be known both subjectively and objectively, from both the first-person perspective and the third-person perspective. Further, it would seem that the extent and location of the overlap is somewhat fluid, and that it varies as a function of how much background knowledge, conceptual sophistication, and recognitional skill the person has acquired. The process is called "coming to understand explicitly what was hitherto inarticulate," and it is entirely to be encouraged. The more epistemic modalities we can bring to bear on any puzzling phenomenon, the deeper our understanding will become. To insist, in advance of real understanding, that a given phenomenon is locked forever within its own epistemic box serves only to block the very research that might dissolve such a prejudicial conception.
VI. A Final Point about Light
In closing, let me return to the opening family of arguments concerning the irreducibility of light. Someone may remark that, with light, I have used an example that is antithetical to my own reductive inclinations in the philosophy of mind. For while light reduces cleanly to EM waves, light is still famous for having escaped the various mechanical reductions (ballistic particle theories, elastic media theories) that everyone in the nineteenth century expected. And it is still famous for having thus emerged as one incarnation of a fundamental and nonmechanical aspect of reality: electromagnetism.
This is quite true, and more than a little instructive. But in the present context it is also instructive (1) that while nonmechanical, light remains an entirely physical phenomenon, and, (2) more importantly, that the modestly special status that light eventually discovered had absolutely nothing to do with any of the considerations urged in the family of antireductive arguments in my opening parody. Light's nonmechanical status emerged primarily as a consequence of Special Relativity, as a consequence of the unity of space-time and the impossibility of a universal elastic aether. It was not a consequence or reflection of any of the arguments offered above. It is ironic that, even though light did turn out, unexpectedly, to be a rather special kind of physical phenomenon, the parodyarguments (A)-(J) did nothing whatever to herald it, and they are, after the fact, quite irrelevant to it.
The parallel lesson about mental states is that, even if conscious phenomena are ontologically special in some way, roughly analogous to the case of light, there is no reason to think that the arguments of Searle, Jackson, and Chalmers do anything to illustrate or establish it. Those arguments are no more instructive about the ultimate nature of mental phenomena than arguments (A)-(J) are instructive about the ultimate nature of light.
Notes
- Cf. Searle, "Intrinsic Intentionality: Reply to Criticisms of Minds, Brains, and Programs," Behavioral and Brain Sciences, III (1980): 450-56, and The Reddiscovery of the Mind (Cambridge: MIT, 1992), pp. 78-82.
- Cf. Searle, "Is the Brain's Mind a Computer Program?" Scientific American, CCLXII, 1 (January 10.1990): 26-31, here pp. 26-27; and The Rediscovery of Mind, pp. 93-95.
- Cf. Searle, "Intrinsic Intentionality," pp. 417-57,11.and "Is the Brain's Mind a Computer Program?" pp. 26-31. This analogy was earlier deployed in Paul M. Churchland and Patricia Churchland, "Could a Machine Think?" in Scientific American, CCLXII, 113. (January 1990): 32-37. We did not then appreciate that the analogy was a member of a systematic and much larger family.
- Cf. Searle, The Rediscovery of Mind, p. 122, and "The Mystery of Consciousness: Part It," The New York Review of Books, XLII, 18 (November 16, 1995): 54-61, here p. 58.
- Cf. The Rediscovery of Mind., pp. 117-18.
- Cf. ibid., pp. 132, 151-56.
- Cf. ibid, pp. 1, 89-93, 124-26; also Searle, "The Mystery of Consciousness: Part II," pp. 55-56.
- Cf. Chalmers, "The Puzzle of Conscious Experience," Scientific American, CCLXXIII, 6 (December 1995): 80-86, here pp. 81-82.
- Cf. Jackson, "Epiphenomenal Qualia" Philosophical Quarterly, XXXII, 127 (1982): 127-36, here p. 130; and Chalmers, "The Puzzle of Conscious Experience," pp. 81-82.
- Cf. Chalmers, The Puzzle of Conscious Experience," pp. 82-83; also Searle, "The Mystery of Consciousness: Part II," pp. 55-56.
- "What Is It Like to Be a Bat?" Philosophical Review, LXXXIII, 4 (1974): 435-50.
- Cf. ibid. p. 437; also Searle, The Rediscovery of Mind, pp. 116-24.
- The Rediscovery of Mind, pp. 124, 112-16. Though here is not the place to mount a systematic criticism, it must be said that Searle's 1992 sketch of the nature and varieties of reduction muddies far more than it clarifies. First, it wrongly assimilates ontological reduction to ontological elimination. Second, there simply is no further category or "half-way house"Searle's so-called "causal reduction"-distinct from ontological reduction. And third, as we just saw, the account attempts to stipulate the closure of certain empirically open questions. To a neutral philosopher of science, Searle's account will appear more as areflection of his peculiar intuitions in the philosophy of mind rather than as an independently motivated attempt to account for the full range of cases throughout the history of science.
- See Austen Clark, Sensory Qualities (New York: Oxford, 1994).
- See Mary Hesse, "Is There an Independent Observation Language?" in Robert Colodny, ed., The Nature and Function of Scientific Theories (Pittsburgh: University Press, 1970), pp. 35-77.