Back to CQ Homepage

Beyond Eliminative Materialism: Some Unnoticed Implications of Churchland's Pragmatic Pluralism

by Teed Rockwell

2419A Tenth Street

Berkeley, CA 94710

510/548 8779


(Revised October 4, 1998)


Paul Churchland's epistemology contains a tension between two positions, which I will call pragmatic pluralism and eliminative materialism. Pragmatic pluralism became predominant as Churchland's epistemology became more neurocomputationally inspired, which saved him from the skepticism implicit in certain passages of the theory of reduction he outlined in Scientific Realism and the Plasticity of Mind. However, once he replaces eliminativism with a neurologically inspired pragmatic pluralism, Churchland 1) cannot claim that folk psychology might be a false theory, in any significant sense 2) cannot claim that the concepts of Folk psychology might be empty of extension and lack reference. 3) cannot sustain Churchland's criticism of Dennett's "intentional stance" . 4) cannot claim to be a form of scientific realism, in the sense of believing that what science describes is somehow realer that what other conceptual systems describe.


One of the worst aspects of specialization in Philosophy and the Sciences is that it often inhibits people from asking the questions that could dissolve long standing controversies. This paper will deal with one of these controversies: Churchland's proposal that folk psychology is a theory that might be false. Even though one of Churchland's greatest contributions to philosophy of mind was demonstrating that the issues in philosophy of mind were a subspecies of scientific reduction, still philosophers of psychology have usually defended or critiqued folk psychology without attempting to carefully analyze Churchland's theory of reduction. This is a serious mistake, for Churchland's theory of reduction, properly understood and purged of certain inconsistencies, is simply not capable of unseating folk psychology with the decisiveness that delights Churchland and frightens his adversaries. Because neither side is aware of this, the battle rages on.

There is a tension between two positions in Churchland's theory of reduction. One of these positions (Which I will call Eliminativism) leads him inevitably to universal skepticism, and is also the basis for his radical dismissal of folk psychology. The second position (Which I will call pragmatic pluralism) saves him from skepticism, but if he adopts it, he 1) cannot claim that folk psychology might be a false theory, in any significant sense 2) cannot claim that the concepts of folk psychology might be empty of extension and lack reference. 3) cannot sustain the criticism of Dennett's "intentional stance" that Churchland outlined in the essay "Nailing folk psychology to its perch" (reprinted in his 1989).

In which we encounter Churchland's conception of reduction as elimination, and how it led him into universal skepticism

Traditionally, reduction was supposed to preserve the truth of one theory within the context of another theory by means of what were called Bridge Laws. Bridge Laws were supposed to set up identities between old scientific concepts and new ones. Because it was assumed that science progressed by new theories building on the foundations established by old ones, bridge laws were supposedly needed to establish reductive unity between the new and the old. Thanks, however, to research in the history of science done by Kuhn, Feyerabend, Laudan, and others, we now know that this kind of continuity between a reduced and a reducing theory usually does not exist. Science usually progresses by revolutionary jumps, which make bridge laws impossible. Awareness of this fact required a new theory of reduction, which Paul Churchland provided in Scientific Reason and the Plasticity of Mind (SRPM.)

For Churchland, the essential goal of a reduction is the elimination of the old theory. The relationship between old scientific theories and new ones is thus seen as essentially the same as that between science and superstition: The old theory is shown to be false by the new one, in the same sense that science falsified existence claims for demons and witches. However, this dismissal is not accomplished by merely an imperious wave of the hand. There does need to be a relationship between the two theories, established by correspondence rules. But these correspondences need not be identities, they can even be (and frequently are) contradictions.

" . . . We find a mapping of one vocabulary onto another, a mapping that preserves certain features thought to be important. .. But the pairing effected therein standardly fail to preserve meaning (SRPM p.81)".

"A reduction . . . provides the basic instructions, as it were, for the orderly displacement of the [old theory] by the [new theory]." ( p.81).

"The correspondence rule pairings need not be construed as identity claims , nor even as material equivalencies. . . We. . . need only the minimal assumption that the second element of each pair truly applies where and whenever the first element is normally thought to apply"(p.83).

An example that I think illustrates vividly how one can use contradiction pairings, rather than identities, to perform something like a Churchland style reduction, is found in the old joke which begins this way:

Aren't you Bob's cousin George, who made a million dollars by investing in oil?

Because the interlocutor in this joke has a false belief, one could dismiss his question thusly, using an elimination without reductive correspondence rule pairings.

No, I'm not!

And then we would have no reduction (and no joke.) But ifi we were to answer this question with the precision required by a scientific reduction, our answer would look something like this:

Not exactly.

A)My name isn't George---- B) it's Jim,

A)his name isn't Bob---- B) it's Bill,

A)he's not my cousin---- B)he's my nephew

A) it wasn't a million dollars,---- B)it was $200,000,

A)it wasn't in oil,---- B)it was in steel

A)and I didn't make it,---- B)I lost it.

We now have what Churchland would call an "equipotent image" of set "A" outlined in set "B", because each of the ordered pairs shows exactly where the mistaken theory was thought to apply, and how it was mistaken. But as you can also see from comparing the members of each pair, these ordered pairs are all contradictions, and there is no need for any identities anywhere.

But don't reductions sometimes produce identities? Churchland is fond of saying that lightis electromagnetic waves or that temperature is mean molecular kinetic energy. Here is where the tensions in Churchland's thinking becomes most apparent. The main thrust of Churchland's argument is that these are only "Contingent Identities" (SRPM P. 82). He says about even these paradigm cases "Synonym pairs these are not. (ibid. p.81)" Nevertheless, they may still be referred to as identities in common and scientific speech, even though the ordinary idiom has technically been eliminated.

familiarity, entrenchment, convenience, and continuity may together counsel a less puritan course. . . . Why be grudging here?" (p.82).

Despite these expressions of tolerance, however, most of the time Churchland wants to claim that this mapping really performs only an anthropological function, to enable people who are used to thinking in the old system to make the transfer to the new system. The new theory in and of itself has no need of the mapping function. Anyone who has grown up with only the new theory would in principle have no need to refer to either the old theory or the reductive connections.

We cannot insist that {a theory} be able to predict how this, that or the other conceptually idiosyncratic culture is going to conceive of that domain. That would be to insist that the new theory do predictive cultural anthropology . . . The demand that molecular theory directly entail our thermal or color concepts is evidently this same unreasonable demand (Italics in original)(NP p.52)

Cross-theoretical identity claims are not part of the reduction proper, and they are not essential to the function it performs."(p.83).

Passages like this reveal Churchland the Eliminativist. They seem to imply that talk of identities between old and new theories is a kind of forgivable carelessness, rather like saying that the sun rises when in fact the earth moves. Strictly speaking, the old theory is dead, the new theory has replaced it, and the fact that the old theory occasionally got a few things right is about as significant as the fact that a broken clock is right twice a day.

However, the claim that reduction is totally eliminative has dangerously skeptical implications. If each theory is falsified by whatever theory replaces it, that next new theory would also be proven false by whatever theory replaces it, and so on ad infinitum. This process can only produce a unified true science if it eventually ends with the discovery of the one true theory. Churchland, however says, I think correctly, that this is very unlikely, given how science has operated in the past.

So many past theories, rightly judged excellent in their time, have since proved to be false. And their current successors, though even better founded, seem but the next step in a probably endless and not obviously convergent journey. (Churchland 1989 p.140)

It is expected that existing conceptual frameworks will eventually be replaced by new and better ones, and those in turn by frameworks better still, for who will be so brash as to assert that the feeble conceptual achievements of our adolescent species comprise an exhaustive account of anything at all?

(ibid. p.52)

So, If each theory is falsified by its successor, and every theory is succeeded by some other theory, therefore all scientific theories that ever existed or will exist are false. I think that Feyerabend succumbed to these skeptical implications of Eliminative Materialism, which is why he ended up writing books with titles like Farewell to Reason, and Against Method. Churchland, however, tempered this skepticism with a variety of evocative metaphors that for our purposes boil down to one point: the line between a true theory and a false one is not as sharp as is commonly supposed.

The Pragmatic Answer to Eliminative Skepticism

Given that the connection between old theories and new ones cannot be accomplished with identity statements, is there any other connective that can do the job? Unlike the logical empiricist's answer, which was straightforward and false, Churchland's answer is every bit as ambiguous as the phenomenon he is attempting to explain. He says that the two theories are "relevantly isomorphic", that the new theory has a "roughly equipotent image" of the old., that the older theory "is just the target of a relevantly adequate mimicry" (all from Churchland 1989 p.49). Yet despite all of this similarity it is still possible for the new theory to be true and the old theory to have "no extension whatsoever" (ibid.) How then do we get identities between two theories? Churchland's answer is somewhat equivocal.

full fledged identity statements are licensed by the comparative smoothness of the relevant reduction (i.e. the limiting assumptions are not wildly counterfactual, all or most of [the old theory's] principles find close analogs in [the new theory] etc.) . . . and thus allows the old theory to retain all or most of its ontological integrity. (ibid p.50)


Note the modest criteria for determining smoothness. Limiting assumptions can be counterfactual as long as they are not wildly so, the principles of the two theories can be different as long as they are close analogs, the old theory need not retain all of its ontological integrity as long as it retains most of it. And yet all of this similarity gets accepted as identity, for all practical purposes. Apparently the binary distinction between real and unreal, true and false, cannot be preserved if we are to avoid skepticism in a post-Kuhnian World. We cannot say, for example, that the "smoothness quotient" for the light -electromagnetic energy relationship is .87 and therefore light is electromagnetic energy , whereas the relationship between caloric and molecular motion has an "S.Q." of only .34 and therefore there is no such thing as caloric.

This continuum between true and false theories is also implied by Patricia Churchland's description of how one theory succeeds another in a post-Kuhnian view of scientific history. In this passage she expands on Paul Churchland's claim that "Reduction may be smooth or bumpy, or anywhere in between" (SRPM p.84)

Theories range themselves on a spectrum . . . some require relatively little correction in order to be reduced. . . but in other cases so much correction is needed that almost nothing save a few low-level homey generalizations can be retained. The spectrum. . . has at one end reduced theories that have been largely retained after the reduction and at the other end theories that have been largely displaced, with sundry cases falling in between. (1986 p.281)

When the smoke clears after the reduction/elimination process, how few "low-level homey generalizations" about X must we have left before we are required to say "there are no X's"? Apparently there is no straightforward answer to this question, which implies that falsifying an old theory is a somewhat ambiguous process. In one passage at least, Patricia Churchland seems almost willing to concede that an old theory is not actually falsified by the one that replaces it , but only abandoned for pragmatic reasons.

We do not identify phlogiston with oxygen . . . Presumably we could do so, inasmuch as there is no strictly formal (logical) reason not to, but there is no theoretical utility in these identifications (p.282){1}

The most decisive step towards pragmatic pluralism occurred when Paul Churchland began to work out the implications of the fact that knowledge is embodied by neural networks. A knowledgeable neural network is one that makes few errors. To say that a network is learning is to say that it is responding to its inputs "in a fashion that systematically reduces the error messages to a trickle" (1989 p.177). It thus undergoes a growth process which might be fancifully described as the network's theories become truer and truer without ever reaching "true", just as we can travel further and further without ever reaching "far".

"nothing guarantees that there exists a possible configuration of weights that would reduce the error message to zero . . . . nothing guarantees that there is only one global minimum. . . perhaps there will in general be many quite different minima, all of them equally low in error, all of them carving up the world in quite different ways. . . these considerations seem to remove the goal itself--a unique truth--as well as any sure means of getting there."(Churchland 1989 p. 194)

Thus, although there is no such thing as "the truth" from a neurocomputational point of view, there is such a thing as learning, which saves us from a Feyerabendian skepticism.{2} Learning is defined not as eliminating error, but as decreasing it. As the growth of science has made it possible for us to do more and more things with less and less error, we are obviously learning a great deal. At any point in the history of knowledge, another theory may be developed which has a much smaller error minimum than its predeceesor, and we may decide to stop using the old theory and go with the new one. But this does not mean that the old theory never had any contact with reality, and is now revealed to have resided only between the ears of the poor dupe who once used it. It makes much more sense to agree with Ruth Millikan that "Knowing must have been something that man has been doing all along" (1984 p.7), for it saves us from having to accept that knowing is something we will probably never get to do. The strongest arguments in favor of pragmatic pluralism are themselves pragmatic: We need a theory of knowledge to help us distinguish our good theories from our bad ones. An epistemology which said that all of our theories are false would be completely useless to us, even if it were "true" by its own solipsistic definition of truth.

The Effects of Pragmatic Pluralism on Eliminativism

If we follow Churchland's suggestion and remove the goal of a unique truth ---and I think we should- we also have to remove it's necessarily implied opposite: an unconditional falsehood which has no extension and no reference to reality at all. And Churchland has not done this: Even as late as his 1992, he reiterates that eliminative materialism is based on "the worry that . . . folk psychology constitute[s] a radically false account of the cognitive activity of humans." (p.420). This is why it is possible for him to build some of his most effective arguments around this format.

1)folk psychology is not significantly different from obsolete scientific theories like Alchemy, Phlogiston etc.

2)Alchemy, Phlogiston, etc. are false, do not apply to reality, and the entities they describe do not exist.


3) It is possible that folk psychology is false, does not apply to reality, and the entities it describes don't exist.

He uses this argument twice in NP.: 1) to show that folk psychology and Alchemy can both be defended by the functionalist multiple realizability argument (NP pp. 12-13) and 2) to show that the Intentional Stance's high predictive powers buy it no ontological status because "It is an objective fact that much of the behavior of metals and ores is predictable in terms of the Alchemical essences. . . . and yet there are no alchemical essences" (NP p. 126)

In all of the commentary on Churchland I have seen, people have tried to attack the first premise, by showing that in some sense or other folk psychology is not a theory. Churchland effectively disposes of all of those, but there is such wide acceptance of the second premise that no one seems to have noticed that Churchland himself cannot accept it. Churchland's pragmatic pluralism requires him to give the intentional stance essentially the same status that Dennett does: a theory with serious shortcomings, which nevertheless has genuine epistemic merit. The comparison to alchemy does not reduce folk psychology to a false theory, for if Churchland denies all epistemic status to alchemy because we now have better theories, we would have to deny epistemic status to our current theories because they will eventually be replaced by better ones. And this would leave us with the universal skepticism described earlier. We can save ourselves from this skepticism by acknowledging that truth and falsity need not be binary, and that alchemy and folk psychology (as well as special relativity theory and whatever theory replaces it) all possess degrees of epistemic virtue. None of these theories are true in the classical sense, but that does not imply they are all false, anymore than the fact that no one possesses vital spirit implies that everyone is dead.

Some may say that any argument that leads one to grant ontological status to the alchemical essences should be taken as a reducto ad absurdum. After all, we have to draw the line somewhere. What about astrology, unicorns and Sherlock Holmes? And surely we don't want to say that truth always admits of degrees. After all, statements like "Paris is the capital of France" and "2 plus 2 equals 4" are 100% grade A true; There are no degrees of truth there. I don't claim to have all the answers to these questions, but to me they seem more likely to be answerable than the various crises caused by the old assumptions. Perhaps that is just because the problems caused by accepting pragmatic pluralism are new, and philosophy has been wrestling for centuries with the problems of accepting various monisms (such as scientific realism). But I prefer going forward to moving backwards, and new challenges to old ones. So here are few tentative suggestions in the direction towards the pluralistic continuum theory of truth which I feel Churchland has been ambivalently prophecying.

Pragmatic pluralism does not require us to refrain from criticizing or comparing theories. Just because a theory is muddle-headed, or prone to errors, or in need of replacement, does not mean that what it is talking about doesn't exist. Astrology is an unjustified and useless conceptual system, but it is still a real objective fact that there are a certain number of Sagitarians in Kansas, and a certain number of Leos. Even if almost everything that astrology says about Sagitarians is false, it still provides us with an observation language that divides up the world in a distinctive way, however crude and misleading, and is thus doing a different job from the entities that inhabit novels and fairy tales. To ignore this difference is to make a serious category mistake.

Entities like Sagitarians, The Crystal spheres, and the Alchemical essences do not belong in the same category as imaginary entities, like unicorns and Sherlock Holmes. Imaginary entities have no existence or extension, because their only function is to entertain us while in the armchair. They have no pretensions of reaching out to the world, and thus do not apply to reality. Alchemy does apply to reality, because it was formulated by a community of living organisms fighting their way through the thick of things, who spent many generations tuning their neural networks so as to minimize errors. According to pragmatic pluralism, this is the only kind of epistemic pedigree that any theory can have, and alchemy does not lose this pedigree just because people stop using it. Consequently, If folk psychology is abandoned the way alchemy was abandoned, it will not mean that folk psychology is false, does not apply to reality, or that the entities it describes don't exist.

Also, the claim that an entire theory cannot be false in a binary sense does not apply to statements made within the context of a theory. Suppose that a shop in medieval Europe sold various exotic items( such as narwhal horns and kudu antlers) as unicorn horns.{3} A customer asks "Are there any unicorn horns in this shop ?". The clerk goes back to the stockroom, and after a casual search which fails to reveal a small box with a few remaining unicorn horns on the third shelf from the left, comes back and tells the customer, "no, there aren't" . The clerk's statement is false, in the traditional binary sense of true and false. Within the context of discourse for that statement, there are unicorn horns in the shop, and the clerk has made an error. But to conflate the observation statement "there are no unicorn horns in the stock room" with the ontological assertion "there are no unicorn horns" is to make another serious category mistake. After all, the discovery of narwhals would not have justified the clerk's saying to the customer "See, I told you we didn't have any unicorn horns" . The fact that this category mistake is so universally accepted as valid has made it possible for Churchland to demonstrate eloquently that within the context of the traditional idea of truth, Folk psychology could be false. Within the context of Churchland's own pragmatic pluralism, however, this "falsehood" is a trivial weakness shared by all theories past and future, possible and actual.

Pragmatic Pluralism and Functionalism

The pragmatic pluralist perspective also avoids a common objection against eliminative materialism regarding the ontological status of the functionally defined terms of common sense. Searle, in his 1992, dismisses eliminative materialism by pointing out that if the entities posited by folk psychology don't exist because they cannot be reduced smoothly to the entities of science, we must also deny ontological status to split level ranch houses, tennis rackets, golf clubs, etc. (p.47){4}. Similar arguments are also raised by Andrew Cling in his 1991, and by Putnam in his 1992. Churchland's reply to Putnam shows that he is still sitting on the fence on this issue, and that he seems none too comfortable there. He makes a distinction between "(legitimately functional) tables and chairs" and (the avowedly non-existent) phlogiston, caloric, and the four principles of medieval alchemy ". To my knowledge, Churchland has never attempted to formulate a distinction between the so called "legitimately functional" and the "avowedly non-existent", and I would advise him not to try. Among other problems, it would not sit well with the thought experiment in NP which is designed to show that the categories of alchemy are functional, and offers this as proof that functionality is no proof of legitimacy. It also seems unlikely that a notion from vulgar discourse like "chair" is going to be more ontologically rigorous than a concept like "phlogiston" which, in its time, was the best that the greatest living scientific minds could produce. If this concept of "legitimately functional" doesn't cash out, the only options available seem to be either denying existence to chairs or granting it to phlogiston: both outrage common sense, but the latter would be much more consistent with the rest of Churchland's philosophy. I find it very encouraging that Churchland referred to phlogiston et al. as being only avowedly non-existent for the first time in his 1992. If he decided to question that avowal, he and I would be on the verge of agreement.

Pragmatic Pluralism and Scientific Realism

This continuum theory of truth also makes it impossible to be a scientific realist, in any meaningful sense of the word. One can still be an admirer of science, but unless one claims that the entities of science are real in a way that other things are not, then strictly speaking one is not a scientific realist. Once we acknowledge that all theories that have pragmatic value have some degree of truth value, the difference between so-called "science" and other cognitive activities is purely a matter of degree, and this distinction becomes more misleading than useful.

One might try to save scientific realism by pointing out that it is an empirical fact that scientific theories are more useful than non-scientific ones. But I think this description has things exactly backwards. I think that because scientific realism is the orthodox ontology of the knowledge industry, we call theories scientific because they are useful. This also means that theories continue to get enough funding and attention to increase their usefulness only if they manage to jury-rig some resemblance to physics, which may or may not have anything to do with their basic nature. From a Scientific Realist perspective this jury-rigging is indispensable. Because physics is unquestionably science, the entities described by physics are seen as the only real ones, so all other terms must be either reduced (shown to be identical) to physical terms, or else eliminated from use all together. This fear of having one's area of study eliminated is, I think, at the heart of most people's objections to Eliminative Materialism. And when you look at the way that Continental philosophy was treated by Analytic departments in the sixties and seventies, the way that Skinnerian Behaviorism treated all other forms of psychology in its heyday, and the way that Cognitive Science treated Behaviorism shortly after that, these fears are fully justified.

From a pragmatic pluralist point of view, these Procrustean tricks are unnecesary. If we agree with Churchland that there could be different, even contradictory, theories of equal epistemic virtue, "all of them equally low in error, all of them carving up the world in quite different ways ," compatability with physics is no longer an essential characteristic of science, and being a science is not the only way of being a respectable cognitive enterprise. Unity with other knowledge-seeking disciplines becomes one of many epistemic virtues, rather than the only characteristic that frees a theory from the prison of our subjective speculations. Reduction thus becomes a way of making knowledge a bit more straight-forward and coherent, but not a way of connecting it back to the one true description of reality.

When Richard Rorty made the transition from Eliminative Materialism to his own brand of pragmatism, he rejected the goals of reductionism altogether, for reasons somewhat similar to these. (See Rorty 1982 p. xlvi footnote 31). He suggested that instead of trying to use reduction to create a single unified science, we should "let a thousand vocabularies bloom and then see which survive." (1970 p.230) Letting vocabularies bloom can not be justified by a binary theory of truth like scientific realism, unless there is a final weeding process that enables us to find the one true theory. If this process is permitted to go on indefinitely, it would be seen as a license to tolerate falsehoods and illusions. But fortunately for pragmatic pluralism, this principle has been in effect in scientific circles for some time, regardless of its conflicts with the orthodoxies of scientific realism. Churchland has always been scrupulous in reminding us that even his beloved neural computational theories are largely based on models that are in many ways biologically implausible, and that the similarities between natural and artificial networks are bridged largely by guesswork. Nevertheless, these theories still maintain a grip on the structure of our cognitive practices, because a theory does not have to be True with a capital T in order to be epistemically worthwhile, provided it has some isomorphism with biology and/or cognitive experience. When we stick our necks out and and posit a theory, we discover what errors that theory is heir to, readjust the theory accordingly, and thus scientific growth is acheived. There is no need to puritanically abstain from theorizing merely because there is a risk of error, for error is an inevitable part of all theorizing.

The obvious next step in this movement towards pluralism is to extend this tolerance to non-scientific cognitive strategies--like folk psychology, for example. Within science itself, it has long been recognised that there are qualities inherent in the subject matter of its various subspecialties that require us to relax the rigours of industrial strength laboratory physics. The theory of evolution has never been proven in the laboratory by replicable experiment, nor could it be. Yet it has been accepted as scientific truth, because it is obvious that this is an impossible demand to put on any theory that attempts to explain a processs that took place over thousands of years. For similar reasons, folk psychology can not be held to the same standards as laboratory neuroscience, even if we recognise that both are theoretical structures that can and should evolve together. There will always be a domain of experience for folk psychology which is distinct from neuroscience, even if the former ends up relying heavily on concepts borrowed from the latter. Folk psychology must make sense out of the brouhaha and hurly-burly that we folk live and die in, and when the concepts cultivated in the hot house atmosphere of the laboratory are transplanted to that rough and tumble environment, they will be inevitably forced to adapt. Our inability to recreate laboratory conditions in the outside world produces hugh gaps that must be filled with metaphors and speculations, which extrapolate from the evidence in ways that would never be acceptable in the laboratory. These are what make up the inevitably non-scientific part of folk psychology, and they will always be with us.

On the last two pages of his 1995, Churchland points out that folk psychology has already been profoundly changed by the influx of Freud and "Psychobabble" and then asks:

"What would happen . . . if a conceptual framework came along that had some real integrity? Some real correspondence with the pulleys and levers of our cognitive and emotional activity? . . . if, unlike its hollow precursors, it brought with it a real grip on the structure of cognitive reality, it just might stick. (Churchland 1995 p.324. )"

I would suggest that to some degree the possibility that Churchland envisions here has already begun to be actualized. A great deal of what is dismissed as "Psychobabble" is already heavily infiltrated with neurological data. Marilyn Ferguson, author of The Aquarian Conspiracy, is also the author of the Brain Revolution, and publishes a regular newsletter called the Brain-Mind Bulletin. And those cocktail party conversations that used to be peppered with Freudianisms are now full of references to drawing on the right side of your brain. Such neurally inspired talk varies from the vacous to the profound, as do the various books that inspire it. But the fact that most of such talk is probably full of biological errors does not mean that devout rationalists should refrain from it. It has brought some real clarity to our experience of ourselves, and the neurologically informed should not dismiss it as lies and illusions. The writings of Churchland and other neurophilosophers will undoubtedly be useful in correcting and refining this neurological "psychobabble", but ultimately they (we) will have to see those psychobabblers as colleagues, not as superstitous savages. Some of the vagueness in such writing is a result of genuine muddleheadedness, and it should be exposed as such. But we still need to remember that any attempt to apply laboratory science to human experience can never be as rigorous as the science itself, and that something like poetry is the only kind of truth we can have for that messy domain that folk psychology attempts to comprehend.

Nor is there any reason to assume that even if neurologically inspired poetry does bloom in the garden of folk psychology, the other 999 flowers in the garden must therefore be considered weeds. Beliefs, desires, superegos, inner children and Oedipus complexes will no doubt remain useful for making sense out of daily experience despite their numerous incoherencies, and according to pragmatic pluralism, anyone who has any desire to continue to refer to such entities has every epistemic right to do so. As long as there are bureaucrats and lawyers, there will be a use for Freud's concept of anal retention, no matter how much we learn about the brain. Churchland's neurocomputational philosophy of mind can take its place amongst the numerous contradictory conceptual structures tangled within folk psychology, but it will almost certainly not completely replace the rest of them no matter how useful it becomes. Folk psychology has always tolerated inconsistency, and probably always will.

The Churchlands are fond of quoting Quine quoting Neurath that philosophy and science are in the same boat. Now we have to take this one step further and include the butcher and the baker, as well as the physicist, in the same boat. We can't use scientific realism as a canonical justification for throwing useful conceptual systems overboard. If we say that epistemic virtue should be measured on a continuum, or perhaps over an intersection of several continua, we cannot use a binary category that divides cognizers into "scientists" who have "the truth", and everyone else, whose beliefs are false (or at best unjustified) whenever they are unsupported by that truth.

Reducing Reductionism

If we follow Rorty's lead and reject reductionism, can we do a reduction of reductionism itself? Can we explain within the context of pragmatic pluralism why scientists and philosophers have always felt that reduction is so important ? I think part of reductionism's appeal is that it invalidates the assumption that individual sciences can be autonomous, which has so frequently lead to errors and sterility. By defending reductionism, Churchland has encouraged communication between disciplines, and all of science has benefited from this. The reason reductionism gives for these benefits is that all science should be unified with the one science which has the firmest foundations. The reason that pragmatic pluralism gives for these benefits is that no science has very firm foundations, so all of the sciences need as much help from each other as they can get.

Churchland has never fully let go of his dream that someday neuroscience will strip away the veil of illusion which is folk psychology and "depict all of us, at last, as we really are". These are his exact words in the last paragraph of his most recent (1995 ) book, even though he acknowledges in his 1992 that the best that a neurocomputational philosophy of mind can hope for is "to be a good theory, better than any of its predecessors" (p.423). It seems to be difficult for Churchland to remember that his definition of a good theory has changed profoundly as his philosophy has matured, and that if we are not careful we will not be able to recognise how close we have already come to realising the modern downsized version of his ideal. The kind of careful correspondence pairing that a good reduction or elimination produces is still necessary in pragmatic pluralism, because this enables our choices between theories to be more precisely motivated, and better informed. But pragmatic pluralism says that we cannot pretend to ourselves that the choice goes away once we have decided to rely primarily on one particular theory for something. Now that Churchland recognizes that there could be different, even contradictory, theories of equal epistemic virtue, "all of them equally low in error, all of them carving up the world in quite different ways ," compatibility with a single unified science is no longer an essential characteristic of a respectable cognitive enterprise. Unity with other knowledge-seeking disciplines becomes one of many epistemic virtues, rather than the only characteristic that frees a theory from the prison of our subjective speculations.

One could say as, John Bickle does, that my evidence and arguments against reductionism could be taken as evidence in favor of some kind of "new wave reductionism". (1992a 1992b) My rhetorical question about such a move is "Can a process legitimately be called reduction if it doesn't reduce? i.e. doesn't result in fewer entities when it is finished." I don't think so, but I'm willing to admit that this question doesn't have an easy answer. How far can we stretch the concept of reduction before it snaps-- before we are compelled to say that it bears no significant relationship to the concept originally defined by Hempel and Nagel? With abstract concepts like reduction, the metaphor of stretching and snapping breaks down--there is no fact of the matter. If reduction is abandoned as a scientific activity, science will have to come up with a different activity that accomplishes most of what reduction used to accomplish, and decide that whatever else reduction used to do is not worth doing. Should this new activity be called new wave reductionism? Only if you want to stress the continuity between the old and the new. Should it be called something else? Only if you want to stress the discontinuity between the old and the new. The considerations which guide us in making these decisions are pragmatic, when they are rational. If we are going to grant truth value to alchemy, we cannot deny it to reductionism.


{1}If we wished to be pedantic, we could point out that she obviously meant to identify the absence of phlogiston with oxygen. [Back]

{2}Some people have claimed that Feyerabend himself was not a skeptic in this sense either, at least in his later writings. This may be true, but this kind of skepticism is so widely attributed to Feyerabend that I think "Feyerabendian" is a convenient label for this position, even if Feyerabend himself did not fully endorse it. [Back]

{3}In this case, we have a fictional entity which is providing the conceptual basis for a fragment of an observation language, much as Santa Claus is both a character in stories and a theoretical entity posited to explain the existance of presents under Christmas Trees. Does this means that fictions and theoretical entities are the same sort of thing put to different uses? This would fit well with Quine's well known claim that there is no difference in kind between our myth of the existence of objects and the Gods of Homer. But I don't claim to be able to deal with all of these issues in a paper devoted primarily to another subject. I only claim that granting positive truth value to statements that presuppose the existence of phlogiston and Santa Claus is not an unthinkable absurdity. [Back]

{4}Searle, however, blithely ignores all of the philosophical problems raised by Sellars and Churchland about "the myth of the given", and embraces the crudest sort of naive empiricism to defend the ontological status of these ordinary language concepts. "Beliefs and desires, unlike phlogiston and caloric fluid, were not postulated as some special theory, they are actually experienced as part of our mental life. Their existence is no more theory-relative than is the existence of ranch houses, cocktail parties, football games, interest rates, or tables and chairs." ( Searle 1992 p.60). It's hard to believe that Searle is claiming that there is no theoretical content to the concept of football game, or that the concept of cocktail party is immediately given to direct experience like the Cartesian Cogito. But that is exactly what he is saying, although he waffles indecisively in the next two paragraphs with some other arguments that vaguely paraphrase functionalism. (a position he dismisses as misguided in the rest of the book). Churchland decisively critiques this distinction between the theoretical and the empirical in Chapter 8 of NP. [Back]


Bickle, John (1992a) "Mental Anomaly and the New Mind-Brain Reductionism" Philosophy of Science 59 217-230

____________(1992b) "Multiple realizability and Psychophysical Reduction" Behavior and Philosophy 20 47-58.

_____________(forthcoming) "New Wave Psychological Reductionism and the Methodological Caveats" Philosophy and Phenomenological Research.

Churchland, Patricia. S.(1986) Neurophilosophy Cambridge, Mass. MIT press.

Churchland, Paul M. (1979) Scientific Reason and the Plasticity of Mind London Cambridge University press.

________________(1987) "How Parapsychology could become a science" in Inquiry 30 227-39

________________(1992) "Activation Vectors vs Propositional Attitudes: How the Brain represents Reality" in Philosophy and Phenomenological Research 62 199-204

_______________ (1989)A Neurocomputational perspective Cambridge, Mass. Bradford books.

_______________ (1995)The Engine of Reason, The Seat of the Soul Cambridge, Mass. Bradford books.

Cling, Andrew (1991) "The Empirical Virtues of Beliefs" in Philosophical Psychology vol 4 #3

Dupre, John (1993) The Disorder of Things Cambridge Mass. Harvard University Press

Feyerabend, Paul (1962)"Explanation Reduction and Empiricism", in Minnesota studies in the Philosophy of Science vol 3. ed. Feigl and Maxwell Minneapolis University of Minnesota Press.

_______________(1970) Against Method: Outline of an Anarchistic Theory of Knowledge in Radner and Winokur eds. Analyses of Theories and Methods of Physics and Psychology . Minnesota studies in the Philosophy of Science vol 4 Minneapolis University of Minnesota Press.

Kuhn, Thomas (1962) The Structure of Scientific Revolutions Chicago University of Chicago Press.

Millikan, Ruth (1984) Language, Thought and other Biological Categories Boston MIT Press.

Nagel, Ernst (1961) The Structure of Science New York Harcourt, Brace, and World.

Putnam, Hilary 1992 "Truth, Activation Vectors, and Possession Conditions for Concepts." in Philosophy and Phenomenological Research 62 211-227.

Quine, W.V. O. (1963) "Two Dogmas of Empiricism" in From a Logical Point of View Cambridge, Ma. Harvard University Press.

Rorty, Richard (1965) "Mind-body Identity, Privacy and Categories" in Rosenthal, D.M. (ed.) (1971) Materialism and the Mind-body Problem Englewood cliffs, N.J. Prentice-Hall

_____________ (1970) "In Defense of Eliminative Materialism" in Rosenthal, D.M. (ed.) (1971) Materialism and the Mind-body Problem Englewood cliffs, N.J. Prentice-Hall

_____________ (1982) Consequences of Pragmatism Minneapolis University of Minnesota Press.

Searle, John (1992) The Rediscovery of the Mind Cambridge, Mass. Bradford books.

Sellars, Wilfrid (1963) "Empiricism and the Philosophy of Mind" in Sellars Science ,Perception, and Reality London Routledge and Kegen Paul.