A Scientific Philosophy of Science:
Some Comments on Keith Campbell

[In 1986 a friend of mine named Kirk sent me five xeroxed pages from an unidentified work by the philosopher Keith Campbell. I believe these pages were from Campbell's book Metaphysics: An Introduction, but I am not certain. (I haven't seen that book, which has apparently been out of print for some time.) Campbell was born in New Zealand, and according to a blurb in another of his books, he teaches philosophy in Australia. These are some comments I wrote in June 1986 in reaction to Campbell's work, and in the context of an on-going discussion that Kirk, another friend Greg, and I were engaged in.

Kirk remarked that Campbell "advances a view of metaphysics as 'distinguished from science... only by attempting a more comprehensive and more systematic basic theory'. Scott will welcome this, since it is in accord with his view of metaphysics as 'the most general and abstract science.' However, Campbell goes further and ties all levels of knowledge in principle to theoretical science—'A theory is accepted and retained so long as it remains an integral part of the best overall explanatory account of the phenomenon we encounter.... The best for the time being.'") To the extent that Scott holds elements of his metaphysic to be immutable in the face of any possible scientific development, he will find this unattractive, whereas Greg and I will welcome it, as well as the following remarks on dogmatism as '... an enemy of reason wherever it appears.'"

Please note that in what follows everyone, including me, is using the term 'metaphysics' in the way it is normally used in bourgeois philosophy, i.e. as that branch of philosophy concerned with the very most general features of the world and our knowledge of it. This differs totally from the usual Marxist usage of the term to mean views which are undialectical or opposed to dialectics.]

First, I would like to thank Kirk for pointing out this passage to me. I had no idea any analytic philosophers were this sensible on the topic of metaphysics. (But Kirk, you neglected to give a citation for the extract!) So, Kirk is correct in his first assertion; I do welcome this.

However, I dispute the next assertion that Campbell "goes further", since I (and Marxism) also tie "all levels of knowledge in principle to theoretical science". And as for the Campbell quotation ("A theory is accepted and retained..."), I accept it unequivocally. In fact, this is one of the things I've been trying to say all along. Since I do not hold any elements of what Kirk calls "my metaphysic" to be immutable in the face of any possible scientific development, I do not at all find Campbell's comment "unattractive".

As far as dogmatism being "an enemy of reason wherever it appears", I can also wholeheartedly agree. I would only add that there are other enemies of reason as well which sometimes masquerade under the banner of anti-dogmatism (such as incoherence, inconsistency, dogmatic agnosticism, subjectivism, etc.). Both kinds of errors are to be guarded against. (Marxists frequently call errors diametrically opposed to dogmatism "revisionism"—because a scientific theory (Marxism) is being improperly (unscientifically) revised. It is a poor term because every scientific theory needs to be scientifically revised; revision in and of itself is neither a positive nor negative sign. Non-Marxists, not understanding this terminological peculiarity, tend to interpret Marxist attacks against "revisionism" as inevitably a sign of dogmatism. It is interesting that the language does not seem to have any ready term for errors diametrically oposite to dogmatism. But they do exist.)

[In his note with the Campbell pages, Kirk also wrote: "Campbell goes on to discuss the recent trends in the philosophy of science which are problematic for a scientific metaphysic, which he correctly characterizes as 'antimetaphysical trends in the philosophy of science' rather than calling them 'antiscientific'—a characterization which I consider, quite frankly, to be puerile. He, like Scott, is finally unconvinced of their cogency, but is not prepared to defend this in detail.]

I presume that Kirk's comment that anyone who characterizes such "antimetaphysical" trends as "antiscientific" is being puerile is directed at me. I find the comment a little puzzling though. Puerile means something like childish, simpleminded, patently silly, or the like. What is it about such a characterization that deserves such a response?

If you are defending a scientific "metaphysic" then anything opposed to that "metaphysic" will of course be anti-scientific, unless it is itself a scientific attack, proposing a new or modified scientific "metaphysic". Though this does not seem to me to be a very difficult point, neither does it seem completely childish or silly.

I suspect something similar to the following may be going on here: Imagine someone who takes the position that "rational argument" itself should not be accepted "on faith" as the way to resolve disputes, but says that it itself must be "justified". The question then arises, how does one justify it? Through rational argument? But that might seem to our irrationalist to beg the question (even if it is on a meta-level). He might even say something like "How puerile of you to reject any other way of settling the issue except by the very means which is at issue!".

It actually seems more correct to me to say that the sorts of trends that Campbell mentions, are 'anti-scientific' than it does to call them 'anti-metaphysical'. They are not, afterall, all opposed to any metaphysics (though some of them may be)—but specifically to a scientific metaphysics.

As far as Campbell being unconvinced by the cogency of these "recent problematic trends", one does indeed wonder why not. He puts forward his opponents' criticisms and then gives only the briefest of replies (at least in the passage Kirk reproduced). Why such a puny defense of his own position? Is this any way to conduct a serious argument?? He even claims that these issues are still "unresolved" and that no one can be sure how the dispute will turn out! I suppose I should not have expected too much from him, but to give up with scarcely a fight after such a fine beginning is more than a little disappointing.

Since Campbell decided to play devil's advocate and leave the angels' reply blank, I will try to say a few things below to knock down his "recent trends".

As stated above, and even to a greater extent than Kirk expected, I like the general tenor of Campbell's comments. Among his comments that I can whole-heartedly agree with is: "Metaphysicians must abandon the hope of finding a theory of the world which can be guaranteed certain. We cannot expect to arrive at any knowledge more surely founded than the self-corrective yet still provisional theories which scientific method yields. Metaphysical truths are synthetic. And they are not necessary in any tough logical sense. They cannot be guaranteed by the principle of noncontradiction."

Given our earlier discussions, however, I wonder if Kirk thinks that this means "that we can't be sure of anything". In fact, through scientific methods we can become very certain of many things. There is no "guaranteed certainty", no "absolute certainty". But there is a large stock of knowledge (theory and fact) which we can rely on in order to make our way in the world. I don't believe Campbell would in any way deny this. Making the same point in yet another way: Just because we recognize that all theories are provisional (in the scientific sense), is no reason not to rely on them to accomplish our goals. Scientific facts and theories are the only things we have to rely on as a guide to action, and to reject this reliance is to reject the whole point of scientific investigation.

Campbell's very first sentence (in the extract) is fascinating, even though I can't fully accept it: "The attempt to set metaphysics apart from other inquiries has been the root cause of the crises in its history." The true root cause is the role of metaphysics as class ideology. However, I would go along with Campbell to this extent: The attempt to set metaphysics apart from other inquiries has indeed been one of the main means by which the root cause of the crises in its history have been manifested. (Or one could get into "real causes", "effective causes", etc., which I would just as soon not do.)

Some of Campbell's formulations are at slight variance with others. When he says that "the relation of metaphysics to science is one of collaboration, not contrast", he seems to be less on the mark than later on. Campbell seems to want to say that metaphysics is just like science, but not that metaphysics is a branch of science. Why he wants to stick to this fragile distinction I don't know. (This is just one indication that Campbell is a bit of a wimp, and afraid to "go all the way".)

Campbell says "A theory is accepted and retained so long as it remains an integral part of the best overall explanatory account of the phenomena we encounter. And precisely the same is true of responsible metaphysics. Space, time, God, events, casual determinism, spiritual minds, properties, classes, numbers, substances all have a title to a place in our metaphysic to the extent that they belong in the best total theory." This is certainly correct. But I would have emphasized the words 'to the extent', and then gone on to say, "This is why things such as God, souls and spirits have no place in a "responsible" metaphysics—because they are demonstrably not part of the best total theory, and are in fact inconsistent with such a theory. This would have been clearer had Campbell chosen to speak of a "scientific metaphysics" rather than a "responsible metaphysics". So here is another sign of his wimpishness or half-heartedness, and one of its typical consequences.

Campbell goes on to say that "It is, of course, possible to insulate any metaphysical belief from the risk of refutation. This is done by always explaining away every consideration which seems to refute it.... This process of insulating favorite beliefs is familiar enough. It is dogmatism and is an enemy of reason wherever it appears." While dogmatism is in fact an enemy of reason wherever it appears, Campbell is saying more than that here. And part of what he is implying is wrong.

Patching up a theory is not necessarily dogmatism. Even providing ad hoc explanations for anomalies, is not necessarily wrong. It depends upon how far the process goes, and on whether anyone can come up with a better theory that requires no (or fewer) ad hoc explanations for things that don't seem to fit. So Campbell is stating his point too strongly here, too absolutely. If one attempts to "insulate" his pet theory beyond all reason, then of course he is a dogmatist. But no theory could ever be developed at all if the existence of some anomalies or loose ends were never allowed, nor ad hoc explanations which require further investigation to see if they are really justified.

If nothing would count as a refutation of a theory, then it is not a scientific theory. However, if even the tiniest flaw, or unexplained point is considered sufficient to shoot down any scientific theory, then scientific theories are impossible—which is to say that science is impossible. Once again the hard task is to find a rational middle ground between anti-scientific dogmatism and equally anti-scientific... what-should-I-call-it-here? Maybe I should characterize it here as "perfectionist idealism". In any case it is an anti-theoretic stand which is attempting to require far more from scientific theories than they can ever provide. It is to misunderstand the point of scientific theories; they are put forward not as the complete, final, and perfect explanations for a group of phenomena, but as the best overall account we can come up with so far. (Campbell himself says something like this, but doesn't seem to recognize all its implications.)

Near the top of page 19 of the excerpt, Campbell says that we should view metaphysics "as continuous with science". This is a better formulation than he used earlier, but you can see how it gets him into big trouble right in the next sentence: "Metaphysics is the attempt to interpret all experience, including science, so as to form one coherent and systematic view of the world and man's place in it." The phrase "all experience, including science" implies that some experience is not explanable in scientific terms. What in the world is this supposed "experience"? (The answer is that it is not in the world; it doesn't exist.) This is why I say again that if you don't understand that any "rational metaphysics" must be in fact a scientific metaphysics, you will inevitably get trapped into at least implying some anti-scientific views or theories. I would like to ask Campbell: If there is "experience" which is not covered (actually or potentially explained by) science, then is this "experience" also "continuous with" science? If not, then how can a metaphysical theory (which you say is "continuous with science") encompass phenomena which is neither scientific nor "continuous" with it?

Another way to put all this is that it seems pretty obvious to me that if Campbell wants to say that metaphysics is only "continuous" with science, and not actually science, then first it is incumbent upon him to explain a little better what this "continuous" stuff amounts to, and second, he will find himself defending a theory which is needlessly clumsy.

Campbell talks about his attitude towards metaphysics as being the "restoration of metaphysics". As someone adhering to a trend which has held this attitude all along (albeit more thorough-goingly) I find this a bit amusing. Like most bourgeois philosophers, when they occasionally stumble in a scientific direction, he seems to have no idea that he is reinventing the wheel. In this case, the revolution to a scientific world-view ("metaphysic") and the rejection of non-scientific metaphysics was carried out nearly a century and a half earlier! [I am referring to the early writings of Marx and Engels, of course, such as The German Ideology.] (I am not implying here that there has been no scientific progress in "metaphysics" since then. But that is the time of the original breakthrough.)

Now on to the "new challenge to metaphysics", as Campbell characterizes it.

Campbell first notes that only on the assumption that there can be (and has been) scientific progress, can his "rational" metaphysics be accepted and validated. I'll go along with that. But for Marxists it is very hard to take seriously the claim that there has been no progress in science. It sounds completely stupid. But let me try to take it seriously enough to say a few words in reply.

In this century, especially, there has been a whole trend of bourgois pessimism brought about by wars and revolutions and the prospect now of even the extinction of the human species. Primarily for this reason, but also because bourgeois ideologists imagine that the capitalist system is timeless and has existed at least as long as civilization itself, the most popular view has come to be that there is no such thing as social progress. Changes, perhaps, or cycles, or ups and downs, but no permanent progress for sure. This is one of the main themes of bourgeois ideology in its declining period. (How different it was when the bourgeoisie was young and on the rise!) It is only to be expected that there might be those who would try to extend this ridiculous view to the sciences as well. (Is someone who is consistent in their stupidity to be congratulated or condemned?)

If scientific "progress" were impossible, then science itself would be impossible. Moreover, the means of showing that there is progress in science is the very same as the means of showing that science is possible, i.e., that it "works". How do we know that science works? Because we can employ it to accomplish our goals (we can use it to build a bridge or get to the moon). How do we know that there has been scientific progress? Because we can use it to do things now that we could not do before. If this does not satisfy the skeptics, then it is up to them to specify more clearly exactly what the hell they can possibly mean when they say that there is no progress in science.

Now the next aspect of this attack on science arises from drawing totally wild conclusions from a number of observations, such as that scientific opinion is constantly changing, and many of the reasons for these changes are haphazard and even irrational. From this it is correct to draw the conclusion that the advance of science is not an "orderly march"; it is more like a disorderly march. But it is a march! It is a great exaggeration to draw the conclusion that scientific change ("advance") is "a lurching, directionless course from one poorly based, hazily conceived, and woefully incomplete position to another". And not only that; even if this were literally true, it would still not show that no progress has been made. Because we see lurking behind this criticism once again an unrealistic, idealist picture of what scientific theories are supposed to be. It does not matter if they are poorly based, hazily conceived, and woefully incomplete. If they help explain anything then they are still of value despite this! And if they explain more than their predecessor theories, then they are a real advance, no matter what their shortcomings may still be. (Note: Here I am using explanatory scope as shorthand for all the things that it is rational to look for in scientific theories, including coherence, simplicity, etc.) The critics are still looking for perfection, or close approximations to it. It is a totally idealist approach.

Is it any big deal that it has been "discovered" that "education, fashion, habit, prejudice and accident seem to determine change as much as anything rationally compelling"? No, not really, and I am surprised that anybody finds anything surprising in this fact. All these factors are indeed involved; they always have been, and to some degree they always will be. But what the "skeptics" are forgetting is that some aspects of rationality are also involved in every genuine scientific advance, and in the long run this is all that matters, regardless of how small a role rationality may have had at the beginning. (And by the way, the role of rationality is seldom as minimal as the skeptics are claiming. It takes a hell of a lot of rational argument to convince the general scientific community to change their collective mind on any substantial theory.

If we look back in hindsight and see lots of irrational elements (confusion, etc.) in previous scientific advances, this just shows that the full rationality behind the advance was only appreciated over time. This does not show that rationality is absent, only that further investigation and consideration bring out a clearer and more complete picture of the real rationality behind the theoretical change, and it is not necessary that every rational consideration in favor of the theory change be appreciated at the time the change is first made. The fact is, we do not usually go back to the old theory after more exhaustive rational investigations of the reasons for going to the new theory are made. Newton will never supplant Einstein, regardless of how rash the jump from Newton to Einstin may have been. (And actually it was not that rash at all; it took some very important experimental evidence plus a mountain of argument (most of it rational) to get the physics world to adopt relativity theory.)

In Campbell's one good comment on this point, he says "it must be shown... that despite the vagaries of the actual course of science, later theories are rationally superior to earlier ones...". But he makes it sound like this is yet to be done! This (in general) is a completely off-the-wall implication. The fact is that science is constantly reconsidering old points of view, and sometimes switching back to them on the basis of a higher theory. (One interesting example I just came across: Richard Feynman, in his excellent little popular book QED, points out that according to quantum electrodynamics, light is not a wave afterall. It is a particle, and only a particle. Wave/particle dualism should now be abandoned, he says.) It is completely wrong to think that the rationality of previous scientific advances is not being reexamined; and once that examination shows anything but that the change was rationally justified, serious proposals are immediately advanced to return to the more rational stance.

Onward to the notion, now commonplace, that facts are only facts in terms of a theory, or in other words that there are no "facts" independent of theory. This is a position which I hold, and in no way do I see it (as evidently some people do) as implying that we cannot rationally choose among different theories. The skeptical worry, as Campbell expresses it, is that "the game is rigged in favor of the theory" when we attempt to test a theory by checking it against "the facts" since what is a fact and how it gets expressed are all determined by the theory itself.

But actually, all this shows is that one does not really test a theory by comparing it against "the facts". One really tests theories by comparing them against other theories. The procedure is to see how each organizes and explains the phenomena it covers, and which does the better job of it. It is true, but irrelevant, that the "phenomena" (the "facts") are characterized differently in different theories, but if the theories are in fact truly competing, there will be a substantial overlap, at least, of phenomena. (If this is not true, we are dealing with two independent sciences.) Remember again that nothing can shoot down a theory, no matter how poor it is, except a better theory. So this skeptical objection is also completely off track, in that it presupposes an erroneous idea of what is involved in "validating" a scientific theory.

On this issue Campbell only comments that "...it must be shown... second that despite the links between theories and the data they account for, theories can be objectively tested by comparing their consequences with the actual run of affairs...". This is ok as far as it goes, but it comes off rather lame when the point I raised in the previous paragraph about comparing one theory with another is totally missing. (One begins to see why Campbell thinks a lot of work needs to be done to see if the skeptical arguments can be disposed of. He is merely ignorant that the work has already been done by others.)

Finally we come to the third and last of Campbell's skeptical assaults on scientific metaphysics, namely the issue of the "validity of the criteria we use in judging a scientific theory". Campbell remarks that these include "the precision and simplicity with which the theory accounts for the phenomena, the extent to which the theory is confirmed by new facts beyond those it was invented to explain, and how naturally it fits into the general view of the world which we favor". (The way he puts that last point is questionable. It would be better to say something like "compatibility with more general scientific theories".) I have already mentioned some additional criteria, such as explanatory scope, coherence and consistency. We could probably come up with one or two more.

Now Campbell himself remarks that the authority of these criteria "developed gradually and painfully in the course of the development of science itself". This is an excellent comment. But he immediately proceeds to spoil it all with his ridiculous (devil's-advocate?) question, "But what guarantee could there be that the criteria we now use will select theories which are always getting closer to the truth?" One wants to yell out, "Pay attention, Campbell, to what you yourself have been saying!". Campbell is concerned that we must be able to show "that there is no vicious question-begging involved in the development of a self-critical standard of judgment for theory within the scientific tradition". But he fails to point out that the demand for a "guarantee" that the criteria we now use will select theories which are always betting closer to the truth, is in fact to beg the question in the other direction. Remember: "guarantees" of this sort are not scientific; they can only be idealistic metaphysical principles, of the sort which Campbell supposedly rejects.

The plain and simple fact is that there is not, and cannot be, any absolute (or "metaphysical") guarantee that scientific methods must work, or that there will be scientific progress, or that our criteria for judging scientific theories are the best. All we can say is that scientific methods have been found to work (i.e. humanity has been able to come up with scientific theories which allow us to control the world to a degree), that scientific progress has in fact been found possible (i.e. our ability to control the world has increased as scientific theories have changed and developed), and that these things have been possible because we have used the criteria for judging scientific theories that the skeptics now question.

In short, the world is the test of scientific theory and of the criteria for judging scientific theories. What anti-scientific philosophers demand is that at least the criteria for judging scientific theories must be established by non-scientific means. But why? The real reason is that they don't understand or trust science in the first place. The stated reason is that there is "vicious circularity" involved. But I have yet to see anybody spell out this supposed viciously circular reasoning precisely. To repeat, these philosophers are anti-scientific because they want to establish something about the world by non-scientific means. (In this case, it is the criteria for scientific theories that they want to establish. But this is still something about the world; it is not a question of formal logic for example.)

Quite a bit earlier I suggested the philosophical thought experiment of someone rejecting "rational argument" as a means of resolving disputes because the only means of establishing this as a valid method seems to be through the use of rational argument, which is "obviously circular". There is clearly something wrong with this argument, but what? One could cook up a number of theores, such as a theory of meta-levels of argument, and so forth. But the real answer here is that the use of rational argument does not need to be justified. It is true there is "no arguing" with anyone who rejects rational argument on principle, but so what? (Why would you want to argue with such a person?)

Rejecting a scientific world view is a lot like this. In fact, it is really the same thing. (Part of being rational is being scientific.) Requiring that the criteria for judging scientific theories be justified by non-scientific means is just like requiring that the criteria for rational argument be justified by non-rational means.

In actual fact there need not be any circularity involved in determining the criteria for rational argument via rational argument. It can simply be a question of raising principles to a higher level of abstraction; or perhaps of formalizing principles which are at first amorphous or ill-defined. (Recognizing the important role of abstraction in scientific theory is one of Marx's great achievements. See for example his comments on this in the introduction to volume 1 of Capital.)

Similarly, using scientific techniques to determine the criteria for "validating" scientific theories is not circular if what is really involved is just refining, or extending, or modifying criteria which are already usable. Sure there is an element of bootstrapping involved here. But why cannot scientific methods be applied to science itself, and to the improvement of scientific techniques? There is really no theoretical difficulty here at all.

Campbell makes the remarkable statement that "the problems raised by the new philosophy of science [i.e., the anti-scientific skeptics] are not spurious". It seems to me that that is just what they are. He makes the claim that it is "not yet known" whether this skeptical attack can be answered. This to me sounds almost disingenuous. (One can't help but wonder which side Campbell is really on.) The truth of the matter is that he (and his school presumably) have not yet been able to give a good answer, but Marxism has answered these points long before Campbell was even born. It is a pity that bourgeois philosophy is unable to see this and has to struggle over the same ground again.

I started out by saying that I was impressed by Campbell, and surprised that anything this sensible was being written about metaphysics by a bourgeois philosopher. Then I proceeded to attack him at length (excessive length no doubt). But I still am impressed that he does at least seem to be on the right track. I have recognized for some time that there are scientific strains in analytic philosophy (and no doubt bourgeois philosophy in general), even though these strains pale in significance when compared to dialectical materialism. It even seems to me that the scientific strains in analytic philosophy have been getting noticeably stronger over the past couple decades. Moreover there are certain topics which they address on which Marxism has had little to say, and which therefore constitute new ground for scientific philosophy. For the most part, however, these are in highly technical areas (philosophy of language, formal logic, etc.) that are really becoming sciences independent of philosophy, and usually they concern small corners of the subjects, not the overall theory of the subject. Analytic philosophy can never become fully scientific, but a trend in that direction is not to be totally despised.

—Scott H.
   June 1986 (With some slight editing on 9/5/99)

— End —

Philosophy Home Page on MASSLINE.ORG
Scott H.'s Home Page on MASSLINE.ORG