[ Preface | Part One | Part Two | Notes | Bibliography | Cover ]
Much more is known now than was known fifty years ago, and much more was known then than in 1580. So there has been a great accumulation or growth of knowledge in the last four hundred years.
This is an extremely well-known fact, which I will refer to as (A). A philosopher, in particular, who did not know it, would be uncommonly ignorant. So a writer whose position inclined him to deny (A), or even made him at all reluctant to admit it, would almost inevitably seem, to the philosophers who read him, to be maintaining something extremely implausible. Such a writer must make that impression, in fact, unless the way he writes effectively disguises the implausibility of his suggestion that (A) is false.
Popper, Kuhn, Lakatos, and Feyerabend, are all writers whose position inclines them to deny (A), or at least makes them more or less reluctant to admit it. (That the history of science is not "cumulative", is a point they all agree on). Yet with a partial exception in the case of Feyerabend, none of these writers is at all widely regarded by philosophers as maintaining an extremely implausible position. On the contrary, these are the very writers who are now regarded by most philosophers as giving an account of science more plausible than any other. So if what I have said is true, they must write in a way which effectively disguises the implausibility of their position. My object in Part One of this book is to show how they do it.
Of course I do not suppose that these authors, or even any two of them, agree on every point. Feyerabend argues persuasively, indeed, that in the end Lakatos's philosophy of science differed only in words, not in substance, from his own more openly irrationalist one . And Kuhn has no difficulty in showing the very great amount of agreement that exists between himself and Popper . Lakatos and Popper, on the other hand, are at pains to magnify any distance separating them from Kuhn , and would be still less willing to acknowledge affinities with Feyerabend; and Popper is almost equally anxious to distinguish Lakatos's position from his own . To an outside philosopher, indeed, the differences of opinion among the four must appear trifling by comparison with the amount of agreement that unites them. But it is in any case sufficient for my purposes that they all agree so far as to share a certain reluctance to admit the truth of (A).
Everyone would admit that if there has ever been a growth of knowledge it has been in the last four hundred years. So anyone reluctant to admit (A) must, if he is consistent, be reluctant to admit that there has ever been a growth of knowledge at all. But if a philosopher of science takes a position which obliges him, on pain of inconsistency, to be reluctant to admit this, then his position can be rightly described as irrationalism or relativism. Lakatos and Popper were therefore right in applying these epithets to Kuhn's position . There were further right, I believe, in the suggestion, which is a major theme running through their comments on Kuhn, that this irrationalism stems from the conflation, in Kuhn's writings about science, of the descriptive with the prescriptive: from his steady refusal to distinguish the history or sociology of science from the logic or philosophy of science .
Kuhn, of course, `admits the soft impeachment' and defends his practice in this respect . (Feyerabend likewise rejects the distinction between description and prescription ). But Kuhn also retorts that in any case Popper and Lakatos do exactly the same thing themselves . This was a very palpable hit, quite impossible to deny. That he confused the logic with the history of science was a common complaint against Popper, and one only too well-founded, long before Kuhn mentioned it in his tu quoque; and to try to defend Lakatos from the same reproach would be even more idle. But if it is true, as these critics of Kuhn alleged, and as indeed it is, that the source of irrationalism in his case is the conflation of the history with the logic of science, then the same cause cannot fail to have the same effect in their own case as well.
The question from which I began may therefore be replaced by a more general one. I asked in effect, "How do these writers manage to be plausible, while being reluctant to admit so well-known a truth as (A)?" But in view of what has just been said we are entitled to ask instead: "How do they manage to be plausible, while being in general so irrationalist as they are? For example, while being reluctant to admit (A)?"
It is easy enough to answer this question, I think, in general terms. The answer lies in what I have just referred to: the constant tendency in these authors to conflate questions of fact with questions of logical value, or the history with the philosophy of science. That this tendency is present, indeed inveterate, in all these writers is, as I have just indicated, quite widely recognized, and is no more than one could gather, if he could not see it for himself in each of them, and from the things they say about one another. And this tendency is a cause sufficient to explain the phenomenon of plausible irrationalism. For it is so powerful in us all, and so productive of confusion where criticism does not check it, that it is easily equal to the task of making irrationalism about science plausible. It has imposed on philosophers grosser absurdities than that before now: for example, it enabled Mill to find plausible his `proof' for the principle of utility. For my own part, at any rate, I have no doubt that this tendency is the main part of the answer, in general terms, to my question.
But it is a deficient answer just because it is in such general terms. What we want explained is a specific phenomenon of literary history: namely that some philosophy of science, which is irrationalist enough to generate reluctance to admit (A), is nevertheless made plausible to thousands of readers who would have no patience at all with an open assertion that no more is known now than in 1580, or that no one ever knows anything. Between such specific facts as this to be explained, and the very general tendency so far offered (and correctly, as I think), in explanation of them, there is too wide a gap.
To fill this gap what is required, clearly, is to show in detail how the general tendency to conflate the history with the philosophy of science is carried out in the writings of our authors, in such a way as to disguise their irrationalism and make it plausible. We need a catalog of the actual literary devices by which this trick is turned. It is this which I attempt to supply.
If you wish to recommend a philosophy of science to readers who are sure to find the irrationalism in it implausible, then your literary strategy must clearly be a mixed one. Irrationalism which was open and unrelieved would be found hopelessly implausible. So your irrationalist strokes must be softened, by being mixed with others of an opposite kind, or again by being disguised as themselves of an opposite kind. All our authors, accordingly, employ a strategy which is mixed in this sense; and in fact many forms of it.
An extreme form of mixed strategy is, simple inconsistency: that is, assert an irrationalist thesis, but also assert others which are inconsistent with it.
Popper furnishes many examples of this, of which the following is one. He staggers us by denying that positive instances confirm a universal generalization, but reassures us by allowing that negative instances are, as we always thought they were, disconfirmatory (so that for example "(x) (Raven(x) -> Black(x))" is disconfirmed by "Raven(a) and not(Black(a))", but not confirmed by its negation). For he adopts a criterion of confirmation  (one which I have elsewhere called the `relevance criterion' of confirmation ), which is well known to have the consequence that p confirms q if its negation disconfirms q.
A strategy which is mixed in the above sense while falling short of inconsistency, can take the form of stating as the aim of science something which common-sense would agree to be at least one of its aims; while also saying other things which imply that it is impossible to achieve this aim.
Popper and Lakatos both do this. They say the aim of science is to discover true laws and theories. But they also say, concerning any law or theory, that because it is universal, its truth is exactly as improbable, even a priori, as the truth of a self-contradiction : in other words, impossible.
A further form that a mixed strategy can take is this: embrace a methodology which is common-sense as far as it goes, but also say other things which imply that (even if it is possible) it is pointless to comply with it.
Popper does this. He enjoins our utmost efforts to establish empirically the falsity of any proposed law or theory. Yet no labor could be more pointless, if he is right in telling us that (for the reason mentioned in the preceding paragraph) the falsity of any such proposition is already assured a priori.
Yet another form that a mixed strategy can take is, of course, equivocation: leave them guessing what it is you really believe, the irrationalist bits, or the other ones.
Kuhn, for example, says that the world is the same after "paradigm-shift" as it was before ; that scientists working within different paradigms are nevertheless studying the same world ; etc. etc. Well, of course! He is not some kind of crazy Berkelian, after all, and these things are just common knowledge, like the proposition (A) from which I began, only more so. But Kuhn also uses every literary means short of plain English to suggest that these things are not so: that on the contrary, the world is somehow plastic to our paradigms.
Of course it is not always easy to tell equivocation from downright inconsistency. Take Kuhn again. Some one may tell me that he is not reluctant, at all, to admit my historical claim (A). And certainly it would be easy to point to many passages in his writings which support this interpretation. All those passages, for example, in which he says that normal science, operating under the guidance of a paradigm, solves problems. No doubt, in particular, Kuhn would admit that normal science has solved a great many problems since 1580. Well, if it has solved those problems, then those problems have been solved, haven't they? We know Kuhn says that a new paradigm "replaces", "destroys", the old one. But he never says that every solution of a particular problem, achieved under the old paradigm, somehow is "destroyed" or becomes an un-solution under the new. Indeed, how could that be? What would it even mean, to say so? If a problem has been solved then it really has been solved. But if this tautology is not denied, then solutions to problems (unless they were, for example, forgotten) would accumulate through successive paradigms. But what then becomes of Kuhn's famous rejection of the cumulative view of the history of science?
This may be another example, then, of our authors' mixed strategy issuing in an actual inconsistency. But on the other hand it may only be another case of equivocation. When Kuhn speaks of science as having solved problems, he no doubt often uses this phrase in the sense in which people normally understand it: which, whatever it is, may certainly be called an absolute sense. But---the idea naturally suggests itself---perhaps he sometimes also uses it in another and weaker sense: one which is more consistent with his repeated assertion that what constitutes the solution of a problem is relative to the paradigm, the group, and the time.
This suggestion (although I will not pursue it in connection with the phrase "solving problems") seems to me to furnish the key to the two main literary devices by which our authors make irrationalism about science plausible.
The first of these devices I call neutralizing success-words. A homely example will explain what I mean.
Nowadays in Australia a journalist will often write such a sentence as, "The Minister today refuted allegations that he had misled Parliament", when all he means is that the Minister denied these allegations. "To refute" is a verb with `success-grammar' (in Ryle's sense). To say the Minister refuted the allegations is to ascribe to him a certain cognitive achievement: that of showing the allegations to be false. "To deny", on the other hand, has no success-grammar. So a journalist who used "refuted" when all he meant was "denied" has used a success-word, but without intending to convey the idea of success, of cognitive achievement, which is part of the word's meaning. He has neutralized a success-word.
When journalists do this, no doubt they mostly do this inadvertently, from mere ignorance. But imagine the same thing done by a journalist who does know the meaning of the two words, and who believes that in fact the Minister only denied the allegations; but who feels for some reason obliged to use language which, in his own opinion, exaggerates the cognitive achievements of Ministers. (Perhaps the reason is that he thinks his readers will listen to nothing but good about Ministers). Then we would have what I believe is a very close parallel indeed to the way our authors use language to write about science.
For they use the language of success about science---words importing more or less cognitive achievement, such as "knowledge", "discovery", "facts", "verified", "understanding", "explanation", "solution (of a problem)", and a great many more besides---they use this language quite as freely as do any of those older historians of science whom they despise. They clearly must do so, at least to some extent, for they would forfeit all plausibility if they were to write about science without ever using any success-words at all. Their substantive philosophy, however, is not really consistent with applying, to science, such words in their ordinary success-implying sense. So while they use the language of science, they neutralize it. Not all the time, of course: sometimes they use these words in the ordinary sense, despite the inconsistency involved in doing so. But often enough for such neutralized success-words to be a prominent and distinctive feature of the English that they write.
This device is clearly one which, if it were used, would help enormously towards making irrationalist philosophy of science plausible. For in this way you can have, as thick as you like on every page, all the optimistic words of the old historiography and philosophy of science, reassuring the reader (who needs, after all, to be weaned gradually from whiggish notions of science) while all the time, nothing inconsistent with irrationalism need be said at all.
I now have to substantiate my suggestion that the device of neutralizing success-words is characteristic of our authors.
Before coming to cases it will be worthwhile to notice a passage in which the truth of this suggestion of mine is indirectly admitted at once, by one of our authors himself. This is a remarkable paragraph, occurring early in Against Method, in which Feyerabend, who is of course more openly irrationalist than our other authors, tells us that (to put it in my language), whenever he applies success-words to science in that book, they are never to be taken in their ordinary sense, but are intended to be always understood as neutralized.
The context is this: Feyerabend has just been expounding his `anarchist' maxim that anything goes: by which he means that any principle of theory-preference (induction, counter-induction, Tarot-card, or whatever) may on a given occasion advance science more than any other would. Then he adds the following:
"Incidentally, it should be pointed out that my frequent use of such words as `progress', `advance', `improvement' etc., does not mean that I claim to possess special knowledge about what is good and what is bad in the sciences and that I want to impose this knowledge upon my readers. Everyone can read the terms in his own way and in accordance with the tradition to which he belongs. Thus for an empiricist, `progress' will mean transition to a theory that provides direct empirical tests for most of its basic assumptions. Some people believe the quantum theory to be a theory of this kind. For others, `progress' may mean unification and harmony, perhaps even at the expense of empirical adequacy. This is how Einstein viewed the general theory of relativity. And my thesis is that anarchism helps to achieve progress in any one of the senses one cares to choose. Even a law-and-order science will succeed only if anarchistic moves are occasionally allowed to take place" .
It is surely obvious that this addendum to the anarchist methodology is not (what it seems meant to be) an addition which makes the anarchist methodology still more permissive. It just makes it totally pointless. We should perhaps think well of a man's heart, if he gives a million-dollar prize for an advance towards a cure for cancer, and says that anything goes as to what means (scientific, magical, or any other) are taken to that end. But if he adds that, `incidentally', anything goes, too, as to what counts as an advance towards a cure for cancer---that "everyone can read these terms in his own way"---then it will be impossible to think well of his head. It is not as though the second piece of permissiveness is an extension of the first: it simply takes all point out of it.
For my purposes, however, the main importance of this passage is this. It is an admission that Feyerabend's philosophy of science, if it were to be consistently expressed, would require that the success-implication of words like "knowledge" and "discovery", as well as of the weaker success-words he mentions himself, be always taken out. It is therefore a strong advance indication that at least in his philosophy we will find that those implications are often taken out; that is, that such words will be neutralized. And since what requires their neutralization there is the irrationalism he shares with our other authors, it is also an indication that they too will be able to be caught neutralizing success-words.
To be sure, Feyerabend does not do what he said he would. Having undertaken to neutralize all success-words, he promptly forgets all about his undertaking, and when it suits him, as it often does, writes about the history of science like any mere Sarton, Wolf, or Pledge. "It is now known that the Brownian particle is [...]"  etc., etc. That is, he often uses words like "known" with their ordinary success-grammar. This was to be expected. It is just another instance of that mixed strategy which all our authors are obliged, as I have said, to employ.
But to come to details.
One way to neutralize a success-word is to put it in quotation-marks. Thus, in certain circumstances a journalist might write "The Minister `refuted' the allegations", meaning, and being understood to mean, that the Minister did not refute but only denied them. This might be thought a device too unsubtle for authors such as ours to have made use of. It is not so, however. In any case some variations on the device are not altogether without subtlety. One such variation is what I call "suspending" success-grammar: putting a success-word in quotation-marks, not necessarily in order to neutralize it, but just with the intention, or at least the effect, of leaving the reader uncertain whether you have neutralized it or not. (This is the effect momentarily produced by signs advertising `fresh' fish). Another variation is, using the same success-word several times in close succession, and sometimes putting it in quotation-marks and sometimes not, but with no reason that the reader can discover for so doing. Such variations as these can achieve, partially or gradually, that separation of a success-word from its success-meaning, which quotation-marks sometimes achieve completely and abruptly. They are devices, therefore, which are not at all too unsubtle, nor yet too subtle, to be of some use to a philosopher interested in making irrationalism about science plausible. It would be no use for such a philosopher, and everyone now knows it would be no use, to cry "stinking fish" about science. But it may well be some use for him to praise science as "`fresh' fish"; especially if he does it often enough.
Lakatos has certainly done it often enough. Enclosing success-words in quotation-marks was in fact a kind of literary tic with him. He could scarcely have gone to more extravagant lengths in the use of this device, if he had been trying to bring it into disrepute; which, however, he certainly was not.
Take his Proofs and Refutations. The first word in this title is of course a success-word. In the book it is subjected countless times to neutralization or suspension of its success-grammar by quotation-marks. Often, of course, perhaps equally often, Lakatos uses the word without quotation-marks. But what rule he goes by, if he goes by any rule, in deciding when to put quotation-marks around "proof" and when to leave them off, it is quite impossible for a reader of that book to discover. Nor does the reader know what meaning the writer intends to leave in this success-word. He knows that the implication of success is often taken out of it; or rather, he knows that on any given occurrence of the word in quotation-marks, this implication may have been taken out of it. But what meaning has on those occasions been left in it, he is entirely in the dark. Indeed, by the end of the book, or even half-way through it, the reader no longer dares attach success-grammar to "proof" or "proved", even when they occur without quotation-marks. Will any reader of Proofs and Refutations undertake to say what the first word of the title means in the book?
By the time Lakatos came to write about empirical science, his tic had got worse. I draw an example from `Falsification and the Methodology of Scientific Research Programmes'. One short example will suffice, because Lakatos's English is everywhere much the same, and anyone familiar with it will recognize in the following a representative specimen of it.
"One typical sign of the degeneration of a programme which is not discussed in this paper is the proliferation of contradictory `facts'. Using a false theory as an interpretative theory, one may get---without committing any `experimental mistake'---contradictory factual propositions, inconsistent experimental results. Michelson, who stuck to the ether to the bitter end, was primarily frustrated by the inconsistency of the `facts' he arrived at by his ultra-precise measurements. His 1887 experiment `showed' that there was no ether wind on the earth's surface. But aberration `showed' that there was. Moreover, his own 1925 experiment (either never mentioned or, as in Jaffe's  misrepresented) also `proved' that there was one (cf. Michelson and Gale  and, for a sharp criticism, Runge )" .
Here, in the space of seven lines of print, Lakatos manages to neutralize by quotation-marks three success-words, two of them twice each: "facts", "showed", and "proved".
The effect on the reader is characteristic. An episode in the history of science has been described to him, and it is described, as we see, entirely in words importing cognitive achievement. Yet by mere dint of quotation-marks, every single implication of cognitive achievement has at the same time been neutralized or suspended. The reader, remember, almost certainly has no knowledge of his own of the episode as would enable him to object, for example, that Michelson really did show one of the things that Lakatos says he "showed". Nor has the reader any idea, as I said before, how much if anything of the ordinary meaning of the various success-words the writer is leaving in them: he only knows that their success-implication has been, or may have been, taken out. What, then, will the reader be able to carry away from this passage? Nothing at all; except a strong impression that despite all the success-words used in describing it, there was, in this presumably representative episode from the history of science, no cognitive achievement whatsoever.
This passage is a very model of irrationalist philosophy of science teaching by example, and being made plausible by example. Yet it depends entirely for its effectiveness on a device at first sight so trivial as the use of quotation-marks to neutralize success-words.
Where Lakatos raises storms of neutralizing quotation-marks, Feyerabend, in Against Method, just keeps up a steady drizzle of them. For this reason short passages cannot be quoted from him to such effect as they can be from Lakatos. Feyerabend, as we saw, does not keep his promise to neutralize all success-words, but still he often does neutralize them; and when he does it is often by means of quotation-marks. The word "facts", for example, is often thus neutralized: for example, on pp.40,41,46,47. But he does not neutralize only strong success-words. Any success-word, however weak its success-implication may be, or any word which has even an indirect connection with cognitive achievement, he is likely to sprinkle with this soothing balm. For example, "success", pp.44; "truth", pp.28,171; "progress", pp.27,296; "objective", pp.19,181; "rational", pp.154,190,198. There will be no hint left in science of anything so hurtful and undemocratic as success, if Feyerabend can help it.
Popper has always made a certain amount of use of quotation-marks for neutralizing success-words. It is well-known, for example, that though he has always been sure that scientific theories can be disconfirmed, he is still not sure, after fifty years, whether our best-confirmed theories are confirmed, or only `confirmed' . He knows that when he puts quotation-marks around `confirmed', he incurs the suspicion of Humean irrationalism. But then, if he leaves them off it, people may suspect him of believing in non-deductive logic. Hence his indecision.
Kuhn hardly ever resorts to quotation-marks when he wants to neutralize a success-word.
The easiest way, however, to neutralize a success-word, is---just to do it: "bald neutralizing", I will call it. That is, just to use a word which implies cognitive achievement, as though it did not. Set at defiance all mere logicians, Oxford philosophers, accurate speakers, and pedants generally. It is even money, after all, whether your solecism will even be noticed; and with luck it may even catch on.
For the sake of plausibility, of course, you should not do this all the time. At any rate not all the time to all words which, like "knowledge", have the strongest implication of success and are at the same time nearly indispensable for writing about science. If a word is comparatively dispensable, or has comparatively weak success-grammar, you may be able to get away with baldly neutralizing it every time.
Bald neutralizing is, in Lakatos, subordinated to his main weapon for the destruction of scientific success, the quotation-mark. But in all our authors it is common, and is one of the distinguishing features of their English. I will shortly prove this by examples, in connection with two of the strongest success-words, namely "knowledge" and "discovery". And if this can be done then I can fairly be excused, I think, from documenting the execution our authors do on the weaker and more defenseless members of the success-tribe: "confirmation", "explanation", "understanding", "scientific progress", and the like. The execution is terrific, as may be imagined. When the most emphatic of success-words, such as "knowledge", can be murdered with impunity in open day, as they are by our authors, then the quiet extinction of weaker ones will never attract criticism, or even attention. But to document this process in detail would clearly take far too long.
First, then, Kuhn on knowledge. He says that on the cumulative view, "in the evolution of science new knowledge would replace ignorance" , but that this is quite wrong. What really happens is that one paradigm replaces another, and then "new knowledge [...] replace[s] knowledge of another and incompatible sort" . Kuhn writes, therefore, as though some knowledge can be incompatible with other knowledge; and indeed, on his views, such incompatibility must not only be possible, but common in the history of science and even of the essence of it. It is not possible, however: this is just baldly neutralizing the word "knowledge". Knowledge implies truth, and truths cannot be incompatible with one another.
Again, Kuhn simply takes the truth-implication out of the word "knowledge" when he writes, for example, in his most overtly relativist vein, that every scientific theory now discarded (such as Ptolemaic astronomy) possessed in its heyday "the full integrity of what we now call sound scientific knowledge" .
The word "discovery", too, Kuhn baldly neutralizes at his pleasure. To discover what is not true, or what does not exist, is certainly no mean feat; or rather, it is a simple logical impossibility, forbidden by the success-grammar of the verb "to discover". Yet the history of science as Kuhn recounts it, contains "discoveries" of what is not true; and again, such things must be in fact extremely common on his views. Here is one example, not the only one. "Given Galileo's paradigms, pendulum-like regularities were very nearly accessible to inspection. How else are we to account for Galileo's discovery that the bob's period is entirely independent of amplitude, a discovery that the normal science stemming from Galileo had to eradicate and that we are quite unable to document today" .
Feyerabend's promise to neutralize all his success-words in Against Method is carried out baldly enough in some cases. For example, on the word "facts" on pp.29--33, and on the word "knowledge" in the following representative passage.
"Knowledge so conceived is not a series of self-consistent theories that converges towards an ideal view; it is not a gradual approach to the truth. It is rather an ever increasing ocean of mutually incompatible (and perhaps even incommensurable) alternatives, each single theory, each fairy tale, each myth that is part of the collection forcing the others into greater articulation and all of them contributing, via this process of competition, to the development of our consciousness" .
The "so-conceived" in the first line here means, "as I, Feyerabend, conceive it". Words meaning what they do, however, his `conception' is mere nonsense. It may be true, or at least intelligible, to say that `knowledge' is an ocean of incompatible etceteras, or that what passes for knowledge is an ocean of incompatible etc. But it makes no sense to say that knowledge is an ocean of incompatible etc., or even (what presumably Feyerabend meant) that the objects of knowledge are an ocean of incompatible etc. Knowledge entails truth, and truth entails possible truth, and possible truth entails compatibility. These are facts about the meaning of common English words, and facts which are, in themselves, not especially important. They are facts, though, and because they are, you might as well say that knowledge is a poached egg, as say Feyerabend says about it here.
All our authors except Popper, it should be understood, not only exercise but more or less openly claim the right to talk nonsense. Feyerabend would exclude no one from this right. (But then he is all heart and "would not hurt a fly" ). He thinks that talking nonsense is just good for you, like many other things which are familiar to us all, and the value of which no well-disposed person denies: like rotation of crops, state control of scientists, and turning yourself into a wolf and back again. Lakatos is far more exclusive. Talking nonsense, when it is done by people he approves of, he calls "language-breaking", and he hints that all the very best people do it. Certain great scientists, he implies, have possessed this gift for language-breaking , and it would be a dull reader indeed who could not name one other person that Lakatos thinks is gifted in the same way. Kuhn in his more demure style merely warns us in his Introduction that what he says "strains customary usage" ; which, when you think about it, is at any rate not more than the truth.
Yet I have no doubt that Kuhn and Lakatos (Feyerabend may be different) would react just as any other philosopher would, if they were told by someone else, such as a mere undergraduate in an essay, that knowledge is a poached egg; or, say, that knowledge entails falsity; or that belief entails knowledge. But what could they or anyone say to such a student, except that, words meaning what they do, what he has written either makes no sense at all, or at best is necessarily false? And that is what we must often say about what they write.
For sheer bald neutralizing of success-words, however, Popper remains in a class of his own. It is reasonable to believe, indeed, in view of his extensive influence on our other three authors, that it was from him that they learned what skill they have in this art. Anyway Popper has left monuments of the art which are not likely ever to be excelled. He actually seems to prefer neutralizing the very strongest of success-words, and to prefer to do it as publicly as possible: that is, in the very titles of his books and articles.
The title of his most famous book in its English translation is a uniquely daring instance of the use of the old optimistic language of the historiography and philosophy of science (the rationalistic and authoritarian language, Lakatos and Feyerabend would say) to introduce a book which, by its actual contents, did far more than any other intellectual cause to discredit that language, and to inaugurate the irrationalist revolution in the historiography and philosophy of science. Thousands of readers have noticed this fact, so far as it concerns the use in that title of the word "logic"; and even Lakatos remarks on the "paradoxicality"  of the title in that respect. But my present concern is with the other part, because of the success-word it contains.
"The Logic of Scientific Discovery", indeed! There is scarcely a word in it, or in anything else Popper ever wrote, about scientific discovery, and the reason is as simple as it is sufficient. "Discovery" is a success-word, and of the strongest kind: it means the same as "discovery of what is true or of what exists". The history of science, therefore, to the extent that it has been a history of discovery---as it has been so markedly in the last four hundred years, for example---is a history of success. But that is not the way that Popper sees the history of science, far from it. For him the history of science is a succession of `problems', `conjectures and refutations', Socratic or Pre-Socratic dialogues, `critical discussions'. It is all talk. In this context any vivid reminder of an actual scientific discovery would be as out of place as a hippopotamus in a philosophy class. The only thing worse would be a reminder (though this would be too horrible) of what whig historiography used so often to bracket with scientific discoveries: inventions. Popper is perhaps the first person to see, in the glorious history of scientific discovery, nothing more productive and exhilarating than a huge W.E.A. philosophy class, and one which, to add to its charms, might go on forever. Does anyone suppose that Popper ever wrote or meant to write a book for which a non-misleading title would have been "The Logic of Scientific Discovery of Truth, or of what Exists"? Yet that is a purely analytic extension, only objectionable on aesthetic grounds, of his actual title. But clearly this title would belong, in the history of thought about science, in the heyday of the `whig supremacy', probably somewhere between J.S.Mill and Samuel Smiles, and it sounds a good deal more like the latter than the former.
No, the right title for that book---and it is of some importance to realize that I am here only saying what everyone familiar with its contents has been at least half-conscious of all along---would have been "The `Logic' of Scientific `Discovery'". But of course that would have been too openly irrationalist. Better to let the word "discovery" stand, and trust to the contents of the book, rather than to quotation-marks in the title, to neutralize the unintended implication of success. Which duly happened, and never a word said.
It is the word "knowledge", however, which was the target of Popper's most remarkable feat of neutralization. This word bulks large in his philosophy of science (much larger than "discovery"), and in recent years, in particular, the phrase "the growth of knowledge" has been a favorite with him and with those he has influenced most. Some people have professed to find a difficulty, indeed, in understanding how there can be a growth-of-knowledge and yet no accumulation-of-knowledge. But then some people cannot or will not understand the simplest thing, and we cannot afford to pause over them. Let us just ask, how does Popper use the word "knowledge"?
Well, often enough, of course, like everyone else including our other authors, he uses it with its normal success-grammar. But when he wishes to give expression to his own philosophy of science he baldly neutralizes it. Scientific knowledge, he then tells us, is "conjectural knowledge". Nor is this shocking phrase a mere slip of the pen, which is what anywhere else it would be thought to be. On the contrary, no phrase is more central to Popper's philosophy of science, or more insisted upon by him. The phrase even furnishes, he believes, and as the title of one of his articles claims, nothing less than the "solution to the problem of induction" .
In one way this is true, and must be true, because any problem clearly must yield before some one who is prepared to treat language in the way Popper does. What problem could there be so hard as not to dissolve in a sufficiently strong solution of nonsense? And nonsense is what the phrase "conjectural knowledge" is: just like say, the phrase "a drawn game which was won". To say that something is known, or is an object of knowledge, implies that it is true, and known to be true. (Of course only `knowledge that' is in question here). To say of something that it is conjectural, on the other hand, implies that it is not known to be true. And this is all that needs to be said on the celebrated subject of "conjectural knowledge"; and is a great deal more than should need to be said.
In all our authors there is another misuse of language, and one which is even of an opposite kind to that of neutralizing success-words, the explanation of which is nevertheless furnished by that very process. The most striking instance of it is Popper's misuse of the word "guess".
He says that "we must regard all laws and theories [...] as guesses" . Taken on its own this would be an inexplicable thing for anyone to say. For who is so ignorant, or so irrationalist, as to believe that? Recall what a guess is. A paradigm case of guessing is, when captains toss a coin to start a cricket match, and one of them `calls', say "heads". This cannot be a case of knowledge, scientific knowledge or any other, if it is a case of guessing. If the captain knows that the coin will fall heads, it is just logically impossible for him also to guess that it will. More than that, however: guessing, at least in such a paradigm case, does not even belong on what may be called the epistemic scale. That is, if the captain, when he calls "heads", is guessing, he is not, in virtue of that, believing, or inclining to think, or conjecturing, or anything of that sort, that the coin will fall heads. And in fact, of course, he normally is not doing any of these things when he guesses. He just calls. And this is guessing, whatever else is.
Now, does Popper believe that what he calls "the soaring edifice of science"  is built out of cricket captains' calls of "heads" and "tails"; or of other things in the same epistemic, or rather non-epistemic, boat? Presumably not. Then what has happened?
Simply this. If, when you talk about science, you insist on neutralizing success-words, depressing them (as it were) on the epistemic scale, then in the interests of plausibility you will find yourself obliged, as if by a kind of hydraulic compensation, to elevate in the epistemic scale some non-success-words, and even to promote onto the epistemic scale other words which do not belong there at all. The former is what has happened to some extent to the word "theory" in all our authors. The latter is what has happened to the word "guess" in Popper.
This compensatory process is not confined to philosophers, but appears to have already affected for the worse the language of scientists themselves. An especially common instance is this: a scientist will say that p is consistent with q, when what he means, and is understood by other scientists to mean, is that p confirms q. For example, when what he means and is understood to mean is that the red-shift of light from other galaxies confirms the hypothesis that those galaxies are receding from ours, he will say instead that the red-shift is consistent with galactic recession. The absurdity of such a remark, if "consistent with" has its usual sense, is only too evident. Obviously red-shift is consistent with the hypothesis of galactic recession; it is consistent with the absence of galactic recession too; almost every proposition, come to that, for example "Socrates is mortal", is consistent with galactic recession; and with its negation. But it is evident enough, too, what the pressure is, that leads to such statements being made. The influence of Popper's irrational philosophy in particular, and of `modern nervousness' in general, is so widespread and powerful, that even scientists now prefer to steer clear of a word which is even so weakly suggestive of cognitive success as "confirms". So, although "confirms" is what they mean and are understood to mean, they say instead "is consistent with".
This phenomenon, although it is of course yet another abuse of language by our authors and those whom they influence, is one which, as far as it has gone, is rather encouraging than otherwise. For it suggests that, at least until the final triumph of irrationalist philosophy of science, language will to some extent obey a `law of conservation of success-grammar'. That is, if you empty all success-grammar out of certain words, some of it is going to seep into other words, including some which were quite devoid of it before. Still, this process has not gone very far up to the present, and is not likely to go further than it already has. On the whole our authors' efforts to eradicate belief in scientific success have been remarkably successful.
In all our authors except Popper there is yet another process which is a natural complement of that of neutralizing success-words, or rather is a simple logical extension of it. This is the neutralizing, though in the opposite direction as it were, of failure-words when they are applied to science: words like "error", "mistake", "is refuted", "is falsified", and so on. With failure-words, or at least with the strongest ones such as those just mentioned, neutralization typically consists, of course, in removing the implication, which is part of the meaning of such words, of the falsity of the proposition in question.
This kind of neutralization can be done, just like the other, by means of quotation-marks, or again baldly. It is done in both ways by our authors. For an example of the former the reader need only look back to the quotation to which footnote 17 above is appended. There he will see, what by now he could predict, that when Lakatos used the phrase "experimental mistake", he was quite unable to keep his quotation-marks off it.
With failure-words as with success-words, it was Popper who showed the way ahead. For he labored long to persuade scientists that no professional stigma attaches to their being refuted. Nor did he labor in vain, but rather to such effect that he succeeded in persuading some of the sadder Popperian scientists that to be refuted was actually the goal of all their endeavors. (They appear to have had rather successful careers). Yet Popper had only a Pisgah-view of this matter, because he never neutralized the implication of falsity in saying, not of a scientist but of a proposition, that it "is refuted" or "has been falsified". On the contrary, that implication was for him the whole point of such phrases.
To take this great leap forward has been left to Lakatos and Feyerabend. As other public benefactors removed the social stigma from illegitimacy, these two reformers have removed the stigma of falsity from refuted propositions. It is true that in these early days (just as happened when Popper first neutralized success-words in the very first days of the revolution), irreconcilable bourgeois elements complain that they do not know what meaning is left in "refuted" and the other strong failure-words, now that the implication of falsity has been taken out of them. Well, no doubt these temporary shortages of meaning are unfortunate; but then, a revolution in the philosophy of science is not a tea-party. That the falsity-implication has been taken out of failure-words, however, the reader can satisfy himself on scores of pages of Lakatos and Feyerabend. Or if the reader prefers to take his instruction in newspeak neat, here it is in so many words: "If a theory is refuted, it is not necessarily false" . For these authors at least, the era of "falsism", as the odious old practice of discrimination against refuted propositions is now called (or soon will be), is over.
On failure-words Kuhn is, as he usually is, less open, and far more chilling, than even Lakatos and Feyerabend.
If the reader looks back to the question to which footnote 22 above is appended, there is one thing he can hardly fail to notice: the distance that Kuhn will go about in order to avoid saying that some one working with a paradigm since replaced, as Galileo was, believed, because of that paradigm, something which is not true. Kuhn himself draws attention, three pages from the end of his book , to the fact that up to that point he had not once used the word "truth". This avoidance seems understandable. "Truth" is the most stripped-down of all success-words, and therefore does not lend itself to being neutralized in its turn. So the only thing an irrationalist philosopher of science can do with it is to avoid it as long as he can.
But wait a moment: is this so understandable? Kuhn is as lavish as the next man in applying to science words like "knowledge" and "discovery", which all import truth anyway: so what can be the point of this fantastic punctiliousness about avoiding the word "truth"? But you only need to recall, first, that along with our other authors Kuhn is engaged in taking the success-implication out of those success-words which he does not simply avoid; and second that, on pain of losing all plausibility, you cannot take the success-grammar out of, or avoid, all the success-words all of the time. Then this puzzling fragment of literary history will become entirely understandable once more.
But Kuhn is not content to avoid expressions like "not true" and "false" himself. When one of our authors permits himself to employ, in connection with scientific theories, words which imply falsity, Kuhn is up in arms against him. And for what? Why, for his misusing language! For his "odd usage" or words; for saying something "difficult to understand"; for perpetrating a "gross anachronism"; for permitting questions to be raised which are not even "sensible". Truly, there is nothing in any of our authors, not even Feyerabend's belief in lycanthropy, more staggering than this.
The quotations in the preceding paragraph, and in the present one, are all from Kuhn [1970a] pp.10--13, where he is criticizing Popper for using, in connection with science, phrases such as "trial and error" and "learning from our mistakes". "Mistake" and "error" both, you see, import falsity. Kuhn thinks that to call an out-of-date scientific theory, such as the Ptolemaic astronomy, "mistaken" or "a mistake", will "immediately seem an odd usage". Kuhn says: "it is difficult for me to understand what Sir Karl has in mind when he calls that system, or any out-of-date theory, a mistake;" Indeed, Kuhn says, to call any out-of-date theory mistaken would be "so gross an anachronism" that no one of "sound historic instinct" is likely to be guilty of such a lapse. To speak in that way, he says, is to invite questions which are not even "sensible", such as: "What mistake was made, what rule broken, when and by whom, in arriving at, say, the Ptolemaic system?"
Was there ever effrontery, if that is what this is, at once so bold and so hollow? Or a challenge so easily met?
Here is one of the many mistakes which was made in arriving at the Ptolemaic astronomy: the belief that the sun goes round the earth each day. As to propriety of language, consider the sentence: "The Ptolemaic system of astronomy is false". There is no `odd usage' in that: it is good common English. Nor is there any anachronism in it. (One cannot help wondering what Kuhn thinks the word "anachronism" means). What this sentence says is, quite obviously, what Popper `had in mind' when he called Ptolemaic astronomy a mistake. What it says is, moreover true and well-known to be so. And if Kuhn really believes that no out-of-date scientific theory---no theory, that is which was accepted once but is not now---can properly be called mistaken, then he must also accept the consequence: that every scientific theory which can properly be called mistaken is accepted now if it ever was. A consequence none too consistent with that `growth of knowledge' which he, like our other authors, is always willing to acknowledge, at least in words.
We should for a moment try, though it is almost impossible, to take in the full grotesqueness of the contemporary situation in the philosophy of science. We have already encountered Popper, a grown man and a professor, implying that it is a guess---that is, something like a cricket captain's call of "heads"--- that the sun does not go round the earth every day. But here is Kuhn, perhaps the most learned and certainly the most influential of living historians of science, writing in such a way as to imply that, like a great many people in 1580 and a few uncommonly ignorant ones even now, he does not know that it is false that the sun goes round the earth every day! And implying too, what is far worse still, that to say that he or anyone else does know this would be a glaring misuse of language.
I am sure that this was not effrontery. It is simply a revelation, and all the more terrifying for having been made inadvertently, that Kuhn has simply lost contact with the meaning of common English words (such as "false"), and now knows only the vocabulary of his own irrationalist philosophy of science.
This must be very close, at least, to the end of the line. Non-cognitivist philosophies of morals are one thing; but here we have a non-cognitivist philosophy of science. Our philosophy of science, as I remarked earlier, lost contact long ago, at least as early as Popper, with the refreshing realities of scientific discovery and invention: with the actual objects of science. But with Kuhn even the intensional objects of science, the propositions of science, have vanished into thin air, and with their disappearance, of course, the cognitive aspect of science vanishes too. Science, it turns out, whatever may be believed to the contrary by the vulgar and by whig historians, is really as intransitive as sleep.
This conclusion is willingly embraced, or at least is implied, by all our other three authors. Thus Feyerabend loves to write such phrases as "science, religion, prostitution and so on" , and says that science has no more "authority [cognitive or other] than any other form of life" . The same non-cognitive conception of science is also what allows Feyerabend to demand that scientific laws be put to the vote ; and again is what authorizes him, in the middle of mystifying his readers of science, or rather, as a part of that process, to bore them about art . The same conception of science as having no cognitive aspect, nothing to do with knowledge or belief, is implied by Popper and Lakatos. For while knowledge entails belief, they both insist that science has nothing to do with belief .
His non-cognitivism is of course the reason why Kuhn can, and even must, sentence all present and future philosophers of science to the torments of the damned: that is, to reading the sociology of science. If he were right there would indeed be nothing else for them to do. This prospect inspires in our other authors the horror it should. Yet they could not without inconsistency deny the justice of the sentence.
Helps to Young Authors (I)
Neutralizing success words, after the manner of the best authorities
How to rewrite the sentence: Cook discovered Cook Strait.
[ Previous: Part One | Next: Chapter II | Table of Contents ]