by Antony Flew
3.6 I write both “apriori” and “aposteriori” as single, unitalicized words deliberately. It is chauvinistically, you might even say bigotedly, purist to treat as unnaturalized aliens immigrants first landed over a quarter of a millennium ago. But it may perhaps help in remembering the meanings of these two contrasting terms to say that both “apriori” and “aposteriori” derive from Latin, the former meaning from before or earlier and the latter meaning from after or later.
3.7 If we put things in terms of this fundamental distinction between the apriori and the aposteriori, then the No-true-Scotsman Move consists in responding to the falsification of a contingent proposition by covertly so reinterpreting the words in which it was originally formulated that these now become the expression of an arbitrarily constructed necessary truth. This maneuver always involves either a high or a low redefinition of a crucial term. Where the qualifications for membership of the class are increased, we have a high redefinition. Where they are reduced, we have a low redefinition. To rate as a true Scot you have to be not merely a Scot but a Scot who is not a sex maniac. To score—if that is the word—as one of the hateful exploiters you have to be not merely a Caucasian but a Caucasian who is not a Latin American Communist. What must have been an example of the low redefinition of the word “socialist” was provided by the British Queen Victoria’s eldest son, the future King Edward VII, when he—of all people—said in a speech at the Mansion House, in the City of London, the financial center of the then–British Empire: “We are all socialists nowadays” (November 5, 1895).
3.8 The temptation here is not just to slide, under the pressure of falsification, from an interpretation of some form of words which construes it as expressing a contingent to one which construes it as expressing a necessary proposition. It is to fail to recognize what has happened and, hence, to be apt to slide back again into the original interpretation immediately after that pressure is removed. Our Scottish nationalist is all too likely to go away believing that his original overconfidence in Scottish superiority has been vindicated; while Mr. Carmichael’s starkly confrontational picture of the world was in his own mind probably preserved without the revision which he himself had in his own way admitted that the existence of the Castro regime made necessary. It is important to be aware of the possibility of analogous equivocations even where we have not been alerted by a flagrant No-true-Scotsman Move (see paragraphs 1.14–1.15).
3.9 Suppose, for instance, that someone assures us that all criminal behavior or all sexual deviation or all suicidal desires and decisions are symptomatic of mental disease. Such things are said frequently nowadays, and not least by people paid to know about mental disease and mental health. The first question to press upon such psychiatric spokespersons, before we begin to ask whether what is said is true, is What sort of thing actually is being said? Are we really confronted by a genuine contingent claim, or is it all ultimately a matter of definition and of more or less arbitrarily chosen criteria? That is to say, are these spokespersons claiming to have made a discovery; the discovery perhaps that all those who commit crimes are also, by quite independent criteria, mentally diseased? Or is it that they are taking the committing of a crime as itself a criterion of mental disease, and in effect insisting that such behavior constitutes a logically sufficient condition (see paragraphs 2.18–2.21) of being mentally diseased? In that case it would for them be contradictory to say that a man is committing a crime and yet is not mentally diseased.
3.10 In short, is it as it used to be when suicide under English law was still a crime? Then juries, without hearing any expert psychiatric evidence, regularly returned the verdict: suicide, while the balance of the mind was disturbed. The reason for returning these unevidenced verdicts was to protect the suicides’ heirs from the loss of their inheritances. For it was illegal for anyone to benefit in consequence of either their own or of someone else’s crime. (For substantial discussion of the concepts of mental disease and mental health see, for instance, Flew 1993, chapter 7.)
3.11 Consider another example, which again illustrates the importance of distinguishing between analytic necessary and synthetic contingent interpretations of the same form of words. It is provided by an argument which has been widely current since as early as 1651, the year in which Thomas Hobbes (1588–1679) published his Leviathan. Whether or not Hobbes himself did ever present this argument, it certainly was thereafter generally credited—or perhaps we should say discredited—to him. The premise is that all actions must as such be motivated. But in that case, it is urged, that all actions are performed and could only be performed because the agent wants to perform them. Since to do always exactly what you want to do regardless of the proper claims of anyone else is to be utterly selfish, it appears to follow that there is and could be no such thing as an unselfish action.
3.12 Wait a minute! The premise is presumably construed as analytic and necessary. Certainly no other grounds are offered why we should accept it as true. But the conclusion is equally certainly put forward as a scandalous and demoralizing revelation. This by itself is enough to show that there must be something wrong. It cannot be right to infer a substantial conclusion from a merely tautological premise. Exactly what is going wrong and how it is going wrong can best be seen by displaying the ambiguity of the crucial term “want.” In the premise to want is simply to have a desire or a motive, any desire or motive at all. It is only this equation which permits us to go on to say that all actions “are performed and could only be performed because the agent wants to perform them.” But in the conclusion “to do always and only exactly what you want to do” has to be taken in a different and stronger sense: the sense in which someone might claim to have gone to visit a sick relative although that person did not at all want to do so. It is only in this interpretation that the proposed conclusion becomes exciting and surprising and carries the shocking implication that “there is and could be no such thing as an unselfish action.” So construed it is, however, not supported by the premise; and in any case, as my illustration must have suggested, it is known to be false.
3.13 Here it is desirable also to distinguish between self-interest and selfishness. To say that someone’s conduct was selfish is to say more than that it was self-interested. For selfishness is always and necessarily to be condemned. But surely self-interest is not? For example, when two healthy siblings eagerly eat their dinners, it would presumably be correct to say that each is in this matter pursuing his or her self-interest. Yet this is no sufficient reason to start reproaching them. The time for that would be after brother has grabbed and eaten his sister’s dinner as well as his own or has perhaps in some less flagrant way refused to respect the proper claims of someone else. Even when my success can be gained only at the cost of the failure of others, it would be inordinately austere to insist that it is always and necessarily selfish for me to pursue my own self-interest. For is anyone prepared to assert that all rival candidates competing for appointment to some coveted position are culpably selfish in not withdrawing in order to clear the way for the others?
3.14 The failure, or rather the willful refusal, to observe the distinction between self-interest and selfishness is the prime source of perennial misrepresentations of Adam Smith (1723–1790) and his followers as apostles of selfish greed. Certainly the Nobel laureate Chicago economist George Stigler was absolutely right to describe Smith’s Inquiry into the Nature and Causes of the Wealth of Nations (1776) as “a stupendous edifice erected upon the granite of self-interest.” But even if anyone has tried, certainly no one has succeeded in finding in that work any passage advocating selfishness. And there is, inevitably and obviously, much to the contrary in Smith’s other masterpiece, The Theory of the Moral Sentiments (1759).
3.15 A third example of alternations between analytic necessary and synthetic contingent interpretations of the same or very similar forms of words is provided by Joan Rockwell’s Fact in Fiction (1974), a work boldly subtitled The Use of Literature in the Systematic
Study of Society. In the preface she writes: “My basic premise is that literature neither ‘reflects’ nor ‘arises from’ society, but rather is an integral part of it, and should be recognized as being as much so as any institution, the Family for instance, or the State” (p. vii). So long as this is interpreted as in effect a matter of definition, it is clearly only too true. But in her very next sentence she slips toward a more substantial interpretation of this “basic premise.” In it she asserts: “Narrative fiction is an indicator, by its form and content, of the morphology [shape] and nature of a society. . . .” In the main body of the book we find that first denial completely forgotten. It is no longer false but, it seems, necessarily true that literature “reflects” the society within which it arises: “To say that writers necessarily reflect their own times, which I must repeat is the justification for using their fictions to study the facts of their society, is to say that they are bound to do so, and cannot choose to do otherwise” (ibid., p. 119).
3.16 It is one thing, and quite indisputable, to say that anything which happens in society is, by definition, a social phenomenon: “my basic premise is that literature . . . is an integral part of ” the society in which it is produced. But it is quite another thing, and not one to be accepted without the deployment of a deal of evidence, to say that literary products always contain information about the societies in which they were produced: “writers necessarily reflect their own times.” Maybe they do. Maybe it is impossible for even the best of historical novelists or of science fiction writers to cover their parochial social tracks. But to show this it is certainly not sufficient to appeal to the quasidefinitional truth that the production of literature is a social phenomenon.
3.17 The last thirteen paragraphs (3.4–3.16) have explained the fundamental distinction between logically contingent and logically necessary propositions and have shown how, by failing to observe it, we can be most powerfully tempted fallaciously to infer substantial, logically contingent conclusions from premises constitutionally incapable of yielding conclusions of that kind. The No-true-Scotsman Move is an attempt to evade falsification. Through it a piece of sleight of mind replaces a contingent by a logically necessary proposition. We need next to relate the criticism of such maneuvers with the Popperian ideas presented and criticized in chapter 2 (see especially paragraphs 2.10–2.13), and by these means bring out why any such maneuver is inconsistent with a forthright concern for truth.
3.18 The heart of the matter is that the substance and extent of any assertion always is and must be exactly proportionate to the substance and extent of the corresponding denial necessarily involved in the making of that assertion. To assert a proposition is as such to deny the contradictory of that proposition (p is therefore equivalent to the negation of the negation of p, symbolized as p≡~~p). So the more you want to say, the more you have to stick out your neck. The wider and the more substantial your claims, and the greater the scope and the explanatory power of the theory which you propose, the greater must be the range of theoretically possible occurrences which would, if they actually occurred, falsify your claims or your theory. In the words of a fine Spanish proverb: “ ‘Take what you like,’ said God, ‘take it, and pay for it.’ ”
3.19 The No-true-Scotsman Move can thus be seen as one of too many alternative maneuvers attempting to get hold of a big dollop of substantial truth without paying for it. The original contingent claim was wide and substantial. What it denied made it falsifiable by the misbehavior of any Scot; and we are supposing it actually was falsified by the deplorable conduct of our imaginary Mr. Angus MacSporran. The substitute assertion was not in the same way substantial. Precisely because it could not be falsified by any describable occurrences, it was not really making any assertion at all about what does or does not happen in the universe around us. This is what was meant when, in that enigmatic masterpiece, the Tractatus Logico-Philosphicus (1922), Ludwig Wittgenstein (1889–1951) wrote: “The propositions of logic are tautologies. The propositions of logic therefore say nothing. (They are the analytical propositions.)” (§§ 6.1–6.11).
3.20 The logical observation of paragraph 3.18 constitutes another element in the Popperian philosophy of science, which, as we saw in chapter 2 (see paragraph 2.11), insisted that “The best we have or ever could have can and could only be the best so far.” That surely too absolute rejection of all ultimacy in science, a rejection inspired especially by Albert Einstein’s overthrowing of the Newtonian establishment, constitutes no license to abandon the critical pursuit of truth. Quite the reverse is true. Certainly science requires openness and bold conjectures. But the openness required is openness to the possibility of new and possibly upsetting discoveries of what actually is the case, while the conjectures must be conjectures about what is in fact true. And, as I have been urging, it is the mark and test of our sincere concern for truth that we should be in this way open and ready to accept the falsification of our claims about how things actually are in the universe around us. By withdrawing your attention from flesh-and-blood Scots to talk of true Scotsmen, you show that your concern now is with what you would like to be, rather than what actually is, the truth.
3.21 These considerations led Popper to maintain that falsifiability is the proper criterion of demarcation between science and nonscience. It is, that is to say, an essential mark of a scientific hypothesis that it should be in principle falsifiable: that there should be intelligibly describable phenomena which, if they were to happen, would by their actual occurrence show that that hypothesis was false. A theoretical structure, like that of Isaac Newton’s Principia (1687), which is eventually shown to have been false, is not thereby shown to have been unscientific: “We cannot identify science with truth, for we think that both Einstein’s and Newton’s theories belong to science, but they cannot both be true, and they may well both be false” (Popper, quoted in Magee 1973, p. 28).
3.22 What must disqualify theories or theoreticians as unscientific is rather that these theories or theoreticians refuse to allow for any things which would constitute falsifications if they were to occur. If a statement is to be substantial, then it has to deny something, something the past, present, or future occurrence of which would constitute, or would have constituted, falsification. And if a theory is to explain why this happens, then it must explain why it is this and not something else which happens. So when we learn from one of the classics of anthropology of all the many cunning intellectual devices by which the Zande ensure that no describable occurrence ever could constitute a falsification of their witchcraft beliefs, then this very discovery constitutes a sufficient reason for not awarding to this thought system the diploma title “Zande science” (Evans-Pritchard 1937, passim). Observing that Sigmund Freud (1856–1939) and his followers always seemed eager and able to show, subsequent to the event, that any and every apparently falsifying fact was after all what their own theories should have led us to expect, Popper himself contended that this constituted a very good reason for doubting whether, whatever else it may be, psychoanalysis is a science.
3.23 Either or both of these particular examples may be contested. But the general moral remains. Whenever we are uncertain, as certainly we often should be, how much, if anything, is actually being said or whether we really are being offered a genuinely explanatory and scientific theory, then we ought to press home the Falsification Challenge: “Just what, please, would have to happen, or have to have happened, to show that this statement is false or that this theory is mistaken?”
3.24 For the wife of that doomed soldier outside Stalingrad belief in God was—at least as far as the present test goes—scientific: “If there is a God . . . then He will bring you back to me soon and healthy” (see paragraph 2.1). Her hypothesis was falsifiable and in fact false. It is not equally obvious that the same could be said when that original religious hypothesis is enriched, but at the same time qualified, in the fashion favored by the father of English writer Edmund Gosse (1849–1928): “Whatever you need, t
ell Him and He will grant it; if it is His will” (Gosse 1907, chapter 2). Fear of falsification has led many a brash, decently falsifiable hypothesis to be progressively eroded away—the death by a thousand qualifications (Flew and MacIntyre 1955, pp. 96–100).
3.25 The No-true-Scotsman Move is typically made by people who are concerned to avoid admitting that an objection brought against the truth of some proposition to which they are themselves strongly committed does indeed reveal that the proposition, at least as originally understood and maintained, is simply false. There is another somewhat similar controversial move that should be mentioned here, albeit briefly and a little distastefully. This move is typically made by people forced reluctantly to recognize deficiencies in some regime or organization or whatever else which they themselves are vehemently committed to promoting. The move is to maintain that the regime or organization or whatever else suffers from those deficiencies only because of some regrettable but no doubt unavoidable features of its history, and to take it that this somehow enables us to discount those reluctantly admitted deficiencies. Since this move is made as an objection to the apparently undeniable contention that the perceived deficiencies are actually present, let us introduce the label: the It-isn’t-true-only-because-of-why-it-is Objection.