by Antony Flew
2.25 There is another sort of distinction between necessary and sufficient conditions that is of great practical importance. When we say that p is a logically sufficient condition of q, then we are saying that we could not assert p and deny q without contradicting ourselves. But when such-and-such is said to be not the logically but the causally sufficient condition of this or that, then what is being said is: In the world as it actually is, with the laws of nature what they actually are, such-and-such could not be produced or occur without producing this or that. When such-and-such is said to be not the logically but the causally necessary condition of this or that, then what is being said is: In the world as it actually is, with the laws of nature as they actually are, this or that could not occur or be produced without such-and-such first occurring or being produced.
2.26 Suppose that heavy smoking is, for people in certain circumstances, people having certain constitutions, and people who do not happen to die of something else first, a causally sufficient condition of eventually contracting lung cancer. Then this means that heavy smoking guarantees eventual lung cancer for those people, in those circumstances, and always assuming that they do not die of something else first. But it does not mean that heavy smoking is either a causally necessary or a causally sufficient condition guaranteeing that every heavy smoker will eventually contract lung cancer if he or she does not die of something else first. For there certainly are—although this appears very unjust to many of my fellow never-smokers—some lung cancer victims who have never smoked at all. Equally certainly our Greek contemporaries, who are the heaviest smokers in Europe, nevertheless have one of the lowest lung cancer rates in the world. (It is perhaps just worth noting as a sign of our times that investigation of the possible causes of this Greek immunity appears to have been inhibited as politically incorrect. It is no doubt thought in some quarters that those who rebelliously persist in defying official campaigns against smoking deserve to die of it.)
2.27 It is, however, not contradictory to suggest that what happen to be in our world either causally sufficient or causally necessary conditions would in a different universe be neither. Or, putting exactly the same point in different words, there is nothing inconceivable about the idea of a universe governed by quite different laws from those which seemingly govern our actual universe. Or, in yet other words again, the occurrence of an event which must be, according to these actual laws, physically impossible is by no means by that token logically impossible (see paragraphs 1.25 and 1.26). Indeed there seems to be no contradiction in the suggestion of a universe not governed by any laws at all—a point made, to the great scandal of the philosophical public, by David Hume (1711–1776) in A Treatise of Human Nature (I [iii] 3).
2.28 In order to put this distinction between causally necessary and causally sufficient conditions to work at once, let us consider an argument offered by a pair of contributors to a widely circulated volume of papers on educational theory. They were attacking the proposition that heredity matters. The argument concerns all kinds of achievement and not, therefore, only one kind of ability or disposition. “The model, and especially [Francis] Galton’s version of it, of course denies the possibility of change . . . if 80 percent of adult performance is directly dependent upon genetic inheritance, how have the styles of our lives and patterns of our thinking changed to the extent that they have?” (Richardson and Spears 1972, p. 74).
2.29 This objection, which its proud authors obviously regarded as knockdown decisive, is simply irrelevant. For what their opponents were maintaining was that some appropriate level of ability is a causally necessary condition of every kind of achievement and that what abilities are available to each individual is largely determined—to the extent of about 80 percent—by that individual’s heredity. They were most certainly not maintaining that having the necessary inherited measure of ability is a causally sufficient condition of reaching the appropriate level of achievement. To put it coarsely, what those opponents held was not (1) that silk purses are born not made, but (2) that you cannot make silk purses out of sows’ ears. The objection offered here would be decisive against that first thesis. But against the second it has no weight at all.
2.30 The citing of this supposed refutation here is not, however, a contribution to the ongoing debate about how far the differences in performance between individuals and between sets of individuals is to be explained by reference to their genes and how far by reference to their environments. Rather, it is to use this supposed refutation as an example of a dispute in which one party maintains an extreme position and the other does not, and to indicate the temptations to which the parties in such disputes are exposed.
2.31 If we ourselves are committed to an extreme position then we are apt to assume or to pretend—to borrow the phrase General Robert E. Lee employed to refer to his opponents in the Union armies during the American Civil War—that “those people,” our opponents, also are. Thus those inclined to hold that everything is environmentally determined become, for that very reason, tempted either to assume or to pretend that anyone who disagrees with them must hold that nothing is environmentally determined. Certainly, if this polarization assumption were universally true, it would make it easier for all spokesmen of extreme positions to dispose of the thus correspondingly extreme positions of their opponents, hence the strength of the temptation. But the controversial world is not so conveniently arranged, with this sort of ideal balance. In the present case, for instance, there never have been—and for very good reason—any extreme hereditarians maintaining that our genetic constitutions are the sufficient conditions of all that we are and can be. But there certainly are at this time many environmentalist ultras. These people do want to hold that everything about us—or as near as makes precious little matter everything—is determined by our environments. (For some citations from spokespersons for such extreme environmentalism see, for instance, the introduction to Flew 1997.)
2.32 These points about the temptations of extremism neither depend upon nor warrant any general contention that we ought to strive for moderation in the content of all our opinions, or that truth is always in the middle. Suppose that it had been, say, President John F. Kennedy and not, as it was, Senator Barry Goldwater, who said: “. . . extremism in the defense of liberty is no vice . . . moderation in the pursuit of justice is no virtue” (Acceptance speech at the Republican National Convention, July 16, 1964). Then all of America’s liberals—and not they alone—would have endorsed these splendid words. Such hypothetical endorsers would have been right. We should be right to concur anyway. Again, it actually was Abraham Lincoln himself who warned: “Let us not be diverted by more of these sophistical contrivances wherewith we are so industriously plied and belabored—contrivances such as groping for some middle ground between the right and the wrong” (Speech at the Cooper Union, New York City, February 27, 1860).
2.33 The alleged truism that truth is always in the middle is not merely false but demonstrably false. If, as this silly and unprincipled principle requires, the truth really were, for all values of A and Z, midway between A and Z; then it must also be, according to the requirements of the same principle, halfway between the halfway point between A and Z and Z; and so on, indefinitely. This conclusion is incoherent and absurd. The same must therefore apply to the principle from which it is thus validly deduced: The Truth-is-always-in-the-middle Damper.
2.34 This abstract theoretical demonstration has concrete practical relevance. For there are in fact a great many people who adopt as one of their main guiding principles the principle that they should, with regard to all controverted issues of belief, strive to position themselves equidistant between whatever they see as the most extreme opposing standpoints currently adopted. Not only does this principle, for reasons just now stated or suggested, guarantee commitments to error of all kinds. It also—paradoxically—exposes its protagonists to manipulation from those very extremes which it professes to eschew. Persuasive operators, discerning this common aff
ection for whatever can be presented as the middle ground, can and do find ways of shifting the apparent center in whatever direction they desire.
2.35 Such persuasive operators strive to make out that the position which they want the subjects of their persuadings to adopt is located roughly halfway between two opposite extremes. This sort of manipulation of people who want always to be soberly and levelheadedly in the middle is in the sphere of belief and persuasion closely analogous to a universally familiar and too often successful bargaining strategy. This is the strategy whereby one side makes inordinate demands in order that any splittings of the difference fifty-fifty shall yield a result going further its way than the way of the other.
2.36 Whatever may be the case in bargaining situations, where the object just is to get more or less of whatever it may be, it is quite different where the aim is—or ought to be—to discover truth. When discovering truth is the object, then considerations of moderation or extremism are as such neither here nor there. They can in consequence find purchase in some particular context only insofar as we happen to have some substantial independent reason to believe that in that context either a moderate thesis or an extreme one is more likely to be true.
2.37 There are, as we have just been noticing (see paragraphs 2.32–2.35), insidious dangers and temptations in what presents itself or is presented as the middle ground. The temptation of the extremist, on the other hand, is to assume or to pretend that everyone else is committed to some equally extreme opposite position (see paragraph 2.31). The particular extremists who occasioned this comment believed (see paragraphs 2.28–2.29), or at least wanted to believe, that All human differences are determined by the environment. To disagree with this it is sufficient to assert no more than that Not all human differences are determined by the environment. The first of these two propositions is a universal proposition and the second is its contradictory. But, as we have seen (e.g., paragraph 2.10), any universal proposition can be decisively falsified and its contradictory equally decisively verified by the production of even one single genuine counterexample. (A proposition which is the contrary proposition or the contradictory proposition of another proposition is always known simply as its contrary or its contradictory.) Yet our extremists, perhaps on the basis of an invalid argument, and for certain mistakenly, assumed that their opponents are committed to the correspondingly extreme and diametrically opposite thesis that No human differences are determined by the environment.
2.38 This third proposition, which Logicians somewhat misleadingly dub the contrary of the first, is like that first, itself, but unlike the second, universal. It, too, therefore, is exposed to definitive falsification by the production of a single genuine counterexample. So the protagonists of the first proposition, All human differences are determined by the environment, are bound to see their controversial task as easier if they believe or pretend to believe that they have to refute only the third proposition, No human differences are determined by the environment, rather than the second proposition, Not all human differences are determined by the environment. Nevertheless, as must by now be obvious, it is fallacious to argue that because someone denies the first proposition and hence is committed to the second, therefore, they must in consistency assert the third. For the contradictory is not the contrary.
2.39 The distinction between contraries and contradictories is crucially important in debates in which at least one party is maintaining that something is for that party a matter of principle. The first thing to get clear, and this is not always easily done, is precisely what the supposed principle actually is. If it really is a principle, then the persons proclaiming it must be maintaining either that all or that no so-and-sos ought to be such-and-such. For some people, for example, it is a matter of principle that all schools, or at any rate all tax-funded schools, should be comprehensive (i.e., nonselective) neighborhood schools. By this principle, which these people may well wish to describe as the principle of comprehension, they necessarily become committed to opposing the establishment or, if already established, the continuation of any schools of any other kind.
2.40 The temptation for these people is to argue that any opponents who favor the establishment or, if already established, the continuation of any schools not of their own single approved kind thereby concede what they themselves see as the principle of comprehension. But opponents of that principle may be, and most likely are, simply contradicting it. In that case, all that they are denying is that all schools ought to be of that one, or perhaps any, particular kind. They are not necessarily, and probably are not, in fact, maintaining the contrary principle: that there should be no such schools.
2.41 As a more stimulatingly controversial example, consider the principle to which Davy Crockett was appealing when he told his colleagues in the Congress of the United States: “We have the right, as individuals, to give away as much of our own money as we please in charity; but as members of Congress we have no right to appropriate a dollar of the public money.” That principle, of course, was that Congress possessed no rights or powers other than those granted to it by the Constitution. And Davy Crockett, like President James Madison, could not “undertake to lay my finger on that article of the Constitution which grants a right to Congress of expending, on objects of benevolence, the money of their constituents.”
3.1 Imagine some aggressively nationalistic Scotsman settled down one Sunday morning with his customary copy of that shock-horror tabloid The News of the World. He reads the story under the headline, “Sidcup Sex Maniac Strikes Again.” Our reader is, as he confidently expected, agreeably shocked: “No Scot would do such a thing!” Yet the very next Sunday he finds in that same favorite source a report of the even more scandalous ongoings of Mr. Angus MacSporran in Aberdeen. This clearly constitutes a perfect counterexample, one which definitively falsifies the universal proposition originally put forward. (“Falsifies” here is the opposite of “verifies”; and it therefore means “shows to be false.”) Allowing that this is indeed such a counterexample, he ought to withdraw, retreating perhaps to a rather weaker claim about most or some Scotsmen. But even an imaginary Scot is, like the rest of us, human; and none of us always does what we ought to do. So instead he amends his statement to: “No true Scotsman would do such a thing!”
3.2 An equally simple but actual example of this No-true-Scotsman Move was provided by Stokely Carmichael during the early days of the Black Power movement. On a visit to London he was arguing the thesis that the world is now divided between exploiting white and exploited colored people. “What about [Fidel] Castro?” asked one member of his audience. “What about Che Guevara?” “I don’t,” retorted Mr. Carmichael, “consider them white.” (He did not, at least on that occasion, follow his Black Power associate James Baldwin by explaining that, for him, “whiteness is a state of mind, not a complexion.”)
3.3 In these two textbook examples it is immediately obvious what is going on, and what is wrong. A bold, indeed reckless, claim about all those who happen to be members of a certain category is being surreptitiously replaced by an utterance which is, in effect, made true by an arbitrary redefinition. If anyone who satisfies all the ordinary requirements for being accounted a Scot behaves like the Sidcup sex maniac, then our aggressively nationalistic Scot will take that fact as by itself sufficient reason to disqualify that person from rating as a true Scot. True Scots are, by his implicit definition, not sex maniacs. Likewise, if any people who would conventionally be accounted Caucasians commended themselves to Mr. Carmichael, then that fact alone became for Mr. Carmichael a decisive reason against considering those exceptional Caucasians to be truly white.
3.4 If all examples of the No-true-Scotsman Move were as simple and straightforward as these two, then there might perhaps have been little need to introduce such a label. But they are not. So more needs to be said. The essence of the move consists in sliding between two radically different interpretations of the same or very similar forms of words. In one, in this cas
e the original interpretation, what is asserted is synthetic and contingent. In the other, in this case the later and highly factitious interpretation, what we have is a custom-built necessary truth. This fundamental distinction can be brought out and fixed securely in mind with the help of William Shakespeare. Questioned about what the ghost said, Hamlet replies:
There’s ne’er a villain dwelling in all Denmark
But he’s an arrant knave.
His friend Horatio responds:
There needs no ghost, my lord, come from the grave To tell us this. (Hamlet 1.5)
3.5 There needs no ghost because Hamlet’s proposition is analytic and logically necessary. What makes it analytic is that its truth can be known simply by analyzing the meanings of all its constituent terms. What makes it a logically necessary truth is that to deny it would involve self-contradiction: Its contradictory would be a logically necessary falsehood. It is an apriori truth, one which could be known to be true without reference to what does or does not in fact happen in this universe as it actually is. Instead of this, and very reasonably, Horatio wanted a synthetic and logically contingent proposition; one the truth or falsity of which could not be known merely by fully understanding its meaning, and the contradictory of which would be neither necessarily true nor necessarily false. He wanted not an apriori but an aposteriori truth. Had Hamlet claimed—however anachronistically and irrelevantly to Shakespeare’s dramatic purposes—that all Danish villains were the products of maternal deprivation, then his proposition would have been not analytic but synthetic, not necessary but contingent. It would also have been aposteriori, inasmuch as it could be known to be true—if it were true—only by reference to some actual study of the deprived home background of Danish villains.