by Antony Flew
4.18 For example, the EEOC does this when it discovers what it calls indirect discrimination to have occurred when some minority subset of the total national population is found not to be proportionately represented either in some kind of employment in some particular geographical area or among those employed by a single employer throughout the country. The employers concerned are then presumed, until and unless they are able to prove their innocence, to have been guilty of hostile discrimination against members of the minority subset in question. It is notoriously hard to prove a negative. So, paradoxically, the only sure way for employers to provide acceptable proof that they were innocent of hostile discrimination against members of some minority subset is for them to operate a quota policy of positive preferential discrimination in favor of members of that minority subset. Presumably the EEOC recognized that discrimination as a kind of choosing is essentially intentional. This was no doubt part of the reason why the EEOC decided that what it was proposing to combat as a supposed further form of the evil of discrimination must be described not as unintended but as indirect discrimination.
4.19 The economist Lester Thurow, however, on the second page of his Poverty and Discrimination (1969) simply defines “discrimination” as his word for all the actual differences with respect to economic prospects and economic achievements among members of the various sets which he there chooses to distinguish. This is done altogether without any reference to causes from which these differences may have resulted. Thurow then proceeds to provide an abundance of statistics showing the extent of these differences, a proceeding which, according to his definition, constitutes the provision of statistics showing the extent of discrimination. All intergroup differences in cultural orientation toward education, work, family, risk, self-employment, enterprise, and everything else have been banished from consideration by definition. “Discrimination” here becomes a word for nothing but statistical results, although the very reason why most of us are concerned about discrimination in the usual understanding of that word is that, in that understanding, it refers to a particular and unlovely kind of intentional behavior.
4.20 By this verbal maneuver Thurow not only abandons inquiry into the causes of these observed and recorded intergroup differences, causes of which hostile or favorable (direct) discrimination is certainly only one cause although perhaps in certain cases the most important cause. At the same time, he attempts to redirect the moral disapproval of his readers from such discrimination as a possible cause of these intergroup differences to the intergroup differences as such. In so doing he appears to be trying to persuade us to abandon the ideal of equality of opportunity, which forbids discrimination upon grounds which are properly irrelevant, and to abandon it in favor of the very different, and in practice incompatible, ideal of equality of outcome.
4.21 Thurow presents this altogether different ideal of equality of outcome in an oblique and very misleading way, as if its realization could be achieved by systematically preventing all defections from the entirely different ideal of nondiscriminatory equality of opportunity. Thurow, like everyone else, is fully entitled, if he so wishes, to proclaim his personal ideals straightforwardly, and to endeavor to persuade others to join him in pursuing those ideals. But to present what appears to be his own personal ideal as if it were a very different ideal, and that one which is today almost universally shared, that is a different matter. (For some discussion of the relations or lack of relations between these two different ideals of equality see, for instance, Flew 1981, chapter 2.)
4.22 Paragraphs 4.13–4.14 brought out that the Subject/Motive Shift is fallacious, whereas paragraph 4.15 suggested a salutary disciplinary drill. The Subject/Motive Shift now needs to be distinguished from three other moves that are perfectly proper and on occasion absolutely necessary. Certainly it is fallacious to urge that simply because someone has a vested interest in the truth of a proposition or in the validity of an argument, the proposition or argument is, or very probably is, therefore, false or, as the case may be, invalid.
4.23 But in the first place suppose that the proposition has been asserted by someone giving evidence in a court. It remains always relevant that a particular proposition may be both true and known by the person who asserts it, notwithstanding that the person has every kind of powerful vested interest in both its truth and the assertion of its truth. But that it is asserted by someone who is in a position to know and has no reason (motive) for trying to deceive us, is for us, who are not in that person’s position to know, better evidence for believing that it is true than the same assertion made by someone in an equally good position to know, but with opposite interests. There are, of course, other dimensions of complexity in the reasonable assessment of testimony. The present point is, however, simple and uncontroversial. For instance, it is taken for granted by everyone who in a question of foul play would be inclined, all other things being equal, to accept the testimony of a neutral spectator rather than that of any of the contestants.
4.24 The second entirely legitimate kind of move which needs to be distinguished from the fallacious Subject/Motive Shift is what I call the On-your-own-principles Maneuver. It is entirely legitimate and perfectly proper to point out to opponents in argument logical consequences of their own contentions; in doing this one does not necessarily commit oneself to either accepting or rejecting those contentions. Indeed, this is the typical form of rational moral argument. For if one wants to persuade people rationally to alter or abandon what they hold as their moral principles, how else could this be done save by revealing to them logical consequences of those principles which they themselves are not prepared to accept as moral? (For further discussion of the nature of rational moral argument see, for instance, R. M. Hare 1952; and compare paragraphs 5.41–5.46, below.)
4.25 People against whom this On-your-own-principles Maneuver is employed may be tempted to complain that its employment constitutes an illicit argumentum ad hominem. This Latin expression means, literally, an argument addressed to the man. It is an expression we need to understand but ought never to employ ourselves. For a start it is unacceptably sexist. But, much more the present point, it is an expression that has been and still is regularly employed to refer not only to a kind of fallacy, but also to a sort of argument which is perfectly proper. The fallacy in question may be seen as a variant of the Subject/Motive Shift. The premise attributes some discreditable characteristic to the other party in the argument, whereas the conclusion purports thereby to establish the falsity of some thesis propounded by that party. The perfectly proper form of argument is that of executing the On-your-own-principles Maneuver.
4.26 It cannot be too often or too strongly emphasized that those who execute this maneuver are not thereby and necessarily committing themselves to the actual or pretended principles of those against whom they perform it. To emphasize this is indeed the main point and purpose of refusing to speak of ad hominem arguments and introducing instead the expression “On-your-own-Principles Maneuver.” For example, suppose that some set of politicians were found to have been profiting hugely from deals of a kind which they had themselves repeatedly and loudly denounced as utterly illicit. And suppose that opponents then pointed to the inconsistency between the actual behavior of that set of politicians and their own frequently stated principles. Then it would be preposterous for these politicians to respond by claiming that since their opponents themselves were well known to have benefited as much or more by making the same kind of deals, they were just as hypocritical. For we are assuming that those opponents had never maintained that such deals are wrong and had never pretended not to be making them.
4.27 There is a third perfectly legitimate move which needs to be distinguished from the fallacious Subject/Motive Shift. There is one condition on which it is entirely legitimate to raise and to pursue questions about interests and motivations. If and only if the original questions of truth and validity have been settled, then—at least from the point of view of soundness of think
ing—it is perfectly legitimate and it can often be very illuminating to raise and pursue such questions. For, given that some individual or some set of individuals is mistaken and now known to be mistaken, and given that we have no evidencing reason to believe that individual or set of individuals to be in general intellectually challenged, then we may reasonably expect to discover what it is which is misleading that person or set of people. Yet it will not do—notwithstanding that it is all too often done—to offer more or less speculative answers to such consequential questions as a substitute for, rather than as a supplement to, the direct examination of whatever were the prior issues.
4.28 It appears that this offense is more prevalent today than ever before. Certainly the temptation has been much increased by the proliferation of psychoanalysis and the sociology of knowledge. Consider a statement by a leading British Freudian, Charles Berg: “To achieve success the analyst must above all be an analyst. That is to say he must know positively that all human emotional reactions, all human judgments, and even reason itself, are nothing but the tools of the unconscious; that such seemingly acute convictions which an intelligent person like this possesses are but the inevitable effect of causes which lie buried within the unconscious levels of his psyche” (Berg 1946, p. 190).
4.29 Suppose that the scope of this statement had been explicitly limited to the analytic situation and that its application had been emphatically confined to neurotic patients crying out for psychiatric attention. Still, many lay readers would have inferred that psychoanalysis licenses the absolutely general conclusion “that all human emotional reactions, all human judgments, and even reason itself, are nothing but the tools of the unconscious” (Berg 1946, p. 190). As it is, Dr. Berg was apparently presenting this reckless and unqualified claim as either a finding or a presupposition of psychoanalytic investigations. For what else could be the point of his saying that this is something which the successful analyst “must know positively”? Presumably some rule of professional procedure (i.e., never to attend to the truth or validity value of anything the client says in the analytic hour but always to seek some motive for the client’s saying it) is being confused with, and mistaken to warrant, the conclusion that such utterances—and hence by a somewhat drastic extension all utterances (presumably including the utterances of psychoanalysts reporting what they believe to be their own findings)—do not actually have, and/or cannot be known actually to have, any truth or validity value at all. Nothing like leather becomes, not for the first nor for the only time, nothing but leather.
4.30 Be that as it may with these interpretative conjectures, there is no doubt but that, if psychoanalysis really does carry any such universal consequence, then the entire enterprise must thereby discredit itself. For this is exactly that same disastrous consequence which, as we saw earlier (see paragraphs 4.7–4.8), has often, but wrongly, been said to be implicit in any form of scientific “naturalism.” Let us put this objection in the most pointed and personal way. If “all human judgments . . . are nothing but the tools of the unconscious,” and if the point of this “nothing but” is, as it surely must be, to preclude the possibility of these tools of the unconscious also being known to be true, then this must apply equally to all the judgments of the no-more-than-human analyst, and hence that “all” has to include the judgment which the psychoanalyst himself is here making about all human judgments. What is poison for the goose is poison for the gander (and for the farmer, too).
4.31 It is therefore understandable, if not on that account venial, that usually those who see the putative insights of psychoanalysis as instruments for discrediting opinions apply these instruments only to other people. To lose the initiative would be fatal. It thus has to be the opinions of other people rather than our own that are, with no good grounds given, dismissed as nothing but the expressions of unconscious motivation. Thus, it has to be the motives of other people rather than our own to which the Subject/Motive Shift shifts.
4.32 With appropriate alterations what has just been said also applied to Marxist sociologists of knowledge and still applies if there are still Marxist sociologists of knowledge to whom anything can apply. It had, and perhaps has, to be the opinions of other people that were as such repudiated as nothing but the “false consciousness” of a particular social or economic class. The pretended findings of these Marxists naturally were not in the same way mere ideology, but instead science, perhaps even “scientific socialism.” A similar, more recently fashionable contention, is that of the Deconstructionists. Since in their view all knowledge is socially constructed, books (i.e., texts) are interesting only for revealing their author’s ideology rather than for their apparent content. This contention therefore generates the ideally appropriate rule for interpreting the writings of Deconstructionists themselves, if anyone should think it worthwhile to spend time in this fashion.
4.33 For anyone who sincerely wants to know what’s what, the right moral to draw from these proceedings is that whatever relevant insights psychoanalysis and sociology may be alleged to give us should be applied in the first instance rather than as a last resort to ourselves. It is most salutary to remember and to follow an example set by Charles Darwin. Long before Freud was even born Darwin made it his practice to note down all objections to his theories the moment he met them. That observant naturalist had noticed that we are all more apt to forget what we have some interest in not remembering.
4.34 Another common move which resembles the Subject/Motive Shift is not the same, although it too is unsound. Someone ripostes to a speaker who has put forward a view that is distinctively Christian, Marxist, Conservative, or whatever: “You would believe that because you are a Christian (or a Marxist, or a Conservative, or a whatever).” Such remarks may often have point and value. But they cannot be admitted as objections. For that some view happens to be an essential element in some general position only begins to be relevant to the question of the truth of that particular view when the general position is already known to be in error. But to appeal to this assumption of error in a debate with someone who starts as an adherent of that particular general position constitutes a textbook example of Begging the Question, that is, of taking for granted precisely what is in dispute. Furthermore, even if the general Christian, Marxist, Conservative, or whatever position is in error, it could—and almost certainly does—contain some elements and carry some consequences which are in their own right true.
4.35 Such attempts to refute a view merely by classifying it in some irrelevant way are not the same as attempts to refute a view by referring to the possible motives of its protagonists. One favorite contemporary token of the former type (see paragraph 1.49) is to meet a contention by dismissing it on no other or better ground than that it is well worn, hackneyed, boring, or predictable. It ought to go without saying that none of these facts, even if it is indeed a fact, has by itself any bearing at all upon the question of the truth value of the proposition so characterized. It is a scandal of our trendocratic times that a purely journalistic classification of this kind should so often be allowed to pass as a refutation.
4.36 Another token of the same type was common in the United States during the Vietnam War. Some said that if South Vietnam were to be allowed to fall to the Communists, then Laos, Cambodia, and other countries of Southeast Asia would follow in quick succession. Their position was regularly described as the Domino Theory. Yet the very aptness of this happy description was almost equally regularly mistaken to be a disproof of that theory. Certainly almost everyone at that time whom I challenged to give evidencing reasons for dismissing that theory seemed surprised to be so challenged. Most apparently believed that the description was itself sufficient refutation. So let us do our bit to discourage bad practice by introducing a suitably opprobrious label: the Fallacy of Pseudorefuting Description.
4.37 Today this fallacy is perhaps most commonly committed by peremptorily dismissing hopefully explanatory accounts of human events as conspiracy theories. That is clear
ly not an adequate, or indeed any, refutation. But in investigating the actual outcomes of social and political policies we do need to be aware that these policies always do in fact have consequences which were unforeseen and hence unintended by the promoters of those policies. This is why the social sciences have sometimes been defined as studies of the unintended consequences of intended action. Once we have recognized that it is possible, even likely, that policies may have consequences which were not intended by the promoters of those policies, we ought at least to hesitate and seek clearer evidence before accusing them of conspiring to produce those consequences. (For some discussion of the special difficulties of discussing social facts and social policies see, for instance, Flew 1995. Murray 1984, although by now somewhat dated, is still a book which should be read by anyone who wants to engage in critical thinking about the United States or any similar welfare system.)
4.38 This is the moment to say something about the cry: “We want no witch-hunts.” Certainly there are two reasons, each of which is by itself more than sufficient, for not wanting to see a revival of that ancient and fortunately now long-since-defunct institution. One is that in truth there neither have been nor could be women either naturally endowed with nor able to acquire the powers to do the deeds witches were believed to do. The other is that the hunt for victims allegedly guilty of doing such deeds typically, if perhaps not quite universally, involved the use of torture to extract confessions. (For a historian’s account of the horrors of such actual witch-hunts, consult Trevor-Roper 1956.)
4.39 Since, at least in the Western world, no one in our time believes that there are or could be people actually endowed with the powers the victims of historic witch-hunts were accused of exercising, no one today can be honestly and truthfully accused of being in that sense a witch. So if and when any of our contemporaries are accused of advocating or conducting a witch-hunt, their first response to their accusers should be to ask those accusers what they see as the analogies between whatever it is which the accused are advocating or themselves doing and the witch-hunts of history.