by Antony Flew
6.41 This is one more occasion where we need the fundamental distinction between propositions that are analytic and logically necessary, and propositions that are synthetic and contingent (see paragraphs 3.4–3.17 and 3.29–3.30). A similar fallacious move is sometimes made in Third World First circles. For it is necessarily true that the rich make the poor and the other way around but only in the factitious sense that such correlative terms as “rich” and “poor” depend for their meaning upon the possibility of mutual contrast. From this alone we cannot legitimately infer—what I am not here either asserting or denying—that rich men and rich countries always and everywhere somehow grind the faces and pick the pockets of the poor.
6.42 Having explained the nature of the mistake Leach made, we can now go on to ask how he was misled into making it. It can be seen as a case of sophistication going to the head. The sophistication is to be aware, in general, that even the most honestly and competently compiled figures may not mean what they seem to mean, and, in particular, that they may sometimes tell us more about the method of compilation than about their supposed subject. Yet to assume that this must always be so, or to assume without particular reason that it is so in some particular case, is to be infatuated with one’s own insight. However, there but for the grace of God go we. For when we first become acquainted with some fallacy, as students of the present book need to recognize, it can be almost irresistibly tempting to identify arguments as tokens of this fallacy type when what we really have is either a token of some other type or even no fallacy at all (see paragraphs 1.49–1.54).
6.43 The insight by which Leach allowed himself to be carried away is important, and it provides one of the reasons why we have to ask both where figures come from and upon what principles they were compiled. All U.K. crime statistics at the time when Leach was lecturing were compiled by the police, who were required to include all but only crimes reported to police officers. Since that was also the time when sexual relations between consenting (or active!) adult men had only very recently been decriminalized, students of the criminal statistics were all very much aware that the rise in the number of successful prosecutions recorded in some particular area might previously have indicated not a rise in the number of such offenses committed, but the recent appointment of a notorious homophobe as chief constable for that area.
6.44 Indeed my own favorite headline from that earlier period—“Wave of Buggery in Bootle”—may actually have pointed not to a rise but to a decline in the number of offenses committed. For some potential offenders may have been deterred by news of the successful prosecutions and punishment of offenders. In that case the headline ought to have been “New Broom Sweeps Bootle’s Buggers,” giving discredit where discredit was due.
6.45 Legend also tells, I know not whether truly or falsely, of an eager woman who as a research student carried out a wide and weary psychometric program, coming up eventually with the agreeable finding that the average IQ of men is equal to the average IQ of women. It must have been hard to tell her that one of the specifications for the particular battery of tests she was applying was that the tests chosen to constitute her battery must yield this result. Whether this sexist legend is true or false we do still need to recognize that sometimes batteries of tests and other systems of inquiry may be deliberately designed and required to produce some of the particular findings which they do produce.
6.46 Today the most familiar examples are those test batteries that are designed and required to ensure that different subsets in the populations taking these tests, whether those subsets are racially or ethnically or however else defined, are represented in the outcomes in proportion to their representation in the total populations of which they are subsets. The one thing that needs to be said about these proceedings is that it is viciously circular to maintain, as all too often is maintained, that some particular test battery is biased if you can offer no other or better reason for so insisting than that it fails to produce some desired and no doubt politically correct result.
6.47 That Karl Pearson (1857–1936), a leading statistician and philosopher of science, once published a paper entitled “The Scientific Aspect of Monte Carlo Roulette” is not legend but history. With appropriate qualifications, his conclusion was that “roulette as played at Monte Carlo is not a game of chance.” This result was derived from his analysis of the outcomes of every spin made, day by day and week by week, for many months. These outcomes were all recorded, for the benefit of seekers for systems, in a paper called Le Monaco. It later emerged that the reporter assigned to this dispiriting task had preferred to sit out his days at a nearby café rather than to stand on tiptoe peering at the tables in the casino. So Pearson’s statistical analysis was really not evidence of some hidden and perhaps exploitable physical regularities in those spinning wheels, but evidence for the more general psychological fact that if we try to think up, straight out of our own heads, a genuinely random series, we shall not succeed.
6.48 One mistake in the interpretation of statistics often made by otherwise wide-awake students of social affairs and disputants about social policy is to argue that, because the distribution of some characteristic among persons who are all at one extreme in respect of that characteristic may not account for much of any differences among their achievements, therefore that characteristic may be altogether unimportant for the securing of such achievements. Since top basketball players who are six-foot-ten are not noticeably better than those who are only six-foot-nine, an argument on these lines would leave unexplained the fact that in this field the professionals all tower over the rest of us. Again, “At a college where virtually all students score in the top 10 percent on national tests, just where in that select group a particular student is located probably means much less than . . . motivation, emotional state, and other such personal factors. . . . But to use this as a reason to admit students from the bottom half of the test scores to compete with those at the top is to set the stage for disaster. The 1960s saw this kind of reasoning, and these kinds of disasters, on college campuses across the country” (Sowell 1984, p. 129, emphasis original).
6.49 Certainly it is vital to be alert to the possibility that figures may simply have been conjured out of the air. We need always to be ready with the equivalent of the question asked of strangers in old-time Westerns: “Where you from?” Always, too, we need to be aware of the possibility that what the statistics in front of us now do really reveal is, if anything, something about the conditions of their compilation rather than something about what they seem or purport to record. But these are only occasionally realized possibilities. They are not general necessities. No recognition of the dangers of mendacity, ingenuous error, and self-deceit ought to blind us to the indispensable necessity in so many fields of quantitative evidence and quantitative thinking (see paragraph 6.6). If we are committed to the improvement of our understandings we must categorically reject as obscurantist the pseudosophisticated slogan: “There are lies, damned lies, and statistics.”
6.50 In the first edition of this book the concluding section of chapter 6 contained a couple of pages of examples of how different sorts of financial calculation may go wrong if proper account is not taken of the effects of monetary inflation. It is no longer necessary to provide such a wide range of examples before drawing the moral that all measures of money have become systematically ambiguous. It was Aristotle who first drew attention to certain terms that have different meanings when applied to different sorts of things, yet the same meanings when applied to the same things. Our money measures now have a new variety of such systematic ambiguity: systematic ambiguity in the time dimension. The word for a dollar is the same word anywhere and at any one time. But a dollar (1997) is not, but is very much less than, a dollar (1947).
6.51 One consequence is that conscientious writers or speakers making financial comparisons across any period during which there has been substantial inflation in whatever currency they are employing as their means of measurement now
always add parenthetic dates to the names of those currencies in order to indicate the real values. That is indeed a criterion of their conscientiousness.
7.1 It is a pity that chapter 6 had to stop when it did. Nothing but good could have come from examining more, and more various, ways in which figures can be misinterpreted. It is equally a pity that this chapter will again be far too short. In both cases the examination of more, and more various, examples is the practice that can scarcely fail to improve practice. Yet the first thing is to appreciate that what is needed most and all the time is an unspecialized critical alertness. We must always be ready to discern a fault in argument in any field, and most especially when that argument is our own or seems to support some conclusion we ourselves cherish. Both where figures are involved and more generally where they are not the faithful treatment of a few examples should be sufficient to emphasize the need and indicate the kind of discipline required in order to do better.
7.2 There is a fine tale, probably apocryphal, told of that notoriously merry British monarch Charles II. The story goes that there was a dinner to commemorate the foundation of the Royal Society of London, one of the great scientific institutions. At the end of the evening, “with that peculiar gravity of countenance which he usually wore on such occasions,” he put a challenge to the Fellows: “ ‘Suppose two pails of water were fixed in two different scales that were equally poised, and which weighed equally alike, and that two live bream, or other small fish, were put into either of these pails,’ he wanted to know the reason why that pail, with such additions, should not weigh more than the other pail which stood against it.” Many suggested possible explanations and argued for their own suggestions with more or less vigor. But, at last, one, who perhaps remembered that the Latin motto of that society is “Nullius in verba” (Take no man’s word for it!), denied the assumption: “It would weigh more.” The king was delighted: “Odds fish, brother, you are in the right” (D’Israeli 1814, p. 341).
7.3 The king’s move was an instance of what has traditionally been called the Fallacy of Many Questions. The stock example usually offered is the question, “When did you stop beating your wife?” when this is put to a man who either is not married at all, or else has not started, or perhaps has started but not stopped, beating his wife. It is not clear that what is wrong here is, in the strictest sense, a fallacy. But obviously it is wrong to build false assumptions into a question and to give answers that accept such assumptions. The mistake here—as in the case of the so-called Pathetic Fallacy (see paragraph 1.52)—is that of making unwarranted assumptions. The crux is noticed in chapter 6 of Caroll’s Alice’s Adventures in Wonderland (1865): “ ‘How am I to get in?’ asked Alice again, in a louder tone. ‘Are you to get in at all?’ said the Footman. ‘That’s the first question, you know.’ ”
7.4 Unfortunately these, like all the best textbook illustrations, make it so clear what is going wrong that it may become hard to believe that anyone could make such a mistake seriously and ingenuously. But a crucial and contestable assumption may be very hard indeed to unearth when it is concealed in a longer and more complicated statement. Even when it is embodied in a single and lucid sentence we may well miss it. Consider, for instance, exchanges that may occur in national assemblies on the all-too-rare occasions on which the current administration proposes some reductions in direct personal taxation. Opponents may well complain that too much or too little is to be given away and/or that it is to be given to the wrong sets of taxpayers. Or consider an assertion made in what Ninian Smart believed to be a neutral and uncommitted study of The Religious Experience of Mankind: “We have records of the inaugural visions of some of the Old Testament prophets, of the experiences which taught them something profoundly important about God, and that spurred them on to teach men in his name” (Smart 1971, p. 22).
7.5 After what has been said earlier about the ambiguity of such expressions as “the income and wealth of the nation,” “the national wealth,” “Gross Domestic Product,” and “Gross National Product” (see paragraphs 5.22–5.23), it is easy to appreciate what has to be said about utterances of the first of these two kinds. It is—unless we want to make some radical socialist assumption—that we ought in such contexts to talk not of giving away but of not taking in the first place. The assumption concealed in the second passage is built into the phrase “the experiences which taught them something profoundly important about God.” By writing these words Smart implicitly, and presumably without becoming fully seized of this implication, claims religious knowledge in both the strong and the weak senses distinguished earlier (see paragraphs 5.20–5.21). For there is a decisive difference between, on the one hand, affirming that someone enjoyed or suffered such and such experiences, which led them to believe this, that, or the other and, on the other hand, while still affirming that the experiences occurred, also conceding that those consequent beliefs were in fact true. If the beliefs in question are religious and if you are supposed to be writing a detached essay on comparative religion, then you have no business either to assert or to imply any propositions of the latter sort, not even if you do believe, and maybe rightly, that they actually are true.
7.6 To appreciate fully what has just been said it is necessary to master a distinction between two senses of the word “experience.” It is a distinction essential for epistemology, (i.e., the theory of knowledge). In the everyday philosophical layperson’s sense of “experience,” if we say that someone has had experience of cows or computers or whatever else then there is no question but that there are such entities for them to have experience of and to be familiar with. But the other sense of “experience” is altogether different. In this sense of “experience,” to say that someone had or was the subject of an experience is precisely not to say or imply that the person was in some sort of cognitive contact with something existing altogether independently of their experience of it. To speak of someone’s experience in this sense is to talk exclusively about how things felt or seemed to that someone. It is talk without prejudice to any questions about possible causes of that experiencing external to and independent of its subject. In the first sense of the word “experience” it refers to experience (objective) and in the second to experience (subjective). The problem for anyone writing about experience of God is to provide reliable criteria for distinguishing the merely subjective from the genuinely objective, if any. For as the incorrigible Thomas Hobbes remarked in chapter 23 of Leviathan: “If any man pretend to me that God hath spoken to him . . . immediately, and I make doubt of it, I cannot easily perceive what argument he can produce to oblige me to believe it. . . . For to say that God . . . hath spoken to him in a dream is no more than to say he dreamed that God spoke to him.”
7.7 The use of the word “experience” in the subjective sense is very much a philosophers’ use. So before proceeding to the next fallacy it is worth saying a little more about this philosophers’ idea of experience. Anyone not interested should skip at once to paragraph 7.10. It is generally agreed that the modern period in philosophy began in 1637 with the publication in French by René Descartes of A Discourse on the Method. This Discourse consists of only six short parts. The first three are quietly preparatory and tell of how Descartes came to be meditating “in a room with a stove.” Then, suddenly, in the first two paragraphs of Part 4, he makes an intellectually explosive presentation of the almost all-corroding Cartesian doubt. What follows next (paragraph 7.8) is crucial quotations from those two paragraphs of A Discourse on the Method. (Anyone wanting to read more argument about the arguments of Descartes and other great philosophers is referred to Flew 1989.)
7.8 “I do not know whether I ought to tell you of the first meditations in which I engaged there; for they are so metaphysical and so unusual that perhaps they will not be to everyone’s taste. Yet at the same time if people are to be able to judge whether the foundations which I have laid are sufficiently firm I am in a way forced to speak of them. . . . I thought that I must . . . reject
as if it were absolutely false everything about which I could suppose there was the least doubt, in order to see if after that there remained anything which I believed which was entirely indubitable. So, on the grounds that our senses sometimes deceive us, I wanted to suppose that there was not anything corresponding to what they make us imagine. And, because some men make mistakes in reasoning—even with regard to the simplest matters of geometry—and fall into fallacies, I judged that I was as much subject to error as anyone else, and I rejected as unsound all the reasonings which I had hitherto taken for demonstrations. Finally, taking account of the fact that all the same experiences which we have when we are awake can also come to us when we are asleep without there being one of them which is then veridical, I resolved to pretend that everything which had ever entered into my mind was no more veridical than the illusions of my dreams.”
7.9 One of the first conclusions of these meditations of Descartes was, therefore, that he must take all his experience as being only experience (subjective). Or, rather, that he must take it all only in this way until and unless he can find some reliable means of identifying some of it as being what we have distinguished as experience (objective). The writings of Descartes have been enormously influential. Ever since the first publication of his Discourse all the leading figures in philosophy have recognized that they had or have somehow to come to terms with the challenge of Cartesian doubt. (It should be said, just as an aside, that the Cartesian conception of “the simplest matters of geometry” was probably not the same as yours or mine. For Descartes was also a major creative mathematician. His main mathematical achievement was to have developed coordinate geometry.)
