The Blind Watchmaker

by Richard Dawkins

  Languages, then, evolve. But although modern English has evolved from Chaucerian English, I don’t think many people would wish to claim that modern English is an improvement on Chaucerian English. Ideas of improvement or quality do not normally enter our heads when we speak of language. Indeed, to the extent that they do, we often see change as deterioration, as degeneration. We tend to see earlier usages as correct, recent changes as corruptions. But we can still detect evolution-like trends that are progressive in a purely abstract, value-free sense. And we can even find evidence of positive feedbacks, in the form of escalations (or, looking at it from the other direction, degenerations) in meaning. For instance, the word ‘star’ was used to mean a film actor of quite exceptional celebrity. Then it degenerated to mean any ordinary actor who was playing one of the principal roles in a film. Therefore, in order to recapture the original meaning of exceptional celebrity, the word had to escalate to ‘superstar’. Later, studio publicity began to use ‘superstar’ for actors that many people had never heard of, so there was a further escalation to ‘megastar’. Now there are quite a few advertised ‘megastars’ whom I, at least, have never heard of before, so perhaps we are due for yet another escalation. Shall we soon hear tell of ‘hyperstars’? A similar positive feedback has driven down the currency of the word ‘chef’. It comes, of course, from the French chef de cuisine, meaning chief or head of the kitchen. This is the meaning given in the Oxford Dictionary. By definition, then, there can be only one chef per kitchen. But, perhaps to satisfy their dignity, ordinary (male) cooks, even junior hamburger-slingers, started referring to themselves as ‘chefs’. The result is that now the tautological phrase ‘head chef’ is frequently heard!

  But if this is an analogy for sexual selection, it is so, at best, only in what I have called the ‘weak’ sense. Let me jump now straight to the nearest approach I can think of to a ‘strong’ analogy: to the world of ‘pop’ records. If you listen to discussion among aficionados of pop records, or switch on the mid-Atlantic mouthings of disc jockeys on the radio, you will discover a very curious thing. Whereas other genres of art criticism betray some preoccupation with style or skill of performance, with mood, emotional impact, with the qualities and properties of the art-form, the ‘pop’ music sub-culture is almost exclusively preoccupied with popularity itself. It is quite clear that the important thing about a record is not what it sounds like, but how many people are buying it. The whole sub-culture is obsessed with a rank ordering of records, called the Top 20 or Top 40, which is based only upon sales figures. The thing that really matters about a record is where it lies in the Top 20. This, when you think about it, is a very singular fact, and a very interesting one if we are thinking about R. A. Fisher’s theory of runaway evolution. It is probably also significant that a disc jockey seldom mentions the current position of a record in the charts, without at the same time telling us its position in the previous week. This allows the listener to assess, not just the present popularity of a record, but also its rate and direction of change in popularity.

  It appears to be a fact that many people will buy a record for no better reason than that large numbers of other people are buying the same record, or are likely to do so. Striking evidence comes from the fact that record companies have been known to send representatives into key shops to buy up large numbers of their own records, in order to push sales figures up into the region where they may ‘take off’. (This is not so difficult to do as it sounds, because the Top 20 figures are based upon sales returns from a small sample of record shops. If you know which these key shops are, you don’t have to buy all that many records from them in order to make a significant impact on nationwide estimates of sales. There are also well-authenticated stories of shop assistants in these key shops being bribed.)

  To a lesser extent, the same phenomenon of popularity being popular for its own sake is well known in the worlds of book publishing, womens’ fashions, and advertising generally. One of the best things an advertiser can say about a product is that it is the bestselling product of its kind. Best-seller lists of books are published weekly, and it is undoubtedly true that as soon as a book sells enough copies to appear in one of these lists, its sales increase even more, simply by virtue of that fact. Publishers speak of a book ‘taking off’, and those publishers with some knowledge of science even speak of a ‘critical mass for take-off’. The analogy here is to an atomic bomb. Uranium-235 is stable as long as you don’t have too much of it in one place. There is a critical mass which, once exceeded, permits a chain reaction or runaway process to get going, with devastating results. An atom bomb contains two lumps of uranium-235, both smaller than the critical mass. When the bomb is detonated the two lumps are thrust together, the critical mass is exceeded, and that is the end of a medium-sized city. When a book’s sales ‘go critical’, the numbers reach the point where word-of-mouth recommendations et cetera cause its sales suddenly to take-off in a runaway fashion. Rates of sales suddenly become dramatically larger than they were before critical mass was reached, and there may be a period of exponential growth before the inevitable levelling out and subsequent decline.

  The underlying phenomena are not difficult to understand. Basically we have here yet more examples of positive feedback. A book’s, or even a pop record’s, real qualities are not negligible in determining its sales but, nevertheless, wherever there are positive feedbacks lurking, there is bound to be a strong arbitrary element determining which book or record succeeds, and which fails. If critical mass and take-off are important elements in any success story, there is bound to be a lot of luck, and there is also plenty of scope for manipulation and exploitation by people that understand the system. It is, for instance, worth laying out a considerable sum of money to promote a book or record up to the point where it just ‘goes critical’, because you then don’t need to spend so much money on promoting it thereafter: the positive feedbacks take over and do the work of publicity for you.

  The positive feedbacks here have something in common with those of sexual selection according to the Fisher/Lande theory, but there are differences too. Peahens that prefer long-tailed peacocks are favoured solely because other females have the same preference. The male’s qualities themselves are arbitrary and irrelevant. In this respect, the record enthusiast who wants a particular record just because it is in the Top 20 is behaving just like a peahen. But the precise mechanisms by which the positive feedbacks work in the two cases are different. And this, I suppose, brings us back to where we began in this chapter, with a warning that analogies should be taken so far, and no farther.

  CHAPTER 9

  Puncturing punctuationism

  The children of Israel, according to the Exodus story, took 40 years to migrate across the Sinai desert to the promised land. That is a distance of some 200 miles. Their average speed was, therefore, approximately 24 yards per day, or 1 yard per hour; say 3 yards per hour if we allow for night stops. However we do the calculation, we are dealing with an absurdly slow average speed, much slower than the proverbially slow snail’s pace (an incredible 55 yards per hour is the speed of the world record snail according to the Guinness Book of Records). But of course nobody really believes that the average speed was continuously and uniformly maintained. Obviously the Israelites travelled in fits and starts, perhaps camping for long periods in one spot before moving on. Probably many of them had no very clear idea that they were travelling in any particularly consistent direction, and they meandered round and round from oasis to oasis as nomadic desert herdsmen are wont to do. Nobody, I repeat, really believes that the average speed was continuously and uniformly maintained.

  But now suppose that two eloquent young historians burst upon the scene. Biblical history so far, they tell us, has been dominated by the ‘gradualistic’ school of thought. ‘Gradualist’ historians, we are told, literally believe that the Israelites travelled 24 yards per day; they folded their tents every morning, crawled 24 yards in an east-northeasterly directi
on, and then pitched camp again. The only alternative to ‘gradualism’, we are told, is the dynamic new ‘punctuationist’ school of history. According to the radical young punctuationists, the Israelites spent most of their time in ‘stasis’, not moving at all but camped, often for years at a time, in one place. Then they would move on, rather fast, to a new encampment, where they again stayed for several years. Their progress towards the promised land, instead of being gradual and continuous, was jerky: long periods of stasis punctuated by brief periods of rapid movement. Moreover, their bursts of movement were not always in the direction of the promised land, but were in almost random directions. It is only when we look, with hindsight, at the large scale macromigrational pattern, that we can see a trend in the direction of the promised land.

  Such is the eloquence of the punctuationist biblical historians that they become a media sensation. Their portraits adorn the front covers of mass circulation news magazines. No television documentary about biblical history is complete without an interview with at least one leading punctuationist. People who know nothing else of biblical scholarship remember just the one fact: that in the dark days before the punctuationists burst upon the scene, everybody else got it wrong. Note that the publicity value of the punctuationists has nothing to do with the fact that they may be right. It has everything to do with the allegation that earlier authorities were ‘gradualist’ and wrong. It is because the punctuationists sell themselves as revolutionaries that they are listened to, not because they are right.

  My story about the punctuationist biblical historians is, of course, not really true. It is a parable about an analogous alleged controversy among students of biological evolution. In some respects it is an unfair parable, but it is not totally unfair and it has enough truth in it to justify its telling at the beginning of this chapter. There is a highly advertised school of thought among evolutionary biologists whose proponents call themselves punctuationists, and they did invent the term ‘gradualist’ for their most influential predecessors. They have enjoyed enormous publicity, among a public that knows almost nothing else about evolution, and this is largely because their position has been represented, by secondary reporters more than by themselves, as radically different from the positions of previous evolutionists, especially Charles Darwin. So far, my biblical analogy is a fair one.

  The respect in which the analogy is unfair is that in the story of the biblical historians ‘the gradualists’ were obviously non-existent straw men, fabricated by the punctuationists. In the case of the evolutionary ‘gradualists’, the fact that they are non-existent straw men is not quite so obvious. It needs to be demonstrated. It is possible to interpret the words of Darwin and many other evolutionists as gradualist in intent, but it then becomes important to realize that the word gradualist can be interpreted in different ways to mean different things. Indeed, I shall develop an interpretation of the word ‘gradualist’ according to which just about everybody is a gradualist. In the evolutionary case, unlike in the parable of the Israelites, there is genuine controversy lurking, but that genuine controversy is about little details which are nowhere near important enough to justify all the media hype.

  Among evolutionists, the ‘punctuationists’ were originally drawn from the ranks of palaeontology. Palaeontology is the study of fossils. It is a very important branch of biology, because evolutionary ancestors all died long ago and fossils provide us with our only direct evidence of the animals and plants of the distant past. If we want to know what our evolutionary ancestors looked like, fossils are our main hope. As soon as people realized what fossils really were — previous schools of thought had held that they were creations of the devil, or that they were the bones of poor sinners drowned in the flood — it became clear that any theory of evolution must have certain expectations about the fossil record. But there has been some discussion of exactly what those expectations are, and this is partly what the punctuationism argument is about.

  We are lucky to have fossils at all. It is a remarkably fortunate fact of geology that bones, shells and other hard parts of animals, before they decay, can occasionally leave an imprint which later acts as a mould, which shapes hardening rock into a permanent memory of the animal. We don’t know what proportion of animals are fossilized after their death — I personally would consider it an honour to be fossilized — but it is certainly very small indeed. Nevertheless, however small the proportion fossilized, there are certain things about the fossil record that any evolutionist should expect to be true. We should be very surprised, for example, to find fossil humans appearing in the record before mammals are supposed to have evolved! If a single, well-verified mammal skull were to turn up in 500-million year-old rocks, our whole modern theory of evolution would be utterly destroyed. Incidentally, this is a sufficient answer to the canard, put about by creationists and their journalistic fellow travellers, that the whole theory of evolution is an ‘unfalsifiable’ tautology. Ironically, it is also the reason why creationists are so keen on the fake human footprints, which were carved during the depression to fool tourists, in the dinosaur beds of Texas.

  Anyway, if we arrange our genuine fossils in order, from oldest to youngest, the theory of evolution expects to see some sort of orderly sequence rather than a higgledy-piggledy jumble. More to the point in this chapter, different versions of the theory of evolution, for instance ‘gradualism’ and ‘punctuationism’, might expect to see different kinds of pattern. Such expectations can be tested only if we have some means of dating fossils, or at least of knowing the order in which they were laid down. The problems of dating fossils, and the solutions of these problems, require a brief digression, the first of several for which the reader’s indulgence is asked. They are necessary for the explanation of the main theme of the chapter.

  We have long known how to arrange fossils in the order in which they were laid down. The method is inherent in the very phrase ‘laid down’. More recent fossils are obviously laid down on top of older fossils rather than underneath them, and they therefore lie above them in rock sediments. Occasionally volcanic upheavals can turn a chunk of rock right over and then, of course, the order in which we find fossils as we dig downwards will be exactly reversed; but this is rare enough to be obvious when it occurs. Even though we seldom find a complete historical record as we dig down through the rocks of any one area, a good record can be pieced together from overlapping portions in different areas (actually, although I use the image of ‘digging down’, palaeontologists seldom literally dig downwards through strata; they are more likely to find fossils exposed by erosion at various depths). Long before they knew how to date fossils in actual millions of years, palaeontologists had worked out a reliable scheme of geological eras, and they knew in great detail which era came before which. Certain kinds of shells are such reliable indicators of the ages of rocks that they are among the main indicators used by oil prospectors in the field. By themselves, however, they can tell us only about the relative ages of rock strata, never their absolute ages.

  More recently, advances in physics have given us methods to put absolute dates, in millions of years, on rocks and the fossils that they contain. These methods depend upon the fact that particular radioactive elements decay at precisely known rates. It is as though precision-made miniature stopwatches had been conveniently buried in the rocks. Each stopwatch was started at the moment that it was laid down. All that the palaeontologist has to do is dig it up and read off the time on the dial. Different kinds of radioactive decay-based geological stopwatches run at different rates. The radiocarbon stopwatch buzzes round at a great rate, so fast that, after some thousands of years, its spring is almost wound down and the watch is no longer reliable. It is useful for dating organic material on the archaeological/historical timescale where we are dealing in hundreds or a few thousands of years, but it is no good for the evolutionary timescale where we are dealing in millions of years.

  For the evolutionary timescale other kinds of watch, such a
s the potassium-argon watch, are suitable. The potassium-argon watch is so slow that it would be unsuitable for the archaeological/historical timescale. That would be like trying to use the hour hand on an ordinary watch to time an athlete sprinting a hundred yards. For timing the megamarathon that is evolution, on the other hand, something like the potassium-argon watch is just what is needed. Other radioactive ‘stopwatches’, each with its own characteristic rate of slowing down, are the rubidium-strontium, and the uranium-thorium-lead watches. So, this digression has told us that if a palaeontologist is presented with a fossil, he can usually know when the animal lived, on an absolute timescale of millions of years. We got into this discussion of dating and timing in the first place, you will remember, because we were interested in the expectations about the fossil record that various kinds of evolutionary theory — ‘punctuationist’, ‘gradualist’, etc. — should have. It is now time to discuss what those various expectations are.