The Blind Watchmaker

by Richard Dawkins

  Comparisons between modern punctuationism on the one hand, and catastrophism or saltationism on the other, have a purely poetic force. They are, if I may coin a paradox, deeply superficial. They sound impressive in an artsy, literary way, but they do nothing to aid serious understanding, and they can give spurious aid and comfort to modern creationists in their disturbingly successful fight to subvert American education and textbook publishing. The fact is that, in the fullest and most serious sense, Eldredge and Gould are really just as gradualist as Darwin or any of his followers. It is just that they would compress all the gradual change into brief bursts, rather than having it go on all the time; and they emphasize that most of the gradual change goes on in geographical areas away from the areas where most fossils are dug up.

  So, it is not really the gradualism of Darwin that the punctuationists oppose: gradualism means that each generation is only slightly different from the previous generation; you would have to be a saltationist to oppose that, and Eldredge and Gould are not saltationists. Rather, it turns out to be Darwin’s alleged belief in the constancy of rates of evolution that they and the other punctuationists object to. They object to it because they think that evolution (still undeniably gradualistic evolution) occurs rapidly during relatively brief bursts of activity (speciation events, which provide a kind of crisis atmosphere in which the alleged normal resistance to evolutionary change is broken); and that evolution occurs very slowly or not at all during long intervening periods of stasis. When we say ‘relatively’ brief we mean, of course, brief relative to the geological timescale in general. Even the evolutionary jerks of the punctuationists, though they may be instantaneous by geological standards, still have a duration that is measured in tens or hundreds of thousands of years.

  A thought of the famous American evolutionist G. Ledyard Stebbins is illuminating at this point. He isn’t specifically concerned with jerky evolution, but is just seeking to dramatize the speed with which evolutionary change can happen, when seen against the timescale of available geological time. He imagines a species of animal, of about the size of a mouse. He then supposes that natural selection starts to favour an increase in body size, but only very very slightly. Perhaps larger males enjoy a slight advantage in the competition for females. At any time, males of average size are slightly less successful than males that are a tiny bit bigger than average. Stebbins put an exact figure on the mathematical advantage enjoyed by larger individuals in his hypothetical example. He set it at a value so very very tiny that it wouldn’t be measurable by human observers. And the rate of evolutionary change that it brings about is consequently so slow that it wouldn’t be noticed during an ordinary human lifetime. As far as the scientist studying evolution on the ground is concerned, then, these animals are not evolving at all. Nevertheless they are evolving, very slowly at a rate given by Stebbins’s mathematical assumption, and, even at this slow rate, they would eventually reach the size of elephants. How long would this take? Obviously a long time by human standards, but human standards aren’t relevant. We are talking about geological time. Stebbins calculates that at his assumed very slow rate of evolution, it would take about 12,000 generations for the animals to evolve from an average weight of 40 grams (mouse size) to an average weight of over 6,000,000 grams (elephant size). Assuming a generation-time of 5 years, which is longer than that of a mouse but shorter than that of an elephant, 12,000 generations would occupy about 60,000 years. 60,000 years is too short to be measured by ordinary geological methods of dating the fossil record. As Stebbins says, ‘The origin of a new kind of animal in 100,000 years or less is regarded by paleontologists as “sudden” or “instantaneous”.’

  The punctuationists aren’t talking about jumps in evolution, they are talking about episodes of relatively rapid evolution. And even these episodes don’t have to be rapid by human standards, in order to appear instantaneous by geological standards. Whatever we may think of the theory of punctuated equilibria itself, it is all too easy to confuse gradualism (the belief, held by modern punctuationists as well as Darwin, that there are no sudden leaps between one generation and the next) with ‘constant evolutionary speedism’ (opposed by punctuationists and allegedly, though not actually, held by Darwin). They are not the same thing at all. The proper way to characterize the beliefs of punctuationists is: ‘gradualistic, but with long periods of “stasis” (evolutionary stagnation) punctuating brief episodes of rapid gradual change’. The emphasis is then thrown onto the long periods of stasis as being the previously overlooked phenomenon that really needs explaining. It is the emphasis on stasis that is the punctuationists’ real contribution, not their claimed opposition to gradualism, for they are truly as gradualist as anybody else.

  Even the emphasis on stasis can be found, in less-exaggerated form, in Mayr’s theory of speciation. He believed that, of the two geographically separated races, the original large ancestral population is less likely to change than the new, ‘daughter’ population (on the other side of the mountains in the case of our shrew example). This is not just because the daughter population is the one that has moved to new pastures, where conditions are likely to be different and natural selection pressures changed. It is also because there are some theoretical reasons (which Mayr emphasized but whose importance can be disputed) for thinking that large breeding populations have an inherent tendency to resist evolutionary change. A suitable analogy is the inertia of a large heavy object; it is hard to shift. Small, outlying populations, by virtue of being small, are inherently more likely, so the theory goes, to change, to evolve. Therefore, although I spoke of the two populations or races of shrews as diverging from each other, Mayr would prefer to see the original, ancestral population as relatively static, and the new population as diverging from it. The branch of the evolutionary tree does not fork into two equal twigs: rather, there is a main stem with a side twig sprouting from it.

  The proponents of punctuated equilibrium took this suggestion of Mayr, and exaggerated it into a strong belief that ‘stasis’, or lack of evolutionary change, is the norm for a species. They believe that there are genetic forces in large populations that actively resist evolutionary change. Evolutionary change, for them, is a rare event, coinciding with speciation. It coincides with speciation in the sense that, in their view, the conditions under which new species are formed — geographical separation of small, isolated subpopulations — are the very conditions under which the forces that normally resist evolutionary change are relaxed or overthrown. Speciation is a time of upheaval, or revolution. And it is during these times of upheaval that evolutionary change is concentrated. For most of the history of a lineage it stagnates.

  It isn’t true that Darwin believed that evolution proceeded at a constant rate. He certainly didn’t believe it in the ludicrously extreme sense that I satirized in my parable of the children of Israel, and I don’t think he really believed it in any important sense. Quotation of the following well-known passage from the fourth edition (and later editions) of The Origin of Species annoys Gould because he thinks it is unrepresentative of Darwin’s general thought:

  Many species once formed never undergo any further change …; and the periods, during which species have undergone modification, though long as measured by years, have probably been short in comparison with the periods during which they retain the same form.

  Gould wants to shrug off this sentence and others like it, saying:

  You cannot do history by selective quotation and search for qualifying footnotes. General tenor and historical impact are the proper criteria. Did his contemporaries or descendants ever read Darwin as a saltationist?

  Gould is right, of course, about general tenor and historical impact, but the final sentence of this quotation from him is a highly revealing faux pas. Of course, nobody has ever read Darwin as a saltationist and, of course, Darwin was consistently hostile to saltationism, but the whole point is that saltationism is not the issue when we are discussing punctuated equilibrium. As I
have stressed, the theory of punctuated equilibrium, by Eldredge and Gould’s own account, is not a saltationist theory. The jumps that it postulates are not real, single-generation jumps. They are spread out over large numbers of generations over periods of, by Gould’s own estimation, perhaps tens of thousands of years. The theory of punctuated equilibrium is a gradualist theory, albeit it emphasizes long periods of stasis intervening between relatively short bursts of gradualistic evolution. Gould has misled himself by his own rhetorical emphasis on the purely poetic or literary resemblance between punctuationism, on the one hand, and true saltationism on the other.

  I think it would clarify matters if, at this point, I summarized a range of possible points of view about rates of evolution. Out on a limb we have true saltationism, which I have already discussed sufficiently. True saltationists don’t exist among modern biologists. Everyone that is not a saltationist is a gradualist, and this includes Eldredge and Gould, however they may choose to describe themselves. Within gradualism, we may distinguish various beliefs about rates of (gradual) evolution. Some of these beliefs, as we have seen, bear a purely superficial (‘literary’ or ‘poetic’) resemblance to true, anti-gradualist saltationism, which is why they are sometimes confused with it.

  At another extreme we have the sort of ‘constant speedism’ that I caricatured in the Exodus parable with which I began this chapter. An extreme constant speedist believes that evolution is plodding along steadily and inexorably all the time, whether or not there is any branching or speciation going on. He believes that quantity of evolutionary change is strictly proportional to time elapsed. Ironically, a form of constant speedism has recently become highly favoured among modern molecular geneticists. A good case can be made for believing that evolutionary change at the level of protein molecules really does plod along at a constant rate exactly like the hypothetical children of Israel; and this even if externally visible characteristics like arms and legs are evolving in a highly punctuated manner. We have already met this topic in Chapter 5, and I shall mention it again in the next chapter. But as far as adaptive evolution of large-scale structures and behaviour patterns are concerned, just about all evolutionists would reject constant speedism, and Darwin certainly would have rejected it. Everyone that is not a constant speedist is a variable speedist.

  Within variable speedism we may distinguish two kinds of belief, labelled, ‘discrete variable speedism’ and ‘continuously variable speedism’. An extreme ‘discretist’ not only believes that evolution varies in speed. He thinks that the speed flips abruptly from one discrete level to another, like a car’s gearbox. He might believe, for instance, that evolution has only two speeds: very fast and stop (I cannot help being reminded here of the humiliation of my first school report, written by the Matron about my performance as a seven-year-old in folding clothes, taking cold baths, and other daily routines of boarding-school life: ‘Dawkins has only three speeds: slow, very slow, and stop’). ‘Stopped’ evolution is the ‘stasis’ that is thought by punctuationists to characterize large populations. Top-gear evolution is the evolution that goes on during speciation, in small isolated populations round the edge of large, evolutionarily static populations. According to this view, evolution is always in one or other of the two gears, never in between. Eldredge and Gould tend in the direction of discretism, and in this respect they are genuinely radical. They may be called ‘discrete variable speedists’. Incidentally, there is no particular reason why a discrete variable speedist should necessarily emphasize speciation as the time of high-gear evolution. In practice, however, most of them do.

  ‘Continuously variable speedists’, on the other hand, believe that evolutionary rates fluctuate continuously from very fast to very slow and stop, with all intermediates. They see no particular reason to emphasize certain speeds more than others. In particular, stasis, to them, is just an extreme case of ultra-slow evolution. To a punctuationist, there is something very special about stasis. Stasis, to him, is not just evolution that is so slow as to have a rate of zero: stasis is not just passive lack of evolution because there is no driving force in favour of change. Rather, stasis represents a positive resistance to evolutionary change. It is almost as though species are thought to take active steps not to evolve, in spite of driving forces in favour of evolution.

  More biologists agree that stasis is a real phenomenon than agree about its causes. Take, as an extreme example, the coelacanth Latimeria. The coelacanths were a large group of ‘fish’ (actually, although they are called fish they are more closely related to us than they are to trout and herrings) that flourished more than 250 million years ago and apparently died out at about the same time as the dinosaurs. I say ‘apparently’ died out because in 1938, much to the zoological world’s astonishment, a weird fish, a yard and a half long and with unusual leg-like fins, appeared in the catch of a deep-sea fishing boat off the South African coast. Though almost destroyed before its priceless worth was recognized, its decaying remains were fortunately brought to the attention of a qualified South African zoologist just in time. Scarcely able to believe his eyes, he identified it as a living coelacanth, and named it Latimeria. Since then, a few other specimens have been fished up in the same area, and the species has now been properly studied and described. It is a ‘living fossil’, in the sense that it has changed hardly at all since the time of its fossil ancestors, hundreds of millions of years ago.

  So, we have stasis. What are we to make of it? How do we explain it? Some of us would say that the lineage leading to Latimeria stood still because natural selection did not move it. In a sense it had no ‘need’ to evolve because these animals had found a successful way of life deep in the sea where conditions did not change much. Perhaps they never participated in any arms races. Their cousins that emerged onto the land did evolve because natural selection, under a variety of hostile conditions including arms races, forced them to. Other biologists, including some of those that call themselves punctuationists, might say that the lineage leading to modern Latimeria actively resisted change, in spite of what natural selection pressures there might have been. Who is right? In the particular case of Latimeria it is hard to know, but there is one way in which, in principle, we might go about finding out.

  Let us, to be fair, stop thinking in terms of Latimeria in particular. It is a striking example but a very extreme one, and it is not one on which the punctuationists particularly want to rely. Their belief is that less extreme, and shorter-term, examples of stasis are commonplace; are, indeed, the norm, because species have genetic mechanisms that actively resist change, even if there are forces of natural selection urging change. Now here is the very simple experiment which, in principle at least, we can do to test this hypothesis. We can take wild populations and impose our own forces of selection upon them. According to the hypothesis that species actively resist change, we should find that, if we try to breed for some quality, the species should dig in its heels, so to speak, and refuse to budge, at least for a while. If we take cattle and attempt to breed selectively for high milk yield, for instance, we should fail. The genetic mechanisms of the species should mobilize their anti-evolution forces and fight off the pressure to change. If we try to make chickens evolve higher egg-laying rates we should fail. If bullfighters, in pursuit of their contemptible ‘sport’, try to increase the courage of their bulls by selective breeding, they should fail. These failures should only be temporary, of course. Eventually, like a dam bursting under pressure, the alleged anti-evolution forces will be overcome, and the lineage can then move rapidly to a new equilibrium. But we should experience at least some resistance when we first initiate a new program of selective breeding.

  The fact is, of course, that we do not fail when we try to shape evolution by selectively breeding animals and plants in captivity, nor do we experience a period of initial difficulty. Animal and plant species are usually immediately amenable to selective breeding, and breeders detect no evidence of any intrinsic, anti-evolution forc
es. If anything, selective breeders experience difficulty after a number of generations of successful selective breeding. This is because after some generations of selective breeding the available genetic variation runs out, and we have to wait for new mutations. It is conceivable that coelacanths stopped evolving because they stopped mutating — perhaps because they were protected from cosmic rays at the bottom of the sea! — but nobody, as far as I know, has seriously suggested this, and in any case this is not what punctuationists mean when they talk of species having built-in resistance to evolutionary change.

  They mean something more like the point I was making in Chapter 7 about ‘cooperating’ genes: the idea that groups of genes are so well adapted to each other that they resist invasion by new mutant genes which are not members of the club. This is quite a sophisticated idea, and it can be made to sound plausible. Indeed, it was one of the theoretical props of Mayr’s inertia idea, already referred to Nevertheless, the fact that, whenever we try selective breeding, we encounter no initial resistance to it, suggests to me that, if lineages go for many generations in the wild without changing, this is not because they resist change but because there is no natural selection pressure in favour of changing. They don’t change because individuals that stay the same survive better than individuals that change.