This theory, that generations of superstitious fishermen threw back into the sea crabs that resembled human faces, received new legs in 1980 when Carl Sagan discussed it in his wonderful Cosmos. In his words,
Suppose that, by chance, among the distant ancestors of this crab, one arose that resembled, even slightly, a human face. Even before the battle of Danno-ura, fishermen may have been reluctant to eat such a crab. In throwing it back, they set in motion an evolutionary process . . . As the generations passed, of crabs and fishermen alike, the crabs with patterns that most resembled a samurai face survived preferentially until eventually there was produced not just a human face, not just a Japanese face, but the visage of a fierce and scowling samurai.
It's a lovely theory, too good to die easily, and the meme has indeed replicated itself through the canon. I even found a website where you can vote on whether the theory is true (31 per cent of 1,331 voters), whether the photographs are fakes (15 per cent), whether Japanese craftsmen carve the shells to look that way (6 per cent), whether the resemblance is just a coincidence (38 per cent), or even whether the crabs really are manifestations of drowned samurai warriors (an amazing 10 per cent). Scientific truths are not, of course, decided by plebiscite, and I voted only because I was otherwise not allowed to see the voting figures. I'm afraid I voted with the killjoys. I think, on balance, that the resemblance is probably a coincidence. Not because, as one authoritative sceptic has pointed out, the ridges and grooves on the crab's back actually signify underlying muscle attachments. Even on the Huxley/Sagan theory, the superstitious fishermen would have to have begun by noticing some kind of original resemblance, however slight, and a symmetrical pattern of muscle attachments is exactly the kind of thing that would have provided that initial resemblance. I am more impressed by the same sceptic's observation that these crabs are too small to be worth eating anyway. According to him, all crabs of that size would have been thrown back, whether or not their backs resembled human faces, although I have to say that this more telling source of scepticism had a large bite taken out of it when I was taken out to dinner in Tokyo and my host ordered, for all the company, a dish of crabs. They were much larger than Heikea, and they were thickly encrusted in stout, calcified carapaces, but that didn't stop this superman picking up whole crabs, one by one, and biting into them like an apple, with a crunching sound that seemed to presage hideously bleeding gums. A crab as small as Heikea would be a doddle to such a gastronomic champion. He would surely swallow it whole without batting an eyelid.
My main reason for scepticism about the Huxley/Sagan theory is that the human brain is demonstrably eager to see faces in random patterns, as we know from scientific evidence, on top of the numerous legends about faces of Jesus, or the Virgin Mary, or Mother Teresa, being seen on slices of toast, or pizzas, or patches of damp on a wall. This eagerness is enhanced if the pattern departs from randomness in the specific direction of being symmetrical. All crabs (except hermit crabs) are symmetrical anyway. I reluctantly suspect that the resemblance of Heikea to a samurai warrior is no more than an accident, much as I would like to believe it has been enhanced by natural selection.
Never mind. There are plenty of other examples not involving humans, where animal 'fishermen', as it were, 'throw back' (or don't see in the first place) would-be food because of a resemblance to something sinister, and where the resemblance is certainly not due to chance. If you were a bird, out hunting caterpillars in the forest, what would you do if you were suddenly confronted with a snake? Leap back startled, would be my guess, and then give it a wide berth. Well, there is a caterpillar - to be precise, the rear end of a caterpillar - that bears an unmistakable resemblance to a snake. It is truly alarming if you are frightened of snakes - as I shamefacedly confess I am. I even think I might be reluctant to pick this animal up, despite knowing perfectly well that it is in fact a harmless caterpillar. (A picture of this extraordinary creature appears on colour page 7.) I have the same problem with picking up wasp-mimicking or bee-mimicking hoverflies, even though I can see, from their possession of only one pair of wings, that they are stingless flies. These are among a vast list of animals that gain protection because they look like something else: something inedible like a pebble, a twig or a frond of seaweed, or something positively nasty like a snake or a wasp or the glaring eyes of a possible predator.
Have bird eyes, then, been breeding insects for their resemblance to unpalatable or venomous models? There's one sense in which we surely have to answer yes. What, after all, is the difference between this and peahens breeding peacocks for beauty, or humans breeding dogs or roses? Mainly, peahens are breeding positively for something attractive, by approaching it, while the caterpillar-hunting birds are breeding negatively for something repellent, by avoiding it. Right then, here's another example, and in this case the 'breeding' is positive, even though the selector doesn't benefit from its choice. Far from it.
Deep-sea angler fish sit on the bottom of the sea, waiting patiently for prey.* Like many deep-sea fish, anglers are spectacularly ugly by our standards. Maybe by fish standards too, although it probably doesn't matter because, down where they live, it is too dark to see much anyway. Like other denizens of the deep sea, female angler fish often make their own light - or rather, they have special receptacles in which they house bacteria which make light for them. Such 'bioluminescence' isn't bright enough to reflect any detail, but it is bright enough to attract other fish. A spine which, in a normal fish, would be just one of the rays in a fin, becomes elongated and stiffened to make a fishing rod. In some species the 'rod' is so long and flexible that you'd call it a line rather than a rod. And on the end of the fishing rod or line is - what else? - a bait, or lure. The baits vary from species to species, but they always resemble small food items: perhaps a worm, or a small fish, or just a nondescript but temptingly jiggling morsel. Often the bait is actually luminous: another natural neon sign, and in this case the message being flashed is 'come and eat me'. Small fish are indeed tempted. They approach close to the bait. And it is the last thing they do for, at that moment, the angler opens her huge maw and the prey is engulfed with the inrush of water.
Now, would we say that the small prey fish are 'breeding for' more and more appealing lures, just as peahens breed for more appealing peacocks, and horticulturalists breed for more appealing roses? It's hard to see why not. In the case of the roses, the most attractive blooms are the ones deliberately chosen for breeding by the gardener. Much the same is true of peacocks chosen by peahens. It is possible that the peahens are not aware that they are choosing, whereas the rose-growers are. But that doesn't seem a very important distinction under the circumstances. Slightly more compelling is a distinction between the angler fish example and the other two. The prey fish are indeed choosing the most 'attractive' angler fish for breeding, via the indirect route of choosing them for survival by feeding them! Anglers with unattractive lures are more likely to starve to death and therefore less likely to breed. And the small prey fish are indeed doing the 'choosing'. But they are choosing with their lives! What we are homing in on here is true natural selection, and we are reaching the end of the progressive seduction that is this chapter.
Here's the progression laid out.
1 Humans deliberately choose attractive roses, sunflowers etc. for breeding, thereby preserving the genes that produce the attractive features. This is called artificial selection, it's something humans have known about since long before Darwin, and everybody understands that it is powerful enough to turn wolves into chihuahuas and to stretch maize cobs from inches to feet.
2 Peahens (we don't know whether consciously and deliberately, but let's guess not) choose attractive peacocks for breeding, again thereby preserving attractive genes. This is called sexual selection, and Darwin discovered it, or at least clearly recognized it and named it.
3 Small prey fish (definitely not deliberately) choose attractive angler fish for survival, by feeding the most attractive ones with their own bodies, thereby inadvertently choosing them for breeding and passing on, and therefore preserving, the genes that produce the attractive features. This is called - yes, we've finally got there - natural selection, and it was Darwin's greatest discovery.
Darwin's special genius realized that nature could play the role of selecting agent. Everybody knew about artificial selection,* or at least everybody with any experience of farms or gardens, dog shows or dovecotes. But it was Darwin who first spotted that you don't have to have a choosing agent. The choice can be made automatically by survival - or failure to survive. Survival counts, Darwin realized, because only survivors reproduce and pass on the genes (Darwin didn't use the word) that helped them to survive.
I chose the angler fish as my example, because this can still be represented as an agent using its eyes to choose that which survives. But we have reached the point in our argument - Darwin's point - where we no longer need to talk about a choosing agent at all. Move now from angler fish to, say, tuna or tarpon, fish that actively pursue their prey. By no sensible stretch of language or imagination could we claim that the prey 'choose' which tarpon survive by being eaten. What we can say, however, is that the tarpon that are better equipped to catch prey, for whatever reason - fast swimming muscles, keen eyes, etc. - will be the ones that survive, and therefore the ones that reproduce and pass on the genes that made them successful. They are 'chosen' by the very act of staying alive, whereas another tarpon that was, for whatever reason, less well equipped would not survive. So, we can add a fourth step to our list.
4 Without any kind of choosing agent, those individuals that are 'chosen' by the fact that they happen to possess superior equipment to survive are the most likely to reproduce, and therefore to pass on the genes for possessing superior equipment. Therefore every gene pool, in every species, tends to become filled with genes for making superior equipment for survival and reproduction.
Notice how all-encompassing natural selection is. The other examples I have mentioned, steps 1, 2 and 3 and lots of others, can all be wrapped up in natural selection, as special cases of the more general phenomenon. Darwin worked out the most general case of a phenomenon that people already knew about in restricted form. Hitherto, they had known about it only in the special case of artificial selection. The general case is the non-random survival of randomly varying hereditary equipment. It doesn't matter how the non-random survival comes about. It can be deliberate, explicitly intentional choice by an agent (as with humans choosing pedigree greyhounds for breeding); it can be inadvertent choice by an agent without explicit intention (as with peahens choosing peacocks for breeding); it can be inadvertent choice which the chooser - with a hindsight that is granted to us but not the chooser itself - would prefer not to have made (as with prey fish choosing to approach an angler fish's lure); or it can be something that we wouldn't recognize as choice at all, as when a tarpon survives by virtue of, say, an obscure biochemical advantage buried deep within its muscles, which gives it an extra burst of speed when pursuing prey. Darwin himself said it beautifully, in a favourite passage from On the Origin of Species:
It may be said that natural selection is daily and hourly scrutinising, throughout the world, every variation, even the slightest; rejecting that which is bad, preserving and adding up all that is good; silently and insensibly working, whenever and wherever opportunity offers, at the improvement of each organic being in relation to its organic and inorganic conditions of life. We see nothing of these slow changes in progress, until the hand of time has marked the long lapse of ages, and then so imperfect is our view into long past geological ages, that we see only that the forms of life are now different from what they formerly were.
I have here quoted, as is my usual practice, the first edition of Darwin's masterpiece. An interesting interpolation found its way into later editions: 'It may metaphorically be said that natural selection is daily and hourly . . .' (emphasis added). You might think that 'It may be said . . .' was cautious enough. But in 1866 Darwin received a letter from Wallace, co-discoverer of natural selection, suggesting that an even higher hedge against misunderstanding was regrettably necessary.
My dear Darwin, - I have been so repeatedly struck by the utter inability of numbers of intelligent persons to see clearly, or at all, the self-acting and necessary effects of Natural Selection, that I am led to conclude that the term itself, and your mode of illustrating it, however clear and beautiful to many of us, are yet not the best adapted to impress it on the general naturalist public.
Wallace went on to quote a French author called Janet, who was evidently, unlike Wallace and Darwin, a deeply muddled individual:
I see that he considers your weak point to be that you do not see that 'thought and direction are essential to the action of Natural Selection.' The same objection has been made a score of times by your chief opponents, and I have heard it as often stated myself in conversation. Now, I think this arises almost entirely from your choice of the term Natural Selection, and so constantly comparing it in its effects to man's selection, and also to your so frequently personifying nature as 'selecting', as 'preferring' . . . etc., etc. To the few this is as clear as daylight, and beautifully suggestive, but to many it is evidently a stumbling-block. I wish, therefore, to suggest to you the possibility of entirely avoiding this source of misconception in your great work, and also in future editions of the 'Origin,' and I think it may be done without difficulty and very effectually by adopting Spencer's term . . . 'Survival of the Fittest.' This term is the plain expression of the fact; 'Natural Selection' is a metaphorical expression of it . . .
Wallace had a point. Unfortunately, Spencer's term 'Survival of the Fittest' raises problems of its own, which Wallace couldn't have foreseen, and I won't go into them here. In spite of Wallace's warning, I prefer to follow Darwin's own strategy of introducing natural selection via domestication and artificial selection. I like to think that Monsieur Janet might have got the point this time around. But I did have another reason, too, for following Darwin's lead, and it is a good one. The ultimate test of a scientific hypothesis is experiment. Experiment specifically means that you don't just wait for nature to do something, and passively observe it and see what it correlates with. You go in there and do something. You manipulate. You change something, in a systematic way, and compare the result with a 'control' that lacks the change, or you compare it with a different change.
Experimental interference is of enormous importance, because without it you can never be sure that a correlation you observe has any causal significance. This can be illustrated by the so-called 'church clocks fallacy'. The clocks in the towers of two neighbouring churches chime the hours, but St A's a little before St B's. A Martian visitor, noting this, might infer that St A's chime caused St B's to chime. We, of course, know better, but the only real test of the hypothesis would be experimentally to ring the St A's chime at random times rather than once per hour. The Martian's prediction (which would of course be disproved in this case) is that St B's clock will still chime immediately after St A's. It is only experimental manipulation that can determine whether an observed correlation truly indicates causation.
If your hypothesis is that the non-random survival of random genetic variation has important evolutionary consequences, the experimental test of the hypothesis would have to be a deliberate human intervention. Go in and manipulate which variant survives and which doesn't. Go in there and choose, as a human breeder, which kinds of individuals get to reproduce. And that, of course, is artificial selection. Artificial selection is not just an analogy for natural selection. Artificial selection constitutes a true experimental - as opposed to observational - test of the hypothesis that selection causes evolutionary change.
Most of the known examples of artificial selection - for example, the manufacture of the various breeds of dog - are observed with the hindsight of history, rather than being deliberate tests of predictions under experimentally controlled conditions. But proper experiments have been done, and the results have always been as expected from the more anecdotal results on dogs, cabbages and sunflowers. Here is a typical example, an especially good one because agronomists at the Illinois Experimental Station began the experiment rather a long time ago, in 1896 (Generation 1 in the graph). The diagram above shows the oil content in maize seeds of two different artificially selected lines, one selected for high oil yield, and the other for low oil yield. This is a true experiment because we are comparing the results of two deliberate manipulations or interventions. Evidently the difference is dramatic, and it increases. It seems likely that both the upward trend and the downward trend would eventually level off: the low-yielding line because you can't drop below zero oil content, and the high-yielding line for reasons that are nearly as plain.
Two lines of maize selected for high and low oil content
Two lines of maize selected for high and low oil content
Here's a further laboratory demonstration of the power of artificial selection, which is instructive in another way. The diagram overleaf shows some seventeen generations of rats, artificially selected for resistance to tooth decay. The measure being plotted is the time, in days, that the rats were free of dental caries. At the start of the experiment, the typical period free of tooth decay was about 100 days. After only a dozen or so generations of systematic selection against caries, the decay-free period was about four times as long, or even more. Once again, a separate line was selected to evolve in the opposite direction: in this case the experiments systematically bred for susceptibility to tooth decay.
Was this article helpful?