Thursday, February 12, 2009

Darwinian Nuggets   posted by DavidB @ 2/12/2009 01:34:00 AM

Today is the 200th birthday of you-know-who. I'm sure most people are already sated with Darwiniana, but I can't let the day pass without making my own small contribution.

One of the pleasures of reading Darwin is that one can still find nuggets of insight in unexpected places. Here I want to describe two of these: Darwin's discussion of incest taboos, and his treatment of animal combat. [Added 14 February: On the latter, see the Addendum at the end of the post.]

The obvious place to look for a discussion of human incest in Darwin's works would be in the Descent of Man, but the first edition (1871) does not even mention incest. Instead, Darwin's fullest discussion of the subject comes in the unlikely setting of The Variation of Animals and Plants under Domestication. This is Darwin's longest single work, and, as its title suggests, it is mainly concerned with a detailed analysis of the varieties of domesticated animals and plants, including 100 pages on pigeons alone. But it also includes a mass of information and discussion on heredity in general (including Darwin's erroneous 'provisional hypothesis of pangenesis'). As part of the treatment of heredity, Darwin discusses the effects of cross-breeding and inbreeding, so a discussion of human incest taboos is not quite as much of a digression as it might seem.

In the first edition (1868) Darwin discusses (and rejects) previous explanations for incest prohibitions, and continues as follows:

It would be interesting to know, if it could be ascertained, as throwing light on this question with respect to man, what occurs with the higher anthropomorphous apes - whether the young males and females soon wander away from their parents, or whether the old males become jealous of their sons and expel them, or whether any inherited instinctive feeling, from being beneficial, has been generated, leading the young males and females of the same family to prefer pairing with distinct families, and to dislike pairing with each other. A considerable body of evidence has already been advanced, showing that the offspring from parents which are not related are more vigorous and fertile than those from parents which are closely related; hence any slight feeling, arising from the sexual excitement of novelty or other cause, which led to the former rather than the latter unions, would be augmented through natural selection, and thus might become instinctive; for those individuals which had an innate preference of this kind would increase in number. It seems more probable that degraded savages should thus unconsciously have acquired their dislike and even abhorrence of incestuous marriages, rather than that they should have discovered by observation and reasoning the evil results. The abhorrence occasionally failing is no valid argument against the feeling being instinctive, for any instinct may occasionally fail or become vitiated, as sometimes occurs with parental love and the social sympathies. In the case of man, the question whether evil follows from close interbreeding will probably never be answered by direct evidence, as he propagates his kind so slowly and cannot be subject to experiment; but the almost universal practice of all races at all times of avoiding closely-related marriages is an argument of considerable weight; and whatever conclusion we arrive at in regard to the higher animals may safely be extended to man (1868, vol.2, pp.123-4)

In the second edition (1875) this passage is heavily revised, to take account of new evidence. Darwin's own son George had made a statistical investigation of the effects of first-cousin marriage (in which the Darwins had a close personal interest!), and concluded that the effects were slight. Alfred Henry Huth had also recently (1875) published a book, The Marriage of Near Kin, which concluded that there was no instinctive feeling against incest in man. Darwin expressed some reservations about Huth's evidence, but accepted the main conclusion. As a result, in the second edition his treatment of incest was revised and shortened as follows:

It has been clearly shown by Mr Huth that there is no instinctive feeling against incest in man any more than in gregarious animals. We know also how readily any prejudice or feeling may give rise to abhorrence, as shown by Hindus in regard to objects causing defilement. Although there seems to be no strong inherited feeling in mankind against incest, it seems possible that men during primeval times may have been more excited by strange females than by those with whom they habitually lived; in the same manner as according to Mr Cupples, male deerhounds are inclined towards strange females, while the females prefer dogs with whom they have associated. If any such feelings formerly existed in man, this would have led to a preference for marriages beyond the nearest kin, and might have been strengthened by the offspring of such marriages surviving in greater numbers, as analogy would lead us to believe would have occurred (1875, vol.2, pp.103-4)

It will be seen that the theory is somewhat weakened, since it no longer accounts for 'abhorrence' of incest (which Darwin, following Huth, now believes is not instinctive), but only a 'preference' for more distant marriages. To account for 'abhorrence' (which Darwin still believes is widespread, though not instinctive), Darwin has to appeal to reinforcement by social customs and beliefs. This subject is covered more fully in the second (1874) edition of the Descent of Man, which contains an extended discussion of social conventions. Particularly relevant to incest is the following:

The breach of a rule held sacred by the tribe, will thus, as it seems, give rise to the deepest feelings, - and this quite apart from the social instincts, excepting in so far as the rule is grounded on the judgement of the community. How so many strange superstitions have arisen throughout the world we know not; nor can we tell how some great and real crimes, such as incest, have come to be held in abhorrence (which is not however quite universal) by the lowest savages. It is even doubtful whether in some tribes incest would be looked on with greater horror, than would the marriage of a man with a woman bearing the same name, though not a relation....[quotes examples from E. B. Tylor] We may therefore reject the belief, lately insisted on by some writers, that the abhorrence of incest is due to our possessing a special God-implanted conscience.(1874, p.176)

Darwin thus presents not one but two theories of incest prohibitions, the second of which is close to that usually attributed to Edward Westermarck, writing some twenty years later. Whether either of them is correct I will not consider. The subject of incest prohibitions has been extensively discussed by modern sociobiologists, but so far as I am aware Darwin's accounts in the Variation have not been noticed.

My second 'nugget' concerns Darwin's treatment of animal combat, specifically combat between males to obtain access to females. In this case Darwin's treatment does indeed come in the expected place - the part of the Descent of Man concerned with sexual selection - but the passage I want to focus on occurs only in the second (1874) edition. This used to be the standard version, but since the reprinting of the first (1871) edition by Princeton UP, the second edition is probably now seldom read. In the first edition Darwin gives a long discussion of combat between male mammals, such as rams, bulls, and stags. He attempts to deal with the problem that some animal weapons, such as the long backward-pointing antlers of the Oryx, do not appear very efficient for fighting. In dealing with this he emphasises that the effectiveness of the weapon needs to be considered in relation to the existing weapons and fighting methods of the species concerned. In the case of the Oryx, once they had acquired the method of kneeling down, lowering their heads, and attempting to insert their antlers under the rival's neck, then a gradual evolution of longer and longer antlers could be advantageous (1871, vol.2, p250.) He then considers the branching antlers of most deer species, and discusses how they are used. But he is doubtful that they would always be as effective as simpler antlers, and speculates that they may have a subsidiary advantage for purposes of display (p.254). He then quotes a recent report from America describing a variety of Cervus Virginianus, known as the spike-horn variety, with a simpler, straighter antler, which appears to be increasing in numbers, so that the species is being modified by sexual and natural selection (p.255).

This is all good stuff, but a critic objected that if simpler antlers were more effective, branched antlers should not (on Darwin's principles) ever have evolved. In the 1874 edition he answers this objection as follows:

To this I can only answer by remarking, that a new mode of attack with new weapons might be a great advantage, as shown by the case of the Ovis cycloceros [described earlier] who thus conquered a domestic ram famous for his fighting power. Though the branched antlers of the stag are well adapted for fighting with his rivals, and though it might be an advantage to the prong-horned variety slowly to acquire long and branched horns, if he had to fight only with others of the same kind, yet it by no means follows that branched horns would be the best fitted for conquering a foe differently armed. In the foregoing case of the Oryx leucoryx, it is almost certain that the victory would rest with an antelope having short horns, and who therefore did not have to kneel down, though an oryx might profit by having still longer horns, if he fought only with his proper rivals. (1874, p.783)

I think this shows a sophisticated grasp of the concept of frequency-dependent selection, and might well be regarded as anticipating modern evolutionary game theory. In case this seems fanciful, consider the following:

As everyone knows, male deer have branched antlers. During the breeding season, two stags fight by lowering their heads so that their antlers interlock. Each then attempts to force the other backwards, until at last the weakest is forced to break away and flee. Because of the branching structure of the antlers it is rare for a stag to be pierced by its opponent's antlers. Occasionally, however, a stag grows antlers without branches; such a stag may wound and kill its adversary.... Why should natural selection have favoured a device - the branching of antlers - which appears to reduce the chances a stag may have of winning fights? (1972, p.8)

This passage is not from Darwin, but from an essay by John Maynard Smith which contains the first application of game theory to animal behaviour, predating the better-known paper by Maynard Smith and George Price. Maynard Smith goes on to stress, like Darwin, that the optimal 'strategy' in a given situation will depend on what other animals are doing. Maynard Smith gives no source for his stag example, and he does not cite Darwin. Yet it is virtually certain that Maynard Smith, an expert on sexual selection, had at some point read the second edition of the Descent of Man (the only one easily available at that time). I suspect that the stag example had unconsciously lodged in his memory, and may even have helped to stimulate the train of thought that led to evolutionary game theory itself.

[Addendum: The connection of game theory with Darwin may actually be more through Price rather than Maynard Smith. Price wrote a long paper in 1969 titled 'Antlers, intraspecific combat, and altruism' and submitted it to Nature, who sent it to John Maynard Smith as referee. JMS recommended acceptance in principle, subject to substantial shortening. Price did not resubmit it and nothing further was heard from him. In 1972 JMS incorporated some of Price's ideas in his essay cited above, with due acknowledgement to Price. Price and JMS then got back in touch, and wrote their joint paper (1973). Price's long original manuscript has not been found, but according to a description by Price it specifically refered to Darwin's discussion of deers' antlers. For all this see (1995)]


1868. Charles Darwin: The Variation of Animals and Plants under Domestication, first edition.
1871. Charles Darwin: The Descent of Man, and Selection in Relation to Sex, first edition.
1874. Charles Darwin: The Variation of Animals and Plants under Domestication, second edition.
1875. Charles Darwin: The Descent of Man, and Selection in Relation to Sex, second edition, quoted here from a single-volume edition of 1901.
1972: John Maynard Smith, 'Game Theory and the Evolution of Fighting', in On Evolution.
1973: John Maynard Smith and George Price, 'The Logic of Animal Conflict', Nature, 246, 15-18.
1995: Steven A. Frank, 'George Price's Contributions to Evolutionary Genetics', J. Theoretical Biol., 175, 373-88.