In his earlier book, The Seven Daughters of Eve, Bryan Sykes described how mitochondrial DNA, passed only through the female line, reveals the descent of different populations from ancestral women. Here he turns to the other side of the coin: Y chromosomes, passed only through the male line, reveal descent from ancestral men. Surnames, at least in the system used traditionally in the English-speaking world, behave in the same way, being passed to children by their fathers, but not by their mothers. So does the information deduced from Y chromosomes agree with that implied by surnames? Do all of the people called Sykes, for example, descend from a particular man who adopted this surname in the 13th century when surnames started to be used systematically in England? There are at least two reasons why the answer might turn out to be no: numerous different unrelated men might have adopted the same name independently, leaving numerous different unrelated groups of descendants; or perhaps over several centuries sufficient sufficient women may have had sons by men other than their husbands to have eradicated the evidence. It turned out that about half of the men in England with the names Sykes today have the same (or almost the same) Y chromosome, a surprisingly high proportion that implies about 99% fidelity of transmission in each generation.
Sykes assures us that his figure of 1.3% "non-paternity events" (births of children whose biological and legal fathers are different) is a maximum, but he is forgetting a simple and plausible mechanism that would allow a much higher frequency. In past centuries married women suffered great social pressure to produce children, but what could a fertile woman married to a sterile man do to satisfy her parents-in-law? By far the least risky thing to do, even before anyone knew about Y chromosomes, would be to have a child by a brother-in-law. Such a "non-paternity event" would preserve the appearances, would keep everything within the family, and would be more likely to be tolerated and even approved of by the husband and his family than any other solution.
Much of the detailed information is fascinating, and the book is easy and enjoyable to read. It has some major faults, however, especially in the second half, where much of the discussion is muddled, speculative and at times dishonest. Some scientists avoid statistical arguments, preferring to base their conclusions on experimental results so overwhelmingly clear that statistical analysis is hardly necessary. This can be wasteful, but it is intellectually defensible and is adopted by some entirely respectable scientists. Bryan Sykes wants to have his cake and eat it, however, using statistical calculations when they suit the conclusions that he wants to draw, and dismissing them when they don't; that is not intellectually defensible. Having satisfied himself that the frequency of his own Y chromosome had increased 10000-fold since the 13th century, despite his belief that "the Sykeses were never wealthy or powerful" (forgetting, perhaps, that three Sykes baronetcies, two dating from the 18th century, do not suggest a family on the edge of poverty), he wonders about the reasons for its success. Rejecting, for no clear reason, the common-sense explanation that even without any selection some genes become fixed and others become extinct, he studies school registers in search of evidence that Sykeses have tended to have more sons than daughters, finding that 212 out of 393 Sykes children were boys. He calculates that a proportion of sons as high this was likely to occur "just under 6 percent" just by chance. (The correct answer is about 6.5%, but let that pass.) He neglects to inform his readers, most of whom cannot be expected to be knowledgeable about statistical conventions, that this would have been reported as "not significant" in any serious analysis ever since modern statistical tests were introduced in the 1920s.
Worse is to come, however. Finding a family with 23 girls and 4 boys he estimates the probability of this occurring by chance as 1 in 5000, which he describes as "very long odds indeed", even though he knows perfectly well that the calculation assumes a randomly chosen family, whereas he has carefully selected a specific family in the expectation of an extreme result. What is dishonest about this is that Sykes knows, and occasionally mentions, the reasons why his calculations are illiegitimate, but he hopes his readers will be fooled anyway. Eventually, coming to arguments that supposed evidence for a gene for homosexuality were unfounded, he refers contemptuously to people who think that statistical arguments should be used properly as "the guardians of statistical integrity" and tells us that in his experience they are usually wrong.
Several chapters in the middle of the book set out the ideas of sociobiology, but it is not obvious why these were included, as they add little to the rest of the book, and nothing to the far more expert accounts that can be found elsewhere, in the writings of Richard Dawkins, Mark Ridley, Matt Ridley, and others. At least we are spared, however, the seven chapters of romantic fiction that came close to ruining The Seven Daughters of Eve, and the purple passages dealing with Gaia are mercifully short.
The subtitle of the book promises a "future without men", and the blurb assures us that "men are slowly, but surely, headed for extinction". For most of the book this seems to be little more than advertising hype, and the evidence, when it eventually comes is thin and speculative. It is also confusing and muddled, because although some problems may be specifically human - maintaining the testicles too warm, excessive exposure to mutagens, etc. - the main theme that Sykes develops is of a fight to the death between the man-hating mitochondria and the women-hating Y chromosomes, which, he thinks, will eventually be won by the mitochondria. However, if this fight is going on it must affect not just people but all of the great array of other animals that determine sex in the same way, including virtually all mammals.
30 or 40 years ago popular books about the human species, such as those of Robert Ardrey or Desmond Morris, tended to stress Man the Mighty Hunter, with an obsession that Elaine Morgan rightly ridiculed as the Tarzan school of biological thinking. Fashions change, however, and if one may judge from this book and Steve Jones's recent "Y: the Descent of Men" the current vogue is for books about Man the Pathetic Wimp, with Y chromosomes in bad shape, a "graveyard of rotting genes", and sperms of low and diminishing mobility, unlike those of our cousins the chimpanzees, 100% of whose sperms are "in excellent shape". No doubt this explains the success of chimpanzees in spreading themselves over the entire earth and restricting humans to tiny relics of their original range; but maybe I have missed a key step in the argument somewhere.