of genes, a small minority will be beneficial and will become, with time, more common. So common, in fact, that it is hardly fair to refer to them as ‘mutations’, and instead we call them ‘variants’ or, more technically, ‘polymorphisms’. In Africa, the Δ32 polymorphism of the CCR5 gene is currently increasing in frequency because it confers resistance to human immunodeficiency virus and so to AIDS. This is something new, but many polymorphisms are ancient. They are the stuff from which human diversity is made. They give us variety in skin colour, height, weight and facial features, and they surely also give us at least some of our variety in temperament, intelligence, addictive habits. They may cause disease, but mostly the diseases of old age such as senile dementia and heart attacks.
How common does a mutation have to be before it becomes a polymorphism? The answer is a bit arbitrary, but if a variant sequence has a global frequency of 1 per cent or more it is assumed that it cannot have caused much harm in its history, and may even have conferred some benefit to its carriers. By this criterion, at least one polymorphism has been detected in about 65 per cent of the human genes in which they have been sought, but some genes have dozens. This variety should not overwhelm us. Most human genes have one variant that is far more common than all others, and it is quite sensible to speak of that variant as being normal, albeit only in the statistical sense.
Perfection is far more problematic. The only reason to say that one genetic variant is ‘better’ than another is if it confers greater reproductive success on those who bear it; that is, if it has a higher Darwinian fitness than other variants. It is likely that the most common variant is the best under most circumstances, but this cannot be proved, for the frequencies of gene variants are shaped by history, and what was best then need not be best either now or in the future. To prefer one polymorphism over another – or rather to prefer the way it surfaces in our looks – is merely to express a taste. By this I mean the sort of claim made by the great French naturalist George Leclerc Buffon when he asserted that, for their fair skin and black eyes, the women of the Caucasus Mountains were lovelier than all others. Or when Karen Blixen eulogised the beauty of the Masai
morani
. Recognition of, even a delight in, human genetic diversity does not, however, commit us to a thorough-going genetic relativism. Many of the mutations that batter our genomes do us harm by any criterion.
Each new embryo has about a hundred mutations that its parents did not have. These new mutations are unique to a particular sperm or ovum, were acquired while these cells were in the parental gonads and were not present when the embryo’s parents were themselves embryos. Of these hundred mutations, about four will alter the meaning of genes by changing the amino acid sequences of proteins. And of these four content-altering mutations, about three will be harmful. To be more precise, they will affect the ultimate reproductive success of the embryo, at least enough to ensure that, with time, natural selection will drive them to extinction.
These are uncertain numbers: the fraction of deleterious mutations can only be estimated by indirect methods. But if they are at all correct, their implications are terrifying. They tell us that our health and happiness are being continually eroded by an unceasing supply of genetic error. But matters are worse than that. Not only are we each burdened with our own unique suite of harmful mutations, we also have to cope with those we inherited from our parents, and they from theirs, and so on. What is the total mutational burden on the average human being? The length of time that a given mutation will be passed down from one generation to the next depends on the severity of its effects. If we suppose that an average mutation has only a mildly