No more blondes? No!
Snopes today has an interesting article about a persistent urban myth that human blondes are gradually going extinct. However they do not actually go into any of the scientific and mathematical arguments that show why this is extremely unlikely. Two mechanisms are proposed in the myths - that blondes are going extinct because blondeness is recessive, and that blondes are going extinct because they are selected against. Neither holds up under scrutiny.
First, some background. Genes generally come in pairs, one from each parent. A common example is eye colour. Suppose your parents are both brown eyed, and both have one brown-eyed gene and one blue-eyed gene. You'll get one of their genes at random from each, so your genes will either be brown-brown, blue-brown, brown-blue or blue-blue. But because brown is dominant and blue is recessive, only the last possibility actually is expressed as having blue eyes.
Now, it seems like if there's only a one-in-four chance of the child having blue eyes, that naturally over time, blue-eyed-ness will drop out of the population. But that's a thoroughly specious analysis! In actuality, there's a three-in-four chance that each child of these parents has at least one blue eyed gene, and that saves the blue eyed people from dying out.
Don't believe me? Let's make up some numbers. Let's call the brown-eyed gene B and the blue eyed gene b. We'll give the gene from the father first, followed by the gene from the mother. Children with BB, Bb, bB are brown eyed, children with bb are blue eyed. Let's suppose that these are the only two eye colours, and that the distribution in the population in the current generation is:
BB: 25%
Bb: 25%
bB: 25%
bb: 25%
So only a quarter of the population is blue eyed, but three quarters has at least one blue eyed gene.
Let's make some reasonable assumptions:
• there is no difference in eye colour distribution among men and women
• mating is random insofar as this particular gene is concerned - that blue-eyed people do not seek out other blue-eyed people for the purposes of furthering the blue-eyed race, but rather that they have other reasons for choosing a mate.
• the net mutation rate is zero - that random mutations are as likely to turn B into b as they are to turn b into B.
• the population is large and inbreeding is rare
• all people are equally likely to have children, irrespective of their eye colour genes
• the population as a whole is not shrinking
What will happen?
Suppose there are 64 million people -- 32 million couples each of whom is going to have on average two children and then die. There are 2 million of each of the following kinds of couple:
Father Mother Children BB BB 4 million BB BB Bb 2 million BB, 2 million Bb BB bB 2 million BB, 2 million Bb BB bb 4 million Bb Bb BB 2 million BB, 2 million bB Bb Bb 1 million BB, 1 million Bb, 1 million bB, 1 million bb Bb bB 1 million BB, 1 million Bb, 1 million bB, 1 million bb Bb bb 2 million Bb, 2 million bb bB BB 2 million BB, 2 million bB bB Bb 1 million BB, 1 million Bb, 1 million bB, 1 million bb bB bB 1 million BB, 1 million Bb, 1 million bB, 1 million bb bB bb 2 million Bb, 2 million bb bb BB 4 million bB bb Bb 2 million bB, 2 million bb bb bB 2 million bB, 2 million bb bb bb 4 million bb
For a grand total of 16 million BB, 16 million Bb, 16 million bB, 16 million bb! The exact same proportions as before! Holy cow! How'd that happen?
Now, you might be wondering if I'm just fiddling with the figures here. Maybe it's just because of the initial conditions I picked. Well, yes and no. You can pick any distribution you want for the first generation and you'll discover that sometimes in the second generation, the proportions may have changed. Am I pulling the wool over your eyes here by picking an example that just happens to work?
No. No matter what proportions you pick for the first generation, the third generation proportions are always the same as the second generation. It only takes two generations to achieve a stable equilibrium for any recessive trait. Proving this fact is a relatively simple bit of algebra, which I leave as an exercise. (This fact is now known as Hardy's Law, after the mathematician who somewhat snarkily pointed it out to his biologist collegues.) The percentages need not be all 25% -- any set of percentages such that the number of BBs times the number of bbs is equal to the number of Bbs times the number of bBs will have this property. For example, 36%, 24%, 24%, 16% is also stable because 36 x 16 = 24 x 24. (Notice that the Bb and bB percentages are for practical purposes always the same, because we assumed that men and women are equally likely to have any combination. Therefore the number of people with one of each should be split evenly between the ones who got the recessive from their father and ones who got it from their mother.) The recessive gene can be very seldom expressed but still have a relatively large population of carriers -- from this analysis you can see that if the number of bbs is 0.1% then there will be about 7% of the population who are carriers with one recessive and one dominant.
So recessiveness of a trait does not cause it to die out - in fact, it preserves it by distributing the recessive gene amongst a large population of people carrying the dominant gene! The argument that blondes are dying out because blondeness is recessive does not hold water.
What about the idea that blondes are selected against? Maybe blondes (that is, people with two of the recessive blonde gene) are less attractive as mates and therefore blondeness is being selected against, so the percentage of blondes in the world is gradually declining.
Maybe there's something there. Let's make it a little stronger. Suppose there were hardly any blondes at all, and that the selection pressure was extremely strong. That is, something is preventing blondes from successfully reproducing. Would blondeness be disappearing rapidly?
We know from observing the world that this is not the case, because we can replace the word "blonde" with "haemophiliac". Haemophilia, a disease which prevents blood from clotting and therefore causes people to bleed to death from small cuts, is a rare recessive genetic disorder which historically has frequently killed children before puberty. None of the "bb" kids are reproducing. Surely if selection pressure was gradually reducing the incidence of the recessive haemophilia gene, it would have disappeared from humans millions of years ago when the last one bled to death after getting a scratch. But again, its very recessiveness is what keeps it in the population - that there are many carriers with only one copy surviving and passing their recessive copies along to half their children.
Clearly blondeness is not disappearing because of either being recessive or being selected against. But haven't I just made an argument that no traits ever change then? Surely that's not the case - that would be an argument against the fact of evolution itself.
And indeed I haven't. Blondeness, or any other trait, could change in distribution if any of the assumptions I've made above are violated. If there was some selection pressure that allowed people carrying a recessive blonde gene to be selected against, that would decrease the percentage of carriers and hence the percentage of blondes. If random mutations tended to destroy blonde genes more than other hair colour genes, then blondeness would be declining. If some segment of the population is small and inbred then the percentage of blondeness would increase in that segment (and indeed, the percentage of haemophiliacs amongst the inbred royal families of Europe is enormous compared to the general population.) And if the population as a whole is consistently shrinking over the long term then some day blondes will go extinct along with the rest of the humans.
Blondes are safe. Don't worry.
First, some background. Genes generally come in pairs, one from each parent. A common example is eye colour. Suppose your parents are both brown eyed, and both have one brown-eyed gene and one blue-eyed gene. You'll get one of their genes at random from each, so your genes will either be brown-brown, blue-brown, brown-blue or blue-blue. But because brown is dominant and blue is recessive, only the last possibility actually is expressed as having blue eyes.
Now, it seems like if there's only a one-in-four chance of the child having blue eyes, that naturally over time, blue-eyed-ness will drop out of the population. But that's a thoroughly specious analysis! In actuality, there's a three-in-four chance that each child of these parents has at least one blue eyed gene, and that saves the blue eyed people from dying out.
Don't believe me? Let's make up some numbers. Let's call the brown-eyed gene B and the blue eyed gene b. We'll give the gene from the father first, followed by the gene from the mother. Children with BB, Bb, bB are brown eyed, children with bb are blue eyed. Let's suppose that these are the only two eye colours, and that the distribution in the population in the current generation is:
BB: 25%
Bb: 25%
bB: 25%
bb: 25%
So only a quarter of the population is blue eyed, but three quarters has at least one blue eyed gene.
Let's make some reasonable assumptions:
• there is no difference in eye colour distribution among men and women
• mating is random insofar as this particular gene is concerned - that blue-eyed people do not seek out other blue-eyed people for the purposes of furthering the blue-eyed race, but rather that they have other reasons for choosing a mate.
• the net mutation rate is zero - that random mutations are as likely to turn B into b as they are to turn b into B.
• the population is large and inbreeding is rare
• all people are equally likely to have children, irrespective of their eye colour genes
• the population as a whole is not shrinking
What will happen?
Suppose there are 64 million people -- 32 million couples each of whom is going to have on average two children and then die. There are 2 million of each of the following kinds of couple:
Father Mother Children BB BB 4 million BB BB Bb 2 million BB, 2 million Bb BB bB 2 million BB, 2 million Bb BB bb 4 million Bb Bb BB 2 million BB, 2 million bB Bb Bb 1 million BB, 1 million Bb, 1 million bB, 1 million bb Bb bB 1 million BB, 1 million Bb, 1 million bB, 1 million bb Bb bb 2 million Bb, 2 million bb bB BB 2 million BB, 2 million bB bB Bb 1 million BB, 1 million Bb, 1 million bB, 1 million bb bB bB 1 million BB, 1 million Bb, 1 million bB, 1 million bb bB bb 2 million Bb, 2 million bb bb BB 4 million bB bb Bb 2 million bB, 2 million bb bb bB 2 million bB, 2 million bb bb bb 4 million bb
For a grand total of 16 million BB, 16 million Bb, 16 million bB, 16 million bb! The exact same proportions as before! Holy cow! How'd that happen?
Now, you might be wondering if I'm just fiddling with the figures here. Maybe it's just because of the initial conditions I picked. Well, yes and no. You can pick any distribution you want for the first generation and you'll discover that sometimes in the second generation, the proportions may have changed. Am I pulling the wool over your eyes here by picking an example that just happens to work?
No. No matter what proportions you pick for the first generation, the third generation proportions are always the same as the second generation. It only takes two generations to achieve a stable equilibrium for any recessive trait. Proving this fact is a relatively simple bit of algebra, which I leave as an exercise. (This fact is now known as Hardy's Law, after the mathematician who somewhat snarkily pointed it out to his biologist collegues.) The percentages need not be all 25% -- any set of percentages such that the number of BBs times the number of bbs is equal to the number of Bbs times the number of bBs will have this property. For example, 36%, 24%, 24%, 16% is also stable because 36 x 16 = 24 x 24. (Notice that the Bb and bB percentages are for practical purposes always the same, because we assumed that men and women are equally likely to have any combination. Therefore the number of people with one of each should be split evenly between the ones who got the recessive from their father and ones who got it from their mother.) The recessive gene can be very seldom expressed but still have a relatively large population of carriers -- from this analysis you can see that if the number of bbs is 0.1% then there will be about 7% of the population who are carriers with one recessive and one dominant.
So recessiveness of a trait does not cause it to die out - in fact, it preserves it by distributing the recessive gene amongst a large population of people carrying the dominant gene! The argument that blondes are dying out because blondeness is recessive does not hold water.
What about the idea that blondes are selected against? Maybe blondes (that is, people with two of the recessive blonde gene) are less attractive as mates and therefore blondeness is being selected against, so the percentage of blondes in the world is gradually declining.
Maybe there's something there. Let's make it a little stronger. Suppose there were hardly any blondes at all, and that the selection pressure was extremely strong. That is, something is preventing blondes from successfully reproducing. Would blondeness be disappearing rapidly?
We know from observing the world that this is not the case, because we can replace the word "blonde" with "haemophiliac". Haemophilia, a disease which prevents blood from clotting and therefore causes people to bleed to death from small cuts, is a rare recessive genetic disorder which historically has frequently killed children before puberty. None of the "bb" kids are reproducing. Surely if selection pressure was gradually reducing the incidence of the recessive haemophilia gene, it would have disappeared from humans millions of years ago when the last one bled to death after getting a scratch. But again, its very recessiveness is what keeps it in the population - that there are many carriers with only one copy surviving and passing their recessive copies along to half their children.
Clearly blondeness is not disappearing because of either being recessive or being selected against. But haven't I just made an argument that no traits ever change then? Surely that's not the case - that would be an argument against the fact of evolution itself.
And indeed I haven't. Blondeness, or any other trait, could change in distribution if any of the assumptions I've made above are violated. If there was some selection pressure that allowed people carrying a recessive blonde gene to be selected against, that would decrease the percentage of carriers and hence the percentage of blondes. If random mutations tended to destroy blonde genes more than other hair colour genes, then blondeness would be declining. If some segment of the population is small and inbred then the percentage of blondeness would increase in that segment (and indeed, the percentage of haemophiliacs amongst the inbred royal families of Europe is enormous compared to the general population.) And if the population as a whole is consistently shrinking over the long term then some day blondes will go extinct along with the rest of the humans.
Blondes are safe. Don't worry.