I meant to write this as a new topic a while ago, but never quite got round to it. There’s a lot of confusion around about things like Mitochondrial Eve and what she means for our understanding of the world – problems compounded by typically poor science reporting in the media and by sloppy language from those that should know better. I thought it would be useful to have one post briefly summarising what we know about these common ancestors, how we know it, and what it all means for the EvC debate. Here seems like an appropriate thread!
Mitochondiral DNA is useful to study, because it is inherited wholly from the mother. Sperm don’t have any mitchondria, so all the mitochondria in your body are directly cloned from the mitochondria in your mother’s egg. Most DNA is a mixture from both parents, so your particular DNA will be unique, but, without any mixing with the father’s DNA, your mitochondrial DNA should be exactly the same as your mother’s, and her mother’s, and her mother’s (and so on ad inifinitum).
Of course, it’s not exactly the same, because mutations occur, changing it slightly. You and your sister, then, would have ever so slightly different mitochondrial DNA. Your cousins who share a maternal grandmother would be slightly more different again. Your more distant cousins, with whom you share your mother’s mother’s mother, would be slightly more different still.
By working out the average rate of mutations per generation, scientists can take mitochondrial DNA from different people, and estimate how long it’s been since the direct-line of female ancestors of these people separated. And that’s what was done – taking the DNA of people from different ethnic groups all over the world to see how far back the last common matrilineal ancestor may have lived. The estimate arrived at is 200,000 years ago.
Note that this research isn’t about establishing that we had a matrilineal common ancestor – this is assumed as obvious. It’s about trying to work out how recent she was.
Y-chromosome Adam works on similar principles, except this time its DNA passed down only by men. Y-chromosomes come only from your dad, and the DNA isn’t mixed with any from the woman at all. So, using the same idea, we can look at how different the Y-chromosone DNA is between different people from all over the world, and come up with an estimate as to how recently they all shared a patrilineal ancestor. The estimate for this is much more recent – about 60,000 years ago.
Why so much more recent? It’s because of the difference in reproductive success between men and women. Men have to put a lot less effort into making a baby, so they can have far more children than a woman ever could. There are only so many women to go around, however, so if some men are having piles of kids, a lot of men are having none. The Y-chromosome lineage of every man who fails to produce a son is lost, just as the mitochondrial lineage of every woman who fails to produce a daughter is lost. Throughout history, many more men have failed to produce a son than women have failed to produce a daughter.
The most recent common ancestor
The important thing to remember about this Adam and Eve, is that they are not our most recent common ancestor, nor our most recent common male or female ancestors. This is obvious once you think about. Look at my hastily-drawn diagrams of Dave and Zebedee’s recent ancestors below.
There are eight different ways Dave and Zebedee could share an ancestor in the top generation. A could be the same person as E, A could be the same person as G, B could be the same person person as F, B could be the same person as H etc. etc. There are four ways each that they could share a common male or female ancestor.
However, for them to have the same matrilineal ancestor, there is only one way. A has to be the same person as E. If we continued to join the trees backwards, the number of people in each generation would double, and so there’d be ever more and more people we could look through to find a common ancestor. In each new generation, though, there’d only be one new direct-matrilineal ancestor, or one new direct patrilineal ancestor.
What all this means for the Creation/Evolution debate
As noted about, the fact that we have common ancestors does not support creationism. It’s a prediction of evolutionary theory that all people have a common ancestor (as, indeed, does all life), so it would be evidence against this that different races had been created separately, as some people used to think.
This evidence does provide problems for a young-earth creationist view, though. If all people are descended from one woman 6,000 years ago, then there have only been 6,000 years for one mitochondrial genome to mutate into all the different versions of it in the world today. Similarly, if all people are descended from one man 6,000 years ago, then there has only been 6,000 years for one Y-chromosome to mutate into all the different versions knocking around today. As we’ve seen, at current rates of mutation, these figures are about 30 and 10 times too short, respectively.
The actual problem is worse as, according to the literalist biblical account, the population was reduced dramatically by the flood, just over 4,000 years ago. All the men that survived the Ark were a father and his three sons, so diversity was reduced back to one, almost identical Y-chromosome, which must have explosively diversified at rates totally unheard of today, to make it look like it underwent 60,000 years of mutation in just 4,000. And, sadly for the classic creationist response that ‘things were different before the Flood’, all of this had to happen after the Flood.
This has gone on plenty long enough, so I'll leave it there.
my question is how can you limit the impact of assumptions by using more assumptions?
Because you use different assumptions and see if it varies the results significantly.
Imagine that we're doing a calculation, but there's a variable A that we're not sure of. You take what seems to be a reasonable figure for A, and do your calculation to get an estimate of whatever it is you're trying to calculate.
But somebody could rightly point out that, since you don't know what number A should really have been, you don't know how likely your estimate is to be correct. It could turn out to be way off.
So, you vary the assumption and try a bunch of different numbers for A. If it turns out that you get a very similar answer, even with very diferent numbers for A, then the estimate is robust. You can be confident in it, because it's approximately right even if you got A quite significantly wrong.