"[On the :] It's a bit of a sterile conversation. I suppose I think of it as the cell. That's not to say that it can't act at the level of s. Of course it can. It does all the time. Any selfish gene is acting in it's own interest. I think the trouble with looking at selection only at the level of genes is it tends to downplay the importance of genetic conflict in a strange way... [I]f you have levels of selection you can have, for example... mitochondria... They were bacteria once. They're the power packs inside eukaryotic cells... [O]nce they get inside another cell, inside another originally, then they have an agenda of their own. They're making copies of themselves, and it's the speed at which the bacterium as a whole is making a copy of itself that means whether it tends to dominate in the population or not. It's not the individual genes. They will tend to throw away genes that they don't really need. And the host cell itself has got its agenda. It needs to make sure that it's getting benefits from this symbiont. It's not being taken over. It's not being eaten, and so it's... more intuitive to think of the interests of the cells themselves. And if you simply think of all of them as genes then you don't have that discrimination between the layers. Again, if you're thinking about s at the origin of life, the unit of selection in my mind is, "Can a cell make a copy of itself?" If you have a pure RNA world..."