Selfish information replicators?

There is a 'spat' in the current edition of New Scientist between Richard Dawkins and Edward O Wilson. It is the ongoing argument as to whether there is such a thing as group selection, distinct from kin selection, as set out in this times article.

As I see it, Dawkins' line is an extreme reductionist stance, reducing all evolution to selection of the basic replicator, the gene. I heard Dawkins do his 'selfish gene' talk at Oxford, when I was there c.1978, and it is pretty convincing and certainly elegant. But, is there another stage in the reduction? Is it the information that 'wants' to replicate? This is what computer viruses do, of course. So if it is the information, is that information necessarily only residing in the gene? What if information resides in the group? Could there not be selection for that?