On the information question, by the way, Jan, let's consider one more example. Start with that hypothetical snippet of DNA from above: ATAGTACCT Suppose that a random mutation changes one "letter" of the code, with the result: ATAGTATCT Does this, in your opinion, constitute a loss of information? Let's suppose that the first snippet codes for some useful trait, while the second leads to something that harms the carrier of the gene. Is the mutation then a loss of information? Now consider an organism that starts with the second sequence. A random mutation, same as before, changes one letter of the code, producing (at random) the first sequence. Note that this process is exactly as likely to occur as the first change, above. In this second case, if we are to assume that information was lost in the first example mutation, then it seems to me that we are forced to conclude by the same reasoning that information was gained in the second mutation. This would constitute an example of increase in the information content of the genome, according to this picture of what "information" is (which I am not saying I agree with, by the way).