By 'meta-evolution', we refer to speculations concerning 1) general constraints that might lead the pace of Evolution and 2) general conditions that made Evolution possible. It is commonplace to dismiss the attempt at finding general principles of evolution; Darwin is said to have reduced the history of life to mere contingency (see the complete works of Stephen Jay Gould for a reminder). Such a dismissal appears to us as a neglect of ecological demands on organisms (and human societies) : environmental constraints seem to promote a handful of good solutions to universal problems, so that one can expect a convergence towards these few solution.

Yet, many biologists, especially those concerned with the new synthesis emerging between Evolution and Development (Evo/Devo) are questionning this assumption; natural selection of blind mutations can only do so much: had not nature provided it with a material of variation sufficiently rich, flexible, easily removable, in a word evolvable, then Evolution would not even have started.

So it seems that a set of minimal constraints is necessary to evolvability; these are very simple (and do not require any mythical agency to appear), but neglected until the last fifteen years. One of them is called 'modularity' or 'decomposability': it is the ability for a complex system (be it an organism, a mind, a machine, or whatever), to undergo a change in one of its parts without other parts being affected (or minimally so). Ikea furnitures are often modular: they can be rebuilt or replaced piecemeal. One can see why mudularity is important for evolvability, since it allows for change, but not for chaos. This property is actually not so frequent in nature; organisms and artefacts (or so the theory goes) are supposed to maximize it. A well-known and much debated theory in cognitive science has it that brains, too, are modular. How does the first assumption relate to the second?

Click here for a bibliography.