The central claim of the book is that the biosphere cannot be deduced from the physical laws of the universe. Of course, he cannot prove this (over the centuries countless things were deemed to be unprovable and most of them now have a simple physical explanation). Ditto for human society. Kauffman's reasoning is that Physics only admits "happenings" and not "doings", whereas Biology is about "doings". Agency emerged from the evolution of life: "we act on our own behalf". That, of course, is questionable. A strict materialist would reply that we are just robots programmed to react (not "act") to stimuli from the environment.

His starting point is an article by Philip Anderson, "More is Different" (1972), that depicted symmetry breaking (one of the fundamental features of this universe) as an emergent phenomenon. Anderson used that idea to imagine layers of reality at each of which new laws apply that cannot be reduced to the laws of another layer. Kauffman mixes this with Godel's theorem (that any formal system contains at least one statement that cannot be proven true) and with the multiple-realization argument (that, for example, a computation can be performed using different media and therefore cannot be "reduced" to the medium, although a philosopher would object that Kauffman is confusing "reduction" and "implementation"). Kauffman details several examples of "emergent" views, from Robert Laughlin to Leo Kadanoff. His main argument, however, is that it is just impossible to write down all the equations for all the particles that compose the biosphere and deduce from those equations the biosphere as it is. That's a claim that, of course, he cannot prove. It just sounds plausible to him that no collective effort and no supercomputer would ever be able to achieve this feat. He then claims that the evolution of the biosphere yielded agency, value and consciousness. Since the biosphere cannot be reduced to the physical then these "things" cannot be reduced to the physical either. Agency, in particular, means that we affect the evolution of the universe. This becomes a reoccurring theme of the book: there are phenomena that influence each other, thereby creating nonlinear systems of equations that have no solution.

Because the biosphere cannot be "predicted" deterministically, Kauffman points out, there is endless room for creativity. This is the view of the emergent universe.

Another recurring theme is that of critical networks, poised between order and chaos, that would also maximize the chances of emergence and creativity.

So the source of the endless creativity of the universe, of the biosphere and of the human mind would lie in the Physics itself, but not be reducible to it.

From this weak premise, fairly equivalent to the dogmas of all religions, Kauffman proclaims that science is not the only path to knowledge (just like all religions have always done). His main objection to the materialistic scenario is that he believes that "values" are real features of the universe as much as particles. The materialistic scenario pictures a world with no values.

His theory of consciousness is equally based on "plausibility" and not real scientific demonstrability (but then it's implicit in his work that it cannot be reduced to ordinary demonstrability). Kauffman argues that the ideal state for the emergence of consciousness is a critical state poised at the border between quantum coherence and decoherence. The coherence of the quantum world derives from the deterministic Schroedinger equation that states how quantum "possibilities" evolve in time. The decoherence arises whenever we perform a particular experiment to measure a value: then the "possibility" becomes a number. The possibility becomes an actual event. Kauffman thinks that consciousness is an intermediate state that still propagates quantum coherence while decohering to the reality of objects that we perceive.

The weakest part of the book is, inevitably, the last one, that deals with ethics and with the deity itself. The scientific foundations here are shaky at best and the conclusions sound to me either implausible or simplistic.

The two main schools of Western ethics are the deontological one (notably Kant's categorical imperative, that good actions are those that one would want as universal laws ) and the consequential one (notably utilitarianism, according to which good actions are those that maximize the collective good). Kauffman finds both inadequate although he doesn't really prove why. He mentions that evolution itself has set a bar for morality, having introduced group selection (altruism) but then suddenly jumps into Pareto optimality (a concept introduced by an Italian economist) to deal with the fact that there is no absolute moral law (a` la Aristotle) and that the ethics that makes sense is an ethics of multiple moralities.

The god that Kauffman invokes is basically Spinoza's pantheistic god, the god that equates with the unfolding of Nature. Kauffman's main divergence from Spinoza is that Kauffman does not believe that Physics can explain the unfolding of the universe. Kauffman argues that such a god heals the split between reason and faith. In my opinion, it simply removes faith from the picture, as Science has been doing ever since.