In _The Evolution of Reason_, William S. Cooper shows that, using a simple decision theoretical model of evolution and another of the environment you can find the seeds of classical analytical thought. He does this by "reduction," showing that "evolution" requires decision theory, which in turn requires probability theory, and so on down to basic propositional logic. Although this is precisely the opposite of what is usually meant by reduction in mathematics, Cooper is trying to give a "scientific" or empirical underpinning to the rules of rational thought rather than to begin with a priori assumptions about truth, such as the law of the excluded middle, or the most elegant -- that is to say, smallest -- batch of primitive assumptions. It should be clear that every step except the first is valid.
This is only half of the book, however. In later chapters, he examines complications to his model to see how they affect his derivation, and upon this basis makes some suggestions about how logic as a discipline should be practiced. Although a simplification, I think it is fair to say that he would like further reasoning about logic to be descriptive, to show how we should think in light of biology instead of some
The major annoyances from this book come are in its tone: the author constantly compares himself with Copernicus, Newton, and Einstein and calls his thesis "scientific," even though it is based in no way upon empirical data, or even "evolutionary theory," as he claims. Like many theorists with radical conjectures, he chides (unnamed) people for trying to think of situations in the world that could actually test the hypothesis, and the empirical arguments he does refute are obviously straw men.
The least satisfactory part of the book is the identification of an ill-defined concept of fitness (which the author himself says he doubts) with subjective utilities. In particular, he postulates intelligence to be evolution by other means by imposing on human judgment that the limit of subjective probability assessment be objective probabilities and the limit of utility (happiness) assessment be evolutionary fitness (more or less, fertility rates).
Were this book less expensive, I would think about using it in a freshman or sophomore-level logic class in conjunction with Flew's _Thinking About Thinking_. Its earlier sections are a good introduction to the connections between differently levels of mathematics, its various discussions of plausible inference and subjective probability are accessible, and its later interpretations show logic to be alive in a way that could encourage students.
Usually, I wouldn't recommend it to the general reader, however I think that because of currently fashionable academic trends, it is likely that Cooper's theory or one very much like it will become very popular in the coming decades, and it might behoove an interested party to be familiar with it beforehand.