User Review - Flag as inappropriatePatrick Maher has reversed, in an interesting way, the usual steps in the construction of verisimilitude measures. He starts from a Bayesian framework, where it can be proved in a representation theorem that a scientist will have a cognitive utility function.

He gives conditions (resembling TR1, TR2, and TR3) for the cognitive utility u(h, x) of accepting proposition h when x is the true state of the world. This function is 'subjective' in the sense that it depends on the scientist's cognitive interests and preferences.

The truthlikeness v(h, x) of h relative to state x is then defined by v(h,x) = (u(h,x) - u(t,x))/(u({x},x) - u(t,x)) where t is tautology. This function is normalized so that v({x}, x ) = l and v(t, x) = 0. On the basis of v(h, x), Maher defines measures of information c(h) and distance from being true d(h, x), which allow him to express the truthlikeness measure in a combination of the form v(h, x) = yc(h) - d(h, x), where y > 0 which is a generalization of Levi's epistemic utility in the direction of Niiniluoto's min-sum measure.

Maher's book is wonderfully clear and accessible to readers with little mathematical sophistication. His proof of his representation theorem is more accessible than those written by and for mathematicians. By developing a concept of rational cognitive decision making, Betting on Theories opens promising and exciting research programs for both decision making theory and the philosophy of science .In short it is a tour de force for research of this genre and further proof, if any were needed, of Maher's combination of word economy, and lucid realization of his subject matter.

WGP