Dopamine and Reward Prediction Error: An Axiomatic Approach to Neuroeconomics