Illuminating the Three Dogmas of Reinforcement Learning under Evolutionary Light