Paradoxes have always been important and controversial issues in both philosophy and logic. Paradoxes of acting rationally establish a sub-class of paradoxes: Problems in which we usually encounter a dilemma of making one of two inconsistent decisions, while there are apparently rational and persuasive arguments for both. Thus we will be in a dilemma or a paradoxical situation. Newcomb’s Paradox is the most celebrated example of this kind of Paradoxes. The aim of studying paradoxes and attempting to solve them is to defend Rationality.
In Newcomb’s Paradox, there is a predictor who predicts the decision of the agent and also determines a part of the situation of the problem, which depends on his prediction. In this paradox we also encounter an apparent (alleged) inconsistency between two principles of choice, namely: Dominance Principle and Expected Utility Principle; because it seems that though the agent’s decision and the predictor’s prediction are causally independent, there is an epistemic dependence between them. In order to solve this paradox, we need to consider and study the possible mechanisms of predicting the future and then solve the problem in each case. We will show that the Backwards Causation interpretation is logically impossible and the Common Cause interpretation is the only possible one. Then we will consider Newcomb’s Paradox in two cases: 1- a common cause problem with an infallible predictor; 2- a common cause problem with a fallible predictor. We will argue that in the first case, no rational decision is available and in the second case there is no conflict between two principles of choice and choosing two boxes is the rational decision.