Emotion as Rationality

Aaron Sloman has reduced the argument about emotions to simple mathematical terms. Let us look at an "agent" (whether a human, an animal, a robot or a piece of software), which is limited and intelligent and must act in a complex environment. A "complex" environment may well include a very large number of factors. In fact, it may be made of an "infinite" number of factors, if one starts counting every little detail that may have an influence. Our agent, which is limited, would never reach a conclusion on what to do if it blindly analyzed all factors. Therefore, in order to survive (to move at all, actually) it must be endowed with mechanisms that cause emotions. In other words, emotions are the result of constraints by the environment on the action of the intelligent being.

An emotional state is created by a situation, through a somewhat mysterious chemical reaction in the nerve system. A cognitive state is created by a number of situations and by a thinking process that relates those situations and draws some kind of conclusion. The relation between emotional states and cognitive states is reduced to the need to draw conclusions when cognition would face combinatorial explosion of possible reasoning threads.

Ronald DeSousa expresses this fact in a different way: emotions play the same role as perceptions, i.e. they contribute to create beliefs and desires. Beliefs and desires are necessary elements of any logical system: one attempts to satisfy desires by acting in the environment according to one's beliefs about the environment. DeSousa believes that emotions are learned like a language. And, like any language, they have their own grammar, i.e. their syntax and semantics (an idea also advanced by Sloman). Just like the meaning of words ultimately derives from the sentences in which they can be used, the semantics of emotions derives from the scenarios in terms of which they have been learned. Emotions can therefore be studied in a formal way just like any other language. The complementarity between reason and emotion becomes what he calls "axiological rationality", yet another way to express the fact that emotions determine what is salient, i.e. can restrict the combinatorial possibilities that reason has to face in the real world.