The construction of emotions and what language has to do with it.

Experiencing emotions is complicated, but so is understanding them from a scientific perspective. The problem with emotions – like with many other phenomena we experience daily – is that we think we know what we are talking about. But when delving into the scholarship on emotion it becomes very clear that emotions are complicated, even defining them is, or deciding whether a given experience should or should not be classified as an emotion. Surprise, for example, is not considered an emotion by everyone but is instead sometimes viewed as an epistemic state (which can trigger an emotion). In fact, not everyone in the world of affective science assumes that there are primitives that are to be classified as "an emotion". This is because they have reasons to think that emotions are constructed (a theory typically associated with Lisa Feldman Barrett and her colleagues). This theory of constructed emotions has helped me to understand an astounding characteristic of human language: No language has a grammatical category which is dedicated to encoding emotional states. Of course there are words to describe emotions (I am sad.), and words that express emotions (Ouch!), and I can express my feelings by modulating my intonation (like yelling in anger). But what is conspicuously missing is a dedicated grammatical category that is used to encode emotions. That is, we frequently find categories like tense, person, or number, but not anger, fear, or happiness. But why should this be so? According to a common view the development of grammatical categories is rooted in language use and that they reflect common human interests. If this was all there is to the sources for grammatical categories, emotions would be expected to be among them, as they surely are of common human interest. So why are such categories absent? (Surprisingly, this question has never been asked before!). And my answer is simple: it is because emotions are constructed and the content of grammatical categories is by necessity a cognitive primitive. Thus, linguistic evidence presents us with a novel argument for the theory of constructed emotions.