Various scholars have pointed out that propositional attitudes (PA) are not good candidates for studying emotions. An example is the fact that if emotions were PA it would entail propositional contents in creatures that don't have language, namely small children or animals, as it has been shown. Despite well-known objections, the idea that emotions are PA remains because it has some theoretical advantages. Notably, it allows the mapping of emotions with other kind of cognitive phenomena like judgments or desires, both entities involving propositional attitudes. My proposition is that emotional content can sometimes emerge in response to propositional content, but is not of propositional nature in itself. I argue for a conception of emotions as cognitive products in the sense developed by Twardowski (1912). New perspectives are provided by a clear distinction between an act and the content it produces. Moreover, this distinction echoes the way neurosciences consider emotions and their productions.

The notion of cognitive product allows for a better understanding of what emotions are,. What is more, they permit a better encompassing of the complexity of emotional occurrences and their properties. Indeed, the intentional and temporal parts of emotions are not quite resolved in fields like Philosophy or Linguistics. Emotions and their manifestations are impacted by socio-cultural factors and it add complexity ot their study, but there are universals. From them I build a comparative reflexion between the cognitivist hypothesis and the notion of cognitive product.