The definition of emotion has commanded only limited consensus in its philosophical and psychological history. In part this has been due to differential stress on the components of emotion. Starting with pre‐Socratic and later Greek notions of its somatic and bodily aspects, Aristotle then added cognitive elements to emotional experience. These three characteristics in various combinations defined emotion then and later. With Descartes, a serious attempt started to specify basic emotions and their contribution to other emotional concepts, though again consensus was not achieved. With the nineteenth century, classification became a dominant theme, punctuated by the constructivist attempts of the James‐Lange approach and its derivatives. The dominance of William James lasted into the twentieth century. His hegemony was accompanied and followed by new contributions to emphases and distinctions between central vs. peripheral processes. By mid‐twentieth century, cognitive factors once again became a dominant emphasis, which dissipated into a spreading net of competing theories, some again centered on psychic/mental, others on somatic/visceral bases of emotional experience and behavior, with conflict theories holding the middle ground. With a variety of contesting approaches available at the beginning of the twenty‐first century, William James' question, “What is an emotion?” still has no unchallenged answer.