This dissertation offers a theory of emotion called the primitivist theory. Emotions are defined as bodily caused affective states. It derives key principles from William James’s feeling theory of emotion, which states that emotions are felt experiences of bodily changes triggered by sensory stimuli (James, 1884; James, 1890). However, James’s theory is commonly misinterpreted, leading to its dismissal by contemporary philosophers and psychologists. Chapter 1 therefore analyzes James’s theory and compares it against contemporary treatments. It demonstrates that a rehabilitated Jamesian theory promises to benefit contemporary emotion research. Chapter 2 investigates James’s legacy, as numerous alterations of his theory have influenced the field of emotion research over the past fifty years, including so-called neo-Jamesian theories. This chapter argues that all these variations of the Jamesian theory assume that emotions require mental causes, whether in the form of evaluative judgments or perceptual contents. But this demand is not present in James’s theory. Nor, as Chapter 3 demonstrates, is this assumption necessary or even preferable for a fecund theory that explains human and non-human emotions. Thus, Chapter 3 offers the details of the primitivist theory of emotion: emotions are affective states that contribute to perceptual states by affectively representing relationships between the sensed environment and the sensing organism. Rather than relying on prior perceptual contents as triggers, emotions operate concurrently with, and as influencers of, exteroception. The information they carry can be conceptualized according to the theory of affordances proposed by ecological psychologist James J. Gibson (1979): emotions inform emoters of their potential responses to ecological concerns. Chapter 4 then explains how the primitivist theory is compatible with uniquely human emotion episodes, namely instances in which we identify and conceptualize our emotional experiences according to introspective and contextual cues. It argues that one resource available for categorizing our emotional episodes is a felt bodily map: different emotion concepts correspond with patterns of increased or decreased activity across the body. Finally, Chapter 5 situates the primitivist theory within the debate about natural kinds in psychology. While emotion constitutes a natural kind, discrete emotions do not.