The disorders of affect have not contributed much to the diagnostic definition of mental disease, and their phenomenological description has never achieved the richness of the psychopathology of perception or cognition. This paper shows how the subordinate role played by affectivity in the Western concept of man led to the early and enduring view of mental illness as an exclusive disturbance of intellect. Attempts by nineteenth-century alienists to challenge this notion were only partially successful, due to the conceptual unmanageability of most forms of affective behaviour and the terminological redundancy that this engendered. These efforts were frustrated by the rebirth of Associationism, the rise of brain localization experiments, the peripheralist definition of the emotions and, finally, by the unfolding of Darwinism. As a result, no autonomous psychopathology of affectivity was ever developed. The eventual recognition of the so-called 'primary' disorders of mood has not led, however, to a refinement in the semiology of the experiences themselves. This has been impeded by the use of descriptive behavioural surrogates or by metapsychological accounts of affect as a form of energy or as a driving force. None of these developments has contributed to the clinical description of the mood disorders.