Abstract

In order to avoid reluctance on the part of subjects to discuss mental illness, this paper argues that a study of traits rather than symptoms or syndromes is not only less charged, but also more satisfactory for the scientific study of the relevant semiology. The relation of traits to the classification and etiology of bipolar illness is discussed in four vignettes, each from a distinct vantage. Two study methodologies are proposed that would confirm and expand upon that which anecdotal evidence already suggests, namely, that the role of stress remains grossly under-appreciated and that the bipolar "personality" is pretty much the DSM equivalent of the common cold, a possibility requiring intensified research.