Conventional rules encourage an investigator to focus primarily on quantifiable data. In psychophysiological studies, however, the most relevant data may be nonquantifiable, at least at the present time. Behavior, visceral or general, does not necessarily depend on the quantity of a particular stress, but rather is the result of a complex interactive central processing of afferent signals, reflecting more the relevance of actuating factors than their quantitative features. Thus, the rules of evidence must ask not only for measurement, when possible, but for context as well and for data on related intangibles that determine the significance of an experience to an affected individual. Such an inquiry requires the perspicacity of a skilled and disciplined observer. As the practiced ear of a well-trained cardiologist can draw reliable anatomical inferences from listening to the heartbeat, so the eyes and ears of an educated physician should be able to determine the significance of events to his patient with a fair degree of reliability. Both the physician and the cardiologist, however, must deal with the potential distortion brought on by their own biases. Despite problems of bias, lack of replicability, and the need to persist in efforts toward precise measurement, progress continues in understanding the contribution of psychosocial stresses to a variety of gastrointestinal, cardiovascular, respiratory, cutaneous, and other bodily disturbances and the pathways and neurohumoral mechanisms whereby they are mediated. The agenda now calls for developing new strategies for dealing with the powerful intangibles that activate the mechanisms. Meanwhile, as Robert Morrison once cautioned, "We must not expect a science to know more than it does."