Attempted suicide in social networks.

It is generally agreed that a suicidal attempt commonly represents a bid by the patient to com- municate his distress to key figures in his intimate social group (Stengel and Cook, 1958; Farberow and Shneidman, 1961). An implication of this view is that such acts may function as a 'language' employed within such a group in situations where other groups would use less pathological forms of communication. The 'language' component of the act can be recognized only as a social and not as an individual phenomenon, and this leads to the hypothesis that suicidal attempts should be found to be concentrated in socially-linked clusters of individuals. More specifically, it could be predicted that, among the close acquaintances of patients attempting suicide, the number of suicidal attempts would be greater than would be expected from the age, sex, and demographic characteristics of these acquaintances. This study was designed to investi- gate this hypothesis.