To the Editor.—Ìý It took medicine a long time to realize that alcoholism ran in families not because unhappy families caused alcoholism, but because the risk of alcoholism was genetically transmitted. This error of attribution could only be clarified through the careful experimental study of adopted and step-parented children.More recently a similar error of attribution is being made in the opposite direction. In the March 1989 issue of the Archives, Buydens-Branchey and co-workers¹ suggest, as have many before them,² that alcoholism can be divided into two types. They suggest that type 1 alcoholism occurs late in life, and is accompanied by guilt and binging but not by alcoholic relatives. Type 2 alcoholism, on the other hand, occurs early in life and is associated with positive genetic loading, impulsive aggressive behavior, and little sense of guilt.However, just as advocates of environmental transmission of alcoholism did not