Sociologists in the interactionist tradition have focused on societal reaction to deviance, but much of their research has emphasized "official" labeling by a social control agency (e.g., police, mental hospital, agency for the blind). The initial identification and informal labeling of deviance has received considerably less attention. However, one area in which there has been notable research is the identification of mental illness (Goffman 1961; Mechanic 1962; Sampson, Messinger, and Towne 1962; Yarrow et al. 1955). All of these studies concur that it is not necessarily a change in behavior that causes deviance to be identified, but rather that the change can be found in the significant audience. Marion R. Yarrow et al. (1955) note that wives "normalize" their husbands' deviant behavior until third parties enter the scene and redefine the situation. David Mechanic (1962) points out that the basic decision about mental illness is made by community members and not professionals and may be traced to the characteristics of the social situation where the deviant and his labelers interact, rather than characteristics of the deviant himself. Harold Sampson, Sheldon L. Messinger, and Robert D. Towne (1962) describe how families tend to identify and accommodate to deviance until there is a change in the situation of accommodation and the deviant comes to the attention of official labeling authorities. A summation of this research is that identification, definition, and labeling can be seen as distinct, though sometimes overlapping, phases of audience reaction.