This article analyzes data derived from the first detailed long-term follow-up of a sample of serious adolescent female delinquents and similarly situated males. Neither marital attachment nor job stability, factors frequently associated with male desistance from crime, were strongly related to female or male desistance. A symbolicinteractionist perspective on desistance is developed as a counterpoint to Sampson and Laub's theory of informal social control, and life history narratives are used to illustrate the perspective. This cognitive theory is generally compatible with a control approach but (a) adds specificity regarding underlying change mechanisms, (b) explains some negative cases, and (c) fits well with life course challenges facing contemporary serious female (and more provisionally male) offenders.In a series of recent analyses, Robert Sampson and John Laub highlight the importance of marital attachment and job stability as key factors associated with desistance from crime (Laub, Nagin, and Sampson 1998;Laub and Sampson 1993;Sampson and Laub 1993). While the delinquents they studied were more likely than others to continue to offend as adults, there was considerable variability in the success of their adult transitions and in the timing of movement away from a criminal lifestyle. Sampson and Laub develop a social control explanation that emphasizes the gradual

A symbolic interactionist perspective on the emotions is presented that highlights their social character, forges links to cognitive processes, and suggests ways in which emotions influence long-term patterns of criminal involvement. This neo-Meadian perspective contrasts with theories of desistance that focus on the role of informal social controls and develops the view of an emotional self that flourishes somewhat independent of the major role transitions typically emphasized in sociological studies of the life course. The authors also explore ways in which attention to the emotional realms of experience adds to traditional treatments of the impact of adult transition events (e.g., the "good marriage effect"). Interviews with male and female adolescent offenders and two waves of adult followup data document general patterns of association and support the argument that a social view of emotional processes is critical to a comprehensive understanding of life course patterns of criminal continuity and change. 1 We would like to thank Ross Matsueda, Emma Engdahl, and the AJS reviewers for helpful feedback on several drafts of this article, and Josh Rossol, who contributed significantly to the analysis.

Institute of Mental Health, Center for Studies of Antisocial and Violent Behavior. We would like to thank two anonymous Criminology reviewers for their helpful comments. CRIMINOLOGY VOLUME 25 NUMBER 2 1987 295

No abstract

While there is considerable evidence that blacks experience school in qualitatively distinct ways from whites, there has been a general failure to examine racial variation in the impact of school variables on juvenile misconduct. The purpose of this research is to describe the manner in which school bonding affects delinquent conduct, focusing in particular on the role of the school in the delinquent involvement of black youths. Our orientation is primarily a control theory one that suggests that the greater the degree of school bonding the lesser the likelihood of involvement in delinquent activities. Our review of the literature leads us to expect differential levels of bonding by race and across varying racial environments of schools, with resulting differential effects on delinquency. On the basis of a neighborhood sample of 942 adolescents, we identijj seven distinct dimensions of school bonding. The analysis reveals that blacks are at least as strongly bonded to the school as whites, that our model explains comparable amounts of variance in delinquency across race-sex subgroups, and that the racial composition of the school is generally unimportant in conditioning the effect of school bonding on delinquency. While our findings are generally supportive of control theory, a model that purports to be invariant across race, gender, and socioeconomic boundaries, we caution that such a conclusion may be both premature and mistaken. We discuss the implications of these findings and suggest that they be interpreted within a framework that also considers family and peer bonding.School and school-related variables assume prominent roles in most major theories of delinquency: disorganized schools (cultural deviance theory), school environments that frustrate students' goals and aspirations (strain theory), a lack of attachment and commitment to the school (control theory), an educational system that provides an inferior education to poor and minority children (conflict theory), and the negative labeling of school failures and "troublemakers" (societal reaction theoryeall increase the likelihood of delinquency (Empey, 1982:289; Siegal and Senna, 1988:300) and/or school dropout (cf. Bachman and OMalley, 1978; Elliott, 1966; Elliott and Voss,