facet of the campaign. NCI has begun an experimental evaluation of a community intervention trial that combines all of the just-discussed efforts to deter tobacco use. Preliminary evidence from the evaluation of a community-wide campaign to affect cardiovascular fitness indicates that such programs can have a beneficial impact (Perry, personal communication). A key feature of the latter campaign is its focus on getting community leaders to adopt the goals of the campaign as their own. Such formal placement of these issues on the agenda of the community probably enhances the effectiveness of each part of the campaign. For example, efforts to get health care providers to encourage nonaversive family practices may be aided if the social context for the providers includes a good deal of public discussion about the importance of positive family interactions and evidence that other individuals and agencies are working toward the same goal.
Aversive practices are a critical feature of most family problems. A good deal of progress has been made in understanding how to ameliorate aversive exchanges through modification of the context for individual family members. However, efforts to change the context for entire families have been limited primarily to interventions by therapists. Analysis of the larger context for family aversive practices suggests that the impact and scope of these efforts could be enhanced through programs in other facets of the context for families, such as government, schools, health care agencies, community service agencies, and the media. Only careful evaluation of the effects of such interventions will tell if they can, in fact, add to our ability to ameliorate these harmful interactions. The task is certainly formidable, since such interventions will be more costly than clinical work with individual families and substantial methodological problems will need to be overcome. However, interventions of this sort have the potential to affect more people than clinical interventions and will affect them in ways that might not be possible in therapeutic interventions. Given the scope and importance of the problem and what we already know, the time has come to take this next step.
Is it reasonable for society to fund such an undertaking? We have tried to show that it is. However, from a contextual point of view, research itself may be seen as a cultural activity occurring in a context. According to Harris ( 1979), activities such as scientific research will have an impact on other aspects of the culture only to the extent that they are congruent with the requirements for production and reproduction. If research on ways to reduce aversive practices in families is going to be supported by the society and is going to affect actual practices, it will be important to articulate the ways it could contribute to the productive processes. There are several ways in which such a contribution could occur. First, we may be able to reduce the costs associated with aversive family interactions. For example, Elder et al. ( 1986) have documented that males who