In Selby's novel, Marion is a painter, and her dream is to open a store that sells clothing based on her sketches. Marion is terrifically cultured, and her mind is filled with ideas about Italian museums, Renaissance music, and light: 'All that summer and fall she painted, mornings, afternoons, evenings, then walked around the streets that were still echoing the music of the masters, and made out of interior monologues, and a film is necessarily more visual, more exteriorized. Section II

Behavioral Couples Therapy:

Harry and Marion

It is clear from watching the movie that the extent and depravation that results from the profound acts of violence and addiction that plagues the beautiful young couple could have been saved by intervention therapies. However, there are a number of cautions and clarifications about BCT and partner violence in treating substance abusing patients. First, it is important to realize that BCT was not designed as a treatment method for partner violence. BCT is a couples-based treatment for alcoholism and drug abuse. We simply learned through our clinical experience treating couples and through our research that male patients seeking help for substance abuse problems are a high risk group for perpetrating partner violence, and that the violence should not be ignored.

Second, data currently available support the use of BCT for a specific subgroup of men with co-occurring problems of addiction and partner violence. BCT is recommended for married or cohabiting male substance abusing patients who have sought help for their substance abuse problem if there is not an acute high risk of severe, injurious, or lethal violence (as already described above). BCT is not recommended for substance abusing patients seeking help but not currently living with a partner. A dual-focused intervention program may be better suited to the substance-abusing, violent male who is not in a partner relationship as one means of prevention of future domestic violence (Palmer, et al. 2002). BCT has not been tested on batterer clinic male patients with substance abuse problems, and is therefore not currently recommended for this population.

Third, we do not know why partner violence is reduced after BCT. Several possible explanations exist (Gorney, 2007). First, violence may be reduced because alcohol and drug use are reduced or eliminated. Second, violence may be reduced because one or both members of the couple learn constructive communication skills that prevent arguments from escalating to violence. Finally, a combination of these factors may explain the violence reduction associated with BCT (Giles-Sims, 1983).

Some results seem to support the importance of reduced substance use after BCT in reducing violence risk. For example, in one study for both the first and second year after BCT, violence was significantly reduced; further, the extent of violence and of clinically elevated verbal aggression levels were associated with the extent of the alcoholics' drinking (Healey, et. al 2007). Frequency of post treatment drinking was positively correlated with frequency of violence and verbal aggression, and remitted alcoholics no longer had elevated violence and verbal aggression levels when compared with matched controls, while relapsed alcoholics did. These results were observed even after baseline violence levels were taken into account (Gondolf, 2003).

Several studies that show reduced violence and an association between substance use and continued violence after individual (not couple) treatment also seem to support the importance of reduced substance use in reduced partner violence after treatment. One study of drug abusing men with comorbid alcohol problems found that partner violence was significantly reduced from the year before to the year after receiving individually-based substance abuse treatment (Shultz, 2004). This study showed the same pattern of results found with BCT. The greatest violence reductions occurred among patients who were remitted after treatment; and those remitted after treatment experienced similar levels of violence as did a nonalcoholic normative control group. Downs, (2006) also found that, in the year after individually-based treatment, the likelihood of male-to-female violence was 18 times higher on days when the man used alcohol or illicit drugs than on days when he did not. These findings and greater violence among relapsed than remitted patients remained significant and of similar magnitude when baseline violence levels were controlled. Finally, two other longitudinal studies of partner violence after individual alcoholism treatment reported high levels of violence before treatment that were significantly reduced in the year after treatment (Shultz, 2004).

Other results suggest that couple relationship factors may be important in reduced violence after substance...

For example, a randomized study of male drug abusing patients found that BCT was more effective than individual treatment in alleviating partner violence (O'Farrell, et al. 2000). A second study, which investigated differences between partner violent and nonviolent male alcoholic patients, found that relationship distress and alcohol problem severity had independent associations with partner violence (Downs, 2006). Further, relationship adjustment remained considerably associated with partner violence, while alcohol trouble severity did not, after controlling for demographic variables and patient antisocial traits.
Researchers have noted that BCT is contraindicated if there is an acute high risk of severe violence that is potentially injurious or lethal (Giles-Sims, 1983). However, once cases with acute risk of serious injury or death have been eliminated, it is not completely clear where to draw the line on the violence severity continuum when considering the use of BCT. For example, in two studies using the Conflict Tactics Scale (Healey, et. al 2007) definition of severe violence (i.e., kicked, bit, or hit with fist; hit with something; beat up; threatened with knife or gun; used knife or gun), 20%-30% of male alcoholic patients entering and accepted for treatment into BCT have engaged in severe violence toward their female partner in the year before BCT (Downs, 2006). Prevalence of severe violence is significantly reduced to 8%-12% in the two years after BCT in these studies, suggesting that some cases of severe violence can be helped by BCT.

Another issue in any form of couples' therapy is whether being in therapy has an impact on participants' perceptions regarding the integrity of the relationship and their decision-making regarding its possible dissolution, as well as how responsibility is construed. In BCT, participants do not need to have an open-ended commitment to maintaining the relationship. In fact, many couples enter BCT as a last chance to salvage their relationship; often the non-addicted spouse has made it clear that if violence or serious substance use recurs, then the relationship is over (Hardy, et al. 1998). However, in our work on BCT, participants both need to be willing to work to see if the relationship can be improved and to agree to refrain from threatening separation or divorce in anger (Gondolf, 2003). Couples promise not to continue to make threats of separation or divorce in the heat of anger at home because such threats usually sabotage the couple's progress and can lead to heightened anger than can escalate to violence or substance use. However, they also agree to discuss serious thoughts they may have about possible separation or divorce during BCT sessions where they can get help from the therapist in dealing with this issue. In this regard, BCT therapists are careful to stress that the spouse's role of assisting the male substance abusing patient's recovery does not mean that the female partner is responsible for the male's substance use or violence.

Evidence, Rather Than Ideology, Should Guide Interventions

The problem of domestic violence is too weighty, both as public health and social issues, to be content with intervention approaches that sound good or look right based on some conceptual model, but that have no empirical support for their efficacy (Gorney, 2007). Unfortunately, this has been the tradition regarding interventions in this field-many earnest appeals and opinions, very little data (Ganley, 2008).

It is important in conceptualizing interventions for partner violence to distinguish between interventions that are inefficacious and interventions that are harmful. The limited available evidence suggests that a lot of partner violence interventions, both lawful and psychosocial, may be relatively inefficacious when scrutinized and compared to no-treatment managed conditions (Hassan, et al. 2000), although combined legal and clinical interventions may have small, additive effects on outcome variables such as criminal recidivism (Healey, et. al 2007; Giles-Sims, 1983). To date, no intervention has been shown to have large, powerful effects in ending partner violence in controlled studies. Thus, the search for highly efficacious intervention approaches remains a major and urgent priority in the field.

With respect to the harmfulness of interventions, although a great many articles have warned of the harmfulness of couple's interventions for partner violence, and many states explicitly forbid the use of couple's interventions for court-mandated abusers, we know of no empirical evidence indicating that such interventions are, in fact, harmful. Two controlled trials have found couples interventions to be no less effective than the widely promoted gender-specific interventions for wives with a violent husband (Hassan, et al. 2000). O'Leary and…