The friend who picks the abuser

If you have ever tried to intervene with a friend who is dating a person who hurts her, you know that the textbook intervention does not work. The textbook says: name the behaviors, express concern, offer concrete help. You do all of that. She defends him. She stops calling you. The intervention literature has known this for forty years. The relevant research is on the leaving curve, not the staying. A woman in an abusive relationship leaves an average of seven times before she leaves for good — the figure comes from the National Domestic Violence Hotline's clinical data and shows up consistently in academic reviews. Each of the first six leavings is not a failed attempt. Each is part of the leaving. The friend who treats the third return as the moment to stop trying is misreading the curve. What you can do, that actually moves the curve: stay reachable. Don't lecture. Don't issue ultimatums about the relationship. Issue concrete invitations — 'come for dinner on Thursday' — that do not require her to denounce him to attend. Keep a duffel bag for her at your place. Tell her you have one. Do not bring it up again. The seventh leaving is the one that takes. Your job is to still be a person she can call on the seventh. If you cut her off on the third because you were tired, you will not be that person. The point of the intervention is not to win an argument. The point is to keep the phone in service. The seven-leavings number deserves more attention than it gets. The number is not a folk statistic. The number comes from clinical data collected across decades by the National Domestic Violence Hotline and corroborated by academic researchers including Anne Ganley and Lenore Walker. The specific number varies in different studies between five and nine, depending on how the researchers operationalize 'leaving' and which populations they study, but the general finding is robust: leaving an abusive relationship is, in most cases, a multi-step process that includes several incomplete leavings before the final one. The incomplete leavings are not failures. The incomplete leavings are the structure of how leaving actually happens. Why is this. Several reasons. The first is that abusive relationships have produced, over the course of the relationship, a specific psychological dependency that Lenore Walker named 'learned helplessness' in 1979. The dependency is not what the casual observer assumes. The dependency is not love, exactly, or weakness of character. The dependency is the result of a relationship in which the victim's nervous system has been conditioned to read the abuser as both the source of threat and the only available source of relief from the threat. The conditioning is not voluntary. The conditioning is the predictable response of mammalian nervous systems to specific patterns of intermittent punishment and intermittent kindness. The conditioning produces, in the victim, a profound psychological attachment to the abuser that is structurally similar to the attachment infants develop to primary caregivers. The attachment is hard to override from inside. The attachment operates beneath conscious decision. The victim, asked why they returned, often cannot produce a coherent answer. The incoherence is not a sign of low intelligence. The incoherence is a sign that the return was driven by a psychological mechanism that does not speak in language. The second reason is the operational difficulty of leaving. Leaving an abusive relationship is, in concrete terms, a logistically demanding operation. The victim, in most cases, needs to secure housing, income, transportation, child care if there are children, legal representation, and physical safety from a person who has demonstrated the willingness to use violence. The infrastructure for these needs is real but is not instantly available. Many of the early leavings are interrupted not by the victim's psychological attachment but by a specific logistical failure: the shelter is full, the family cannot host, the job does not pay enough, the legal process is too slow. The third reason is the specific danger of leaving. The most dangerous moment in an abusive relationship, documented across many studies of domestic-violence homicide, is the period immediately following an attempted leaving. The abuser, perceiving the loss of control, often escalates. The escalation can be lethal. The victim, who has lived inside the relationship for years, often has a more accurate read on the escalation risk than the friends and family who are urging the leaving. The return to the relationship, in this frame, is not irrationality. It is the calculated assessment that the moment is not safe. What does the friend actually do, given all of this. Not lecture. Not issue ultimatums. Not deliver the speeches that the friend has been rehearsing about how the victim deserves better. The lecturing, the ultimatums, and the speeches have all been tested in the intervention literature and found, consistently, to be either ineffective or counterproductive. The interventions that work are quieter, longer-term, and structured around the principle that the friend's job is not to make the victim leave. The friend's job is to remain available for the leaving when the leaving becomes possible. Available, in this context, has a specific operational meaning. Available means the phone is answered. Available means the couch is offered. Available means the duffel bag exists and the victim has been told it exists. Available means the invitation to dinner continues to be issued, with no requirement that the victim denounce the abuser to attend. Available means the victim's small disclosures are received without escalation — the friend does not, on hearing about a specific bad weekend, immediately demand that the victim leave that night. The friend listens. The friend thanks the victim for telling them. The friend lets the disclosure stay where it was placed. What is happening, in this kind of presence, is the slow building of an alternative psychological resource. The victim has been trained, by the relationship, to read the abuser as the only available source of connection. The friend is, by remaining reachable across years of difficult behavior, providing an alternative. The alternative is not, in the early years, competitive with the abuser's apparent centrality in the victim's life. The alternative is background. The alternative becomes foreground when the leaving begins. Some friends find this kind of presence intolerable. The intolerability is not a moral failure. Watching a friend you love be hurt by someone, repeatedly, across years, is psychologically expensive. Some friends, after a certain number of returns, withdraw. The withdrawal is understandable. The withdrawal is also the specific failure mode the literature identifies as preventable. The friends who manage to stay reachable for the duration are, in the data, the ones whose friends, eventually, actually leave. Stay reachable. Stay quiet about the relationship. Be the person on the other end of the call that comes at midnight in year seven, when the leaving finally takes. The work is long. The work is not glorious. The work is what saves the specific life that the interventions did not save.