Why couples therapy fails (and when it doesn't)

John Gottman's research, run out of his lab at the University of Washington since the 1970s, claims a 90% accuracy rate at predicting which couples will divorce — based on watching them argue for fifteen minutes. The predictor is not what they argue about. The predictor is the presence of four behaviors: contempt, criticism, defensiveness, and stonewalling. He calls them the Four Horsemen. Contempt is the worst. Contempt is when one partner has decided the other is, at the core, beneath them. Eye-rolling is contempt. Sarcasm is contempt. 'Of course you forgot' is contempt. Once contempt is in the room, the room is almost certainly lost. Couples therapy fails when at least one partner has already crossed the contempt line and is going through the motions to prove they tried. It works — Sue Johnson's Emotionally Focused Therapy has the strongest evidence base, with around 70-75% of couples showing recovery — when both partners are still in the building. Not happy. Not in love. Just still in the building. If you are wondering whether your relationship is in the salvageable range, honest self-assessment helps. There are diagnostic tools — stale.love is one that walks couples through the contempt-vs-frustration distinction — that can give a starting picture. The therapy is the work. The diagnostic is just the front door. The literature on what makes therapy work, for couples, has converged in the last twenty years on a small set of findings. The first is that not all therapies are equal. The American Psychological Association's review of evidence-based couple therapies has consistently identified three approaches as having strong empirical support: Gottman Method Couples Therapy, Emotionally Focused Therapy (Sue Johnson's work), and Integrative Behavioral Couple Therapy (Andrew Christensen's work). The other approaches, including the unstructured 'let me hear from both of you' work that many therapists practice, do not show the same outcomes in trials. If your couples therapist is not using one of these three models, the therapy is, from the evidence, less likely to work. The second finding is that the timing matters. Couples on average wait six years between the first emergence of serious problems and the seeking of therapy. The six years is, in most cases, six years of accumulated contempt, of habituated bad patterns, of stories the two partners have developed about the other that are increasingly hard to revise. The earlier the therapy, the more workable the substrate. Couples who arrive in year three have, in many cases, a different prognosis than couples who arrive in year nine. The third finding is that the modality matters less than the honesty in the room. Couples who show up to therapy committed to winning the room — to demonstrating to the therapist that they are the wronged party — produce, almost without exception, no useful therapy. The therapist becomes the judge in a court case rather than the facilitator of a difficult conversation. Couples who show up committed to seeing themselves honestly, including the unflattering parts, produce work that the modality can operate on. The modality is the scaffolding. The willingness is the building material. Why does therapy fail when contempt is established. The Gottman lab has been clear about the mechanism. Contempt, unlike anger or frustration, has already done the work of categorizing the partner as fundamentally deficient. The category is sticky. Once the category is in place, every new interaction is interpreted through it. The partner does something nice; the contempting partner reads it as manipulation. The partner apologizes; the contempting partner reads it as performance. The contempt has made the relationship interpretively closed. New evidence cannot enter. The therapy can introduce new evidence; the evidence will be filtered through the category and emerge with no effect. Sue Johnson's Emotionally Focused Therapy works, when it works, by operating one level underneath the contempt. The model assumes that contempt is, in most cases, a defensive structure built over an underlying attachment injury. The partner who has contempt is, in most cases, a partner who at some point stopped feeling safe in the relationship. The contempt is protection. EFT tries to produce, in the therapy room, the experience of being responded to safely by the partner — the experience that the underlying attachment wound has been waiting for. When the experience occurs, and is repeated enough times, the defensive structure of contempt becomes less necessary. The partner can soften. This works, statistically, for roughly seventy to seventy-five percent of couples who enter EFT and stay through the full course. It does not work for the remaining quarter. The remaining quarter, often, are couples in which one partner has already decided to leave and is, as Gottman would say, no longer in the building. The therapy cannot work on a partner who is not present. The presence is the precondition. How do you tell, before you spend twenty thousand dollars on therapy, whether your partner is in the building. Several signs. The first is the willingness to schedule the appointments without argument. A partner who has left the building, psychologically, generates scheduling friction for therapy. The appointments conflict with work. The appointments are rescheduled. The appointments are missed. The friction is not, in most cases, conscious sabotage. It is the body voting on its own behalf. The second is the willingness to engage between sessions with the small homework the evidence-based modalities assign. Gottman gives specific exercises. EFT gives specific exercises. Partners who are in the building do the exercises. Partners who have left, do not. The homework completion is a strong tell. The third is the willingness to be honest with the therapist about the things one is ashamed of. The partner in the building, given enough time, will tell the therapist about the affair they almost had, the contempt they catch themselves feeling, the fantasy of the alternate life. The partner not in the building withholds. The withholding is interpreted, by the therapist, as self-protective in the wrong direction. The partner is protecting their exit rather than protecting the marriage. If you are in therapy and you suspect your partner has left the building, the useful move is to name it in the room. Not as an accusation. As a diagnostic question. 'I want to ask, honestly, whether you are still trying. I will not be angry if the answer is no. I would rather know.' The asking, in a supervised room with a trained facilitator, can produce the answer that the marriage has been avoiding for years. When the answer is yes, the therapy can do its work. When the answer is no, the therapy turns into a different kind of work: the work of ending the marriage with the help of a skilled third party. This is not failed therapy. This is therapy doing what therapy can do for couples that have arrived too late, which is to help them part with as little further damage as possible. Sometimes that is the available outcome. Sometimes that is enough.