Two models of couple therapy
5th April 2026
There are two broad positions in relationship and psychosexual therapy, and they lead to very different clinical choices. The first, exemplified by the Tavistock Clinic, treats the couple as the client. The other, taught and practised by Relate, sees the client as two individuals who happen to be in a relationship.
The difference is sharpest around confidentiality. Relate sees both partners individually as well as together, respecting each partner's confidentiality as you would any client. The Tavistock holds that the therapist cannot hold secrets between partners — on the grounds that doing so risks replicating the couple's existing dynamic. It is a principled position. But I think it rests on a mistaken premise, and I find myself — having trained under the late Richard Newbury at Relate — firmly in the two-individuals camp.
1. The couple-as-client position echoes a mythology of sexual union — the idea that
sex is, at its best, an expression of love and togetherness, co-created between two people who know each other deeply.
But this is precisely what Esther Perel challenges in Mating in Captivity. The problem in long-term relationships is not that partners don't know each other — it's that they know each other too well. Merger and familiarity are the enemies of desire, not its flowering. Treating the couple as a single unit with a shared erotic life can inadvertently reinforce this — turning sex into a joint possession, something we do, rather than something that happens in the charged space between two irreducibly separate people with their own private erotic worlds.
Lacan puts this more starkly still: there is no sexual relation. Not that sex doesn't happen, but that the two desires never fully coincide. There is always a remainder, a gap, a non-fit. This is not a problem to be solved — it is the very structure of desire, and arguably the source of its energy. The two-individuals position honours this. The couple-as-client position risks smoothing over exactly what needs to remain a little rough.
2. We never know how a client's speech operates — as fantasy, desire, or fact — and our job is not to be an arbiter of truth.
The "no secrets" rule assumes that what a partner discloses is essentially information — reportable, factual, something the other partner has a right to know. But clinical speech doesn't work like that. When someone tells me their partner is emotionally unavailable, they are not filing a report. They are giving me access to their inner world — their projections, their history, their fantasy of what love should look like. Treating this as fact to be shared or withheld is a category error.
Confidentiality is not a courtesy. It is the condition that allows speech to be free — exploratory, contradictory, not yet fully known even to the speaker. Remove it and you get managed disclosure: each partner editing themselves for the other. Which is precisely what many couples are already doing at home