The identified patient

Posted on: 29 September 2026

Saturday night in London, dinner with people who spend their days between trading desks and investment funds, and at some point the conversation drifted onto how we deal with conflict inside our clients' teams. Across the table sat a young woman who works for a multinational entertainment company and who was living the problem that very month, since her team had reached the stage of talking to each other only by email with half the company in copy.

Explaining what my business bruvs and I do is always a small disaster, because the moment I say we combine algorithms, social design, group dynamics and incentive alignment in complex environments I can watch the lights go out even in the sharpest eyes at the table. So, helped along by a very decent pint of Camden Pale, I said the first thing that came to mind, which was that we do group therapy for companies.

On Sunday morning I thought about it again and realised it held up, and held up rather better than I had given it credit for at dinner.

In the 1950s in Palo Alto a group of therapists working with families noticed something awkward, namely that the child brought in as "the problem" was often just the place where the family's problem had become visible. They called that child the identified patient, because treating the child alone achieved nothing and as soon as he started to improve the symptom would move to a sibling or to one of the parents. Don Jackson, the psychiatrist in the group, first presented the idea in January 1954 at the Palo Alto veterans' hospital, which was in fact in Menlo Park. He published it three years later under the name family homeostasis. The family, in effect, defended its own balance even if the price was keeping somebody ill.

The idea travelled well, and nowhere more so than in Milan, where Mara Selvini Palazzoli made it the starting point of the work that became Paradox and Counterparadox, first published in Italian in 1975. In the same year, on the other side of the Atlantic, the article I come to further down appeared in print. Pure coincidence, but I like it.

The mid-sized companies we work with go through exactly the same thing with their processes, which even when they are well designed tend to run into someone who doesn't adopt them or quietly works around them. Management treats that person as the case to be fixed, first with training and then, when training fails, by replacing them, only to find a few months later that the new hire behaves exactly like the one before.

Replacing the person is the cheapest move available to whoever runs the company, because the cost is visible and contained and above all nobody has to reopen the rules that management wrote in the first place. It is Jackson's homeostasis applied to an org chart, since the system gets rid of the symptom precisely so that it doesn't have to change.

Almost always the person resisting is responding quite rationally to what the system actually asks of them, perhaps because the new process exposes mistakes that used to stay hidden or because it makes them work harder for a result that will be credited to another department. Steven Kerr put it in the Academy of Management Journal in 1975 under a title that is already a diagnosis, On the Folly of Rewarding A, While Hoping for B. One of his examples was team sport, where everyone preaches collective play while what counts is the points scored, because baskets are easy to see and measure and the pass that made them possible vanishes from the statistics.

At dinner I had talked about ego and I had put it badly, because ego does come into it but it arrives later, once defending your own patch is the only thing that gets rewarded, and at that point ego is simply the shape that defence takes.

Our job is to bring these frictions to the surface with the whole group in the room, because seen from inside everyone's reasons make sense and the trouble lies in the way they fit together. AI helps us get there sooner, as it lets us see much faster where people are getting stuck, but the rest of the work stays human and consists of redesigning the rules until collaborating pays each person better than guarding their own patch.

If I had to answer the woman from Saturday today, I would tell her to look with a little less irritation at the person on her team who seems to be ruining everything. There is a good chance that person is the one showing everybody where the problem is.