Article Text


Double binds and strange loops
  1. John Launer
  1. Correspondence to Dr John Launer, London Deanery, Stewart House, London WC1B 5DN, UK; jlauner{at}

Statistics from

One of the most famous moments in modern literature occurs in Joseph Heller's novel about the Second World War, Catch 22. The hero of the book, Yossarian, realises that his commanding officer in the air force is so incompetent that every single fighter pilot in their squadron is getting killed in action. He goes to the commanding officer and says that his fellow pilots must be completely insane to follow orders to go on flying missions. The response of the officer is to say: “I completely agree with you: if you've realised this you are absolutely sane. And we need sane men like you to go on flying missions, not the insane ones who are going out at the moment!”

The unsettling quality of this conversation is strangely familiar. It shows an extreme example of something that probably goes on a great deal. Many people would call such an exchange a “double bind”. Communications theorists have paid quite a lot of attention to these kinds of exchanges and have generally given them a different name: “strange loops”.1 They have also suggested ways that we can deal with them in real life.

Basically, their explanation goes as follows. They propose that everything we say automatically carries a set of contexts with it, whether we realise it or not. Thus any statement is (among other things) a response to somebody else, a contribution to a longer conversation and part of a relationship. It also represents a set of social and cultural rules about conversations. These contexts all nest inside each other like Russian dolls. Sometimes both parties have an unspoken understanding of what the different contexts are and how they fit inside each other, but sometimes the parties are at odds without being aware of it. Their conversations then get into terrible muddles, although neither party quite understands why this is.

Unexpressed conflict

According to this argument, strange loops occur when there is a conflict over contexts and their relative importance. In Catch 22, for example, Yossarian is basically saying: “For me, the decision over whether or not to fly in this squadron is a defining context for anyone's sanity”. His commanding officer then reverses the contexts by saying: “For me, sanity is a defining context for the decision that someone should definitely fly”. The only effective way of dealing with this would be to stand outside the whole interaction and comment on it from what is called “a higher context”. For example, Yossarian might say: “You don't understand me, Major. Because I'm sane, I'm absolutely refusing to fly”. By saying this, Yossarian would be assigning a higher context to the conversation (misunderstandings), and then an even higher context for their relationship (the right to say no to irrational orders). The name for such a correcting tactic is a “charmed loop”.

Strange loops seem to happen quite a lot in medical consultations. We may be aware of them only through somatic discomfort or a heightened sense of anxiety rather than through any logical analysis. Doctors probably correct them intuitively as well, restoring charmed loops without realising exactly what they are doing at an intellectual level. Here is a typical example. Supposing a young fit man comes to see me because of back pain. Implicitly, there is an assumption that I will understand the cause of his pain because: (a) I know something about the body; (b) I am a good doctor; (c) I am a caring human being. One could see these three assumptions as contexts that nest inside each other or are at increasingly high levels. Each higher context governs the lower ones, so that being a human being is the cause for becoming a doctor, which is a cause for understanding the body. Equally, each lower context confirms the higher ones, so that my knowledge of the body demonstrates my identity as a doctor which in turn demonstrates to my patient that I am a caring person.

Stress related

Supposing, then, that in the consultation I reach the conclusion that the patient's back pain may be related to stress, and accordingly I start asking him questions about his work, his home life and so on. Implicitly in my mind there is a new context, nestled inside the other ones. This new context is my belief that psychological factors can influence the body and someone's experience of it. As I start to ask these questions, the patient may well go along with the process quite happily at first, expecting that, in due course, I will return to the higher context of the body and offer an explanation of the pain that may incorporate some of the “lower context” information about stresses in his life and so forth.

But supposing I go on asking such questions to the point where my patient starts to feel anxious, uncomfortable or even angry. He may conclude that I have raised the status of my psychological inquiry to a level where his body has now become a lower consideration. While I carry on happily inquiring into his private life and sexual habits in the belief that this will eventually shed light on his low back pain, he has now decided (a) that I am completely uninterested in the body and obsessed with people's private lives because (b) I am an inattentive and incompetent doctor and (c) I am a nosy and uncaring individual. In theoretical terms, the misunderstandings about the lower levels of context have had serious implications for his perceptions of me at all the higher levels.

Restoring harmony

In a consultation like this, the likelihood is that I would actually spot his discomfort fairly early on and attempt to respond to it, probably at an intuitive level. I might, for example, apologise if my questions had seemed intrusive but explain that sometimes this kind of information can be relevant to understanding back pain. (The theoretical term for this is “setting a context marker”.) If I then confirm the higher importance of the body by carrying out a careful physical examination, the patient may suddenly recall occasions in the past when I have made an accurate diagnosis (therefore I am really a good doctor) or treated another member of his family well (therefore I am really a caring human being). Harmony between contexts has been restored, and the strange loop transformed into a charmed loop.

It is easy to assume that all strange loops are bad while all charmed loops are good. This isn't necessarily the case. Strange loops can be used constructively as well. The most common case of this is when someone says something such as, “I never really set aside time for leisure…I can only remember a few occasions in the last few months when I've really had fun”. An astute questioner might then ask: “So what made it possible for you to create the conditions for having fun?”. Instead of demonstrating how difficult it is to set aside time (because this is so rare), the exceptions are used to demonstrate the opposite. In effect, the few episodes of fun are redefined as the higher context rather than the lower one.

Funny feeling

The best reason for identifying and understanding strange loops is that they often hamper conversations when the doctor and patient are working from different assumptions about what is actually going on. A “funny feeling” that you aren't quite on the same wavelength as the patient is usually an indication that this is happening. You then need to ask a question to re-establish an agreement at a higher level of context. An example of this kind of question might be: “You started by talking about your depression, and now we're concentrating on the way your boss is mismanaging you: do you want to focus now on your symptoms or on what your boss is doing to you?” Another example might be “When we started this conversation you were clear there was a definite problem to look at: can I check with if you think that's still the case—and if so what the problem is”.

Manoeuvres like this are also helpful whenever you feel that you are going round in circles, or have lost your bearings. In some cases it may be worth checking out your basic assumption that your patient still wants to go on with the conversation. Sometimes they may be trying to answer questions in a way that shows you they have had enough, while you are blithely responding to each answer as a prompt for a further question about the “problem”. This is a classic—and common—strange loop.


View Abstract


  • Competing interests None.

  • Provenance and peer review Commissioned; not externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.