Paradoxical interventions: when to use them and how to deliver them ethically

Ask an insomniac to try harder to fall asleep and you have described the problem back to her. Ask her to stay deliberately awake as long as she can manage, and the pressure keeping her awake often drains away, because the effort sustaining it is no longer pointed in the same direction.

What makes a case a candidate

Paradoxical interventions fit a narrow band: problems maintained by voluntary effort applied to something involuntary. Trying hard not to blush. Trying hard to relax. Trying hard to stop a thought. In each, the trying is the machinery.

If the map shows a client fighting a symptom head-on and losing ground with every attempt, telling her to fight harder adds nothing, since that is already what she does. Asking her to produce the symptom deliberately, on a schedule, in a specified way, strips out the involuntary quality that was fuelling the fight.

It is a poor fit anywhere the clinical priority is stabilisation: acute risk, compromised capacity to consent, or a “symptom” that is actually self-harm. In those cases the answer is simply no.

A milder form: restraining and predicting relapse

Not every paradoxical move asks for the symptom itself. The gentler variant asks the client to slow down, or to expect the old pattern back for a week or two before change settles. Duncan (1992) describes a case where language of this kind validated a client’s ambivalence about leaving her husband instead of pushing her through it, and argues the effect came from her feeling accurately understood in her hesitation rather than from any cleverness in the technique.

The ethical version and the effective version converge on the same point more often than the literature admits.

Delivering it without deceiving anyone

The objection to paradox is not a formality. Duncan contended that MRI-style interventions can shade into manipulation and an implicit claim to therapist power, and he is fair about covert delivery, where the client is kept ignorant of the therapist’s reasoning.

The answer is not to abandon the technique but to be honest about the frame while withholding only the mechanics. “I’m going to ask you to do something that will sound backwards. The logic is that fighting this directly has been feeding it, so we’re going to change the fight” respects the client’s capacity to consent. Spelling out every predicted micro-effect in advance would blunt the intervention for no ethical gain.

O’Connor’s (1983) case of a ten-year-old boy with an obsessional fear of vomiting shows the restraint in practice. The boy was given a rationale he could understand, framed around taking charge of when and where the feared thing happened, rather than a covert manoeuvre disguised as something else. The full case is in the study of breaking an obsessional ritual.

Three questions before you prescribe paradox

Answer these out loud, even if only to yourself.

Has a direct request honestly been tried and failed, or was it skipped because paradox felt more elegant? Can this client tolerate a brief frame of “this will sound strange, and here’s roughly why” without it collapsing the intervention? Is there any safety concern that rules the technique out entirely?

If any answer is uncertain, a more direct approach to resistance is the safer next move, and a straightforward prescription is frequently enough on its own.

Write down what you told them

Record the rationale you gave the client, in the words you gave it, in the clinical note.

Two reasons. If the intervention goes wrong, or the client later reports feeling manipulated, the record shows what she was actually told and what she agreed to. And rereading your own framing a fortnight later is a reliable test of whether it was honest. A rationale that looked fine when you said it and reads badly on the page was probably not one you should have used.