Using Somatic Experiencing to enhance Ulysses Contracts

What each one is actually for

Somatic Experiencing

A precommitment device works because you accept, in advance, that there will be a version of you who cannot be reasoned with. Somatic Experiencing is a way of working on the condition that version of you is in, rather than on the options in front of them. Its central claim, and it is a claim rather than an established fact, is that a threat response which never got to complete leaves the body holding a charge that keeps the nervous system tuned for danger long after the danger is over. An animal that escapes a predator will often shake violently for a minute or two afterwards and then go back to grazing; Levine's proposal is that human beings frequently interrupt that discharge, and that the interruption is a large part of what persists. So the work is not to talk through the event. It is to track physical sensation directly, at a pace slow enough that the system stays inside the range where it can still process rather than tipping into flood or shutdown. In practice a session looks like this. You notice one specific physical detail, not the whole feeling, just the tightness in the front of the throat. You stay with it for a short while and watch what it does, because sensation left alone rarely stays still: it moves, spreads, sharpens, or releases into a shiver or a long breath. Then, before it builds too far, you deliberately go somewhere else, the weight of your feet on the floor, the light in the room, and let the system come back down. Then back. The alternation is the method, and Levine's names for its two halves are titration for the small dose and pendulation for the going back and forth. What it is trying to produce is not insight and not catharsis; it is a nervous system that has had repeated experience of going up a little and coming back down, and a system with that experience is a different negotiator at the moment the contract is being tested.

Ulysses Contracts

A Ulysses contract is an arrangement you make while thinking clearly that removes an option from the version of you who will be thinking badly. The distinguishing feature is that it does not rely on you being persuaded later; it relies on the option not being there. The everyday forms are unglamorous. You hand your car keys to a friend before the party. You put the savings in an account with a withdrawal penalty. You post money with a service that gives it to a cause you dislike if you fail to hit a target, which is the model behind deposit contracts like the ones stickK was built to hold. The formal clinical version carries the name explicitly: a psychiatric advance directive, sometimes called a Ulysses contract in the bioethics literature, is a document a person writes while well, specifying what treatment they consent to during a future episode in which they may refuse it. What the whole family of devices is aimed at is a specific and well-documented feature of how people value things across time. Under most conditions a reward available now is weighted far more heavily than a larger one available later, and the ranking can reverse as the moment approaches, so a choice that looked obvious on Sunday looks quite different on Thursday evening. Precommitment does not try to fix that bias. It accepts it as a given and simply arranges for the Sunday version of you to have already acted.

What you get out of pairing them

The reason to bring these together is a specific, unglamorous problem that anybody who has done body-based work alone will recognise. You sit down to do twenty minutes of tracking. Something opens. It is interesting, or it is finally moving after weeks of nothing, or it hurts in a way that feels like it is going somewhere, and you keep going. Ninety minutes later you are on the floor, wrung out, and you do not go back to the practice for three weeks. Or the mirror image: five minutes in, something starts to build, you get a flicker of what is underneath it, and you find a reason to stop that is entirely plausible, you should eat, you should check that thing, and you never quite return. In both cases the dose got decided by the state you were in, which is exactly the state the practice exists to work on. That is the whole argument for putting a binding rule around it. A stopping rule set on Sunday afternoon, when nothing is activated, is made by the version of you who can actually think about dose. The version of you at minute forty cannot, and more to the point does not know that it cannot, because activation does not announce itself as impaired judgement. It announces itself as clarity. What you get from the pairing is not more control over the practice; it is less need for control at the moment control is least available. And there is a second, quieter payoff. When the end is fixed in advance, you stop spending attention on the question of whether to stop, which is a surprising amount of the attention available in a session, and that attention goes back into the tracking.

What Somatic Experiencing rests on

• Interoceptive attention randomised trials
  Directing sustained attention to internal bodily sensations increases awareness of the body's state and changes how that state is experienced.
• Graded exposure reduces avoidance randomised trials
  Repeatedly approaching a feared or avoided stimulus or action within a controlled, tolerable structure reduces the fear and avoidance response it triggers.
• Somatic discharge completes stress cycle clinical practice
  Physically completing a build-up-then-release cycle in the body discharges stored tension rather than leaving it held.

What Ulysses Contracts rests on

• Implementation intention closes intention action gap randomised trials
  Pre-committing to a specific situational cue paired with a specific response lets the behavior fire automatically when the cue occurs, without depending on in-the-moment willpower.
• Removing future optionality protects present intent randomised trials
  Binding a future choice through an external cost or barrier prevents a later moment of weaker willpower from overriding a decision made while thinking clearly.
• Temporal discounting erodes willpower randomised trials
  People systematically undervalue future consequences relative to immediate temptation which is exactly the bias a binding precommitment is designed to neutralize.

Where they actually meet

Look at what each practice is actually protecting against and the same shape appears. Precommitment exists because a person in a state of craving is not the same decision-maker as the person who set the goal, and the craving version reliably wins any argument conducted during craving. Titration exists because a person in a state of high activation is not the same decision-maker as the person who sat down, and the activated version reliably wins any argument conducted during activation. Both traditions have arrived, from opposite directions, at the conclusion that a decision cannot be trusted to survive a change of state, and that the fix is to move the decision out of the state where it is unreliable. Where they differ is in what they do next, and the difference is what makes them useful to each other. Precommitment moves the decision out into the world, into a document, a deposit, a person holding your keys, so that it is beyond your reach entirely. Body-based work moves in the opposite direction: it tries to make the state itself more legible, so that the version of you who is activated can notice what is happening earlier and with more resolution. One reduces the need for judgement; the other improves it. A person doing only the first ends up with a set of external rules and no felt sense of what they are for, which is why so many deposit contracts get grimly honoured and then abandoned. A person doing only the second has a beautifully detailed read on their own state and no protection from the specific moment when the read stops being accurate.

The principle underneath both

Say it plainly: the you who decides and the you who acts are frequently not in the same condition, and any plan that requires them to agree is a plan that fails at exactly the moment it matters. Both practices are built on that observation and neither of them owns it. It shows up in the surgical checklist, which exists because a competent surgeon under time pressure is a different operator from the same surgeon planning the case. It shows up in the fire drill, which exists so that nobody has to decide anything while a building is filling with smoke. It shows up in the pre-flight card. What all of these have in common with a Ulysses contract and with a session boundary is that a specific response has been attached to a specific trigger in advance, so that when the trigger arrives, nothing has to be worked out. In the psychology literature this attachment is called an implementation intention, meaning a plan of the form when this specific situation happens, I will do this specific thing, and it is one of the better-supported findings in the area: a large meta-analysis pooling nearly a hundred studies found a medium-to-large improvement in goal attainment from making plans in that if-then form rather than holding a general intention. That evidence is about starting behaviours rather than stopping practices, which is a real gap and not one to paper over. But the mechanism it identifies, that specifying the cue and the response in advance removes the need to generate either under pressure, is the mechanism this whole pairing runs on.

What that makes the real goal

If the working part is getting the decision out of reach of the state that will distort it, then the goal states cleanly and belongs to neither lineage: arrange things so that the person who will be flooded, craving, exhausted or activated does not have to make the call, and cannot overturn it if they try. That goal is separable from Levine's method and from any particular contract. It does not require a therapist, a timer, a deposit or a legal document. It requires three things, and they are worth naming because most attempts at this get one of them wrong. First, the rule has to be specific enough that there is no interpretive room, so twenty-five minutes on a timer qualifies and I will stop when it feels like enough does not. Second, the rule has to be set while you are demonstrably in the other state, which in practice means not on the same day and not in the hour after a session. Third, and this is the one people skip, the rule has to bind in a way the later you cannot dissolve, which usually means a person or a device rather than an intention, because an intention is precisely the thing the later you will renegotiate. Once the goal is held that way, the question stops being how much titration is right and becomes what is the smallest arrangement that would take this decision away from me on a Thursday.

Other practices that serve the same goal

Other practices aim at the same goal, and several of them carry harder evidence than either half of this pairing. Deposit contracts are the most rigorously tested: a randomised trial in the Philippines by Gine, Karlan and Zinman offered smokers a savings account whose balance was forfeited to charity if they failed a nicotine test six months later, and the offer produced a meaningful increase in the proportion who quit, verified biochemically rather than by self-report. That is real evidence for the binding mechanism itself, in a domain where the target behaviour is unambiguous. Implementation intentions, the if-then plans described above, are the lightest-weight route and the best supported across many domains, and unlike a deposit they cost nothing and can be revised. Dialectical behaviour therapy, an approach originally developed for chronic suicidality by Marsha Linehan, contains a version built for exactly this population: the crisis survival plan is written when the person is well, names the specific early signs, and specifies the actions and the people to contact, precisely so that nothing has to be worked out during a crisis. And the psychiatric advance directive is the formal legal instrument in the same family, available in many jurisdictions, with a body of research on uptake and on the fact that people who complete one report a greater sense of control over their own care. What is worth noticing across all four is that the ones with the strongest evidence are the ones where the target behaviour is countable. Nobody has built a deposit contract around a body-tracking session, and it is not obvious what the verifiable outcome would be.

A tradition that already does this

The tradition already did this, and it did it structurally rather than by writing it down. The fifty-minute hour is a precommitment device. Its function, in every depth-oriented therapy that keeps it, is that neither the client nor the therapist gets to decide mid-session whether to continue, because the decision was made by the container before the session started, and because a person in the middle of difficult material is not a good judge of whether to go further into it. Judith Herman's phase model of trauma treatment formalises the same instinct at a longer timescale: safety and stabilisation come first, and the work with the traumatic material itself is deliberately deferred until the person has demonstrated the capacity to come back down. That ordering is not caution for its own sake. It exists because the clinical tradition learned, expensively, that a person can be persuaded in the moment to go further than they can afford, that the persuasion often comes from inside them rather than from the therapist, and that the cost shows up afterwards rather than during. Somatic Experiencing inherits this: it is taught as a therapist-held practice with a titration discipline built into the clinician's role, and a person doing it alone has quietly removed the part of the system that was holding the boundary. Setting an external stopping rule is, on this reading, not an addition to the method. It is a replacement for something the method assumed was there.

Using body-tracking to design a contract that fires early enough to matter

Precommitment devices fail in a specific and predictable way, and improving your read on your own physical state is a genuine fix for it. The failure is timing. Most contracts are keyed to a clock, a place, or an event: no drinking after eight, keys with the host on arrival, phone in a drawer at nine. But the state that will overturn the contract does not start at eight. It starts considerably earlier, usually as something bodily and unremarkable that never gets named, and by the time the clock trigger arrives you have already spent an hour building a case. Sustained attention to internal sensation is the skill that catches that earlier. Here is the concrete move. For two weeks, do nothing but notice, in the fifteen to thirty minutes preceding the choices you keep losing, what the body is actually doing. Not the feeling and not the story about it. The physical facts: shallow high breathing, a hollowed feeling under the sternum, restless legs, jaw set, a specific kind of scanning where your eyes will not settle on anything. Write them down immediately, because they are almost impossible to recall afterwards. What you are building is a personal list of the signs that reliably precede the state, and it will be idiosyncratic; the useful signs are rarely the obvious ones. Then rewrite the contract so the trigger is one of those signs rather than the clock. When my breathing goes high and shallow and I notice I am pacing, I text the person and I hand over the thing, whether or not I currently believe I need to. The whole point is that the rule fires before the case has been built, which is the only window in which handing something over does not feel like a defeat. There is a real limit here worth stating: some people, particularly after prolonged threat, have a genuinely unreliable read on their own internal signals, and for them this route will produce confident false alarms and confident misses in roughly equal measure. If two weeks of logging produces no stable pattern, the clock trigger is the better instrument and there is no shame in it.

Using a binding stop to make small doses actually stay small

Run it the other way and you get the fix for the dose problem. The concrete move is a written session contract, made at least a day in advance, and it has four parts. A hard duration, on a timer that is across the room rather than in your hand, because a timer in your hand is a timer you will extend. A named landing activity that starts the moment the timer ends and is physical and dull, washing up, walking to the end of the road, folding something, so that the transition is occupied rather than left open for you to drift back into the material. A stated floor as well as a ceiling, which is the part everyone omits: I will stay for the full fifteen minutes even if nothing is happening, because leaving early is the more common failure and it is invisible, since nobody notices the sessions they quietly cut short. And a rule about what to do with anything that opens near the end, which should always be the same rule: it gets written in one line and left for the next session, never followed. Two details make the difference between a rule that binds and one that does not. Tell someone. A stopping rule you can quietly extend is not a contract, it is a preference, and the whole reason the Odysseus story involves rope and other people is that a private resolution is exactly what the later state dissolves. And write the rule down in your own handwriting where you will see it during the session, because the version of you at minute twelve will produce genuinely persuasive arguments for continuing, and the only thing that reliably interrupts a persuasive argument is a sentence you already wrote saying you would not listen to it. What this changes about the practice is not obvious until you have done it for a few weeks: sessions that end on time stop being events, and once they stop being events you do them far more often, which is where whatever benefit is available in this method actually lives.

Minute fourteen, and it is finally moving

You are fourteen minutes into a twenty-minute session, sitting on the edge of the bed because the chair felt too formal. For three weeks there has been almost nothing, which is why you nearly did not sit down tonight. Then something in the left side of your chest, which had been a dull unremarkable tightness, changes. It gets sharper and more specific and it moves up towards your throat, and along with it comes a flash of a room you have not thought about in a long time, and your hands are cold. And here is the thing you should notice, because it is the whole reason for the rule: what you feel right now is not distress. It is relief. Three weeks of nothing and finally something is happening, and the obvious thing, the sensible thing, the thing any reasonable person would do, is to stay with it. The timer is across the room and it has six minutes left. You know from your own notes that the last two times you extended, you were up until two and useless the following day, and that in both cases you did not go back to the practice for over a fortnight. So at twenty minutes the timer goes and you stand up, and it feels wrong, and you write one line on the pad, left chest to throat, cold hands, the upstairs room, and you go and wash the dishes. The dishes take eleven minutes and by the end of them you are back in the room you are actually standing in. That is the move. You did not process anything, and the thing that opened is still there and will still be there on Thursday. What you protected was your capacity to come back on Thursday, which the version of you at minute fourteen was not weighing at all, because that version was not in a position to weigh anything.

When this is worth reaching for

You're doing body-based work outside a session with a practitioner, and taking the practitioner away without replacing the function they served as your stopping rule is the usual cause of trouble. A boom-and-bust practice — long intense sessions followed by weeks of avoidance — is a dose problem, not a motivation problem, and treating it as motivation makes it worse. If the sessions you remember are the dramatic ones, that's a sign intensity has quietly become the thing you're going for. The opposite case counts too: stopping at four minutes and telling yourself the practice doesn't suit you, since a floor binds just as usefully as a ceiling and is far less often used. And there's the pattern you may already know about — short with people, sleeping badly, unable to work the day after a heavy session — because that downstream cost is invisible from inside minute fourteen and is exactly what a rule written on Sunday can take into account.

Where this does not apply

There are situations where adding a binding constraint to this kind of work is the wrong move, and one of them is serious. If you have a dissociative response, meaning that under pressure you go somewhere else and lose contiguous access to what is happening, then a rule that makes stopping costly or that binds you to continue is dangerous, and the floor half of the contract should simply be dropped. The general form of that rule holds for everyone: never attach a stake, a penalty or a person's disappointment to stopping. Stopping must always be free. A contract that makes it expensive to stop has inverted the entire purpose of the thing. Second, if you are in an acute crisis, whether that is a fresh loss, an active safety concern, or a period where you are barely holding the ordinary day together, this is not the moment for structured trauma work with or without a stopping rule, and the honest advice is to get a person involved rather than a better protocol. Third, a smaller and more common one: for some people, a timer running turns the whole session into clock-watching, and the sense of being timed produces exactly the low-grade vigilance the practice is trying to reduce. If that is you, a duration is the wrong instrument, and a fixed number of cycles, three rounds of going toward and coming back, or an external anchor like the end of a piece of music, binds just as well without the surveillance. Fourth, and worth saying plainly: none of this addresses whether you should be doing this practice unsupervised at all. Somatic Experiencing is taught as a facilitated method, its practitioner training is long, and a stopping rule is a partial substitute for one function of a practitioner, not for the practitioner.

What the evidence does and does not support

The two halves of this page sit on quite different evidential ground, and the difference matters for how much weight to put on each. Somatic Experiencing has a small randomised literature. The most frequently cited trial, by Brom and colleagues in 2017, randomised people with post-traumatic stress to the method or a waitlist and reported meaningful improvement, but the sample was modest and waitlist comparisons cannot separate the specific method from the general effect of receiving attentive treatment. Reviews of the area consistently describe the evidence as promising and thin. The theoretical claim underneath it, that an incomplete threat response leaves a physical charge which discharging resolves, is a model with clinical observation behind it rather than a demonstrated physiological mechanism, and it should be held as an account rather than a fact. Precommitment has better evidence but in narrower territory: field trials in savings and smoking cessation, where the target is countable and verifiable, show that binding arrangements change behaviour, and the if-then planning literature is one of the more robust bodies of work in applied psychology. What none of that touches is the pairing. To test it you would need to randomise people doing body-based practice to a pre-set binding stopping rule versus their own judgement, and then measure something meaningful, and the measurement is where it gets genuinely hard. Session length is easy to record. Whether the practice was done at the right dose is not, and the two most important outcomes, whether people keep practising over months and whether they are worse off the following day, are both usually captured by self-report from a person whose judgement of their own state is the very thing in question. Until somebody solves that, this pairing is reasoning from mechanism, and should be treated as such.