Second of three. The first page laid out the model. This one locates it — what it shares with established orientations, and where it adds something.
The argument, up front
Here is the claim this whole page rests on, and it’s a claim about the logic of therapeutic technique rather than about any one tradition:
Every restructuring and every attentional redirect steers toward an emotional target. The only question is whether the target is named or covert.
Consider what a thought record actually does. A client is anxious. The therapist works the thought. Something changes. What changed? Not the world — the world is the same. What changed is which emotional system has the wheel. The client has been moved out of threat and into something else. And the therapist had a sense of where the client was being moved, or there would have been no way to know when to stop.
That destination is an emotional target. It was there whether or not you called it one.
This isn’t a criticism of anybody. It’s an observation that the field has, for the most part, done this work without a shared vocabulary for the thing it’s aiming at — and one tradition has already shown what happens when you do name it.
The existence proof: CFT
Paul Gilbert’s founding observation was that self-critical clients could follow the logic of CBT perfectly well, could generate a balanced alternative thought on demand — and would still feel nothing. The thought was correct. The system didn’t move.
His response was to change the target. Not “what’s the accurate thought?” but “what’s the compassionate one?” — restructuring aimed explicitly at the soothing and affiliative system rather than at accuracy. The compassionate thought record. Compassionate self-correction as against shame-based self-attacking.
That is restructuring toward a named emotional target, and it works, and CFT got there decades before I did. It’s the proof that naming the target is both possible and clinically useful. Everything I’m proposing is a generalization of a move Gilbert already made.
The generalization
CFT names one target: compassion, warmth, the soothing system. It names it because self-criticism was the problem it was built to solve, and warmth is the right answer to that problem.
But warmth is not the right answer to every problem.
Sometimes the target is drive — the client is flat, stalled, and needs the seeking system to come online. Sometimes it’s curiosity — they’re locked in threat and need to get interested in something instead. Sometimes it’s assertiveness — a mild heat injection, a spine. Sometimes it’s grounded calm. Sometimes, genuinely, it’s sadness, because there’s a loss that needs the slow-down and the deep review that only sadness gives you, and every attempt to be productive about it is a way of avoiding it.
So the generalization is this: restructuring is cognitive work toward a named proactive emotional state — and the state should be chosen to fit the situation. Compassionate reframing is the compassion-specific instance of a general move.
Which yields a clinical question available at the start of any thought record: what state is this person being steered toward, and is it the right one?
Most clinicians answer that question implicitly already. Making it explicit changes the work. The new thought stops being optimized for accuracy and starts being optimized for sufficiency — is this enough to satisfy the system that’s firing, so that a different system can take over? Those are not the same standard, and they come apart more often than the literature admits.
What each tradition already has, and what this adds
This is offered as complementarity, not ranking. Every orientation below is doing something real, and in most cases doing it better than a general framework could. What follows is what each one contributes and what a systems-and-steering layer would add on top — nothing more.
CBT
What it has. The structural skeleton in near-universal use: developmental history → core beliefs → intermediate beliefs → coping strategies → situation → automatic thought → emotion → behaviour. A rigorous, well-specified restructuring procedure. Collaborative empiricism and guided discovery. And critically, a multi-function method — modern CBT is not “fixing the ten distortions,” it’s a metacognitive tool, a conceptualization tool, an evidence-testing tool, and a Socratic method all at once.
What this adds. The map stays; the arrows change meaning. CBT’s stated mechanism is that belief change is upstream of affect. Mine inverts it: affect is the substrate, and the thought is a lever into it — a restructuring works by satisfying the firing system enough to hand off, not by installing a truer belief. And CBT’s target, accuracy, is itself an unstated affective stance: treating the emotion as an error to be corrected toward neutral is a claim about what the emotion is for. I’d rather make that claim out loud so it can be examined.
To be clear: this is a disagreement about mechanism between two people who use nearly the same procedure. Beck would not accept my account, and I’d rather state that plainly than pretend at a consensus that doesn’t exist.
Emotion-Focused Therapy
What it has. The closest sibling, and the one I’m most in debt to. Affect as primary datum rather than output. Emotion schemes. And the mechanism that most of what I do is a version of: changing emotion with emotion — a maladaptive state is transformed by co-activating a more adaptive one. That is Greenberg’s, not mine, and it’s the deep source of “use emotion to work with emotion.”
What this adds. Two things, both small and both real. EFT sorts emotions into adaptive and maladaptive, and that grading drives its interventions. I don’t grade. No emotion is more trustworthy than another; every one carries a real function and a live impulse, and the work is learning what each means — which is often not its surface message — rather than deciding which to trust. And EFT is chair-work-forward; the steering frame lets the same mechanism run through cognitive, behavioural, and situational levers as easily as through experiential ones.
ACT
What it has. The single best account of why fighting your own content backfires, and a deeply useful move: stop struggling with the thought, turn toward what matters. It converges with the reactive trap almost exactly — we arrive at the same clinical caution from different premises.
What this adds. A friendly disagreement about the affect-neutrality of values. On my account there is no motivation without feeling: you don’t act because something “matters” in an emotion-free way — mattering is affectively charged. So a model that disavows a feeling-target doesn’t escape having one. It has one covertly, and it tends to resurface as covert achievement-striving or introjected shoulds. This is the cleanest case of the covert-target argument, and I offer it as an invitation rather than an indictment.
Also: I permit deliberate suppression. Contain now, reopen later. Chronic containment is the pathology; containment is a skill.
Compassion-Focused Therapy
What it has. The three circles. The mechanism that warmth and affiliative affect down-regulate threat — Gilbert’s, stated directly in his work, and I use it as a skill and credit it as his. And, as above, the precedent for the entire argument on this page.
What this adds. Only the generalization: the target needn’t be warmth. It can be any proactive system the situation calls for.
Internal Family Systems
What it has. Multiplicity, taken seriously. Parts with motives, protective functions, and legitimate standing — and a locus that relates to them rather than being them. The functional architecture is very close to the council.
What this adds. Restraint about the metaphysics. IFS posits a capital-S Self — core, undamaged, essential. I locate the work in conscious awareness and take no position at all on what, if anything, sits behind it. Same clinical function, no essentialist commitment. I don’t assert a self and I don’t deny one; the distinction between the systems and the awareness that hears them is enough.
Schema Therapy
What it has. The best account anywhere of the layer I call strategies-and-patterns — how the coping styles and modes accrete developmentally on top of the drives. The punitive-critic mode. I use this directly and don’t improve on it.
What this adds. A structural clarification with a clinical consequence: the needs are biological base; the strategies are what was built. Which means the critic is a motivational strategy carrying a real self-correction drive — and the full response has two halves. Let the critical thoughts sit in peripheral awareness without engaging them (the steering half), and find another way to meet the self-correction drive they’re carrying (the resolution half). Do only the first and the critic keeps coming back, because its job never got done.
DBT
What it has. Distress tolerance and the whole concept of skills as deliberately initiated, taught and practised. Contain-now-reopen-later has clear DBT cousins.
What this adds. An organizing principle for the skills: each one is the deliberate initiation or de-initiation of a state. Which means the skill set isn’t a fixed list — states outnumber systems, because they include variations within a system and blends across them.
MBCT and the mindfulness traditions
What they have. Decentering. The attentional training. The demonstration that the relationship to a thought matters more than its content.
What this adds. A narrower name for what I’m actually cultivating, and a wider door. I call it curious engagement rather than mindfulness — not because mindfulness reduces to curiosity, but because “mindfulness” is a rich, contested term and I don’t want to inherit that argument to make a smaller point. And presence has two doors, not one: high-arousal absorption — hard exertion, intense sensory engagement — is a legitimate route to presence, not a lesser one. The quiet, inhibitory pathway is not the only pathway.
Behavioural Activation and exposure
What they have. Two of the most robust behavioural technologies we’ve got. Action ahead of motivation; approach ahead of readiness. Inhibitory learning as the modern account of what exposure does.
What this adds. A re-mechanization. On this account, activation isn’t primarily about reinforcement schedules and exposure isn’t primarily about habituation — both are direct work on emotional systems through action: cultivating drive, following an impulse, giving the threat system the corrective data it cannot get any other way. The procedures don’t change. The reason they work does.
Attachment theory and SDT
What they have. Attachment supplies the developmental account of strategy formation. SDT supplies internalization — the extrinsic-to-intrinsic pathway — which underwrites the observation that most practices don’t start as enjoyable and become so.
What this adds. A durability principle for motivation. A motivation is workable fuel when what it pulls for is real and can actually be met. It’s structurally vulnerable scaffolding when satisfaction is hostage to something the person cannot guarantee — a specific person, a status, an outcome. That scaffolding tends to hold for years and then collapse at a transition. The line is met vs. hostage, not self-directed vs. other-directed. Mixed motives are normal and there’s no purity requirement.
The psychodynamic inheritance
What it has. Defence, and specifically compartmentalization, understood as structured rather than merely pathological. And counter-transference — which I use as more than a heritage note: it’s the model for how a person learns to read their own reactions as data to interpret rather than instructions to follow.
What this adds. A temporal rule where there was a pathology judgment, and no unconscious-as-agent.
The one place I don’t collaborate
Psychological constructionism — the account on which discrete emotions are constructed from core affect plus conceptual categorization rather than being evolved systems — is the one live rival, and I’m on the other side of it.
I want to be precise about the boundary, because constructionism is right about things. Its critiques landed: there is no single brain region per emotion, and there are no invariant facial or autonomic fingerprints. I concede both, freely, and the model doesn’t need either. Conceptual and linguistic layering shapes emotional perception — of course it does.
What I don’t accept is the generative claim: that valence and arousal plus concepts produce the discrete emotions. That claim is, as far as I can see, undemonstrated — and the move of excluding animal evidence from consideration, on the grounds that animals can’t self-report, makes it very hard to test against the largest relevant body of evidence we have.
Perception, yes. Generation, no. That’s the boundary, and it’s the only one on this page.
The invitation
The ask here is not that anyone adopt this framework. It’s smaller and, I think, more useful:
Emotional targets, owned explicitly.
A thought record run with a known destination — what state it’s steering toward. A behavioural experiment assigned with a system in mind. A grounding skill taught knowing what it’s making room for.
Every clinician is already steering somewhere. Everyone is. Naming the destination costs nothing and changes what gets built — and if a shared language for those destinations turns out to be useful across orientations, that’s the whole contribution this framework claims.
Next: How it works with existing therapies → Back: The Model →