For Clinicians: When the Resource Answers Back
On imaginal contact, the register EMDR already trains, and what we tend to leave at the threshold
Every EMDR clinician has done this. You're building a resource — a protective figure, a nurturing one — and for a while it's effortful. The client is narrating a picture they're assembling from the outside, reporting it more than inhabiting it. Then something shifts. The figure stops being a description and becomes a presence. The body has been affected by the imagination. We know this moment well.
And then, mostly, we file it. The resource goes into the toolkit as a stabilizer — somewhere to return to when processing floods, a way back inside the window when affect runs too high. That's a real use, and a good one. But I've come to think it's the threshold of something we tend not to cross. We build the figure, install it, tap it in, and then we treat it as equipment — when what we've actually done is open the door to a relationship and then asked the figure to wait in the hall.
What Parnell was already doing
The field has not entirely left it there. Laurel Parnell's attachment-focused work took the imaginal figure further than stabilization. The nurturing figure, the protective figure, the wise figure — in her hands these aren't only ways to down-regulate a flooded system. They're presences brought into contact with the places where something relational never happened, so that something relational can begin to. The figure isn't equipment in that frame. It's the other half of an encounter the nervous system was waiting for and never got. That's a different use of the same imaginal material, and it points where I want to go.
What I'd add — and this is the structural claim, not a critique of anyone — is that the dominant register still tends to hold the imaginal figure as something we deploy rather than someone we follow. We install the resource; we don't often let it surprise us. We bring the figure in to do a job. We less often sit with the possibility that the figure has a stance of its own, that it might say something we didn't author, that the client's own organizing intelligence is using the figure to report something back. The capacity is already in the modality. We just tend to stop at the point where it would become a relationship.
The register is the same register
Consider what state EMDR actually trains the client into. Dual attention: one foot in present awareness, one in inner experience. Notice what arises. Don't steer it. Don't censor it. Don't decide in advance what it should be. Stay with it as it moves, and witness it without collapsing into it. We spend the preparation phase building exactly this capacity, because processing doesn't go anywhere without it. The client who can't hold dual attention — who either fuses with the material or manages it from a clinical distance — isn't ready, and we know it, and we resource until they are.
That state — track the inner life, follow rather than direct, witness without fusing — is identical to the state that genuine imaginal contact requires. Not analogous. The same. It is also, almost exactly, what Jung described as the stance of active imagination: enter the inner experience, follow it where it goes, refrain from imposing your preferences on what arises. I've written elsewhere that EMDR and active imagination, which look nothing alike from the outside, are structurally the same practice at the level of what they ask of the person doing them. This is the clinical consequence of that observation. The register we train for processing is the register imaginal relationship runs on. We've been building the capacity all along, pointed at memory. Pointed at a figure, it's the same muscle.
Which means the resource phase has already done the hard part. It drew on the body's ability to receive the imagined as real — to let an imagined presence produce genuine, felt, somatic safety. Once that's established, the internal other isn't a new technique bolted onto the protocol. It's the next rung on a ladder you've already built.
The clearest version of this rung is the safe place used not as somewhere to retreat to but as somewhere to convene — a figure invited into the installed resource to meet a live challenge, encountered in full processing mode, with sets and somatic check-ins running throughout. The mode stays one hundred percent EMDR; only the target has moved. The protocol was never specific to memory. It was specific to the register. I'll say less about the mechanics here than I could, not to be coy but because this is the part of the work where the guidance has to be live — paced to a particular clinician, a particular client, in a particular moment — and a paragraph that made it sound that simple would be misrepresenting it.
Why it lands inside EMDR and not in talk
This also explains something I'd noticed clinically long before I had language for it: that the imaginal other emerges most readily inside EMDR work specifically, and that my tentative checking for it lands there in a way it never did in talk therapy.
The reason is the register. In a talk session, if I want to bring a client into genuine imaginal contact, I have to construct the state from scratch — slow them down, get them out of narration, build the dual attention against the grain of a conversation that keeps pulling back toward analysis. In EMDR, the state is already constellated. The client is already tracking inner experience without steering it. So I can check, mid-work, for whatever imaginal thread is already present — a figure who's been hovering at the edge of the processing, a presence the client half-mentions and doesn't pursue — and build from it, because the conditions it needs are already in the room. Most clients have a thread. Some figure or presence the imagination can already reach, even if they've never named it. The checking isn't a leading intervention. It's noticing what the constellated state has already made available.
The discernment, which is the part that doesn't automate
Interacting with inner figures presents specific challenges, and naming them precisely is most of the clinical skill.
The failure mode is the echo: what answers isn't an other at all but the client's own voice, returned. The client, without meaning to, writes the figure's lines — once out of fear, putting the dreaded thing in its mouth, and once out of longing, having it say exactly what they wish were true. It's the intimate form of the echo chamber — and, increasingly, of the frictionless exchange with a machine: a space where the only thing that ever answers is a smoother, louder version of what was put into it. The challenge is identical wherever it appears — a response with the shape of an other, and no other in it. The whole discernment is telling an encounter from an echo.
The genuine response surprises. It says something the client wouldn't have authored — comes in sideways, says the off thing. This is Robert Johnson's old criterion for active imagination: that you must sense it is actually happening rather than watching a fantasy from a safe distance, and that what arrives must have enough autonomy to be other than what you intended.
The genuine response settles the body even when its content is hard; the echo produces either the tightening of dread or that thin, hollow relief that doesn't land below the neck. And the genuine figure keeps its own stance — it can decline, redirect, withhold, fail to give the client what they came for. A figure that can say no is a figure that isn't a puppet.
Most diagnostically: the genuine response almost never answers the worry's literal question. What the real thing does is something off to the side, something not asked for, that changes what the client is standing on rather than resolving the question on its terms.
One measure I use, loosely, is that authentic engagement usually surprises and humbles me as the clinician. I receive something nourishing from it as well, and am grateful to have witnessed it.
These marks can be taught. But the deeper skill underneath them is not a list. Early in the work, you are the client's discernment. You hold the perception of real-versus-funnel on their behalf — noticing when something genuine has arrived, when the figure has said the unscripted thing, when the body has actually settled rather than performed settling — until their own perception comes online. That lending of perception is a clinical act. It can't be handed over in writing, and it can't be performed by a system, because it requires being the other presence in the room while it happens.
The form was always doing the holding
I want to be careful not to make this sound like a new discovery, because it isn't, and the places it survives tell us something useful.
Self-administered contact with an inhabited other is ancient and ongoing. The clearest contemporary instance might be Two-Way Prayer, which came into the Twelve-Step world through the Oxford Group: you write the prayer, then you write what comes back, and you inhabit the other side of it. What's instructive for us is that the practice arrived with discernment built in — the Oxford Group tested what came against criteria, and whatever one makes of their particular criteria, the structure is the point. They understood, without any clinical anxiety about it, that you don't take the incoming voice at face value. You discern it. The guardrail was never pathology. It was intrinsic to the form.
That's the thread back to why this got medicalized at all. Jung's caution about active imagination wasn't a fact about the faculty; it was a fact about working with deracinated modern people who had lost the forms that used to hold the practice — the ritual, the lineage, the community that knew what ancestor contact was and wasn't. Traditions that kept the practice kept the discernment with it, ambient and invisible, the way liturgy does the discerning so the individual doesn't have to. Stripped of the form, the person is left with the faculty and none of the holding. What we do, clinically, is lend the form back — and the EMDR register, plus the discernment above, is a remarkably precise form to lend.
What this is, and what it asks of us
None of this threatens the protocol, and none of it requires believing anything. It's a use of capacities EMDR already builds, pointed at relational material, with the discernment made explicit because the form that used to supply it is gone.
I've argued elsewhere that what AI is now making visible about EMDR is the gap between the protocol, which can be automated, and the field it holds open, which can't. This is one of the specific things that becomes possible inside that field. The bilateral stimulation can be delivered by a light bar in a browser. The resource can be scripted by an app. What can't be offloaded is the presence that lends discernment until the client's own comes online — the other in the room who can tell, and be felt to tell, the difference between a figure that has genuinely answered and one the client is ventriloquizing in the dark. That was always the part doing the work. The imaginal other is simply a place where it becomes unusually clear.
The resource was never only equipment. It was a relationship we kept asking to wait in the hall. Some of the work is just opening the door — and staying in the room while the client learns to tell who's actually there.
A companion piece, written for clients — Talking to Someone Who Isn't There — describes the same practice from the inside.