What I Built. What I Built On.
A clear-eyed account of where this work stands. Each original move placed against its closest precursor in the existing literature. Each one named honestly as either clinical phenomenological discovery or synthesis.
Published April 2026. Living document. Revisions noted at the bottom.
I get the same question often.
Are you not just restating what others have said? Is this not just established science with a new vocabulary?
It is a fair question. It deserves a careful answer.
Yes. The foundations are established science. I did not invent the substrate. Anyone who claims to have done so, in this field, is either reinventing something that already exists or hiding from the literature. I have done neither.
What follows is the cleanest account I can give of where this work stands. What I built on. Why I built it. What the original moves are, and how each one stands against its closest precursor in the existing work. What I would not claim. And what would have to be true for me to change my mind.
I have tried to be precise enough that a careful reader can decide for themselves. I would rather lose a claim under scrutiny than keep one that does not deserve to stand.
This work is not for everyone.
The popular psychology of the last thirty years has given the field extraordinary tools. Trauma-informed care. Polyvagal theory. Internal Family Systems. Schema therapy. Somatic frameworks. The shame and vulnerability work. The addiction-as-wound work. Each of them has helped a lot of people. Some of them have helped me.
There is a population they do not reach.
People who have done all the work. The therapy. The retreats. The men's groups. The right partner. The right job. The books. The protocols. And underneath all of it, the body still cannot rest.
This work is for them.
Not because the other work is wrong. Because their problem is at a layer the other work was not built to address. If you have done everything and you are still not okay, you are not failing. The work was not built for the layer where your lock-in is.
That layer is what this framework names. The capacity to remain present with internal experience. The source of regulation. The architecture underneath the patterns the other frameworks describe.
Every framework rests on the work that came before it. This one rests heavily.
Stephen Porges
Polyvagal theory. Neuroception. The autonomic substrate of safety and threat. The floor under everything I have written.
John Bowlby and Mary Ainsworth
Attachment theory. Internal working models. The architecture of how early experience shapes adult relational behavior.
Mary Main
The Adult Attachment Interview. The recognition that the coherence of an adult's narrative predicts a child's attachment quality more reliably than almost any other variable.
Allan Schore
Right-brain development. The science of how early caregiver attunement shapes the regulatory architecture of the developing brain.
Bessel van der Kolk
The body keeps the score. Trauma reenactment. The integration of trauma research with somatic understanding.
Richard Schwartz
Internal Family Systems. Parts as protective adaptations rather than pathology. Even where I depart from IFS, I depart from a foundation I could not have built myself.
Jeffrey Young, Janet Klosko, and Marjorie Weishaar
Schema therapy. Schema chemistry. Schema perpetuation. The direct precursor to most of what I say about dyadic systems.
Karl Friston, Andy Clark, Jakob Hohwy
The free energy principle. Active inference. The brain as a prediction machine that acts on the world to make its predictions come true. Without their work, the Reconstruction Mechanism is not even possible.
Daniel Stern
Representations of Interactions Generalized (RIGs). The early articulation of how repeated infant-caregiver interactions become the building blocks of internalized self-with-other knowledge.
Edward Tronick
Mutual regulation. The Still Face Experiment. The recognition that infants and caregivers regulate each other through micro-interactional repair, and what happens when repair fails.
Myron Hofer
Hidden regulators. The physiological substrate of separation distress. What gets called grief is, in part, the loss of a co-regulating system.
Onno van der Hart, Ellert Nijenhuis, Kathy Steele
Structural dissociation. The substrate beneath my secondary containment identities.
Peter Levine and Pat Ogden
Somatic experiencing. Sensorimotor psychotherapy. The recognition that trauma cannot be metabolized through cognition alone.
Gabor Maté
The shift from why the addiction to why the pain. The Reach as a frame in this work would not exist without him.
Daniel Siegel
Interpersonal neurobiology. The Window of Tolerance. The integration of attachment, neuroscience, and clinical practice.
Deb Dana
The clinical translation of polyvagal theory.
Object relations
Fairbairn, Winnicott, Guntrip, Kernberg. The internalization of relational configurations as the building blocks of inner experience.
If any part of my work proves useful to you, please send the gratitude in their direction first. They built the floor I am standing on.
I did not set out to build a framework. I set out to find one.
Twenty years inside this literature. I read the books. I sat in therapy. I sat in church. I sat in boardrooms. I lived inside several of the systems this framework now describes. Each adjacent framework named something true. Each one helped. None of them, alone or together, answered the questions my body could not stop asking.
- Why do people get stuck.
- Why do patterns repeat after the original conditions are gone.
- Why do people who have done the work, who have insight, who genuinely want to change, walk back into the same relational architecture they swore they would not return to.
- Why does genuine safety register as wrong.
- Why does healing sometimes deepen the wound it was meant to address.
- Why does insight not produce change.
The literature has pieces of each answer. Schema therapy describes how schemas perpetuate once selected. It does not describe what the system does when it cannot find the schema-confirming environment. Repetition compulsion names the phenomenon of return. It does not give it a mechanism. Active inference describes a brain that acts to make its predictions come true. It has not been clinically operationalized for relational and attachment configurations at the resolution I needed. Polyvagal theory describes the autonomic substrate. It does not describe what happens when nervous system states stabilize, repeat, and organize into the identity a person calls "me." IFS describes parts. It assumes a stable Self that many people, including me, did not experience.
I built this because I could not find what I was looking for. Not because the existing work was wrong. Because the answers I needed were at a layer the existing work was not built for.
The framework is what the answers looked like once I started writing them down. It maps a layer. The layer underneath the patterns. The layer where the lock-in lives. The layer most popular work operates above without naming.
What kind of contribution this is.
I want to be honest about the category.
This is not laboratory neuroscience. The substrates are established. I have not discovered a new neural pathway or a new synaptic mechanism. That is not what is here.
But this is not only synthesis.
Some of what is here is clinical phenomenological discovery. The naming of mechanisms, phenomena, and categories that have not been named in the existing literature with this clarity. That is the kind of discovery that founded the fields I build on. Bowlby did not discover that infants attach. He discovered the architecture of how attachment works. Porges did not discover the vagus nerve. He discovered the polyvagal architecture. Schwartz did not discover that people contain multiplicity. He discovered the structure of how parts function. Naming a phenomenon precisely enough that the field can see it is what discovery looks like in this work.
Several moves in this framework are discoveries in that sense.
The Reconstruction Mechanism is one. The specific clinical phenomenon, that survival configurations actively generate environmental friction to recreate missing familiar signal in real time, is not in the literature. Active inference is the substrate. The clinical phenomenon, named with this specificity, is new.
Loss of Signal is one. Genuine safety registering as absence rather than threat is a phenomenological category that has not been named in the existing literature. It is a different clinical object than withdrawal, dissociation, or avoidance.
The 10-stage Containment Loop is one. The four moves locked together (false relief as the mechanism of capture, recommitment as the trap that feels like faithfulness, shame as accelerant rather than regulator, the same analysis applied across system types) constitute a structural mechanism that has not been named at this resolution.
The four-category taxonomy of inner phenomena is one. The disambiguation of Internalized Figures, Survival Roles, System Functions, and Exiled States is not made elsewhere with this clarity, and the clinical implication that different categories require different interventions is novel.
The eight Secondary Containment Identities formed AFTER threat is one. The categorical distinction between primary survival identities and post-threat control overlays is not in the existing literature.
State-Person Fusion is a discovery in the formalization sense. The phenomenon has been gestured at in clinical observation. Naming it precisely enough that it can be measured, with the Fusion Error mechanism (the collapse of person, safety, identity, and regulation state into one), is original.
Identity forms before Self when safety is absent is a discovery of structural assumption. It directly inverts a foundational claim in IFS, and it explains why parts work without preliminary self-attachment building does not produce stable change for many people.
Other moves in this work are synthesis. The Fastest Stability Principle is recombinatorial. The four configurations in the Borrowed Safety Model are a parallel re-typology of attachment styles measured along the regulation source axis. The three-layer regulation arc parallels earned secure attachment trajectories. The 10 identity pattern count is empirically provisional and may collapse to a smaller set under principled coding. These are real contributions but they are integration rather than discovery.
There is one more thing worth naming about the category. This framework operates at a layer most popular work in this field does not. The layer of self-attachment capacity. The layer of regulation source. The layer underneath the patterns, the parts, the schemas, the behaviors. That is the layer where lock-in lives. For people whose problem is foundational, work at the layers above produces relief without structural change. This framework is built to address that gap.
The page below names which is which. The reader who wants to know what is being claimed at which level can see it directly.
The strongest original moves.
I keep this list curated rather than exhaustive. The full list lives in the companion document. These are the moves I would defend most directly under serious adversarial review. For each one, I name the closest precursor in the existing literature and the specific distance.
1. The Reconstruction Mechanism
This is clinical phenomenological discovery. The closest existing accounts are Freud's repetition compulsion, van der Kolk's trauma reenactment, Young's schema perpetuation, and Friston's active inference. None of them say what this work says. Freud named the phenomenon without giving it a mechanism. Van der Kolk's reenactment names the neurobiology of returning to familiar danger. Schema perpetuation describes maintenance, avoidance, and compensation as ways schemas are preserved once selected. Active inference describes a brain that acts on the world to reduce prediction error. The new move is the specific clinical phenomenon: when the environment fails to provide the familiar signal, the survival configuration actively generates environmental friction in patterns predictable by configuration. The Scanner escalates interpretive vigilance in low-threat environments. The Performer generates performance demands where none exist. The Fixer magnifies external instability. The Invisible withdraws strategically. The behavior is observable. The mechanism is named. That phenomenon, at that resolution, is not in the literature.
2. Loss of Signal phenomenology
This is clinical phenomenological discovery. Adjacent work gestures at the experience. Trauma clinicians describe calm states feeling unfamiliar to chronically activated systems. Polyvagal work notes that ventral vagal can feel like unfamiliar territory after sustained sympathetic or dorsal occupancy. Schema therapists describe healthy relationships feeling "wrong." None of them name this as a discrete clinical category. The discovery is the explicit phenomenological claim. Genuine safety registers not as threat (the standard avoidant or dissociative reading) but as absence. Silence on the channel. Boredom. Flatness. A vague sense that something is wrong that cannot be named. This is a different clinical object than withdrawal, dissociation, or avoidance, and it explains why people terminate genuinely healthy relationships and therapeutic alliances by reporting them as wrong rather than as threatening. Naming this as a discrete category is what the literature has not done.
3. The four-category taxonomy of inner phenomena
Internalized Figures, Survival Roles of Self, System Functions, Exiled States. The closest precursors are IFS parts, ego state therapy, and structural dissociation. IFS lumps these together as "parts." Ego state therapy does not make these distinctions. Structural dissociation has ANP/EP but at a different level of analysis. Object relations talks about internalized objects but does not distinguish them from survival roles or system functions. The disambiguation, with the clinical implication that different categories require different interventions, is not made elsewhere.
4. Eight Secondary Containment Identities formed AFTER threat
The closest precursors are IFS protector parts (managers, firefighters) and structural dissociation. IFS does not distinguish primary survival identities from post-threat control overlays. Structural dissociation describes ANP/EP at higher trauma loads. The categorical distinction between "formed under threat" and "formed after threat ends but before reliable safety arrives" is novel, and it explains the spike in self-criticism that follows escape and insight, which the existing literature describes but does not mechanistically account for.
5. The 10-stage Containment Loop sequence
The closest precursors are cult dynamics literature (Lifton's eight criteria, Singer's six conditions, Hassan's BITE model), religious trauma syndrome (Marlene Winell), and Maté on shame and addiction. Cult dynamics describes top-down ideological control. The Containment Loop describes bottom-up regulatory capture, applicable to systems that are not cults (therapy, coaching, well-meaning religious community). The four moves together (false relief as the mechanism of capture, recommitment as the trap that feels like faithfulness, shame as accelerant rather than regulator, the same analysis applied across system types) are not made together with this clarity in any existing source.
6. Identity forms before Self when safety is absent
The closest precursor is IFS, which assumes a stable Self always present, even if inaccessible. This claim directly inverts that assumption. Identity can form INSTEAD of Self when safety is absent during the developmental window. This is a structural reframe with a specific clinical consequence. It removes the shame people feel when they cannot find Self in parts work, and it explains why parts work without preliminary self-attachment building does not produce stable change for many people.
7. Signal Recursion vs Signal Continuity
The closest precursors are general couples therapy on conflict escalation, polyvagal on co-regulation, and Gottman's flooding. None of them name this specific structural distinction. The novel move is naming the difference between solo regulation conditions (closed signal field, where input arrives only from inside) and relational regulation conditions (open signal field, where each person's output becomes the other's input before the previous signal can metabolize). This explains why solo capacity does not transfer to relational contexts, which is one of the most common clinical questions and one the existing literature does not answer at this resolution.
8. The 5-dimension capacity model of self-attachment with hierarchical modulator structure
The closest precursors are the Window of Tolerance (Siegel), distress tolerance (Linehan/DBT), interoception research (Garfinkel, Critchley, Barrett), and affect tolerance (Schore). Window of Tolerance is one-dimensional. DBT's distress tolerance is one-dimensional. Interoception research has multiple dimensions but they target sensory accuracy, attention, and awareness rather than the capacity dimensions named here. The five-dimension model with the hierarchical modulator structure (Signal Integrity and Interruption Resistance gating the operating conditions of the other three) is not in the existing literature.
9. State-Person Fusion as a named, operationalized construct
The closest precursors are schema chemistry (Young), idealization in object relations, and attachment-based love research (Hazan and Shaver). The phenomenon of person-state fusion has been described informally and gestured at clinically. The novel move is naming it as a measurable construct with the Fusion Error mechanism (the collapse of person, safety, identity, and regulation state into one), and putting it inside an operational equation (the Regulation Dependency Index). The mechanism behind disproportionate grief, romantic obsession, and cycles of return is not neediness. It is the collapse of four distinct things into one.
10. The Fastest Stability Principle as governing axiom plus the derivation that insight is not a regulatory pathway
The closest precursors are allostasis (Sterling, McEwen), free energy minimization (Friston), satisficing (Simon), and behavioral economics' bounded rationality. The principle is recombinatorial. It draws on all of these. The novel move is the explicit lock-in clause ("reuse it, even when better outcomes exist") and the derivation it makes possible. From the axiom, the persistence of identity patterns under awareness, the externalization of regulation, the loyalty to harmful systems, and the failure of insight to produce change are derivable rather than empirical observations. The claim that insight does not produce change is a long-standing observation in the trauma literature. The derivation from a single governing principle is what is novel.
These are the ten I would defend first. There are more in the documents and in the companion paper.
Where the work is more synthesis than original.
I want to name where the work is more recombinatorial than original. A reader who knows the literature should not have to look for it.
The Fastest Stability Principle is recombinatorial. It draws on allostasis, free energy minimization, and satisficing. The lock-in clause is doing the additional work, and the derivation of "insight is not a regulatory pathway" is genuine. But the principle as a whole synthesizes existing work rather than originates new ground. The page should not present it as discovery.
The four configurations in the Borrowed Safety Model (Borrower, Hidden Borrower, Distributed Regulator, Defensive Isolator) are a parallel re-typology of attachment styles measured along the regulation source axis. The novelty is in what is being measured (operationalized regulation source rather than behavioral surface), not in the count of categories.
The three-layer regulation arc (Borrowed, Shared, Embodied) parallels research on earned secure attachment and developmental trajectories of attachment differentiation. The operational definitions are mine. The arc itself is not a new shape.
The 10 identity pattern count is empirically uncertain. The Governing Rule (autonomic substrate mapping required for inclusion) is methodologically sound. The specific number ten is provisional. Pleaser, Performer, Fixer, and Chameleon share fawn-adjacent or sympathetic-with-social-engagement substrates. Under principled coding, the count may collapse to a smaller set. The framework allows for this. The page should not treat the number ten as fixed.
The Pattern Formation Loop draws on state-dependent learning and standard conditioning models. The application to identity formation specifically is the contribution, but the underlying mechanism is established.
Naming these honestly does not weaken the framework. It calibrates the work for the reader who is deciding whether to invest the time.
What would update me.
I am not interested in protecting this work from honest contact with reality.
What would update me upward.
- Empirical operationalization of the five capacity dimensions in a clinical sample, with the hierarchical modulator structure showing up in the data. People with high raw capacity scores but low Signal Integrity should show different relational and clinical outcomes than people with high raw scores and intact Signal Integrity. That is testable.
- Behavioral evidence for the Reconstruction Mechanism distinguishable from schema perpetuation and trauma reenactment. In low-signal environments (genuinely safe ones), individuals with strong Survival Configurations should generate environmental friction in patterns predictable by configuration. That is observable.
- Loss of Signal showing up as a distinct phenomenological category in qualitative research with people leaving high-control environments or recovering from chronic relational dysregulation.
- Adoption by clinicians outside my orbit who report that the framework changes their work in ways the existing frameworks did not. Clinicians who can name a clinical move they would not have made without this framework, and an outcome that improved as a result.
- Rigorous differentiation studies showing that the ten identity patterns and the eight secondary containment identities cluster as distinct categories in clinical observation, rather than collapsing under principled coding.
What would update me downward.
- Adversarial review showing that three or four of the operational moves compress back into existing constructs without remainder. That would be a serious finding. It would not invalidate the framework as a whole. But it would require revision.
- Clinical use showing that the ten identity patterns collapse to a smaller set under principled coding. Particularly likely candidates for collapse: Pleaser, Performer, Fixer, and Chameleon may turn out to be functional variants on a smaller number of autonomic configurations.
- The Reconstruction Mechanism turning out to be schema perpetuation in different vocabulary, or active inference applied to relational domain without distinctive predictive content.
- The Fastest Stability Principle failing to produce predictions distinguishable from allostasis or free energy minimization. The principle is doing significant derivational work. If it cannot be distinguished from existing principles in its predictions, it is not earning its place as an axiom.
- The Discrepancy Signature account failing to produce reliable distinctions in self-report data. If the signatures do not show up empirically, that part of the framework cannot stand.
- Loss of Signal failing to be distinguishable from existing accounts of avoidant withdrawal or dissociation under careful phenomenological analysis.
I have not seen any of those happen. I am open to all of them.
What this is not.
- This work is not a diagnosis.
- It is not a personality model.
- It is not a unified theory of human behavior.
- It is not a treatment protocol.
- It is not a tool for explaining other people to themselves.
The constraint is the feature.
The giants gave us the floor.
I built on people who came before me. I have tried to name them honestly. I have tried to name what is mine and what is not. I have tried to name what would change my mind.
If the work is useful to you, take it.
If it is not, leave it.
Either way, the giants gave us the floor. The work above the floor has helped a lot of people. This work names what was underneath, for the people the work above could not reach.
For the longer companion document, see the PDF version of this page.
For clinicians, researchers, and writers.
If you reference this work, please use the following citation:
Charles, R. (2026). The Attachment in Motion Model: Foundations and Origins. Retrieved from https://rosscharles.net/the-attachment-in-motion-model-foundations
Revisions: this page is a living document. Material updates will be noted here as they are made.
Part of a larger architecture.
This page accompanies The Attachment in Motion Model, a complete system for understanding where the nervous system sources its safety and what it takes to move the source inside.
An original framework developed by Ross Charles. rosscharles.net