EXECUTIVE SCIENTIFIC VERDICT
Not one diagnosis, but several processes we can test.
Research does not support codependency as a single validated disorder or clearly bounded trait. It is more informative to decompose the label into family accommodation, repeated reassurance, relationship-contingent self-worth, controlling or overprotective support, low differentiation, and reciprocal influence.
Co-constriction can be retained only as a new, falsifiable process hypothesis—not as a diagnosis or established mechanism. Immediate relief, recipient-side repetition in bounded settings, recurrent joint states, and bidirectional influence have support. Provider-side reinforcement is directly undermeasured; role lock-in, progressive loss of flexibility, and narrowing of perceived futures have not been demonstrated as one longitudinal pathway.
Co-constriction is a proposed relational process in which a narrow set of responses becomes increasingly likely because it solves an immediate problem, while potentially displacing opportunities for autonomous coping, honest feedback, role variation, or alternative action. It is a working hypothesis, not a diagnosis; reciprocal influence does not imply equal power, choice, blame, or responsibility.
HYPOTHESIS TEST
Eight components, nine evidential rulings.
Component two is intentionally split: recipient-side repetition and provider-side repetition are different claims, supported at different levels.
Short-term reduction of distress or uncertainty
Strongly supportedDirectly demonstrated for reassurance, accommodation and several forms of social support in bounded settings.
Recipient-side repetition after relief
Reasonably supportedRenewed reassurance seeking or avoidance is consistent with negative-reinforcement accounts in bounded paradigms. Evidence that relief itself causes the next naturalistic episode remains limited.
Provider-side repetition after relief
Plausible but indirectHelping may reduce conflict, guilt, distress or uncertainty for the helper, but this reinforcement loop has rarely been measured across repeated episodes.
Increasing recurrence of a relational pattern
Reasonably supportedDense dyadic observations demonstrate recurrent joint states, although not the full proposed rescue loop.
Decreasing behavioral flexibility
Plausible but indirectRestricted observed repertoires can be quantified, mainly in parent–child research. Adult caregiving and addiction evidence is indirect.
Greater role rigidity
Plausible but indirectSeveral constructs describe fixed self-sacrificing or dependent positions, but repeated dyadic role switching is rarely measured.
Reduced autonomous coping
Plausible but indirectBounded experimental and clinical findings support a substitution concern; progressive impairment across real relationships is not established.
Narrowing of perceived alternatives
Currently speculativeCompelling as a Flow Hijacked interpretation, but no identified longitudinal dyadic study measures a shrinking set of perceived futures.
Mutual rather than purely individual maintenance
Reasonably supportedBidirectional influence is established as a possible architecture. It does not imply symmetry, equal power or equal responsibility.
RESEARCH ANCHORS
Six clusters that carry the argument.
These are not all 110 sources; they are the most direct anchors for the public reading. The source register and complete bibliography are available below.
Codependency: history without construct unity
The term remains culturally and clinically meaningful, but definitions and measures do not establish one disorder or personality type.
Accommodation, reassurance and reciprocal distress
Immediate relief and symptom association are well supported; temporal direction is mixed and often reciprocal.
Support, autonomy and substitution
Responsive and practical support usually accompanies better goal outcomes; control and overprotection can produce different effects.
Contingent worth, self-sacrifice and differentiation
These constructs make the borrowed-self pattern intelligible without treating it as a discovered unitary entity.
Addiction and the affected family system
Family involvement can improve outcomes and alter local contingencies; relatives should not be blamed for causing addiction.
Dyadic dynamics, recurrence and synchrony
Coupling and recurrent joint states can be measured. Synchrony is not inherently healthy, and an attractor requires an operational model.
TERMINOLOGY AUDIT
Which terms earn their place, and how to use them.
| Term | Publication decision |
|---|---|
| Codependency | Historical or popular doorway only; then replace it with the specific behavior or mechanism. |
| Enabling | Avoid analytically. Name the exact action and contingency. |
| Accommodation | Use technically in the clinical populations in which it was operationalized; elsewhere say accommodation-like. |
| Differentiation | Use technically with the explicit reminder that it means self-position within connection, not detachment. |
| Co-regulation | Specify the signal, direction, timescale and outcome; never use it as a synonym for harmony. |
| Attractor | Technical only when recurrence or convergence is operationalized; otherwise label it as metaphor. |
| Co-constriction | An explicit Flow Hijacked process hypothesis, never a diagnosis, trait or established mechanism. |
PUBLICATION GUARDRAILS
Claims this publication will not make.
- Codependency is a mental disorder.
- Family members cause or sustain addiction.
- Rescuing someone will make them dependent.
- Never reassure an anxious person.
- Natural consequences are always necessary for recovery.
- Differentiation means needing no one.
- Two nervous systems become one.
- Synchrony proves healthy co-regulation.
- Both partners are equally responsible for the loop.
- Co-constriction is the mechanism behind codependency.
METHOD AND DOCUMENTATION
Read the synthesis, inspect every source, and keep disagreeing with it.
This is a critical, mechanism-focused qualitative synthesis. It was designed for close selection and integration of influential, credible, directly relevant literature—not prevalence estimation or exhaustive accumulation. Sources play different roles: empirical result, authors’ interpretation, broader scientific inference, Flow Hijacked synthesis, and new hypothesis. A compelling title or citation count did not substitute for methodological appraisal.
The review joins different populations and levels of analysis; they should not be read as one continuous evidence stream. No included study tested the complete Co-constriction sequence. ‘Psychologically accessible futures’ is not yet an agreed measure in this literature.