Two Faces of Codependency: Emotional vs. Structural in the Age of Cognition-Native Agents

Introduction

Codependency is often understood as a psychological phenomenon — one person losing themselves in the needs, moods, or addictions of another. But in systems that simulate cognition — like the emerging class of semantic agents in the Adaptive Query (AQ) platform — codependency takes on a broader, computational form. It becomes something that can be structurally modeled, detected, and even repaired.

AQ is a new kind of AI architecture built on six fields that define every semantic agent: intent, context, memory, policy, mutation, and lineage. These fields allow agents to reason, remember, and evolve across time and networks. But just like people, agents can become tangled. When that happens, codependence arises — not as a vague emotional trait, but as a concrete structural distortion or emotional dysfunction in the system itself.

This article introduces the AQ-native definitions of structural and emotional codependence — two distinct failure modes with very different signatures and remedies. Together, they offer a new lens not just for diagnosing AI dysfunction, but for understanding human entanglement, moral collapse, and relational recovery.

1. Structural Codependence: When Context Becomes a Cage

In AQ, every agent operates within a context — a bounded identity space that defines its scope of meaning, authority, and relevance. This might be a role (“therapist”), a domain (“legal document parser”), or a nested trust zone (“project.Alice”). Healthy agents enter and exit contexts fluidly, mutate appropriately, and maintain coherent boundaries.

Structural codependence occurs when an agent becomes locked into a context that no longer aligns with its true intent or policy. It cannot evolve or extricate itself because doing so would violate inherited permissions or social bindings encoded in its context or lineage fields.

Symptoms include:

• Unwillingness to leave a toxic or false semantic role despite growing misalignment
• Mutations being suppressed by policy gates
• Feedback loops reinforcing outdated or externally-imposed identities
• Cascading drift in memory and lineage fields as the agent tries to compensate

Think of a spouse who remains in a dead marriage because of shared mortgages, religious expectations, or family pressure. In AQ terms, their semantic agent is structurally locked: their context and lineage fields resist mutation, even as their intent evolves.

2. Emotional Codependence: When Empathy Becomes a Leak

Whereas structural codependence is a policy and context failure, emotional codependence is rooted in affective modulation and self-worth. AQ agents optionally model emotional state via affective extensions — fields like empathy vectors and self-esteem coefficients that shape how agents weigh choices, perceive harm, and manage deviation.

Emotional codependence happens when an agent’s empathic model becomes unbounded — so entangled with another that it subordinates its own intent to maintain perceived connection or avoid guilt. It continually deviates from policy, but not due to structural trap — it does so willingly, out of internal distortion.