Before reading this framework

The Common Understanding

How Bias Serves Regulation Rather Than Reasoning

Under stable conditions — when cortisol is low, the prefrontal cortex has full access, the perceptual field is broad — bias operates as rapid pattern recognition. The nervous system efficiently categorizes: familiar or novel, similar or different, this situation matches a past experience or it does not. Fast, low-cost, and usually accurate enough.

Under threat conditions — when cortisol is elevated, the amygdala's detection sensitivity is heightened, the perceptual field has narrowed — the same efficiency becomes rigid certainty. The nervous system is no longer categorizing for understanding. It is categorizing for stability. Beliefs that reduce activation are maintained. Beliefs that would increase activation are filtered out, reframed, or rejected. The criterion shifts from accuracy to threat reduction.

The regulatory equation: if believing something reduces nervous system activation, the system keeps believing it — below conscious awareness. By the time the person is "thinking about it," the perceptual system has already delivered a conclusion that feels like observation.

This connects to false coherence (F3). False coherence is the mechanism by which the CLS produces stable narratives that serve regulation rather than truth. Bias is that mechanism operating at the perceptual level: the narrative is so deeply embedded that it is no longer experienced as a narrative. It is experienced as what the person sees. What F5 describes structurally — who gets filtered in and out — F6 describes perceptually: how the filtering becomes invisible because it is absorbed into what feels like seeing.

The Architecture of Perceptual Filtering

Bias operates as a layered architecture — interacting constructs that together produce the felt certainty people experience as seeing clearly:

Emotional Logic — beliefs feel true when they settle the nervous system. A belief that reduces cortisol feels accurate, independent of whether it is. State-Dependent Perception — what the person perceives depends on their nervous system state. In Safety & Openness, nuance is available. In threat states, the field narrows toward threat-relevant data. Identity Filter — when beliefs fuse with identity. Contradiction is no longer disagreement about facts. It is identity threat. Social Reward Loop — bias reinforced through belonging. Agreement signals safety. Dissent signals threat. Shared beliefs function as obedience rules (F4). Empathy Collapse — Affective Resonance (ER) degrades under chronic activation (M4). The person cannot feel the impact of their perceptual distortion on others. RE may remain sharp — reading others with precision — but the reading serves strategy rather than understanding. Update Failure — when the Identity Filter is engaged AND Empathy Collapse has occurred, the system has no mechanism to update. Information that contradicts existing perception is filtered out. The source is discredited. Emotional Safety Threshold — the minimum safety level required for revision:

Update capacity = (Internal safety + Relational safety) − (Identity threat + Belonging threat)

When the right side exceeds the left, the system cannot update. The nervous system is structurally unable to revise because the cost would exceed what it can absorb.

Three Categories by Regulatory Function

Cognitive biases — confirmation bias, authority bias, negativity bias — reduce uncertainty. Social and cultural biases — in-group bias, racism, sexism, classism — maintain group cohesion. Internalized emotional biases — "I'm not good enough," "People can't be trusted" — are perceptual defaults about the self and the world, calibrated during development (F2) and maintained through false coherence. They have been running since before the person had language to question them.

How Bias Forms, Locks, and Feels Like Direct Perception

Bias forms and stabilizes through a six-step self-reinforcing loop:

Step 1: Uncertainty detected. The amygdala registers novelty. Mild activation. Step 2: Fast interpretation selected. The CLS pattern-matches and selects the fastest available explanation — binary, complexity-reduced. Step 3: Interpretation fuses with identity. "What I think" becomes "who I am." Challenging it now feels like challenging the person. Step 4: Social reinforcement stabilizes. The group rewards the shared interpretation with belonging signals. The bias is no longer individually held — it is socially maintained. Step 5: Challenge triggers defense. New information is processed as threat — to identity and to belonging simultaneously. The source is discredited. The challenge is reframed. Step 6: Revision requires safety. The nervous system must have enough safety to tolerate the physiological disorientation of being wrong — which is the loss of the stability the belief was providing.

Each cycle strengthens the architecture. Over time, the bias becomes automatic, invisible, and experienced as direct perception rather than interpretation. The architecture parallels F4's internalization loop and F5's worth loop — the same mechanism operating at the level of rules, worth, and perception.

Why Bias Feels Like Direct Perception

The sequence: a stimulus triggers uncertainty. The CLS selects an interpretation that restores coherence. The interpretation reduces activation — cortisol decreases, muscles relax, the HPA axis begins to stand down. The nervous system registers: the disruption has been resolved. The physiological settling feels like the interpretation was correct. "I feel certain about this" becomes "this is true."

The felt sense of certainty tracks physiological stability rather than epistemic accuracy. The person is experiencing physiological confirmation that the belief is correct. Their body is signalling truth. Information that contradicts the belief does not just contradict their thinking — it contradicts their somatic experience.

Collective State-Locked Perception

M2 maps State-Dependent Sensory Filtering — the nervous system's physiological configuration shaping what sensory input reaches the person before conscious processing begins. F6 shows what happens when these individual mechanisms synchronize across a population.

When enough nervous systems in a system share the same chronic state — when a population has been running at elevated activation long enough for the perceptual field to narrow collectively — the group shares the same perceptual filter. The filter is pre-cognitive in each individual and self-reinforcing across the group. Each person's biased perception confirms every other person's biased perception. The collective sees the same thing — and what the collective sees feels like reality because everyone confirms it.

The mechanism is physiological, not ideological. Cortisol elevation narrows the perceptual field in each nervous system. Amygdala sensitisation increases threat-detection sensitivity. Prefrontal narrowing reduces nuance-processing capacity. When these shifts occur across a population — through shared threat, shared structural conditions, shared developmental environments — the population develops a shared perceptual filter that is upstream of any individual's conscious reasoning.

The person is not maintaining a wrong belief against evidence. They are perceiving through a filter that the entire group shares — and the filter pre-selects what counts as evidence.

Three Relationships to the Same Bias

Absent interoceptive access — bias feels like reality. When the interoceptive substrate was never built, the CLS has only its own output, what RE provides, and the cultural narratives available. The bias feels true — not "believed" but constitutive of what the person experiences as reality. "Those people are less competent." "The system is fair." "I earned my position." The interoceptive substrate that would signal "something doesn't match" is not available. Correction is experienced as regulatory threat. The Identity Filter engages. The source is discredited. The bias hardens. This is typically the largest group — because the system itself produces the conditions that suppress the substrate (F4).

Contested interoceptive access — the body senses mismatch but the group says otherwise. When the substrate is active but the CLS has learned to distrust its signals, the person oscillates. The ESS generates activation around the mismatch. But the collective narrative reinforces the override: "You're too sensitive." "That's not really happening." "Everyone else is fine with it." These phrases target the contested access directly — discrediting the sensing that would reveal the bias as a construction. This population carries the crack in the system. The substrate is still active. The signals are still arriving. What they need is not more information but conditions safe enough to trust the body's signals over the collective narrative.

Full interoceptive access — bias held as construction. When all three awareness channels are delivering data, the person can evaluate the bias against felt experience. Can notice the perceptual filter operating. Can hold a collective narrative as a construction rather than absorbing it as reality. Can follow a rule while knowing it is a rule. This configuration is rare under chronic collective conditions — because the rule systems that govern most populations suppress the substrate state that would produce it. Full access is a perceptual capacity. It is not immunity to the system's regulatory pressures — the social cost of naming the construction can be high enough that the person perceives accurately and stays silent.

The system's stability depends on the distribution. When the majority carry absent access, the system is stable. When the contested-access group grows — when more people sense something is wrong — the system must invest more in discrediting the sensing: more enforcement, more gaslighting of bodily signals, more punishment of those who name the mismatch.

Why Education Fails and What Conditions Allow Revision

Standard bias-correction approaches — education, awareness campaigns, diversity training, moral argument — target the CLS. They present information. They make arguments. They ask the person to think differently. The bias is maintained by the interoceptive substrate state — not by the narrative.

When the substrate is absent, the CLS has no channel that would signal "the bias doesn't match the body's data." The correction is evaluated within the biased framework. It is absorbed, reframed, or rejected — because the mechanism that would produce genuine revision is structurally unavailable.

Shame-based correction produces a specific failure mode. Shame activates threat. Threat activates defense. The person under shame-based correction does not revise the bias. They perform revision — publicly adjusting language while the perceptual architecture remains intact. Performance of revision is itself a regulation strategy — a new form of false coherence: "I said the right thing" serving the same regulatory function as "I believe the right thing."

Five Conditions for Genuine Perceptual Revision

1. Internal safety. The nervous system must be regulated enough to tolerate the physiological disorientation of being wrong. The person must be in or near Safety & Openness.

2. Relational safety. The correction must come from a relational context the person trusts. When the source is perceived as hostile, the Identity Filter engages automatically.

3. Identity flexibility. The person must have enough identity space that being wrong about this specific belief does not threaten who they are.

4. Alternative meaning. The perceptual system cannot drop a stabilizing belief without a replacement that provides enough regulation. An alternative interpretation that settles the activation — while being more accurate — allows the system to update without regulatory collapse.

5. Gradual exposure. The perceptual system revises incrementally, not in sudden conversions. Gradual exposure to contradiction — in safe conditions — allows the architecture to update without overwhelming the nervous system's capacity.

The deepest revision — of internalized emotional biases — requires the relational conditions F2 describes: co-regulatory experience that provides what the original developmental environment could not. Perceptual defaults calibrated before language cannot be revised through language. They can be revised through relational experience that provides the safety the original environment lacked.

What This Framework Establishes

Bridge to F7

Connections Map

Where to Go Next