Trust is not a feeling. It is a geometric property of coordination systems. It emerges when structural proportions hold, and it collapses when those proportions distort — regardless of the intentions, competence, or goodwill of the people inside the system.
This means trust is diagnosable. You can read the structural conditions that produce it, identify where they are breaking, and design systems that maintain them. You don't have to wait for trust to collapse to know it's degrading.
The Geometry of Trust identifies five structural dimensions that must be proportionate for coordination to hold. These are not values or aspirations. They are coordination invariants — conditions that any system, in any domain, must maintain for participants to act with confidence rather than monitoring, hedging, or exiting.
The five dimensions are expressed as EARFT: Embodiment, Anticipation, Reciprocity, Feedback, and Tension.
The relationship is multiplicative. If any dimension goes to zero, the product goes to zero. Excellent Reciprocity cannot counterbalance broken Feedback. Each dimension is necessary. None is sufficient alone.
Where risk and consequence actually land in the system. Embodiment asks whether participants are visible — whether they can be seen, known, and accounted for by the system that serves them. When Embodiment is high, the people bearing the consequences are the people making the decisions. When Embodiment is zero, the participant is invisible: a claims number, a risk category, a data point.
Diagnostic question: Can the system see the person it's affecting?
The coherence between what a decision implies and what eventually shows up. Anticipation asks whether outcomes arrive in the form, at the magnitude, and at the time that participants were led to expect. When it holds, the system is navigable. When it breaks, every interaction becomes a surprise, and participants shift from acting to bracing.
Diagnostic question: Can the participant predict what the system will do before they need it?
Whether exchange remains mutual rather than extractive. Reciprocity asks whether what each party gives and receives is proportionate. When Reciprocity holds, participation feels fair — not necessarily equal, but proportionate to contribution, need, and risk. When Reciprocity inverts, the system takes more from the people who need it most. That inversion is the structural signature of extraction.
Diagnostic question: Does the exchange feel proportionate, or is the system extracting?
Whether information returns in time to guide decisions. Feedback asks whether participants can learn from the system — whether signals about cost, quality, outcomes, and consequences arrive before the decision point, not after. When Feedback works, the system is self-correcting. When it's delayed or corrupted, participants are navigating blind, and errors compound without correction.
Diagnostic question: Does information arrive in time to guide the next decision?
Whether strain is held and metabolized within the system or displaced onto participants. Every system generates tension — costs, conflicts, tradeoffs, uncertainty. The question is where that tension lands. When Tension is proportionate, the system absorbs and processes its own strain. When Tension is displaced, the cost of the system's failures is pushed onto the people least equipped to bear it.
Diagnostic question: Who absorbs the cost when the system fails?
The dimensions multiply, they don't add. Any dimension at zero collapses the whole system. Strength in one dimension cannot counterbalance a broken one. This is conjunctive necessity: all must hold.
Excellent Reciprocity cannot counterbalance zero Embodiment. Fast Feedback cannot counteract displaced Tension. Each dimension does its own structural work and cannot be substituted.
The framework reads failure before collapse. When proportions distort, the signals are legible: stress, withdrawal, monitoring, hedging, exit. These are geometric signals, not moral or psychological defects.
EARFT applies to any coordination system: healthcare, banking, education, governance, workplaces, communities. The dimensions are structural invariants, not sector-specific metrics. The same geometry that breaks in a hospital system breaks in a financial institution.
Failure is easier to see than functioning trust. When proportions distort, the signals are loud: delayed Feedback, displaced Tension, broken Reciprocity. When proportions hold, the signal is the absence of noise. Functioning trust is quiet geometry.
The signals are visible: participants monitor instead of act. They hedge. They document everything. They stop volunteering information. They exit when they can. These are not character failures. They are rational responses to unreadable structure.
Participants act without checking. They share information freely. They extend credit — of time, attention, risk — without requiring proof. They stay when they could leave. The system produces an absence of friction, not a presence of warm feeling.
The framework identifies two recurring patterns where systems degrade structurally while appearing to function. These are traps because the system's continued performance masks the structural failure underneath.
A system maintains operational performance while the structural conditions for trust degrade underneath. Transactions still clear. Services still deliver. Outputs continue. But the system has become brittle, expensive, and extractive. It is reliable without being trustworthy.
The Reliability Trap is dangerous because it looks like success. The metrics are fine. The outputs are consistent. But participation is coerced rather than voluntary, the system is sustained by friction rather than alignment, and the cost of maintaining performance keeps rising.
A system responds to structural failure by multiplying procedures rather than repairing the underlying geometry. More forms, more approvals, more review layers, more compliance requirements. Each addition is locally rational — it addresses a specific failure. But cumulatively, the procedures become the system's primary activity, displacing the function they were supposed to support.
The Legitimacy Trap is the structural cousin of the Reliability Trap. Where the Reliability Trap masks degradation with continued output, the Legitimacy Trap masks it with continued process. The system performs the appearance of accountability without repairing the conditions that made accountability necessary.
The Geometry of Trust is not a morality framework. It does not say what systems should do. It says what structural conditions must hold for coordination to be possible. Readability is not morality. A system can be readable and unjust, or just and unreadable. The framework diagnoses structural coherence, not ethical content.
It is not a feeling. Trust in the GOT sense is not warmth, liking, or confidence in another person's character. It is a structural judgment: this system's proportions are coherent enough that I can act without checking. That judgment can be wrong — but it is a perception of structure, not an emotion.
It is not a single-variable model. Trust is not reducible to competence, integrity, benevolence, or any other single factor. It is a property that emerges from the proportion across all five dimensions simultaneously. This is why systems staffed by competent, well-intentioned people still lose trust: the geometry can break even when the people don't.
EARFT functions as a diagnostic lens. Applied to any coordination system, it asks five questions: Is the participant visible? Can they predict what happens next? Is exchange proportionate? Does information arrive in time? Who absorbs the strain? The pattern of answers reveals where the geometry is holding and where it's breaking — before the system collapses.