Viable Terrain Theory (VTT) applies the Constraint Geometry Framework to agency — the problem of how any goal-directed entity navigates toward outcomes under constraint.
VTT introduces two operational measures:
Traction (μ)
Traction measures the degree to which an entity’s actions produce forward progress toward a goal. High traction means actions translate efficiently into goal-approach. Low traction means effort is dissipated — the entity is spinning its wheels.
Traction is not binary. It is a continuous measure that captures the coupling between action and outcome in a constrained environment.
Bearing (p⃗)
Bearing is a vector quantity that captures the direction of viable progress through constraint space. It is not simply “the direction of the goal” — it accounts for the constraint landscape between the entity and its objective.
An entity may need to move away from its goal to maintain bearing — navigating around a constraint barrier rather than pushing against it.
The Navigation Matrix
Together, traction and bearing define a navigation matrix with four operational states:
| State | Traction | Bearing | Description |
|---|---|---|---|
| NAVIGATING | High | Clear | Efficient progress — actions directly advance the goal |
| GRIPPING BUT LOST | High | Unclear | Power without direction — effort is wasted on the wrong heading |
| ORIENTED BUT SLIDING | Low | Clear | Friction — the path is visible but the agent cannot grip |
| FREEFALL | Low | Unclear | Neither direction nor grip — no viable action |
This matrix applies universally: to a robot navigating a physical environment, to an organisation pursuing a strategic objective, to a researcher exploring a problem space, to an AI system pursuing alignment.
Applications
VTT provides a formal framework for:
- Robotics and autonomous navigation — computing viable paths through constrained physical environments
- Strategic planning — assessing whether an organisation has both the direction (bearing) and the capability (traction) to achieve its objectives
- Decision theory — formalising the difference between “knowing where to go” and “being able to get there”
- AI alignment — measuring whether an AI system’s actions are coupled to its stated objectives
Relation to CGF
VTT is not an analogy or metaphor drawn from physics. It is a direct application of constraint geometry to the navigation problem. The same mathematical structures that describe how physical law emerges from constraint accumulation also describe how viable paths emerge from navigational constraints.