Reachability: Why Does an Existing Outcome Not Mean We Can Get There?

Reachability asks not whether a state exists, but whether we can get there from the current state through a sequence of allowed transitions.

Imagine a document system with four states: draft, review, approved and published. The published state certainly exists, but a draft cannot move there directly. If review authority is missing, the target is temporarily unreachable. The problem is not whether the target is reasonable, but whether the starting point, transition rules and path are complete.

Reachability differs from possibility. Possibility says only that an outcome is not logically excluded; reachability also requires an actual path. It also differs from a local optimum, where no nearby step is better. With reachability, even the first valid step towards the goal may be unavailable. Planning, learning and AI all need this distinction: a clearly stated goal cannot replace checking the conditions of transition. Real change often begins by creating a workable path.