We rarely stop to ask how something becomes irreversibly true. Yet this question sits beneath almost everything we take for granted: measurement, memory, entropy, and even the direction of time. Before a coin is flipped, both outcomes are possible. After it lands, one outcome is real and the other is not. What changed?
Most explanations appeal to dynamics, probability, or thermodynamics. But those explanations quietly assume something deeper: that once a distinction is made, it cannot simply be undone. This blog explores a more basic question—what structural condition makes irreversibility possible at all?
The central claim is surprisingly simple: irreversibility does not arise from time, energy, or probability alone. It arises from topology—specifically, from whether information pathways can form closed loops. In systems where all informational paths eventually rejoin (like a tree or a straight line), any distinction can in principle be undone. No choice is final. But the moment a loop exists, a new possibility appears: one alternative can circulate inside the loop while the other continues onward. Once that happens, the alternatives can no longer recombine without breaking the loop. A fact has formed.
This turns irreversibility into a threshold phenomenon, not a gradual one. Below the threshold—no loops—everything is reversible. Above it—at least one loop—irreversible distinctions become possible. There is no halfway state. Either information can be permanently quarantined or it cannot. This sharp transition explains why the smallest possible irreversible act corresponds to exactly one bit of information.
That insight has powerful consequences. The minimum entropy cost of forming a fact emerges naturally as kBln2, not because of heat engines or thermodynamic cycles, but because one excluded alternative must be stored somewhere inaccessible. Entropy, on this view, is not disorder—it is quarantined information. Likewise, the arrow of time does not require special initial conditions. It appears automatically once irreversible distinctions begin to accumulate. Time is not something that flows; it is something that builds—one fixed fact at a time.
Seen this way, facts are not primitive features of reality. They are topological achievements. Reality becomes definite not because the universe “chooses,” but because its informational structure allows some possibilities to be permanently trapped. Where loops exist, facts can form. Where they do not, everything remains provisional.
If this picture is right, then the deep structure of irreversibility—and with it entropy, memory, and time—does not live in equations of motion. It lives in the geometry of information itself.