The Value of Information as the Foundation of Ethics

From the Variational Principle to the Categorical Imperative

Abstract

A pickpocket and his victim render opposite verdicts on the same act. Religious ethics has a referee above both perspectives; secular ethics never found a replacement: Aristotle's "flourishing" is always somebody's flourishing, and utilitarian sums can be rigged to justify the theft. Missing, since Hume stated the gap nearly three centuries ago: a quantity whose growth is good and whose loss is bad regardless of anyone's viewpoint — a non-religious absolute.

The paper argues the quantity exists — that every persisting thing has been extremizing it all along. Every system — planet, cell, person, institution — has a graded capacity to constrain its own future from within. Building on my earlier definitions, I define this capacity as a non-negative, non-additive, intrinsic scalar: a system's value of information. Because that phrase already carries other meanings, the paper proposes a distinct name: *vedance*, from the Indo-European root *weid-*, "to see, to know."

Physics supplies the hypothesis: simple systems extremize classical action, dissipative ones obey entropy principles; the value of information, I propose, is the general quantity of which both are limiting cases. If that is what systems are, Hume's gap needs no leap — or so I argue: a system that can recognize its principle and choose among trajectories turns description into prescription. Good, within a system, is what increases its value of information; right, across systems, what maximizes it overall — one law, two scopes. The quantity does not sum, so this is not utilitarianism. Does a well-organized criminal group count as good? Within its boundary, yes — precisely why good is not the last word: corroding the trust and predictability its neighbors depend on, it fails the stronger test of right.

The framework's categorical imperative, meant literally: when we act badly, we lose the capacity to predict the future.