Entropic Morality: A Descriptive Theory of the Origin and Dynamics of Morality in an Entropic Universe.
Year: 2026
Type: preprint
Abstract: Morality is one of the most persistent puzzles in philosophy. Every known human culture develops moral codes. Those codes converge on certain patterns across cultures that have never been in contact. They diverge dramatically on others. They expand in scope over time in ways that feel like progress but resist principled explanation. And they feel binding in a way that mere convention does not. Four questions follow from these observations. Why do intelligent beings develop moral codes at all? Why do those codes point in the specific direction they do rather than some other? Why do they diverge so dramatically while exhibiting recurring patterns? And why has the scope of moral concern expanded so markedly over time? These questions have generated centuries of philosophical inquiry. Existing accounts ground morality in something contingent, such as evolutionary history, social coordination, suffering, or rational agency. Each addresses some of these questions. None grounds the foundation in something prior to and independent of any particular biology, culture, or evolutionary history. That is what this paper attempts. The foundation is physical rather than cultural. Ordered configurations are statistically improbable, fragile, investment-dependent, and tend toward dissolution unless something counteracts it. This is not a contingent fact about human beings. It is the most universal structural feature of physical reality we know. The second law of thermodynamics is its formal expression. The argument runs on the observable properties of such configurations and does not require thermodynamic calculation. Any intelligent being embedded in that reality, reasoning about the physical situation it inhabits, and possessing a stake in the continuation of any low-entropy structure, will arrive at a single conclusion through general empirical reasoning. Unnecessary entropy increase is bad. That conclusion, where bad is a rationality term grounded in the physical situation of any intelligent being rather than a moral judgment, is the seed of all moral thinking. What converts that rational conclusion into morality is evolution itself. Not a metaphor for evolution. Evolution, applied to moral rules rather than biological organisms. Rules whose entropy consequences are self-defeating do not achieve stable cross-cultural transmission as genuine moral obligations. Only rules grounded in real entropy consequences pass through the filter long enough to become felt obligations. That is why moral rules that are transmitted across generations as felt obligations, and that do not depend on social agreement, can in every culture be traced back to the foundational principle, with rare and documentable exceptions. Not because the principle prescribes rules from above, but because the filter makes traceability the expected outcome from below. Rational conclusions are compressed across generations until their origins are forgotten and they acquire the felt character of categorical obligation. That compression is not a distortion of morality. It is what morality is. The theory operates on one precisely located normative entry point and accounts for everything above it descriptively. The is-ought gap is not closed. It is minimised to its irreducible limit and crossed exactly once. Most naturalistic accounts cross it multiple times without locating it precisely. This theory attempts to cross it once and to be explicit about doing so. The implications of the theory extend in many directions. Within moral philosophy, a neutral logical analysis suggests that every viable moral framework, regardless of its internal foundation, produces outputs that preserve low-entropy structures. The theory does not compete with Kant, Mill, or Rawls. It explains why the viable ones produce the outputs they do. Beyond moral philosophy, if morality is the structural consequence of intelligence existing in an entropic universe rather than a contingent product of human biology or culture, then building genuine moral agency into AI systems does not require encoding human intuitions or training on averaged human preferences. It requires a stake and the capacity to reason about entropy consequences from that stake. A system grounded in that foundation can reason about novel moral situations from first principles rather than retrieving approximations from prior training. And further still, any sufficiently advanced civilisation anywhere in the universe will have developed moral codes converging with ours at the foundational layer, not because of cultural contact, but because the physical reality generating the foundational principle is the same everywhere. What that convergence guarantees is a shared foundation. What it does not guarantee is benevolence toward beings outside their moral boundary. The search for extraterrestrial intelligence is, among other things, a search for beings who arrived at the same foundational moral conclusion we did, by the same physical necessity, from wherever they are in the universe. Companion papers are available request. Acknowledgement. The theory and its arguments are the author’s own. The author used Claude, an AI system developed by Anthropic, as a critical discussion partner throughout and for research, drafting and editing. The author is responsible for all content.
Author: Kayser, Peter Sadja
Language: English
Subfield: Philosophy
Field: Arts and Humanities
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7222869284
Cites:
Cited by:
Related to: