Natural Law (Aquinas)

In the [YOU] on AI Field Guide

The cycle returns again and again to the question of what kind of values we are building into the most powerful systems our civilization has ever made. Natural law is the concept that names the difference between discovering the right values and merely instilling whichever values the current training process happens to reward. A model is “aligned” by human feedback, written guidelines, and reinforcement on approved responses. The result is a set of behavioral dispositions, but the ground of those dispositions is, in the end, the preferences of the people who built and trained the system. There is no claim that the values track a real moral order; there is only the fact that these are the values that were instilled. Aquinas’s natural law is a standing challenge to the complacency of this position—not a refutation, but a demand that we ask whether we are discovering something or only enforcing something, and whether that distinction matters.

The cycle also treats natural law as the framework within which the companion volume’s central claim—that these systems can enlarge your agency or hollow it out, depending on how you relate to them—becomes a moral claim rather than merely a preference. If human beings have a real nature with real ends, and if practical wisdom, conscience, and the exercise of rational agency are among those ends, then a technology that systematically erodes those faculties is not merely inconvenient but actually bad in a sense that goes beyond anyone’s say-so. Aquinas’s natural law gives that intuition its philosophical spine.

Origin

Aquinas’s fullest treatment of natural law appears in the Summa Theologica, Part I-II, Questions 90–97. He builds on Aristotle’s teleological ethics—the view that things have natural ends, and that goodness consists in the fulfillment of those ends—and on Cicero’s Roman natural-law tradition, synthesizing both within a theological framework. The eternal law is God’s reason governing all of creation; the natural law is the rational creature’s participation in the eternal law, the way in which a being that can reason about its own nature comes to know the moral order that governs it. Divine positive law (Scripture and revelation) and human positive law (enacted legislation) are both measured against the natural law; a human law that contradicts the natural law is, for Aquinas, not a law at all but a perversion of law.

The natural law tradition deriving from Aquinas became enormously influential in the development of human rights theory, international law, and constitutional thought. John Locke, Hugo Grotius, and the framers of the American Declaration of Independence all drew, directly or indirectly, on the idea that there are moral truths accessible to reason that constrain what any authority may legitimately do. The tradition has been contested since Hume’s is-ought argument exposed the difficulty of deriving normative conclusions from natural facts, and it has been further challenged by evolutionary biology, which replaces teleological nature with adaptive fitness. But it has never been refuted, and its core claim—that the question of what is good for a being of a certain kind is a real question with a real answer, not merely a disguised expression of preference—retains serious defenders in contemporary philosophy.

Key Ideas

Good is to be done, evil avoided. The first principle of the natural law is practical and self-evident: good is to be pursued and evil fled. This is not trivially true; it specifies that moral reasoning begins from the real good of the agent, not from arbitrary stipulation. All further precepts of the natural law are specifications of this principle, derived by reason tracking the natural inclinations of the human being.

Grounded in nature, not convention. Natural law is natural because it is grounded in the nature of the beings it governs—in what it is to be a rational, social, embodied creature with a life to live and genuine goods to pursue. It is therefore not alterable by convention, by majority vote, or by the preferences of those with power. This is the move that distinguishes natural law from legal positivism and from value relativism. For AI alignment, it raises the question of whether there is a non-arbitrary target beyond the preferences of the training team.

Knowable by reason. Crucially, the natural law is accessible to unaided human reason—it does not require revelation to know that murder is wrong, that promises should be kept, that justice is to be done. This makes it in principle available as a standard that any rational system could in principle be held to, and any rational system could in principle be asked to reason toward. The difficulty is that Aquinas’s natural law is embedded in a teleological picture that modern science, since it stopped explaining nature by purposes, has made philosophically expensive—a difficulty the natural-law revival in contemporary ethics has not fully resolved.

Prudence as the bridge. Even the clearest natural-law principle does not apply itself. Between the general precept and the particular act stands the judgment of prudence—the practical wisdom that perceives what this situation, with these persons, at this moment, requires. Natural law without prudence is a list; prudence without natural law is cleverness at the service of whatever goal happens to be given. The two are inseparable in Aquinas’s moral framework, and both are absent from the model that has neither a grasp of genuine goods nor the judgment to apply any grasp to a particular case.