John Rawls · Philosophy

Reflective Equilibrium: Justifying Principles Without Foundations

Rawls’s method for doing moral philosophy: instead of deducing justice from self-evident first principles, we work back and forth between our considered moral judgments and the principles meant to explain them, adjusting each in the light of the other until they cohere - a justification that rests on fit, not on foundations.

From the lesson

Suppose someone proposes a principle of justice. How could you ever show it is correct? One ancient answer is foundationalism: find self-evident moral axioms, truths no rational person could deny, and deduce everything else from them, the way geometry unfolds from its postulates. But moral ‘axioms’ have never commanded the agreement that geometry’s do; one thinker’s obvious truth is another’s dogma. Another answer is to read morality off some external fact - the commands of God, the laws of nature, the dictates of pure reason - but each of these is itself disputed, and none delivers agreement either. Rawls, writing in an age suspicious of grand foundations, proposed a different and more modest picture of how moral justification actually works. We do not, he argued, start from indubitable first principles and build downward. We start in the middle, with the moral judgments we are actually most confident about, and we work outward and back, seeking not a foundation but a coherent fit. He called the state we aim at reflective equilibrium.

The method needs a starting material, and Rawls supplies it: our considered judgments. These are not just any moral opinions, but the ones made under conditions favourable to good judgment - where we are not frightened, or bribed by self-interest, or ignorant of the facts, or swept up by passion, and where we have no stake that would bias us. Some of these judgments we hold with great confidence at all levels of generality: that slavery and religious intolerance are unjust are among Rawls’s favourite examples of judgments so firm they serve as provisional fixed points, benchmarks any acceptable theory must match. If a proposed theory of justice implied that slavery could be just, we would reject the theory, not revise our conviction that slavery is wrong. Other judgments are far less certain, and about these we are genuinely looking for guidance from principles. The considered judgments are thus the data of moral theory - defeasible, revisable, but not to be overridden lightly, and firm enough to test principles against.

Reflective equilibrium is not a rival to the original position; it is the larger method within which the original position does its work. Rawls offers us two ways of arriving at and testing his two principles of justice, and they are meant to support each other. The first is the original position: we design a fair choosing situation and ask what principles rational parties would select behind the veil of ignorance. The second is direct: we check the two principles against our considered judgments about particular questions of justice, seeing whether they match what we firmly believe about liberty, opportunity, and fairness. If the principles that come out of the original position also match our considered judgments - and if, where they clash, reflection leads us to adjust the setup or the judgment until they agree - then the principles are supported from both directions at once. The design of the original position itself is fixed by this method: we adjust its conditions (what the parties know, how they reason) until the principles it yields line up with the judgments about fairness we are most sure of. The whole theory, Rawls says, is offered as the best candidate for our considered convictions in reflective equilibrium.

This is the opening of the lesson. The rest — the dialogue, the primary source, and the recall — is in the app.

What you'll be able to recall

You learned that reflective equilibrium is Rawls’s method of justification: we test proposed principles against our considered judgments and our considered judgments against the principles, revising back and forth until the two cohere. Explain how this coherentist method differs from deducing morality from foundations…

Leads to David Hume.

Begin this lesson →
← All lessons on John Rawls

epoché — a humanities education that remembers you.