The Euthyphro Dilemma

Two men meet outside the courthouse in Athens. One of them, Socrates, is about to be tried for impiety. The other, Euthyphro, is there to prosecute his own father for the death of a labourer — an act so unusual that it needs justification, and Euthyphro justifies it on the grounds that it is the pious thing to do.

Socrates, facing a charge of not knowing what piety is, has found a man who is certain he does. He asks him to explain. What follows is fifteen pages that end, like most of the early dialogues, with nobody knowing anything — except that a question has been opened which twenty-four centuries have not closed.

Three attempts and a fork

Euthyphro's first answer is that piety is what he is doing right now: prosecuting a wrongdoer. Socrates rejects this on formal grounds. He asked for the thing that makes all pious acts pious, not for one example of a pious act. This distinction — between an instance and the property it instantiates — is doing a lot of quiet work, and it is Plato's, not Euthyphro's.

The second answer is better: the pious is what is dear to the gods. Socrates points out that the gods, on Athens' own account of them, quarrel constantly. If they quarrel, they disagree about what is admirable and just, and the same act will be dear to one and hateful to another. The definition makes an act pious and impious at once.

Euthyphro patches it: the pious is what all the gods love. The patch works. The definition is now consistent. And it is at exactly this point, with the argument apparently repaired, that Socrates asks the question that ruins it.

Is the pious loved by the gods because it is pious? Or is it pious because the gods love it?

The left horn: goodness comes first

Say the gods love piety because it is pious. Then piety is a property acts have on their own, and the gods are responding to it — reliably and perfectly, perhaps, but responding. Their approval is evidence of goodness rather than its source.

For a believer this is uncomfortable but survivable. What it costs is the claim that morality is grounded in the divine. If there is a standard the gods themselves consult, then in principle you could consult it too, and the theology becomes a shortcut rather than a foundation. Euthyphro cannot take this horn, because his whole authority for prosecuting his father was that the gods' approval settles the matter.

The right horn: the command comes first

So take the other one: acts are pious because the gods love them, full stop. The grounding claim is saved. Two things break.

First, arbitrariness. If divine approval is what makes a thing good, then nothing about the thing itself constrains the approval. Had the gods commanded torture, torture would have been good — not tolerable, not excusable, but genuinely good, and our revulsion at it simply mistaken. Most defenders of this horn want to say "but God would never command that," which is an appeal to God's character being good, and that quietly reintroduces a standard the will is answering to. The left horn again.

Second, emptiness. On this horn, "the gods are good" cannot be praise. Unpacked, it says the gods approve of what they approve of. A statement that was meant to be the highest possible commendation turns out to be a tautology.

Why this outlives the gods

Read the dilemma as being about the gods and it is a curiosity for people who already believe. Read it as being about authority and it is inescapable, because the structure has nothing to do with divinity. It applies wherever the goodness of an act is claimed to be constituted by something's endorsement of it.

Substitute the law: is an act wrong because it is illegal, or is it illegal because it is wrong? Take the first and you have no ground from which to call any law unjust — the Nuremberg defence, formalised. Take the second and you have conceded that law answers to a standard outside itself, which is exactly what natural-law theorists have argued for and legal positivists have spent a century resisting.

Substitute the crowd: is it right because we agree on it, or do we agree on it because it is right? The first is cultural relativism, and it forbids you from saying a society ever improved — only that it changed. The second says the consensus is tracking something, and can therefore be wrong.

Substitute a specification. When a system is trained to behave according to a document of stated values, the same fork opens under it: is the behaviour good because the spec approves it, or does the spec approve it because it is good? Answer the first way and "aligned" means nothing more than "conforming" — the system cannot be well-behaved and the spec wrong at the same time, and the entire practice of revising the spec becomes unintelligible. Answer the second way and you have admitted the spec is an attempt at something it does not itself define, which is almost certainly true and considerably harder to work with. Euthyphro's problem is now a design problem.

The ending nobody quotes

Euthyphro does not answer. He tries a few evasions, complains that his arguments keep moving around on him, and then remembers he is late for something and leaves. Socrates is left outside the courthouse, still not knowing what piety is, about to be tried for not being pious enough.

Plato is not subtle here, and the joke is a serious one. The man who was certain enough to prosecute his own father could not say what his certainty was made of — and the city was about to kill someone for the crime of asking.