Plato
The Apology of Socrates — a defence that refuses to plead
Ἀπολογία Σωκράτους · The trial was in 399 BCE, a few years after the democracy was restored. Plato wrote this some time afterwards — the decade is argued over, the 380s are the usual guess — and he says at 34a and 38b that he was in the courtroom.
An Athenian jury of five hundred is asked to decide whether a man has been corrupting the young and importing new gods. He answers by cross-examining his accuser, telling them the god at Delphi put him up to it, and warning them that killing him will damage them more than it damages him. Convicted, and invited to propose his own penalty, he proposes free dinners for life at public expense.
How to read this page — what a person checked, and what the model wrote
The notes beside this text are here so you can read it straight through without stopping to look things up: what the passage assumes you already know, who is named without introduction, when the author is joking, what a social rule cost to break, and where the translator quietly tidied something up.
A language model wrote them, and no specialist has reviewed them. That makes them a companion rather than an authority — useful the way a well-read friend is useful, and wrong now and then in the same way. If you are writing something with a bibliography, cite the translation, named at the foot of this page, and let a note point you at what to check.
The division of labour on this page:
- Checked by a person: the translation, unaltered from the printed book; the source text, from the edition named below; every map dot for a named place, verified by hand against a gazetteer; every quotation, supplied from a hand-built list.
- The model's work: every note, headnote, gloss and piece of background; scene-location pins, which are its judgement and say so; any claim about tone, register or intent — those are arguments rather than facts, and you are free to disagree with them.
Quotations are the one thing the model is never allowed to propose — it may only explain one it has been handed from a hand-verified list, which is why you will see few of them. Spotting an allusion is the most valuable thing a commentary does and the easiest thing to invent convincingly, so that restraint is deliberate and is not going to change.
Spotted something wrong? Tell me — a reader who catches an error is the fastest way this gets better.
Harold North Fowler (1859–1955) did the first Loeb volume of Plato, printed facing the Greek, and it reads like it: close, unhurried, and deliberately flat where the famous translations are ornamental. That plainness is why it is here. Jowett is the celebrated readable Plato, but his Victorian polish smooths the joins in an argument, and Fowler's stiffer English leaves the steps visible — which matters in a text whose whole force is a chain of questions. The costs are real and worth knowing. His ὦ ἄνδρες Ἀθηναῖοι is 'men of Athens', formal where the Greek is ordinary address. The δαιμόνιον, the thing that stops Socrates acting, becomes 'the customary prophetic monitor' or 'the divine sign' — both fair, neither as strange as the Greek, which barely names it at all. And the jokes land softly: Socrates is funny in Greek more often than he is in Fowler.