λᾰβύρινθ-ος

λαβύρινθος (laburinthos, "lah-BOO-rin-thoss") is the labyrinth, and the record gives it a small, tightly-bounded life: 8 occurrences across just 2 works. Almost all of them belong to one place — Herodotus's Histories holds 7, clustered in a single stretch (2.148–2.149), where the word names a real building the historian is describing. Plato's Euthydemus supplies the one remaining occurrence (291). So in this corpus the word lives almost entirely as reportage: a thing Herodotus went to see, and once — in Plato — a figure of speech.

The lexica agree on the object and hesitate on nothing. LSJ (Liddell–Scott–Jones, 9th ed., 1940) gives "labyrinth, maze," and defines it as "a large building consisting of numerous halls connected by intricate and tortuous passages," locating it in Egypt (citing Herodotus 2.148) and in Crete. Beekes (Etymological Dictionary of Greek, Brill 2010) glosses it "labyrinth," a "great building with many corridors and turns," and marks the same geography — Egypt, Crete, Anatolia — before noting its metaphorical use "of complicated thoughts" in Plato, which is exactly where the Euthydemus occurrence sits: the tangle of an argument felt as a maze one cannot walk out of. Chantraine (Dictionnaire étymologique, 1968–80) calls it a "monument compliqué," a complicated monument of buildings joined by winding passages.

On the word's descent, the record is honest and quiet: no etymology pointer is matched here. Beekes's entry does carry one deep trace — a Mycenaean form, da-pu-ri-to-jo, glimpsed on the Linear B tablets — but the brief offers it as the lexicon's own note, not as a reconstructed root, and so it stays a shape rather than a settled origin. The non-Greek look of the word (that -νθος ending, pre-Greek in feel) is left unspoken by the record.

The building Herodotus walked was real; the argument Plato lost his footing in was not. If the same word names both, is a labyrinth a place or a predicament?