For centuries the poems attributed to Homer posed a stubborn puzzle. The Iliad and the Odyssey exhibit a unity of design and a consistency of style that seemed to demand a single controlling author, yet they also bear features - repeated epithets, formulaic half-lines, occasional inconsistencies of detail - that a fastidious writer working on the page would presumably have pruned. Nineteenth-century scholars, the so-called Analysts, resolved the tension by dismembering the poems: the epics, they argued, were composite fabrications, layers of shorter lays stitched together by later editors, and the inconsistencies were the visible seams of the compilation. Against them the Unitarians insisted that only a single artistic intelligence could account for the poems' architecture, and treated the repetitions as deliberate ornament. The debate was transformed in the 1930s by a young classicist who reframed the question itself. Rather than ask whether one poet or many had written the epics, he asked how the poems had been made - and concluded that they had not, in the modern sense, been written at all. The recurring epithets, he showed, were not ornaments but tools: a system of metrically interchangeable phrases so economical that for almost any grammatical position and metrical need the tradition supplied exactly one ready formula, and rarely more than one. Such a system, he argued, could not be the invention of a single writer revising at leisure; it could only be the accumulated inheritance of generations of oral poets who composed in performance, assembling verses from a stock of inherited phrases at the speed of speech. The formulas that had embarrassed the Unitarians and armed the Analysts were, on this account, the fingerprints of an oral technique. To test a claim about a tradition centuries dead, the classicist turned to a living one, recording illiterate singers of epic in the Balkans whose practice displayed the same formulaic economy he had reconstructed for Homer. The parallel was suggestive, but it did not prove identity, and later scholars have pressed the difference. Oral composition, they observe, does not preclude a presiding genius; the Balkan tradition itself produced singers whose performances were manifestly superior to their fellows'. To show that the Homeric formulas are traditional is not to show that the Homeric poems are merely traditional - that no individual shaping hand selected, combined, and surpassed the inherited materials. The theory, in explaining the medium, risked explaining away the art. The most careful contemporary readers therefore decline the choice the older debate had imposed. That the epics were composed orally and that they are the work of a supreme poet are not competing hypotheses but complementary ones; the formulaic system is the language in which such a poet would necessarily have thought, not a substitute for his thought. What the oral theory dismantled was not the poet but a particular picture of authorship - the image of a solitary writer inventing every phrase - which had been silently imported from the age of print. Its lasting achievement was to make visible how much of that picture the older quarrel had taken for granted, and how little the poems themselves require it.