|Rodin, Orphée et Eurydice (1893). The Met.|
And we still have to do this today. People sometimes ask whether the humanities can really create knowledge or make progress. This is a clear instance of both. Sometimes a new piece of evidence turns up, and then you really need the trained specialist who can tell you how it gives us a better reading of what Aristotle or Vergil or some minor Greek poet probably really wrote. It takes a certain kind of person. Not everyone is cut out for this slow and painstaking work. At its most potentially stultifying, textual scholarship involves a good deal of squinting and arguing over near-invisible levels of nuance solely for the sake of fastidiousness—you have to really like arguing about grammar and the faded handwriting of medieval scribes.
But every now and then (more often than you might think), the process turns up some point of contention that alters the meaning of a poem entirely, and then everyone cares a lot. For example, in the seventh poem of the Roman poet Propertius' fourth book of elegies—a transfixingly strange poem that purports to be spoken by the poet's mistress from beyond the grave—there are two variants of one line:
sic mortis lacrimis vitae sanamus amara (Hutchinson)
sic mortis lacrimis vitae sanamus amores (Markland)
(two different readings of Propertius 4.7.69)
The first of these means, "So with the tears of death, we heal the bitternesses of life." The second means "So with the tears of death, we heal the loves of life." The difference in meaning is huge! And according to the rules of Latin poetry, both readings are, strictly speaking, possible. And this is the point where textual scholarship and literary interpretation meet: in coming to decide which of these readings is more likely, we have to bring in our best understanding of similar literature in the time period; of the historical context in which it was written; and sometimes our aesthetic sense of what makes for a better poem.
Good scholarship, of course, is interested mainly in the question of which version is most likely to have been actually written by Propertius. But sometimes scribal error has its own poetry, and there's something haunting about the two versions of this one line. Whether the scribe was paying attention or not, it's a beautiful mistake to have substituted amara for amores in this line or vice-versa—a notional proximity between the concepts of amores, one's loves (or possibly "love affairs" here), and amara, one's experience of bitterness. Anne Carson writes about the twinning of love and bitterness in ancient poetry going back to Sappho at the very beginning of her celebrated essay Eros the Bittersweet—what clever carelessness to have fallen into the sonic similarity between the two in Latin, an almost Finnegans Wake sort of pun. Occasionally, even corruption can generate a kind of beauty. I've heard that in Japanese, there's a special aesthetic term for "the inelegant flaw that makes for an elegant whole." Propertius' Elegies 4.7.69 is like that.