I’m ridiculously busy, but came across — drat, I was interrupted by the phone even as I typed that! — … but I came across a very nice article online about the ancient paragraphos in — drat, interrupted AGAIN! — … about the ancient paragraphos in French here, complete with a very nice photograph of a papyrus.
The papyrus was found in the South-West of the Fayum in Egypt in 1901-2 by a French archaeologist, reused in cartonage, and contained portions of a lost work by Menander. The article links to the announcement of the discovery here, identified by three words in the colophon — Sikuw/nioj Mena/ndrou a)riqmo\j… — which was followed by a numeral indicating 1,000.
The article is not quite correct — the paragraphos, the line in between the lines, indicates that somewhere on that line is a division marker — often a colon, or perhaps a space. In a modern text each speaker would be on his own line. Not so in antiquity.
I always wonder, faced with such comments, how we actually know that this is so. The article tells us that Aristotle mentions the paragraphos in his Rhetoric, and that it is the only punctuation mark he mentions. I was unable to locate the passage in Aristotle, tho, as no reference was given.
UPDATE: I have found that the reference is to Aristotle, Rhetoric, book 3, chapter 8, verse 6 (238-9), rendered here:
A sentence should break off with the long syllable: the fact that it is over should be indicated not by the scribe, or by his period-mark in the margin, but by the rhythm itself.
At Perseus we get this:
ἀλλὰ δεῖ τῇ μακρᾷ ἀποκόπτεσθαι, καὶ δήλην εἶναι τὴν τελευτὴν μὴ διὰ τὸν γραφέα, μηδὲ διὰ τὴν παραγραφήν, ἀλλὰ διὰ τὸν ῥυθμόν.
But the period should be broken off by a long syllable and the end should be clearly marked, not by the scribe nor by a punctuation mark, but by the rhythm itself.