Take this one with a grain of salt, but here is the latest:
Kennedy's breakthrough, published in the journal Apeiron this week, is based on stichometry: the measure of ancient texts by standard line lengths. Kennedy used a computer to restore the most accurate contemporary versions of Plato's manuscripts to their original form, which would consist of lines of 35 characters, with no spaces or punctuation. What he found was that within a margin of error of just one or two percent, many of Plato's dialogues had line lengths based on round multiples of twelve hundred.
The Apology has 1,200 lines; the Protagoras, Cratylus, Philebus and Symposium each have 2,400 lines; the Gorgias 3,600; the Republic 12,200; and the Laws 14,400.
Kennedy argues that this is no accident. "We know that scribes were paid by the number of lines, library catalogues had the total number of lines, so everyone was counting lines," he said. He believes that Plato was organising his texts according to a 12-note musical scale, attributed to Pythagoras, which he certainly knew about.
Do note this:
Kennedy believes his findings restore what was the standard, mainstream view which held for 2,000 years "from the first generation of Plato's followers, up through the renaissance". This held that "he wrote symbolically and that if you worked hard and became wise you could understand the symbols and penetrate his text to his underlying philosophy." Only in the last few hundred years has an emphasis on the literal meanings of texts led to a neglect of their figurative meanings.
It also explains why it is that Aristotle, Plato's pupil, emphatically claimed that Plato was a follower of Pythagoras, to the bafflement of most contemporary scholars.
I used to consider allegiance to this idea (Montaigne, also, for symbolic codes) as one of my absurd beliefs, but maybe now it is looking better. I will have to look elsewhere.