There’s quite a bit of discussion going on about Jay Kennedy’s claim to have “cracked”—not just “discovered”—but cracked a hidden code in Plato’s writings. I’m amazed that he’s knocked surfer-rebel-physicist Garret Lisi’s new paper out of the #1 spot on Scientific Blogging!

Cracked – The Plato Code, Says Historian.

His claim is basically this. Plato’s writings are measured out in twelfths, analogous to the twelve intervals of a musical scale (not a common Greek musical scale but one we can’t prove Plato didn’t know). These twelfth-part of the texts are important. For instance:

• Socrates’ speech in Menexenus takes up 10/12 of the dialogue.
• In Symposium, 3 speeches are each about 1/12th of the work, Alcibiades speech is 2/12th and Socrates long speech 3/12.
• In Phaedrus, Socrates’ second speech is three times as long as his first speech, which is “somewhat longer” than 1/12, so the second is “somewhat longer” than 3/12 (though presumably three times the “somewhat” of the first) and runs from 4/12 of the way through the text to 7/12.

Perhaps. Kennedy doesn’t give much of an interpretation other than to say this emphasis on the harmonies of a musical scale is Pythagorean, that Plato was a crypto-Pythagorean and thus hiding his love of twelfths in plain sight. At any rate it’s good to check some numbers, which turns out to be somewhat interesting.

Kennedy claims that ancient texts were written in columns of consistent length of about 35 letters per line, thus making it easy to count the lines of a text, know how many there were, and know when you were 1/12 of the way through. Let’s just not argue with that for now, or worry about the fact that he quotes his source for this figure as saying “This standard line (Normalzeile) of circa 35 letters therefore … dominated book production unchanged through at least five hundred years from Dionysius’ copy of Thucydides until the time of Justinian,” that is for the 500 centuries beginning about 350 years after Plato died. Kennedy claims that at 35 characters per line, the line counts of Plato’s works would come out surprising close to multiples of 1200. Here are the figures he gives:

On these figures Kennedy says “Using a figure of thirty-five letters per hexameter line, calculations of the total number of lines in the dialogues produce, with about one or two percent accuracy, impressively round numbers involving multiples of the number twelve.”

After you said, “Where’s his data?” you probably said “total number of lines in the dialogues? This is not the dialogues. It’s some dialogues.” And you’re right, so I decided to check to see if he was cherry picking his dialogues. Unfortunately it’s harder to find digital texts of Plato in Greek than I would have thought.  At the Perseus Project, they come in chunks and reassembling the whole would be quite a task. Anyone know a way around this? But I did find some texts on a Greek website.

I didn’t check Protagoras, Cratylus, Philebus, or Gorgias because I haven’t found digital copies of those texts. I didn’t check Republic or Laws because they’re very long. His number for Apology and Symposium check out within about 4% and 3% respectively. Whether this is statistically significant is a question I don’t have the expertise to ask. and which Kennedy doesn’t raise.

The first dialogue not on Kennedy’s list that I checked was Phaedrus, and I must admit that to my great surprise, it fit his pattern. Coming out to 2,400 lines it fits even more closely than the first two. Cratylus was too short but came out to 600 lines within less than 1/5 of one percent! Phaedo, however, is way of and when I saw the figure for Timaeus I thought I was beginning to smell a fallacy.

Without looking at the figures which is closer: 1,251 to 1200 or 3,486 to 3,600?

Imagine you’re an ancient Greek, reading Plato’s Timaeus—one of his most important works—all 3,486 lines of it. Are you going to be able to tell when you’ve passed the 1/12th mark? The 1/6th? the 1/4th? Maybe not as easily as you would reading Apology, if you could figure out that there were about 100 lines to each 1/12th.

About 100 lines. The funny thing is that the margins of that “about” is are narrower for the Timaeus than for either the Apology or Symposium. 3,486 is closer to 3,600 than 1,251 is to 1,200 or 2,478 to 2,400. The last two just look so similar. At least they look so similar in Arabic numerals, but not in Greek numerals where ͵γυπϛ is closer to ͵γχ than ͵ασνα is to ͵ασ (the Greeks were great at geometry but partly due to their number system not so good with arithmetic). Any number between 1,201 and 1,299 looks close to 1,200, and any number between 2,401 and 2,499 looks close to 2,400. What are the chances? If you pick a random number below 1,300, you have better than an 8% chance of it looking close to 1,200. I’d be interested to see how the figures for all Plato’s works stand up.

Thank you Greek Number Converter.