Sunday, May 6, 2007

Joel Spolsky & T S Eliot

Came across an interview of Joel Spolsky on Scoble about a month after the event. Spolsky runs Fog Creek Software, which produces Fogbugz, a program for debugging software, and Copilot, a program that allows you to give tech support by taking over someone's computer and working on it from a remote location.

Copilot works even if both parties are behind firewalls, and it allows for Mac-Windows remote assistance as well as Mac-Mac and Windows-Windows. This means, I think, that it could save publishers from problems like this:

Mr. Eliot's Sunday Morning Service, 2nd stanza, first from Bartleby's with some Greek in the Greek alphabet--

In the beginning was the Word. 5
Superfetation of τὸ ἒν,
And at the mensual turn of time
Produced enervate Origen.

then with the Greek transliterated, from Representative Poetry Online, a poetry site run by the University of Toronto--

5 In the beginning was the Word.
6 Superfetation of to en,
7 And at the mensual turn of time
8 Produced enervate Origen.

The Toronto site adds the helpful comment: Eliot's Greek letters are transliterated and italicized in this edition. The Greek words mean: "the One."

We-e-e-e-e-e-e-e-l-l-l-l-l. Up to a point.

The sound that in English is normally represented by h does not have a letter in ancient Greek. It's written as a fiddly little inward-curving hook above the vowel. Any word that begins with an aspirated vowel (a rough breathing) will have this fiddly hook above the letter. Sadly for the modern typesetter, any word that begins with an unaspirated vowel also has a fiddly little hook -- a fiddly little hook facing the other way.

So if we were writing English words in the Greek alphabet, they'd look like this:

You'll notice just how big the Greek has to be for the direction of the hook to be clear. (Note also that the breathing goes over the second vowel of a two-vowel syllable.) And a syllable may also bear an accent:

The feminine singular of 'one' is μία, and the neuter singular is either

before another word, or

at the end of a sense unit, usually indicated by a punctuation mark.

Long story short, the Greek for 'the One' is either τὸ ἕν or τὸ ἓν, which should be transliterated as, yes, you guessed it, to hen. If the comma at the end of the line is correct, both the accent and the breathing are wrong. And, to get back to extremely fabulous Mr Spolsky, if the typesetter had had a Hellenist with Copilot at the other end of the line, the word could have been printed right for the princely sum of $4.95, and we would not have had misinformation disseminated to millions of Internet users, to the effect that 'to en' is Greek for 'the One'.

It would be silly for a production manager to trawl the pool of typesetters for one who happened to know Greek because a text happened to include a couple of words in Greek. It would not be silly to lay out $4.95 to get someone competent in the language to input the couple of words. The alternative is for the hapless typesetter to have a bash and then fax the page through to the author for proofreading. The fiddly bits that are hardest to get right are precisely the bits that will be illegible in a fax. Copilot, thou shouldst have been living at that hour.

Information on Copilot is available at


Languagehat said...

You might want to fix the Greek, as I just did in the Wikipedia entry -- it should be ἓν διὰ δυοῖν. (I had to hit control-plus three times to up the font size to where I could be absolutely sure I had the breathing right!)

Helen DeWitt said...

You're right of course about the Greek -- and how could I possibly forget this nice example of 2 in the dual? (I am impressed by your restraint in making no comment on dissemination of misinformation to millions of Internet users...) The only thing is, if I'd remembered that the Greek, unlike the late Latin form, used this nice example of the dual, I wouldn't have put it in, because it wouldn't have done what it was supposed to -- provide something a reader new to Greek might recognise if deciphering correctly. I don't think one can reasonably throw such a reader "hen dia duoin" without explaining the dual, and, erm, I think if an example that requires explaining the dual is forced on the reader at this early stage it will confirm all his/her worst fears about the difficulty of the language.

Still, it won't do to disseminate misinformation to millions of Internet users, or even to disseminate misinformation to the tiny band of visitors to this blog, so I have taken hendiadys out altogether -- which does make this exchange of commments incomprehensible to latecomers, but it's better than disseminating misinformation.

This sort of careless mistake is exactly the reason, though, that I would like to see some of the technical obstacles taken out of the typesetting process. Sorting out this post involved a lot of fiddly behind-the-scenes biz, typing in Mellel to get a really big font, taking screenshots, uploading at the right image size, the sort of work that's not unlike what one goes through in the months it takes to get text correctly set using a typesetter, and the part of the brain that should be rejecting hen dia dys as ungrammatical and looking for a workable substitute just goes to sleep.

My friend Guy Deutscher, Assyriologist, linguist and author of The Evolution of Language, said he decided not to have ANY non-roman scripts in the book because he knew what a nightmare it would be to get them into print. This was a book for the general reader; it seems hard on the general reader if a rare chance to see Akkadian in its native cuneiform is taken away because of the bloodymindedness of typesetters.

Languagehat said...

Dissemination of misinformation? Bah, I've done that enough times not to dream of criticizing it in others. The wonderful thing about blogs is that you can edit them and fix the problem! It's those damn old-fashioned books that just sit there misinforming people for years...

I love Deutscher's book and am saddened to hear its style was cramped by such a thing. In this day and age, that shouldn't even be a concern.

Helen DeWitt said...

(Correction. Guy's book is called The Unfolding of Language.)