Betrayal of a Poet and His Greek.

A couple of years ago, at the end of a favorable review of Boris Gasparov’s Beyond Pure Reason, I complained bitterly about “the terrible proofreading and editing of this important scholarly book”; I am happy to see that the TLS has given over an entire page (subscriber-only, I’m afraid) to a similar complaint by the great Robert Bringhurst (who graced this LH thread with his comments) about the way the University of California Press has treated the poet Robert Duncan in its collected edition of his work:

There are, if I’ve counted correctly, just over 200 words of Greek scattered through the two volumes [The Collected Early Poems and Plays and The Collected Later Poems and Plays], and in those words alone I count 130 typographic errors. The errors are new and original, not copied from Duncan’s manuscripts or from previous editions of work. They are also persistently inventive. When a Greek word or phrase is repeated in the text or the editor’s notes, it is misspelled differently each time. The errors stand out in books that are otherwise handsomely produced (typographic errors outside the Greek are mercifully rare). This beats by a huge margin everything else in Duncan’s own error-infested publication history. Even the famously error-prone New Directions editions of Pound’s Cantos come nowhere close. The Greek passages in The Cantos average a mere twenty errors per hundred words. There are well over sixty per hundred in the Duncan.

The errors caught me especially by surprise. The volumes in question are edited by Peter Quartermain, a first-class scholar of modern American literature. When he began, Quartermain sought help from a number of specialists and from several of Duncan’s old colleagues. I was asked to check and correct all of the Greek quotations and the few bits of Chinese. I identified the sources, fixed the errant spellings and provided translations (mostly superfluous, since Duncan included English glosses for almost all his Greek phrases and Chinese glyphs). Knowing that UC Press had no classically trained compositors or editors on staff, I also set all the Greek and Chinese quotations in Unicode-compliant digital type, so that each such passage could be electronically copied and pasted into place by whatever typesetting firm was given the job. I was unaware until recently that when Quartermain presented his vast manuscript to the Press and offered them the corrected digital Greek, they turned him down, offering to handle the work themselves.

Most of the errors in the books could only have been made by someone altogether innocent of Greek. Letters that can only be used at the beginnings or ends of Greek words are placed in the middle, letters with similar shapes are confused with one another, and the diacritics that abound in classical Greek are freely interchanged or replaced with diacritics borrowed from Latin or Cyrillic. The Greek letter nu (ν), corresponding to Latin n, caused particular confusion. The typesetter sometimes replaced it with upsilon (υ), the Greek form of u, but more often with izhitsa (ѵ), a Cyrillic letter formerly used in Russian but now employed only, so far as I know, in liturgical Slavonic.

He sums up by saying “Duncan’s dream of a multidimensional poetry of all poetries has been reduced to alphabet soup.” What a shame! It reminds me the mess Google Books has made of metadata (see this LH post) due to a similar we-don’t-need-your-help hubris.

To wash the bad taste out of your mouth, a nice map: Countries as Named in Their Own Languages.


  1. I’ve been thinking about buying the New Directions hardback version of Pound’s Cantos. Is there a recent printing where these infamous errors have been corrected?

  2. I don’t think so, but I haven’t kept up with Pound studies lately. Good question.

  3. Athel Cornish-Bowden says

    Most of the errors in the books could only have been made by someone altogether innocent of Greek. Letters that can only be used at the beginnings or ends of Greek words are placed in the middle, letters with similar shapes are confused with one another …

    Well maybe, but sometimes a little knowledge can be worse than being altogether innocent of it. I have a book by a distinguished chemist (now deceased) that purports to give the Greek origins of some of the terms. I don’t know how much Greek the author knew, but anyway I’m inclined to blame the publisher (Oxford University Press, which should be deeply ashamed) as he was probably too old (about 90) to do much proofreading. Someone had obviously told the sub-editor that a Greek word couldn’t end with σ, and had probably said to use ς instead, but unfortunately the book used ζ instead, which (to my eyes, anyway) looks much worse at the end of a word than σ. Most of the Greek words had errors (and no diacritics), but the ζ for ς shouted out at me, especially as there are lots of Greek words that end with ς.

    Also, although your map is indeed nice, I’m surprised that its creator ignored the facts that they have other languages apart from German in Switzerland, and others apart from Flemish in Belgium. Maybe the lack of enough space forced a choice of using the majority language in each case. Unfortunately I couldn’t check this hypothesis as the page crashed my browser and I didn’t want to risk trying again.

  4. but unfortunately the book used ζ instead

    This is depressingly common. I don’t know why publishers don’t seem to feel the slightest need to have someone who knows Greek glance at books that include Greek.

  5. AJP "Helge" Avisa says

    Maybe the lack of enough space forced a choice of using the majority language in each case.

    It did. There followed a long thread of angry comments about whether the Falklands ought to be Malvinas, or Ireland Eire etc. Majority language was the decider. I object to The United Kingdom of Great Britain & Northern Ireland. It (or they) can be listed under B for Britain, G for Great, occasionally E for England, or U for United. Anytime I have to fill in an internet form, trying to find where it’s hiding wastes literally seconds.

  6. Occasionally it’s S for Storbritannia (Norwegian airlines).

  7. Or Sassenach.

  8. The name for Singapore was Malay. While Malay is the official language, it really should have been English on the map.

  9. David Marjanović says

    Graece est, non legitur.

  10. David Marjanović says

    it really should have been English on the map

    …It is, at least now.

  11. You’re right. I had trouble picking it out.

  12. With regard to the book, University of California Press has brought a whole new depth of meaning to the expression ‘It’s all Greek to me”.

  13. ‘Sassenach’ is a mildly offensive Scots Gaelic term for a ‘Saxon’ or Englishman, as I’m sure our host knows. It would be just as logical to use Y (for Yankee) for the USA.

  14. ‘Twere just a little joke.

Speak Your Mind