Found in Translation
« previous post | next post »
"Found in Translation" is the latest in the PBS video series Articulate, exploring how "scholarly translations are a constant battle between literal accuracy and literary interpretation."
Connoisseurs of automated translation follies will appreciate the bit about 4 minutes in when Peter Cole copies some Hebrew into Google Translate and gets the English output, "ui on Lbbc stopped Heather." ("Lbbc" is in the Hebrew original for some reason, so this might not be the best translational task.)
(h/t Grant Barrett)
Yuval said,
May 9, 2017 @ 3:59 pm
The ui is also in the input, and the Hebrew words are translated perfectly.
Owlmirror said,
May 9, 2017 @ 5:16 pm
Is it just the video that I watched, or does everyone else see that the Hebrew that appears about 20 seconds in is reversed, that is, left-to-right rather than right-to-left?
If it's meant to be a translation of the English that starts the sequence, I think the word order is wonky as well ("מילות תרגום" should be "תרגום מילות") –although maybe that's supposed to be the point?
Owlmirror said,
May 9, 2017 @ 5:24 pm
If I had to guess, it could be that ui and Lbbc are OCR errors. That is, an OCR program that is supposed to recognize a mixed English and Hebrew text failed to recognize which language a few words were in, and passed the image of the text to the wrong sub-engine, which then made a hash of the text in a language it's not meant to recognize.
I watched this happen just recently, with the OCR software working on a text mostly in English with Greek and Hebrew words in as well.
Yerushalmi said,
May 10, 2017 @ 2:41 am
The sequence is a little bit strange. At 3:58 he is clearly scrolling through the text of the Torah – specifically, Exodus, with both nikkud and interleaved transliteration and translation into English – and yet what he pastes into Google Translate recognizably has no nikkud and is not Biblical Hebrew.
The presence of the word הת'ר (Heather) is the key. Biblical Hebrew didn't have the modern convention of using apostrophes after ת and/or ד to represent θ and ð (not least because Biblical Hebrew, unlike modern Hebrew, probably had θ and ð sounds natively). And the name Heather definitely doesn't appear in the Torah.
So either the entire sequence, including the pasting of essentially random letters into Google Translate, was scripted for dramatic effect — or this is a spectacular failure of OCR even worse than that proposed by Owlmirror.
Bathrobe said,
May 11, 2017 @ 12:58 am
It seems to me that "scholarly translations" and "literary translations" are two different things.
A "literary" translator will try to recreate the work of art in a different language. A "scholarly" translator will try to come as close as possible to the original, including, for instance, annotations.
Ideally there will be middle ground between the two but I think they are essentially distinct approaches to translation. The PBS video seems to have a bias towards the literary side; hence the tendency to be wishy-washy.
For example, the opening sequence which tells us that the Bible did not say "Yeah though I walk through the shadow of the valley of death", but does not make any attempt to enlighten us on this tantalising snippet. The general thrust seems to be that translators should try to "recreate the music". This is fine for a video targeted at the general public, who, I'm afraid, generally know zilch about translation, but it's definitely not scholarly.
BZ said,
May 11, 2017 @ 12:10 pm
I think translating the Bible is a particular challenge because you have competing interests in religious belief, changing language, and tradition. So, for those for who believe that KJV (or the Septuagint for that matter) is divinely inspired, how you separate that from the fact that the archaic wording is not understood by the modern audience? Do you need to translate KJV into modern English? What would that even look like (as opposed to re-translating the Hebrew or Greek)?
Take for example the word "Lord". Sure, everyone thinks they understand it, but do (human) lords exist in the majority of the English-speaking world? And if they do, do they wield the same power they did when the translation was made? And then, I've read in the preface to one Jewish translation that the word "Lord" as referring to God has become too associated with Christianity to be used in a Jewish translation. And it's true, but if it's truly the best translation, how do you get past that? What would it look like if the Bible were a new document never before seen, being translated today? Would it be "I'm the President your CEO, Governor of the multiverse"? Obviously no one would suggest that as a translation given the history, but where is the middle ground and who will accept what?
Owlmirror said,
May 14, 2017 @ 2:54 am
I found the Hebrew wiktionary on the term "צלמות", which suggests that while the understanding that it means "shadow of death" is ancient, it is proposed that it is actually a Semitic term whose root is "צ־ל־מ", and means less poetically "gloom, deep darkness", with no immediate connotation of death.
The wiktionary had no references, but Google Scholar did find, "On The Meaning Of Salmawet", by Giovanni Mazzini, who also references:
W.L. Michel, SLMWT, "Deep Darkness" or "Shadow of Death" , BR 29,1984,5-20, and a few works that reference the latter.
Owlmirror said,
May 14, 2017 @ 3:07 am
[This may duplicate a comment in the moderation queue]
@Yerushalmi:
Incidentally, the text he is scrolling through is this specific PDF of Exodus, available online:
http://www.scripture4all.org/OnlineInterlinear/OTpdf/exo1.pdf
Owlmirror said,
May 14, 2017 @ 9:29 am
And one more reference for the "shadow of death" question:
Thomas, D. Winton. "Tsalmavet in the Old Testament." Journal of Semitic Studies 7 (1962): 191.