Sunday, April 18, 2010


Memiyawanzi (with some help from Thomas Lambdin) makes an excellent point about Bible translations and its usefulness for comparative syntactic analysis which boils down to this: in many instances, Bible translators, regardless of language, quite often slavishly imitate the syntax of the original.

This is true of nearly all translations of sacred texts made before the rise of linguistics in general and translation studies in particular, but even here, there are more extreme cases. One of those is šarḥ, the translation of sacred texts of Judaism into Judeo-Arabic. Ironically enough, the original meaning of the root šrḥ is "to explain, to interpret", but the šarḥ translations are anything but that. Quite the contrary - their language emulates the syntax of the original as closely as possible. In order to do so, the translators - šarḥanim - have gone so far as to introduce new grammatical features to their target language.

One example of such feature is the use of the preposition الي [ilā] to translate the Hebrew direct object marker את or the Aramaic direct object marker ית. Old Arabic (including Quranic Arabic and Classical Arabic) marks the direct object by means of the suffix -a, while Neo-Arabic (which includes the modern colloquial varieties, but also varieties employed by the Arabic-speaking Jewish population of Middle East and North Africa) normally marks the direct object by position. There are exceptions to this, such as Maltese, Cypriot Maronite Arabic and some Syro-Palestinian dialects [1], which use some variant of the preposition l-. In šarḥ Arabic, however, the direct object is marked using the preposition אלא [ilā], which is identical in function to Hebrew את or Aramaic ית. Consider the following example from the Targum to Canticles 3:5:

Targumic Aramaic:

כד שמעו שבעת עממיא דבני ישראל עתידין למחסן ית ארעהון קמו כחדא וקציצו ית אילניא וסתימו ית מבועי מיא וצדיאו קרויהון וערקו

English translation by Jay C. Treat:

When the seven nations heard that the Children of Israel were about to take possession of their land, they rose at once and cut the trees, stopped up the water springs, laid waste their towns and fled.

Judeo-Arabic translation (Iraq, 19th century):

לָמִּן סַמְעוּ סַבִע אֶל אוּמָם אַן בִנִין יִסְרָאִיל מִתְּוּובִדִין ליִוּורְתֹוֹן אֶלָא בִלַדְהוֹם קָאמוּ גִֹמִיעָא וּקַצוּ אֶלא אֶל סִגַֹר וּסַדּוּ אֶלָא מִנָאבִע אֶל מָאיי וכַֹרִבוּ אֶלָא קִרְיָיאתְּהוֹם ואִנְהַזְמוּ

Or to highlight the phrases in question:

Targumic Aramaic



למחסן ית ארעהו

ליִוּורְתֹוֹן אֶלָא בִלַדְהוֹם

take possession of their land

וקציצו ית אילניא

וּקַצוּ אֶלא אֶל סִגַֹר

cut the trees

וסתימו ית מבועי מיא

וּסַדּוּ אֶלָא מִנָאבִע אֶל

stopped up the water springs

וצדיאו קרויהון

וכַֹרִבוּ אֶלָא קִרְיָיאתְּהוֹם

laid waste their towns

And so while in most varieties of colloquial Iraqi Arabic, one would normally render these structures - V + OBJ.M + N - as Verb + Noun, in šarḥ Arabic, the translator feels compelled to produce a verbatim translation and thus translates the semantically empty direct object marker by repurposing the directional preposition אֶלָא [elā].

There are many examples of this in translations from Hebrew and Aramaic, not only in Judeo-Arabic, but also in other Jewish languages, such as Jewish Neo-Aramaic, Judeo-Spanish [2] or Judeo-Persian:

Ruth 4:11 Hebrew:
... יתן יהוה את־האשה הבאה אל־ביתך כרחל ׀ וכלאה אשר בנו שתיהם את־בית ישראל ...

Ruth 4:11 English (NASB):

... May the LORD make the woman who is coming into your home like Rachel and Leah, both of whom built the house of Israel; ...

Ruth 4:11 Judeo-Persian [3]:

... בי דהד כודא מר אן זן אנקי אייא בכאנה תו קון רחל וקון לאה אנקי אודאן כרדנד הר דואן אישאן מר כאנדאן ישראל ...

But unlike in those languages, where the new direct object marker either expanded previous usage (as in Judeo-Spanish אה [a]) or redefined its role both historically and sociolinguistically (as in Judeo-Persian מר [mar] which is normally only found in classical Persian poetry), the Judeo-Arabic repurposing of אלא is a completely different game. Not only did the šarḥanim take a completely innocent preposition and turned it into something completely different, but consider the fourth example from Targum Canticles 3:5:

Targumic Aramaic



וצדיאו קרויהון

וכַֹרִבוּ אֶלָא קִרְיָיאתְּהוֹם

laid waste their towns

So while the Aramaic original does not require the direct object marker, its use has become obligatory in written Judeo-Arabic. As Benjamin Hary notes in his Translating Religion: "... šarḥ created its own Judeo-Arabic grammar and structure" [4].

And I'm thinking: isn't that true, at least in terms of syntax, for every Bible translation and, by extension, of all languages that have been fundamentally influenced by translations of sacred scriptures? How different, I wonder, were real spoken Syriac or Coptic from their varieties recorded in Christian translations and writings?

[1] Borg 2004:46
[2] Hary 1991:605-606
[3] Mainz 1976:21
[4] Hary 2009:165


BORG, Alexander: A Comparative Glossary of Cypriot Maronite Arabic (Arabic-English). With an Introductory Essay. - Ledein: Brill, 2004, xxviii + 486 p.
HARY, Benjamin: On the use of 'ila and li in Judeo-Arabic texts. Pages 595-608 in: KAYE, Alan S.: Semitic studies in honor of Wolf LESLAU on the ocassion of his 85th birthday, November 14th, 1991. Volume I. - Wiesbaden: Harrasowitz, 1991, lxviii + 889 p.
HARY, Benjamin: Translating Religion. Linguistic Analysis of Judeo-Arabic Sacred Texts from Egypt. - Leiden: Brill, 2009, 360 p.
MAINZ, Ernest: Ruth et le Cantique des Cantiques en judéo-persan. Journal Asiatique, 264/1-2, 1976, pp. 9-34
Sefer šir ha-širim ʿim targum ve-šarḥ arvi. Baġdād, 1936/37