Reading Matthew in Hebrew - II

Delitzsch:

אֶל־דֶּרֵךְ הַגּוֹיִם אַל־תֵּלֵכוּ וְאֶל־עִיר הַשֹׁמְוֹנִים אַל־תָּבֹאוּ ׃

כִּי אִם־לְכוּ אֶל־הַצֹּאן הָאֹבְדוֹת לבֵית יִשְׂרָאֵל ׃

Salkinson:

אַל־תָּשִׂימוּ לְדֶרֶךְ הַגּוֹיִם פַּעֲמֵיכֶם וְאֶל־עָרֵי הַשֹׁמְרֹנִים אַל־תָּבֹאוּ ׃

כִּי אִם־לְצֹאן אֹבְדוֹת מִבֵּית יִשְׂרָאֵל תֵּלֵכוּ ׃‎

Shaprut:

בארצות הגוים אל תלכו ובערי השמרונים אל תבואו ׃

לכו לצעאן אשר נדחו מבית ישראל ׃

Please refer to my previous blog post for more background information.

We can pick up the thread here:

Shaprut [lived in Spain in the fourteenth century, but his text is of a type that is much older than that, pre-Vulgate (ca. 400) and pre-Peshitta (ca. 500), and probably even older]:

"Do not go to the lands of the Gentiles, and into the cities of the Samaritans do not enter. Go to the sheep that have strayed from the house of Israel."

First of all, outside of Hebrew Matthew, the verb "strayed" is found only in the Syriac texts (Old Syriac and Peshitta). Both Delitzsch and Salkinson followed the Greek texts in using "lost."

Secondly, it is worth noting that the combination "lands/cities" forms a parallelism, one of the most characteristic features of all Semitic literature. Such parallelisms, along with frequent puns and other wordplay that only work in a Semitic language, have led many to believe that there is a Semitic substratum in parts of the New Testament, notably the Gospel of Matthew, especially in the "sayings" portions. For a parallelism to be stylistically perfect, both of its members should have the same number (singular or plural). That is true here, and they are both plural: "lands/cities." Now "cities" (plural) is quite rare in this passage among all extant textual witnesses. Every Greek text has "road/city" (singular). But the plural "cities" is supported by the Old Syriac (only two manuscripts have survived), the Peshitta (the newer, standard Syriac version), some of the Old Latin manuscripts, and the Vulgate. No recent translation reads "cities," because they have all been made to agree with the Greek texts. As to "lands," I haven't yet found support for that reading, but the parallelism requires a plural.

What is the significance of this situation? In this case, at least, it opens up the possibility that Shaprut's Hebrew Matthew may be very old. If agreement with the Old Syriac and Old Latin is frequent (it is, and is in fact the most frequent among supported variant readings), then we have a text of a very old type.

But we can also consider the possibility that Shaprut's reading with "lands/cities" may be more original than the Greek textual tradition, just on the basis of the likelihood of the reading. Did the Samaritans have more than one city? No doubt. What is "the road" of the Gentiles? Is not "the lands" of the Gentiles more likely? Was there any historical, political, or social reason to suppress such a reading?

(to be continued)