Ibn Ezra's girsa of velo yishama al picha

Summary: Ibn Ezra has a variant girsa of a pasuk in parshat Mishpatim, which lines up with a variant mentioned by the Aruch. It turns out it also matches the Samaritan Pentateuch. Yet I would still side with our Masoretic text.

Post: In parashat Mishpatim, we read {Shemot 23:13}:

13. Concerning all that I have said to you you shall beware, and the name of the gods of others you shall not mention; it shall not be heard through your mouth.

יג. וּבְכֹל אֲשֶׁר אָמַרְתִּי אֲלֵיכֶם תִּשָּׁמֵרוּ וְשֵׁם אֱ־לֹהִים אֲחֵרִים לֹא תַזְכִּירוּ לֹא יִשָּׁמַע עַל פִּיךָ:

This is slightly strange, in that we have the beginning of a new statement in the middle of the pasuk: לֹא יִשָּׁמַע עַל פִּיךָ. And the break is not even at an etnachta, but at a zakef katon. It echoes "לֹא תַזְכִּירוּ" in a poetic manner. Is is clearly wrong? I wouldn't say so, and indeed think it works out rather nicely due to the poetic parallelism.

Still, it might strike the casual reader as somewhat awkward.

This present no problem for Ibn Ezra, who seems to have a variant text of the pasuk:

[כג, יג]
ובכל אשר -עתה הזכיר עבודת כוכבים, והטעם כל מה שאמרתי הם מצותי ומשפטי. ולא כן. משפטי אלוהים אחרים. 
ואמר: לא תזכירו - שלא תזכירו שמותם להישבע בהם גם שלא ישביעו בהם עובדיה. וזה ולא ישמע על פיך שתשביע בו אחרים. שלא תאמר לעובד כוכבים השבע לי באלהיך.

Note that he cites the pasuk with a leading vav. This makes a great halachic difference in our sifrei Torah. We say anan lo bekein bimleiot vachaseirot, that we are not expert in the plene and deficient spelling of words, and such would not invalidate a sefer Torah. Yet to add an extra vav, or in our case, to omit a vav, which is not a vowel letter but a conjunction "and" would have much more halachic significance.

Minchat Shai takes note of this Ibn Ezra:

"In the commentary of Ibn Ezra is written, 'and this is ולא ישמע על פיך, etc.' And the Aruch, entry מטר {? I didn't see it there, so maybe it is a different entry I cannot read}, that until recent generations to now, they would mess up and read ולא ישמע על פיך, and the soferim taught not to read with a vav, end quote. Furthermore, in the masoret, three one would expect ולא and they are לא, and the mnemonic of לא תשימון עליו נשך, and לא ישמע על פיך, and לא תכיר פנים."

I am unsure whether by "read" they meant that Torah texts had it. At any rate, this is what the Samaritans have in their Pentateuch:

They don't only change לא to ולא, but also change the plurality of ישמע to ישמעו and, strangely, תזכירו to the singular תזכר. If I had to guess, these were separate changes trying to make each match the plurality of the other, which crossed over.

(At the bottom of the same page of the above-linked Vetus Testamentum, see a list of Jewish texts which have ולא.)

At any rate, I would not pay any heed to ולא in the Samaritan Pentateuch and how it matches Ibn Ezra. It does not need to reflect the same girsa. Rather, as was spelled out in the masoretic note, one would expect ולא here, and so the Samaritan scribes "fixed" the problem to make the text more fluid. The same awkwardness would inspire the error mentioned in the Aruch and manifest in Ibn Ezra. This is lectio difficilior at play.

In passing, I'll note that Onkelos is like the masoretic text we have before us:

כג,יג וּבְכֹל אֲשֶׁר-אָמַרְתִּי אֲלֵיכֶם, תִּשָּׁמֵרוּ; וְשֵׁם אֱלֹהִים אֲחֵרִים לֹא תַזְכִּירוּ, לֹא יִשָּׁמַע עַל-פִּיךָ.

וּבְכֹל דַּאֲמַרִית לְכוֹן, תִּסְתַּמְרוּן; וְשׁוֹם טָעֲוָת עַמְמַיָּא לָא תִּדְכְרוּן, לָא יִשְׁתְּמַע עַל פֻּמְּכוֹן.

and that so is the Peshitta: