Is there a vav in וְעַל הַכֹּהֲנִים?

Summary: Yes and no. Is there some way of deciding between these competing masorahs?

Post: In shlishi of parashat Acharei Mot, in Vayikra 16:33, we encounter this pasuk:

33. And he shall effect atonement upon the Holy of Holies, and he shall effect atonement upon the Tent of Meeting and upon the altar, and he shall effect atonement upon the kohanim and upon all the people of the congregation.

לג. וְכִפֶּר אֶת מִקְדַּשׁ הַקֹּדֶשׁ וְאֶת אֹהֶל מוֹעֵד וְאֶת הַמִּזְבֵּחַ יְכַפֵּר וְעַל הַכֹּהֲנִים וְעַל כָּל עַם הַקָּהָל יְכַפֵּר:

Note the words וְעַל הַכֹּהֲנִים, and particularly the leading vav in וְעַל. I don't think there was any possibility of taking the mizbei'ach as the thing effecting atonement upon the kohanim etc., even if the vav was not present, because of the et object marker on hamizbeach.

If we wanted, for some reason, to drop the vav, we certainly could. From the perspective of trup, we have:

Note the etnachta on yechaper. This ends the first half of the pasuk. We could then treat it as a semicolon, and have the second half of the pasuk read: "upon the kohanim and upon all the people of the congregation he shall effect atonement." It would be admittedly irregular, but not impossible, and indeed we have just this sort of form elsewhere.

Why am I discussing this theoretical variant? Because it is not so theoretical. The Samaritans have it, and Minchas Shai also discusses a competing masorah that has it.

In terms of the Samaritans:

And on the same page of Vetus Testamentum, a list of Jewish texts which also omit the leading vav:

Relatively few, but still, they exist. All else being equal, I would have expected the Samaritans to insert a vav in this context rather than delete it. This then goes against the grain, which recommends this particular reading. However, here is where the Samaritan harmonizing trend kicks in. Apparently, as we will see from Minchas Shai's words in a bit, וְעַל הַכֹּהֲנִים and  וְעַל הַכֹּהֲנִים appear a sum total of four times in Tanach. And in the other three times, it is just al. Making the phrase consistent across Tanach is something we might well expect a Samaritan scribe to do.

Minchas Shai writes about this:

"In the Masorah Gedolah, maarechet, #20, עַל הַכֹּהֲנִים, there are four. And the mnemonic is וְכִפֶּר אֶת מִקְדַּשׁ הַקֹּדֶשׁ of here, and II Divrei Hayamim 8:14, וְלֹא סָרוּ מִצְוַת הַמֶּלֶךְ, and II Divrei Hayamim 31:9, וַיִּדְרֹשׁ יְחִזְקִיָּהוּ, and Nechemia 12:44, וַיִּפָּקְדוּ בַיּוֹם הַהוּא. End quote. And upon this masorah, many have been confounded. For in all of our sefarim, this pasuk of וְכִפֶּר אֶת מִקְדַּשׁ הַקֹּדֶשׁ has the word וְעַל written with a vav! However, I will tell you that which I found inscribed in a true writing, in a manuscript masorah in an early chumash, and this is it words:


עַל הַכֹּהֲנִים -- there are three, and their sign is וַיִּפָּקְדוּ, and וַיִּדְרֹשׁ, and וְלֹא סָרוּ. And there is one instance of  וְעַל הַכֹּהֲנִים, in וְכִפֶּר אֶת מִקְדַּשׁ הַקֹּדֶשׁ. And there is none like it juxtaposed. And those in the East argue on it, for they say עַל הַכֹּהֲנִים, and in their masoret is four. End quote.


And all questions, with this, are resolves, for in variant texts, we follow the residents of the West {=Eretz Yisrael}, and the mnemonic for this rule is: Ki Mitzion Teitzei Torah, as a sign that there is no Torah like the Torah of Eretz Yisrael."

My take on this: I agree with Minchas Shai in his conclusion, though not for his reason. Rather, a would judiciously apply the idea that maximizing entropy in text rather than smoothing it out will most likely yield the correct original text. And so, that this instance of  וְעַל הַכֹּהֲנִים is a hapax legomenon, while עַל הַכֹּהֲנִים is not, makes me prefer the hapax. Still, what gives me some minimal level of pause is the slight irregularity in not having the connective vav there.