Do Chazal darshen the Samaritan version of Vayikra 20:7?

Summary: In parashat Kedoshim, yet another instance in which a derashat Chazal matches the Samaritan text of the Chumash instead of our Masoretic text. In this instance, however, it is somewhat plausible that Chazal are merely darshening the union of two similar pesukim; and that the Samaritans, as is their wont, harmonized the two similar pesukim. Still, after considering Minchas Shai, Gra, a few suggestions of my own, and considering Talmudic variants, I conclude that Chazal were once again darshening a non-Masoretic text.


Post: In parashat Kedoshim, in Vayikra 20:7, we encounter the pasuk:

ז  וְהִתְקַדִּשְׁתֶּם--וִהְיִיתֶם, קְדֹשִׁים:  כִּי אֲנִי יְהוָה, אֱלֹהֵיכֶם.

7 Sanctify yourselves therefore, and be ye holy; for I am the LORD your God.

There is a similar pasuk in the beginning of Kedoshim, in Vayikra 19:2:

ב  דַּבֵּר אֶל-כָּל-עֲדַת בְּנֵי-יִשְׂרָאֵל, וְאָמַרְתָּ אֲלֵהֶם--קְדֹשִׁים תִּהְיוּ:  כִּי קָדוֹשׁ, אֲנִי יְהוָה אֱלֹהֵיכֶם.

2 Speak unto all the congregation of the children of Israel, and say unto them: Ye shall be holy; for I the LORD your God am holy.

But the much closer match to Vayikra 20:7 is Vayira 11:44, in parashar Shemini:

מד  כִּי אֲנִי יְהוָה, אֱלֹהֵיכֶם, וְהִתְקַדִּשְׁתֶּם וִהְיִיתֶם קְדֹשִׁים, כִּי קָדוֹשׁ אָנִי; וְלֹא תְטַמְּאוּ אֶת-נַפְשֹׁתֵיכֶם, בְּכָל-הַשֶּׁרֶץ הָרֹמֵשׂ עַל-הָאָרֶץ.

44 For I am the LORD your God; sanctify yourselves therefore, and be ye holy; for I am holy; neither shall ye defile yourselves with any manner of swarming thing that moveth upon the earth.

Minchas Shai notes that in Berachot 53b, at the very end of the eighth perek, the gemara states:

דאמר רב יהודה אמר רב ואמרי לה במתניתא תנא (ויקרא כ) והתקדשתם אלו מים ראשונים והייתם קדושים אלו מים אחרונים כי קדוש זה שמן אני יי' אלהיכם זו ברכה: 

Or in English:
Rab Judah said in the name of Rab: some say it was taught in a Baraitha, Sanctify yourselves:  this refers to washing of the hands before the meal; And be ye holy: this refers to washing of the hands after the meal; 'For holy': this refers to the oil; 'Am I the Lord your God': this refers to the grace.

Thus,

Minchat Shai suggests that there is no one pasuk in play, but rather it is a combination of multiple, similar pesukim.

Why say this? Well, they darshen the following elements of the "pasuk", in the following order::

(A) והתקדשתם
(B) והייתם קדושים
(C) כי קדוש
(D) אני יי' אלהיכם

Meanwhile, Vayikra 20:7 has:
(A) וְהִתְקַדִּשְׁתֶּם
(B) וִהְיִיתֶם, קְדֹשִׁים
(D) כִּי אֲנִי יְהוָה, אֱלֹהֵיכֶם

But no (C), ki kadosh. It does have the ki of ki kadosh, but this is irrelevant until we see how the Samaritans have it.

Vayikra 19:2 has hardly any of those elements, While the message is similar in terms of Kedoshim Tihyu, the only phrases it has verbatim are:
(C) כִּי קָדוֹשׁ
(D) אֲנִי יְהוָה אֱלֹהֵיכֶם

Vayikra 11:44 has:
(D) כִּי אֲנִי יְהוָה, אֱלֹהֵיכֶם
(A) וְהִתְקַדִּשְׁתֶּם
(B) וִהְיִיתֶם קְדֹשִׁים
(C) כִּי קָדוֹשׁ אָנִי

This has all the same elements! But, assuming the girsa in our gemara is correct, there is quite a mismatch. (D) is cited and darshened last, rather than first. And the gemara does not have ki as part of (D). And the gemara does not have ani as part of (C). The latter two objections are not that serious, at first glance. After all, the brayta may only be citing the part of each phrase it needs. But likely, that (D) is out of order and that they can be found in the right order (minus one element) driveds Minchas Shai to look to 20:7, and then to other pseukim  to get element (C).

The next pasuk, 11:45, reads:

מה  כִּי אֲנִי יְהוָה, הַמַּעֲלֶה אֶתְכֶם מֵאֶרֶץ מִצְרַיִם, לִהְיֹת לָכֶם, לֵאלֹהִים; וִהְיִיתֶם קְדֹשִׁים, כִּי קָדוֹשׁ אָנִי.

45 For I am the LORD that brought you up out of the land of Egypt, to be your God; ye shall therefore be holy, for I am holy.

And so it has:
(D) כִּי אֲנִי יְהוָה
(B)  וִהְיִיתֶם קְדֹשִׁים
(C) כִּי קָדוֹשׁ אָנִי

These, too, are not precise matches -- (D) misses elokeichem and has an additional ki, and (C) has an extra ani. Furthermore, it is missing element (A). And, it is out of order.

The Gra has a rather ingenious suggestion, in Hagahos HaGra, as well as in Aderes Eliyahu on parashat Shemini. From the prooftext of the last, based on ani Hashem Elokeichem, he removes the words Hashem Elokeichem. Thus:

Thus, to see the gemara again, he would have the brayta, but with the text marked in red removed:

דאמר רב יהודה אמר רב ואמרי לה במתניתא תנא (ויקרא כ) והתקדשתם אלו מים ראשונים והייתם קדושים אלו מים אחרונים כי קדוש זה שמן אני יי' אלהיכם זו ברכה

As far as I can tell, this emendation is based on sevara, rather than any actual manuscript. But in one fell swoop, the Vilna Gaon solves a number of problems. Now, the prooftext, in order, is:
(A) והתקדשתם
(B) והייתם קדושים
(C) כי קדוש
(D) אני

This would then not be the pasuk in Kedoshim, but it would be the first of the two pesukim I cited from parashat Shemini. That pasuk again, Vayikra 11:44, read:

מד  כִּי אֲנִי יְהוָה, אֱלֹהֵיכֶם, וְהִתְקַדִּשְׁתֶּם וִהְיִיתֶם קְדֹשִׁים, כִּי קָדוֹשׁ אָנִי; וְלֹא תְטַמְּאוּ אֶת-נַפְשֹׁתֵיכֶם, בְּכָל-הַשֶּׁרֶץ הָרֹמֵשׂ עַל-הָאָרֶץ.

44 For I am the LORD your God; sanctify yourselves therefore, and be ye holy; for I am holy; neither shall ye defile yourselves with any manner of swarming thing that moveth upon the earth.

If so, the derasha is not based on the phrase כִּי אֲנִי יְהוָה אֱלֹהֵיכֶם in the beginning of the pasuk at all. That accounts for the word כי not appearing in the derasha. Rather, the word ani comes from the last word of the phrase כִּי קָדוֹשׁ אָנִי. Thus, in order in the pasuk, without a single gap in the words, we have:
(A) וְהִתְקַדִּשְׁתֶּם
(B) וִהְיִיתֶם קְדֹשִׁים
(C) כִּי קָדוֹשׁ
(D) אָנִי

Brilliant! Even so, this suggestion is not without difficulty. Slight difficulties are that the derasha works better with ani Hashem Elokeichem as a source for Birkat Hamazon, rather than merely ani; and that it is a rather artificial breakup to split the phrase ki kadosh | ani in this manner, and then darshen it, rather than keeping it intact. But the foremost difficulty, to my mind, is that he had to change the text of the gemara! This would be mitigated a lot if we can find an old manuscript or two which have this change.

From some searches of Talmudic manuscripts on the Internet, the Firenze manuscript is almost there. It has ani Hashem, rather than ani Hashem Elokeichem. Thus,

Well, not exactly. One extra feature this has, that our printed Talmudic text lacks, it that the brayta cites the full range of the pasuk first, before darshening it. That, then, is
וְהִתְקַדִּשְׁתֶּם וִהְיִיתֶם קְדֹשִׁים, כִּי קָדוֹשׁ  אֲנִי יְהוָה אֱלֹהֵיכֶם 

My own proposed nikkud, especially in the word ani, where I put a chataf patach rather than a kametz. This looks and smells like a citation of a pasuk the brayta is engaged in. And then, in terms of the derasha, the brayta would flow as follows, with the text marked in red removed:

והתקדשתם אלו מים ראשונים והייתם קדושים אלו מים אחרונים כי קדוש זה שמן אני יי' אלהיכם זו ברכה

Still, the assumption would be that it is citing a continuous pasuk, and furthermore, that that ani Hashem (Elokeichem) is being darshened. And if so, the eruv parshiyos suggestion of Minchas Shai is difficult. And the suggestion of the Gra does not work with this gemara -- we would need to emend it. That the word elokeichem is omitted later would appear to be a shorthand by the brayta or by the scribe.

Another possibility presents itself, with this girsa in the gemara. That is that the derasha is based on a combination of contiguous pesukim in Shemini, in Vayikra 11:44 and 11:45. That is, we would skip the ki ani Hashem Elokeichem at the beginning of 11:44. Then, 11:44 gives us, in order,

(A) וְהִתְקַדִּשְׁתֶּם

(B) וִהְיִיתֶם קְדֹשִׁים

(C) כִּי קָדוֹשׁ אָנִי

Finally, for (D), we would go to the very start of 11:45, and extract:
(D) כִּי אֲנִי יְהוָה

This recommends itself because 11:45 is missing the word Elokeichem. And if this is not the intent of the brayta, perhaps it was the intent of a scribe, who thought that the phrases should appear in order.

The "problem" with such an interpretation of this manuscript is that the brayta earlier cites it as a seemingly contiguous run of words in a single pasuk. And in that earlier citation, did include the word Elokeichem. And there are extra words here and there, such as ki in (D) and ani in (C). And it is strange to ignore the similar phrase at the start of 11:44 in order to cite the phrase at the start of 11:45. Therefore, I would still fall back to the idea that the sofer or author of the brayta was citing it shorthand, but meant some whole pasuk, even if that pasuk would appear to not exist. And so it remains something of a problem.

Ksav Yad Paris has precisely as in our gemara, where the derasha is in ani Hashem Elokeichem:

So too Ksav Yad Minkin:

But perhaps with enough searching, we might find something parallel to the Gra's suggested emendation.

However, while all of the above were great suggestions, I would conclude that they are all incorrect. In truth, the gemara in Berachos is perfect and needs no emendation. And, as indicated by Ksav Yad Firenze, they are darshening a continuous single pasuk. Now, in our Masoretic text we have no such pasuk. But the Samaritans do have it.

In parashat Kedoshim, the Samaritans have the text on the left, while we have the text on the right:

Note how they add in one word, kadosh. With that added word, added in order to harmonize the other pesukim such as in Shemini, the verse now reads just as in the brayta at the end of the eighth perek of Berachot. Of course, they had to go and partially ruin it. Because the Samaritans also remove the first word, vehitkadashtem. But this does not concern me. I am going to do a palginan diburo. After all, they didn't use entirely Samaritan texts. (Perhaps based on them partially, though.) For as Chazal they said to the Samaritans, ziyaftem Toraschem! But this is enough to demonstrate that at the time of Chazal, a variant of th pasuk in Kedoshim existed, and that (part?) of that variant was the extra word kadosh. And apparently, this was what the brayta darshened, the pasuk word for word as it appears in the brayta in Ktav Yad Firenze. And it accounts fir why ki kadsosh  does not inclue ani, and why ani Hashem Elokeichem does not include ki. Wonderful!

Despite this being the version of the pasuk darshened by Rav, or by a brayta, I believe that our Masoretic text is correct in this instance, under lectio difficilior. The addition of kadosh makes things nice and neat, and harmonized to other pesukim, which is why the Samaritans made this change. Rather, as in a number of instances I've already discussed, this is Chazal darshening non-Masoretic texts. But we should not jump to emend our texts to match, in every instance.

Related, the Samaritan variant in Shemini is also interesting:

Note how they add the word elokeichem in 11:45. If so, the suggestion I had above of the derasha spanning two pesukim, 11:44 through 11:45, would work most nicely. (A), (B), (C), (D) would be in order, and as written in the brayta. Of course, it would still be difficult that certain words like ki and ani aren't cited.

Despite this Shemini possibility, the approach which recommends itself most to me, and which I am 99.44% certain is correct, is the penultimate one. That is, it is a derasha on Kedoshim, on 20:7, but where Chazal had text parallel to the Samaritan text, with the word kadosh inserted.