Variants in the Masoretic Text: From Talmud to Rashi

The Babylonian Talmud tells of how Rabbi Ishmael warned his young pupil, Rabbi Meir, to be careful never to make even a single mistake when writing a Torah scroll:

^{בבלי עירובין יג.} ...אמר רב יהודה אמר שמואל משום רבי מאיר: ... כשבאתי אצל רבי ישמעאל אמר לי: "בני, מה מלאכתך?" אמרתי לו: "לבלר אני."

^{b. Eruvin 13a} … Rav Yehudah said in the name of Samuel who was repeating a tradition from Rabbi Meir: “… When I came to Rabbi Ishmael, he said to me: ‘My son, what is your profession?’ I said to him: ‘I am a scribe.”^[1]

אמר לי: "בני, הוי זהיר במלאכתך שמלאכתך מלאכת שמים היא, שמא אתה מחסר אות אחת או מייתר אות אחת – נמצאת מחריב את כל העולם כולו".

He said to me: “My son, be very careful with your work, since your work is the work of heaven. Maybe you will miss a letter or add a letter, and it will turn out that you are destroying the entire world!”

The rabbis were aware that the biblical text has variants. This is not only true when comparing differing textual traditions, like the Samaritan Pentateuch, the Septuagint, or Qumran scrolls, but was also the case within the Proto-Masoretic tradition itself.^[2]

Rabbi Meir’s Torah

R. Ishmael’s admonition is delivered with rhetorical hyperbole because of the reality of variants. That the scribe in question is Rabbi Meir ironically highlights the fact that Rabbi Meir’s text was known to differ from the standard Masoretic Text in at least three places.

1. The Son or Sons of Dan

In the description of Jacob’s many descendants who end up in Egypt, we find an incongruity between the plural introduction for Dan and the list of only one son:

^{בראשית מו:כג} וּבְנֵי דָן חֻשִׁים.

^{Gen 46:23} The sons of Dan: Hushim.

Noting this problem, R. Abraham ibn Ezra (1089–1164) offers two suggestions:

^{אבן עזרא בראשית מו:כג} יתכן שהיו שנים, ומת אחד מהם, ולא הזכירו הכתוב. או: כן דרך הלשון, [כי הנה נמצא (במדבר כו:ח): ובני פלוא אליאב.]⁠^[3]

^{Ibn Ezra Gen 46:23} It is likely that there were two and one of them died, so the text didn’t mention him. Or this may just be the way that language works, for elsewhere we find (Num 26:8): “And the sons of Palu: Eliab.”^[4]

The text of Rabbi Meir’s Torah, however, has the introduction in the singular:

^{בראשית רבה (תיאודור-אלבק) ויגש צג} בתורתו של ר' מאיר מצאו כתוב "ובן דן חושים."

^{Gen Rab (Theodor-Albeck), Vayigash §93} In Rabbi Meir’s Torah, they found written: “The son of Dan: Hushim.”

While this makes the verse problem free, the principle of lectio difficilior potior (“the harder reading is preferable”) likely applies here,^[5] especially since both the Samaritan Pentateuch and the Greek LXX (Septuagint) also have the plural. The simple explanation is that, given all the other sons of Jacob are listed as having multiple sons, the scribe here simply stuck with the formula. This may be what Radak (R. David Kimchi, ca. 1160–ca.1225) is trying to explain, similar to Ibn Ezra’s second suggestion:

^{רד"ק בראשית מו:כג} לא היה לו אז אלא בן אחד ושמו חושים. ואמר ובני כאלו אמר כל הבנים שהיו לו לא היה אלא זה, וכן (דברי הימים א ב:ח) "ובני איתן עזריה." 

^{Radak Gen 46:23} He only had one son at that point, and his name was Hushim. And it says “sons” as if to say all the sons he had at that time were just this one [son]. And the same for (1 Chron 2:8) “And the sons of Eitan: Azariah.”

It would seem, therefore, that Rabbi Meir’s text is a correction, whether by him or by an earlier scribe.

2. Garments of Light

Before sending Adam and Eve out of the garden of Eden, God makes them clothing:

^{בראשית ג:כא} וַיַּעַשׂ יְ־הוָה אֱלֹהִים לְאָדָם וּלְאִשְׁתּוֹ כָּתְנוֹת עוֹר וַיַּלְבִּשֵׁם.

^{Gen 3:21} And YHWH God made garments of skins (ʿor) for Adam and his wife and clothed them.

Genesis Rabbah again reports that Rabbi Meir’s Torah contained an alternative text:

^{בראשית רבה (תיאודור-אלבק) בראשית כ} בתורתו של ר' מאיר מצאו כתוב "כותנות אור."

^{Gen Rab (Theodor-Albeck) Bereishit §20} In Rabbi Meir’s Torah, they found written: “garments of light (ʾor).”

This mystical reading is uncharacteristic of the Bible, and likely a secondary reading. In this period, the two words were homophones: this may lie behind the change. Indeed, the Talmud warns about the problem of pronouncing the guttural ayin and the glottal stop aleph in the same manner:

^{בבלי מגילה כד:} אין מורידין לפני התיבה לא אנשי בית שאן, ולא אנשי בית חיפה, ולא אנשי טבעונין, מפני שקורין לאלפין עיינין ולעיינין אלפין.

^{b. Megillah 24b} People from Beth-Shean, Beth-Haifa, and Tivonin should not be allowed to stand before the ark [to lead services] since they pronounce their alephs like their ayins and their ayins like their alephs.

Thus, the scribe responsible for the change may have been copying down a text from oral dictation and made an error based on the words’ phonological similarity.^[6] In addition, R. Meir—whose name contains the same root word “light”—could have preferred this reading for its ideological message, in line with the imagery used for YHWH in Psalms:

^{תהלים קד:א} ...הוֹד וְהָדָר לָבָשְׁתָּ. ^כד:ב עֹטֶה אוֹר כַּשַּׂלְמָה...

^{Ps 104:1} …You are clothed in glory and majesty, ^104:2 wrapped in a robe of light…

In other words, R. Meir may be suggesting that to allow them to cling to God properly, Adam and Eve were dressed in robes of light, in imitation of God’s own clothing.^[7]

3. Death Is Good

After completing creation in six days, God is satisfied with the outcome:

^{בראשית א:לא} וַיַּרְא אֱלֹהִים אֶת כָּל אֲשֶׁר עָשָׂה וְהִנֵּה טוֹב מְאֹד...

^{Gen 1:31} And God saw all that he had done and behold it was very good…

Yet, Rabbi Meir’s Torah offers a very different message:

^{בראשית רבה (וילנה) בראשית ט:ה} בתורתו שלר' מאיר מצאו כתוב והנה טוב מאד והנה טוב מוֺת.

^{Gen Rab (Theodor-Albeck) Bereishit §9} In Rabbi Meir’s Torah, [in place of] “and behold it was very (meʾod) good” they found written “and behold death (mot) was good.”

Again, Rabbi Meir’s text seems to be secondary, as the statement makes little sense in context. As in the previous example, the origin in this change is likely in the aural similarity between the words, based on two features of pronunciation in that period:

1. The glottal stop aleph was not pronounced.^[8]
2. The voiced dental plosive (d) was pronounced like a voiceless dental plosive (t) at the end of words.

Thus, both words would have sounded like mot,^[9] allowing a scribe to mistake one for the other. The next part of the text, however, seems to imply that this was understood by some at least to be a midrash and not an actual variant:

אמר ר' שמואל בר נחמן: רכוב הייתי על כתיפו שלזקיני ועולה מעירי לכפר חנא דרך בית שאן ושמעתי את ר' שמעון בן ר' לעזר יושב ודורש בשם ר' מאיר והנה טוב מאד והנה טוב מות.

Rabbi Samuel bar Nahman said: I was riding on my grandfather’s shoulders, coming up from my hometown of Kfar Chana through Beth-shean, and I heard R. Simon ben R. Elazar sitting and expounding in the name of Rabbi Meir that [the phrase] “behold it was very good” [should really be read] “behold, death is good.”

It seems, therefore, that at least some sages were unsure whether Rabbi Meir actually had a physical text of the Torah that differed from the standard Masoretic Text, or whether these were all meant as homiletical readings. Indeed, it is possible that in some or all of these instances, the intent is not that Rabbi Meir actually had a different text, but rather he had a midrashic teaching on the standard text, and that the “Torah of Rabbi Meir” means his teaching.

The Three Variants that Were Cancelled

Several rabbinic texts^[10] tell of how three textual variants were found in the scrolls of the Temple courtyard:

^{ירושלמי תענית ד:ב} שְׁלֹשָה סְפָרִים מָצְאוּ בָעֲזָרָה: סֵפֶר מְעוֹנֵי, וְסֵפֶר זַעֲטוּטֵי, וְסֵפֶר הִיא.

^{j. Taanit 4:2} Three scrolls [of the Torah] were found in [storage in] the Temple courtyard. The scroll with maʿonei; the scroll with zaʿatutei, and the scroll with hee.

The Talmud goes on to describe the exact difference in each scroll and how they were dealt with:

1. Same word, but masculine versus feminine spelling

בְּאֶחָד מָצְאוּ כָתוּב מְעוֹן אֱלֹהֵי קֶדֶם, וּבִשְׁנַיִם כָּתוּב מְעוֹנָה אֱלֹהֵי קֶדֶם, וְקִייְמוּ שְׁנַיִם וּבִיטְלוּ אֶחָד.

In one, they found written (Deut 33:27) “a shelter (meʿon, [ms]) is the Ancient God,” and in two were written “a shelter (meʿonah [fs]) is the Ancient God.” So they accepted the version with two witnesses and nullified the version with one.

2 Same term, Aramaic versus Hebrew

בְּאֶחָד מָצְאוּ כָתוּב וַיִּשְׁלַח אֶת זַעֲטוּטֵי בְּנֵי יִשְׂרָאֵל, וּבִשְׁנַיִם כָּתוּב וַיִּשְׁלַח אֶת נַעֲרֵי בְּנֵי יִשְׂרָאֵל, וְקִייְמוּ שְׁנַיִם וּבִיטְלוּ אֶחָד.

In one, they found written (Exod 24:5) “and he sent the young men (zaʿatotei)^[11] of the children of Israel,” and in two, were written “and he sent the young men (naʿarei) of the children of Israel.” So they accepted the version with two witnesses and nullified the version with one.

3 Same word, archaic versus updated spelling

בְּאֶחָד מָצְאוּ כָתוּב תֵּשַׁע הִיא, וּבִשְׁנַיִם כָּתוּב אַחַד עֶשְׂרֵה הִיא, וְקִייְמוּ שְׁנַיִם וּבִיטְלוּ אֶחָד.

In one, they found the term “she” (hee spelled היא)^[12] written nine times, and in two they found the term “she” (hee spelled היא) eleven times.^[13] So they accepted the version with two witnesses and nullified the version with one.

In these three cases, the rabbis are aware that variants existed for the Torah text. In many others, we simply find the rabbis quoting a text that doesn’t reflect the MT as we have it.

Elkanah Followed His Wife

The Talmud criticizes Manoach, the father or Samson, for following his wife—i.e., literally walking behind her—violating the Tannaitic teaching prohibiting this:

^{בבלי ברכות סא.} אָמַר רַב נַחְמָן: מָנוֹחַ עַם הָאָרֶץ הָיָה, דִּכְתִיב (שופטים יג:יא): "וַיֵּלֶךְ מָנוֹחַ אַחֲרֵי אִשְׁתּוֹ."

^{b. Berakhot 61a} Rav Nahman said: “Manoach was an ignorant person, since it says (Judg 13:11): “Manoach followed his wife.”

The Talmud then challenges this conclusion since Elkanah, Samuel’s father, and the prophet Elisha, who were not ignoramuses, also walked behind women:

מַתְקִיף לַהּ רַב נַחְמָן בַּר יִצְחָק: אֶלָּא מֵעַתָּה, גַּבֵּי אֶלְקָנָה, דִּכְתִיב: "וַיֵּלֶךְ אֶלְקָנָה אַחֲרֵי אִשְׁתּוֹ", וְגַבֵּי אֱלִישָׁע, דִּכְתִיב (מלכים ב ד:ל): "וַיָּקׇם וַיֵּלֶךְ אַחֲרֶיהָ", הָכִי נָמֵי אַחֲרֶיהָ מַמָּשׁ?

Rav Nahman bar Yitzhak attacked: “If so, then when it writes regarding Elkanah “Elkanah followed his wife,” and when it writes regarding Elisha (2 Kgs 4:30) “he got up and followed her,” did they also literally walk behind these women?!

The Talmud concludes that, at least in the latter two and perhaps in all three cases, the phrase does not refer to physical walking, but is meant as a metaphor, meaning that the men followed the women’s advice. As already noted by the 12^th century Tosafist commentary on the Talmud (ad loc.), no such verse exists in the Bible regarding Elkanah:

^{תוספות ברכות סא.} שבוש הוא שאין פסוק זה בכל המקרא ול[א] ג[רסינן] ליה.

^{Tosafot Berakhot 61a} This is an error, for there is no such verse anywhere in the Bible, and we do not include this line [as a legitimate part of the Talmudic text].

What the verse in MT actually says is:

^{שמואל א א:יח} ...וַתֵּלֶךְ הָאִשָּׁה לְדַרְכָּהּ וַתֹּאכַל וּפָנֶיהָ לֹא הָיוּ לָהּ עוֹד. ^א:יט וַיַּשְׁכִּמוּ בַבֹּקֶר וַיִּשְׁתַּחֲווּ לִפְנֵי יְ־הוָה וַיָּשֻׁבוּ וַיָּבֹאוּ אֶל בֵּיתָם הָרָמָתָה וַיֵּדַע אֶלְקָנָה אֶת חַנָּה אִשְׁתּוֹ...

^{1 Sam 1:18} …So the woman left, and she ate, and was no longer downcast. ^1:19 Early next morning they bowed low before YHWH, and they returned to Ramah and entered their home. And Elkanah knew is wife, Hannah…

One possibility is that Rav Nahman bar Yitzhak or the editors of the Talmud actually had the phrase “Elkanah followed his wife” in their biblical text either before or in place of the plural “and they returned.” Another possibility, however, is that Rav Nahman bar Yitzhak was working from memory^[14] and the confluence of the singular “the woman went her way” followed by their going together home and then the singular “and Elkanah knew…” implanted the false memory of such a phrase.^[15]

Concubines

Towards the end of Abraham’s life, to clarify that Isaac was the only son inheriting his legacy, he gives his other sons gifts and sends them away:

^{בראשית כה:ו} וְלִבְנֵי הַפִּילַגְשִׁים אֲשֶׁר לְאַבְרָהָם נָתַן אַבְרָהָם מַתָּנֹת וַיְשַׁלְּחֵם מֵעַל יִצְחָק בְּנוֹ בְּעוֹדֶנּוּ חַי קֵדְמָה אֶל אֶרֶץ קֶדֶם.

^{Gen 25:6} But to Abraham’s sons by concubines Abraham gave gifts while he was still living, and he sent them away from his son Isaac eastward, to the land of the East.

In Genesis Rabbah, Rabbi Judah claims that both of Abraham’s secondary wives, Hagar and Keturah, were the same woman, while Rabbi Nehemiah says they were two different women, noting that the verse says “concubines” in plural:

^{בראשית רבה (תיאודור-אלבק) חיי שרה סד} אמר ליה [ר' נחמיה] "והכת[יב] (בראשית כה:ו): 'ולבני הפילגשים אשר לאברהם'?"

^{Gen Rab Chayei Sarah §64} [R. Nehemiah] said to him: “But doesn’t it say (Gen 25:6): “But to his sons by concubines” (i.e. plural)!”

To this, Rabbi Judah takes note of the unusual spelling:

אמר ליה [רבי יהודה]: "'פילגשם' כת[יב]."

[R. Judah] said to him: “But it is spelled pilagsham (i.e., as if it were singular).”

R. Judah’s point is that the word is written in the defective spelling, without the mater lectionis (the yod before the final mem, that indicates the vowel sound “ee” and that the word is a plural). He argues that the word is spelled that way on purpose to hint that Abraham only had one concubine, just at two different points in his life. This midrash is also quoted by Rashi (R. Solmon Yitzhaki, 1040–1105):

^{רש"י בראשית כה:ו} "הפלגשם" – חסר כתיב,⁠ שלא היתה אלא פלגש אחת, היא הגר היא קטורה.

^{Rashi Gen 25:6} “The Concubine(s)”—it is written with defective spelling, since he only had one concubine, Hagar and Keturah are the same person.^[16]

Yet in the MT, the word is spelled plene, with the yod. We can see this already in the commentary of Rashi’s grandson, R. Samuel ben Meir (Rashbam), who spells the word plene and interprets it in that light:

^{רשב"ם בראשית כה:ו} "הפילגשים" – הגר וקטורה.

^{Rashbam Gen 25:6} “The concubines”—Hagar and Qeturah.

This problem was noted by R. Hezekiah ben Manoach (13^th cent.) as well as in the anonymous Tosafist commentary known as Hadar Zekeinim (13^th cent.), both of whom begin by quoting Rashi and then respond similarly:

^{חזקוני בראשית כה:ו} ... חז"ק^[17] לפי[רוש] ר"ש, שהרי הוא מלא בספרים מדוייקים והן הגר וקטורה.

^{Hizkuni Gen 25:6} …There is a difficulty with Rashi’s commentary, since in vetted scrolls, it is written plene, and the reference so to Hagar and Keturah.

^{הדר זקנים בראשית כה:ו} ...וקשה כי בכל חומשים מדוייקים מלא הוא וגם במסורת חשיב עליו במלאים.

^{Hadar Zekeinim 25:6} …But there is a problem, since in all the vetted chumashim (books of the Pentateuch) it appears plene, and even in the Masorah (lists), it is counted as an example of plene writing.

It appears, therefore, that for a while, both spellings existed in the rabbinic tradition of the MT.

A Superfluous “And”

Another example where Rashi has a different text than our standard MT appears in the verse where YHWH explains to Moses the purpose of the ark:

^{שמות כה:כב} וְנוֹעַדְתִּי לְךָ שָׁם וְדִבַּרְתִּי אִתְּךָ מֵעַל הַכַּפֹּרֶת מִבֵּין שְׁנֵי הַכְּרֻבִים אֲשֶׁר עַל אֲרוֹן הָעֵדֻת אֵת כָּל אֲשֶׁר אֲצַוֶּה אוֹתְךָ אֶל בְּנֵי יִשְׂרָאֵל.

^{Exod 25:22} There I will meet with you, and I will impart to you—from above the cover, from between the two cherubim that are on top of the Ark of the Testimony—all that I will command you concerning the Israelite people.

On the final phrase, Rashi comments as if it begins with a conjunction (bold):

^{רש"י שמות כה:כב} "ואת כל אשר אצוה אתך" – הרי ו"ו זו יתירה וטפילה, וכמוה הרבה במקרא...

^{Rashi Exod 25:22} “And all that I will command you”—this “and” is extra and superfluous, and there are many like this in the Bible…^[18]

In this case, R. Abraham ibn Ezra also had this alternative text, and he begins his gloss with וטעם ו"ו ואת כל אשר אצוה “and the reason for the vav in “and all that I will command…”

The Tosafist commentator R. Judah b. Elazar (late 12^th/early 13^th cent.), after quoting Rashi, notes that this is not our text:  

^{מנחת יהודה שמות כה:כב} ... פליאה גדולה נשגבה בעיני שליח צבור מטורויי״ש הר״ר מנחם ז״ל כי בספרים מדוייקים כתוב את בלא וא״ו.

^{Minchat Yehudah Exod 25:2} …It is a great and overwhelming surprise in the eyes of the prayer leader of Troyes, R. Menachem, may his memory be a blessing, since in corrected scrolls it is written without the conjunction.

In this case, the variant lasted at least until the 12^th century.

Small Differences in the Masoretic Text

What we call the Masoretic Text of the Bible—the one used by all Rabbinic Jews from Ultra-Orthodox to ultra-liberal, as well as by Karaite Jews—was finalized in the tenth century C.E. by the Tiberian School of Aaron ben Asher, who built on the millennium-old Proto-Masoretic textual tradition, which was itself relatively stable, even before the Masoretes. Nevertheless, as we can see from the examples above, even this stable text has some small variations that lasted centuries, some even lasting beyond the time of the Masoretes.