Shiur #7a: Nusach Ha-mikra – Accuracy of the Biblical Text

  • Rav Amnon Bazak

A.        Introduction


In previous chapters we discussed the processes by which the body of biblical literature came into being, along with the conclusions drawn by academic research in this area, known as “higher criticism.” In this chapter we will discuss the accuracy and history of the biblical text itself – an area known as “lower criticism.”[1] This type of criticism seeks to explore the emergence of the precise biblical text that we possess and the changes that this text has undergone over the course of generations, by comparing manuscripts, examining textual witnesses, and employing various philological tools. The aim of lower criticism is to locate textual variants and to try to evaluate their relative accuracy. As such, this realm of study is not committed to any particular textual version. This, once again, creates a fundamental gap between the traditional Jewish approach maintained over the generations, and the academic approach. In general, the traditional approach would not dream of addressing the possibility of textual variants:


“There is a concept in Judaism, deeply rooted in the consciousness of the nation, concerning the sanctity of the biblical text, even its very letters. This is usually explained in historical terms: namely, the text, down to the last letter, has reached us in the same original form in which it was first composed. Over the course of many generations, this concept has come to assume something of the validity of a fundamental principle of Judaism, by virtue of many statements surrounding this subject, in both halakha and aggada, as well as in Jewish thought. Thus, any method that casts doubt on the absolute reliability of the transmission causes a believing Jew to recoil.”[2]


In this chapter we will address two main questions: First, is the text that we have today indeed the original text of the Tanakh? Second, if what we have is not identical in every detail to the original text, does this allow for the possibility of proposing emendations to our text? Our discussion will address the history of the text through the generations, as well as the various textual witnesses, including inter alia the Dead Sea Scrolls, the Samaritan Torah, the Septuagint, and other translations.


B.        The Tanakh text during the period of Chazal[3]


From time immemorial, great care has been taken concerning the precise transmission of the text, which has been regarded as having great importance. Josephus, writing in the 1st century C.E., testifies: “The scribes have taken care to maintain extreme accuracy, and – if I may so bold – they will continue to do so (for all generations).”[4] The care taken in preserving the precise text is obvious throughout many rabbinic sources. For instance, R. Meir teaches:


“When I came to R. Yishmael, he said to me: My son, what is your occupation? I told him, ‘I am a scribe.’ He said to me, ‘Be meticulous in your work, for your occupation is a sacred one. If you were to omit or add a single letter, you would thereby destroy the entire world.’” (Eruvin 13a)


                        Chazal were altogether proficient with the entirety of the biblical text, as we may deduce from the corpus of midrashim, in which verses from throughout the Tanakh are treated.[5] In addition, the Sages also concerned themselves with the clarification of the precise text of the Tanakh, and many sources testify to the tremendous care taken concerning textual details. For example,[6] concerning the verse, “Take for us the foxes (shu’alim), the small foxes that spoil the vineyards” (Shir Ha-shirim 2:15), there is evidence of a tradition specifying a difference in spelling between the two appearances of the word “shu’alim” – the first time with a ‘vav,’ and the second time without it: “R. Berakhia said: The first [appearance of the word] ‘shu’alim’ is written in plene form, the second – in defective form.”[7]


Similarly, attention is paid to the tiny yud in the word “teshi,” in the verse “Of the Rock that begot you, you are unmindful (teshi)” (Devarim 32:18),[8] noting it as an unusual phenomenon: “The yud is small…” (Vayikra Rabba 23:14, Margaliot edition p. 548).


It is clear that, in general, Chazal worked with a single textual version that they all shared. Nevertheless, this does not mean that there was a single textual version that was agreed upon absolutely in all its details, as we gather from various sources.


            1. There are instances of questions as to the correct textual version, with the need arising to choose between different versions. In the Sifri Devarim we find:


Three books [of Tanakh] were found in the azara (Temple courtyard): one of the “me’onim,” another of “hi-hi,” and one called sefer za’atutim. [The explanation for these appellations for the books follows:] In one book (the verse in Devarim 33:27) was written, “The Eternal God is a dwelling place (“ma’on”), while in the other two it was written “me’ona.” The Sages [therefore] rejected [the version appearing only in] the one copy, and accepted [the version that appeared in] the other two. In one version the word “hi” (“it” or “she”) appeared nine times, while in the other the word “hi” appeared eleven times. The Sages [therefore] rejected the one copy, and accepted the other two. In one version the text read, “and he sent the young men (za’atutei) of Bnei Yisrael” (Shemot 24:5), and “upon the young men (za’atutei) of Bnei Yisrael…” (24:11), while the other two read, “he sent the young men (na’arei) of Bnei Yisrael” and “upon the nobles (atzilei) of Bnei Yisrael” (Shemot 24:11). The Sages [therefore] rejected the one copy, and accepted the other two.[9]


We may assume that the three books found in the courtyard of the Temple represented the most sacred and most important textual witnesses available to the Sages during the Second Temple Period. No two were identical in every detail. The decision of the Sages to establish the “majority version” in each case of discrepancy created an interesting and surprising reality: the standard text for all Torah scrolls was determined in accordance with these three “models” – and, as a result, all three of them became unfit for use, each owing to the specific “defect” that it contained.


2. In several places in rabbinic literature, mention is made of a special Sefer Torah belonging to R. Meir, which included at least seven divergences from the accepted text.[10]


For example there is a different version of the verse (Bereishit 46:23), “And the sons (benei) of Dan, Chushim” than the one that appears in the Masoretic text. Where the verse as it appears in the Masoretic text displays a lack of correlation between the plural form (benei – “sons of”) and the fact that only one son is named,[11] a different, simpler version is cited as being present in R. Meir’s text: “The Torah of R. Meir was found to read, ‘and the son (ben) of Dan, Chushim.’”[12]


Another example concerns the verse in which Yosef declares, “[God] has made me (va-yesimeni) a father for Pharaoh” (Bereishit 45:8): “The Torah of R. Meir was found to read: ‘[God] has made me a creditor (va-yasheni) as a father [to Pharaoh],’ as it is written, ‘which he lends (yasheh) to his neighbor’ (Devarim 15:2).”[13]


Regarding the latter example, the midrash notes that this is “one of the words written in the Torah that left Jerusalem with the captives and was taken to Rome, where it was hidden in the synagogue of Severus.”[14] The midrash then continues directly with a list of all thirty instances where the text of this Sefer Torah differed from the accepted version,[15] most of them minor variations.[16] With the exception of these thirty instances, the text of the Torah taken to Rome was identical to that of the accepted text of the Sages.


(To be continued)


Translated by Kaeren Fish



[1]The following are some sources offering extensive reviews of the subject: M.Z. Segal, Mavo ha-Mikra 4, Jerusalem 1977, pp. 842-910; S. Talmon (ed.), “Tanakh, nussach,” Encyclopedia Mikrait 8, Jerusalem 5742, columns 621-641; E. Tov, Bikkoret Nussach ha-Mikra, Jerusalem 5750; Z. Talshir, “Le-Toldot Nussach ha-Mikra,” in Z. Talshir (ed.), Sifrut ha-Mikra – Mevo’ot u-Mechkarim 1, Jerusalem 5771, pp. 37-85.

[2]M. Cohen, “Ha-Idea bi-Devar Kedushat ha-Nussach le-Otiotav u-Bikkoret ha-Text,” in U. Simon (ed.), Ha-Mikra va-Anakhnu, Tel Aviv 5739, pp. 42-43.

[3] For the purposes of this discussion, we will define “the period of Chazal” as extending from the Destruction of the Second Temple to the completion of the Talmud, without noting the sub-divisions within this period, which are not pertinent to the matter at hand.

[4] Josephus, Against Apion, 1, 6.

[5] As noted by S. Lieberman, Yevanim ve-Yevanut be-Eretz Yisrael, Jerusalem 5723, p. 189. Admittedly, there were exceptions, as noted already by the Rishonim – “Sometimes they were not proficient in the verses” (Tosafot, Bava Batra 113a); see the anecdote recounted in Bava Kama 54b concerning R. Hiyya bar Abba. However, as Lieberman notes, these were exceptions to the rule.

[6]See D. Rosenthal, “Al Derekh Tipulam shel Chazal be-Chilufei Nussach ba-Mikra,” in Y. Zakovitch and E. Rofe (eds.), Sefer Yitzchak Aryeh Zeligman 2, Jerusalem 5743, pp. 397-398.

[7]R. Berakhia’s teaching is reflected in the Koren edition of the Tanakh, but most manuscripts and printed versions do not follow this rule. For instance, in the Aleppo Codex, the word is missing the vav in both instances (see also Minchat Shai), while MS Leningrad includes the vav in both instances.

[8] Here, too, the “authentic” spelling is a matter of debate: while the letter yud in this verse is mentioned in all Masoretic lists of letters that are written in diminutive form, it does not actually appear in this form in some important witnesses including the Aleppo Codex and MS Leningrad.

[9] Sifri Devarim, piska 356, Finkelstein edition p. 423. For various sources offering this description and the differences between them, see S. Talmon, “Shelosha Sefarim Matz’u ba-Azara,” in: Y.M. Grintz and Y. Liever (eds.), Sefer Segal – Kovetz Mechkarim ba-Mikra, Jerusalem 5725, pp. 252-264.

[10] Some of these variants appear to be “midrashic” in nature, since it is difficult to make sense of them on the plain level in the context of the verse. Especially well-known are two examples. One concerns the verse, “And God saw all that He had done, and behold, it was very good (tov me’od)” (Bereishit 1:31). In Bereishit Rabba (parasha 9,5, Theodor-Albeck edition p. 70) we find, “The Torah of R. Meir was found to read, instead of, ‘And behold it was very good (tov me’od)’ – ‘and behold it was good to die (tov mot).’” The second concerns the verse, “And the Lord God made for the man and for his wife coats of skins (‘or,’ written with the letter ayin), and He clothed them” (Bereishit 3:21). Bereishit Rabba (parasha 20,12, Theodor Albeck edition, p. 196) teaches, “The Torah of R. Meir was found to read, ‘coats of light’ (‘or,’ written with the letter alef).” In both of these cases the discrepancy in the text has a simple phonetic explanation, but given the content R. Meir’s version seems to be aimed at conveying a certain homiletic message, rather than reflecting an actual textual version.

Another variant with conceptual or homiletic significance concerns the national struggle against Rome: the verse in Yishayahu that starts with the words, “The burden of Duma…” (21:11) is substituted in R. Meir’s Torah with “the burden of Rome.” In this case, too, the substitution is easily explained in terms of the graphic similarity between the letters dalet and resh, but here again the orientation seems to be midrashic – i.e., conveying a homiletic message, rather than simply featuring a textual variant.

[11] The commentators note the difficulty in the verse, and Ibn Ezra offers two possible explanations: either Dan had two sons, one of whom died, and therefore the verse mentions only Chushim, or the verse adopts in relation to Dan the same standard formula that appears for each of the other sons of Yaakov (“And the sons of …”) even though in this case there was only one son.

[12]Bereishit Rabba 94,9, Theodor-Albeck edition, p. 1182.

[13]Bereishit Rabba 45, 8 Theodor-Albeck edition p. 209.

[14] Two Roman emperors (Septimius and Alexander), who reigned during the 2nd and 3rd centuries C.E. were called by the name Severus. The reference here is most likely to Severus Alexander (222-235), who was known for his positive relations with the Jews.

[15] The list in the Albeck edition is somewhat corrupted, but with the aid of other manuscripts it is possible to arrive at the full list. Concerning manuscripts, see Talshir (above, n. 1), p. 40, n. 14.

[16] Some of the differences include the versions unique to the Torah of R. Meir; others involve remnants of an ancient script in which no distinction is made between final letters and regular letters (for instance, יומ מותי [Bereishit 27:2]). A small number represent reasonable possible alternatives to the textual version with which we are familiar, such as the verse containing God’s statement about Sedom: “I shall go down now and see whether they have done altogether according to the cry of it (ha-ke-tza’akatah)…” (Bereishit 18:21): the text in this Sefer Torah reads, “whether they have done altogether according to their cry (ha-ke-tza’akatam).”