Nusach Ha-mikra – Accuracy of the Biblical Text (3)
C. The version of Chazal and the Masora version
As was made clear over the course of the previous two shiurim, there exists no complete, clear copy of the Tanakh even from the period of Chazal. The text was handed down from generation to generation in the form of manuscripts and, while the scribes who undertook this work undoubtedly attached great importance to the accurate transmission of the texts and regarded them as holy, they were human – and in the copying of such lengthy manuscripts human error inevitably creeps in. Thus, over the years a large number of manuscripts were produced, with very slight differences among them – until the time of the Masoretes.
The Masoretes lived between the 8th and 10th centuries, and sought to establish a uniform version of the text that would be accepted, from that point onwards, in all Jewish communities. To this end they instituted Masoretic notes (to be discussed below), which served to set down, inter alia, the number of times that certain words appeared in plene or defective form (maleh ve-chaser) throughout the entire Tanakh; they wrote books discussing the exact wording and spelling; and they developed the system of vowels and cantillation marks in order to preserve the proper form of reading. To this day, the Tanakh that we have is based for the most part on the Masoretic version of the text.
Two main textual traditions are mentioned already in the Talmud: that of Babylonia (the "eastern") and that of Eretz Yisrael (the "western"). These differ from one another in their letters, vowelling, cantillation, and "keri u-khetiv" variants (where a word is vocalized differently from the way in which it is written). Thus there developed the "Babylonian tradition" and the "Tiberian tradition,” which reflected these differences. The Tiberian Masora, regarded as more accurate, was the version that was ultimately accepted.
The Masoretes of Tiberias themselves were split between two traditions – that of Aharon ben Asher, and that of Moshe ben Naftali. There were hundreds of differences between them, mostly small discrepancies in matters of vowelling and cantillation. The version of Ben Asher came to be considered authoritative, and books of Tanakh published today are almost identical to it, with the exception of a few minor details. The Masora text was preserved with great accuracy and meticulousness by means of various devices, especially the mechanism of the "Masora Gedola" and the "Masora Ketana,” comprising tens of thousands of notations above, below, and at the sides of the columns of the text, which ensured the precise way in which words were to be written throughout the Tanakh.
In particular, "Keter Aram Tzova" (the Aleppo Codex), a 10th century manuscript of the Tanakh that was kept in Aleppo, Syria from the 14th century until 1947, came to be regarded as an authoritative text, since its notations had been written by Aharon ben Asher himself. It is generally agreed that the Rambam's mention of a certain Sefer Torah in Egypt (where the Codex was kept prior to its move to Aleppo) refers to this manuscript:
"The manuscript that we have relied upon in these matters [the list of 'open' and 'closed' units] is the well-known scroll in Egypt, comprising [all] twenty-four books [of Tanakh], which was kept in Jerusalem for some years for other manuscripts to be checked against, and everyone relied upon it because its notations were written by Ben Asher, who worked on it painstakingly for many years, and checked it many times as he copied it. This is what I relied upon [also] for the Sefer Torah that I wrote in accordance with its laws." (Hilkhot Sefer Torah 8:4)
The main merit of this manuscript lies in the almost perfect correlation between its text and its Masoretic notes, demonstrating the impressive degree of accuracy with which it was written. The Aleppo Codex serves as the basis for the Breuer edition of the Tanakh, and later also Prof. Menachem Cohen's edition, Mikraot Gedolot – Ha-Keter and the Mif'al Ha-Mikra edition published by the Hebrew University.
In general, the Masora version of the text continues the version used by Chazal, and therefore the questions and uncertainties discussed in the previous shiurim still apply. Nevertheless, it should be noted that in dozens of instances the Masora version differs from the text used by Chazal, and sometimes these differences are substantial. We have already mentioned the discussion in Kiddushin (30a) that reflects the discrepancies between the text as cited by Chazal and the text as we know it concerning the middle word and the middle letter of the Torah. Further on in the same discussion, we find the following beraita:
"Our Sages taught: The verses of a Sefer Torah number 5,888; the verses of Sefer Tehillim number eight less than that, and the number of verses of Divrei Ha-yamim number eight more."
In the Tanakh that we use today, the number of verses in the Torah is 5,845 – 43 less than the number cited in the Gemara, while the number of verses in Tehillim (2,527) and in Divrei Ha-yamim (1,556) are even further removed from the gemara’s count. Concerning the Books of Tehillim and Divrei Ha-yamim the Geonim and Rishonim offer various explanations, but it seems in any event that the accepted division of the verses during the time of Chazal was different from that of the Masoretes.
Some differences pertain to matters of content and meaning. Many of Chazal's teachings cite verses that differ from the Masora text. The medieval commentatorsnote that "in several places there was controversy between the redactors of the Talmud and the Sages of the Masora" (Responsa of Ridbaz, III, 594). Sometimes these discrepancies involve plene or defective spelling. For example, the Gemara (Shabbat 55b) cites Rav's opinion that of the two sons of Eli, only Chofni sinned, and not Pinchas. He maintains that although the text states that Eli told his sons that bad rumors were circulating, suggesting that they were "causing (in the plural) God's people to sin" (ma'avirim am Hashem) (Shmuel I 2:24), he was in fact referring only to one of them. Rav proves this by noting that "It is written 'ma'aviram'" (in the singular). In the Masora version the letter 'yud' appears before the final 'mem,' such that the verb is unquestionably in the plural, as noted by the Tosafot ad loc. who state: "The Gemara disagrees with the version as written in our Tanakh." Elsewhere this is formulated the other way round: "We find that the accepted [textual] tradition differs from what the Gemara says" (Tosafot, Nidda 33a). Rabbi Akiva Eiger (1761-1837), in his Gilayon Ha-shas to Shabbat, notes many similar examples.
Sometimes the discrepancies are more substantial. Tosafot on Shabbat 55b quotes the Yerushalmi (Sota 1:8, 17b):
"In one place we read [regarding Shimshon], 'He judged Israel for forty years,' while in a different place we read, 'He judged Israel for twenty years.' Rabbi Acha said: This teaches that the Philistines were as fearful of him [and therefore refrained from attacking] for twenty years after his death as they had feared him for twenty years in his lifetime."
However, as the Tosafot note, in our Masora version the first verse that is cited reads, "He judged Israel in the days of the Pelishtim for twenty years" (Shoftim 15:20).
Another example: In two places in the Gemara we find:
"Rav Nachman said: Manoach was an ignoramus, as it is written, 'Manoach went after his wife' (Shoftim 13:11). Rabbi Nachman bar Yitzchak disagreed: But what about Elkana, concerning whom it says, 'Elkana went after his wife,' and what about Elisha, concerning whom it is written, 'And he arose and went after her' (Melakhim II 4:30)? Are we then to understand that [in these latter instances, too, where the men were clearly knowledgeable and pious,] the man literally walked behind the woman? [Surely not;] it means he followed her words and her advice. Therefore here too, [with regard to Manoach], he followed her words and her advice." (Berakhot 61a; Eruvin 18a)
The verse about Elkana does not appear in the Tanakh as we know it; the verse closest in meaning to the one cited is, "Elkana went to Rama, to his house" (Shmuel I 2:11). For this reason, the Tosafot comment in Berakhot ad loc.: "This is a corruption [of the text], for there is no such verse in all of the Tanakh and we do not follow this tradition"; this view is also upheld by many other Rishonim. However, there are other Rishonim – first and foremost Rashi and Rabbeinu Chananel on Eruvin – who accept the discussion at face value and do not reject the textual version upon which it is based. Indeed, in most of the manuscripts of both massekhtot, the teaching appears as cited above, with no alteration to the verse – even though it does not conform with the Masoretic version. It would therefore seem that these Rishonim believed that this is the version that Chazal had before them, and that for this reason they did not remove the verse.
The discrepancies between the version of the Tanakh that Chazal used and the Masoretic text that we have, give rise to the obvious question: how are we to determine the proper and accurate version of the text? This question occupied Rishonim and Acharonim alike. Many of the Rishonim adopted the principle articulated by Rashba, according to which a distinction should be drawn between instances where the discrepancy between the versions has some bearing on the law that the Gemara derives from a certain version of the verse, in which case "We rely on the Sages of the Talmud,” and instances where the textual variants have no practical bearing on the Halakha, and here "we rely on the Sages of the Masora." However, the rule that was actually accepted in practice, formulated by the Minchat Shai, is that the Masora version is decisive in every instance:
"In every place where the Gemara or midrash differs from the Masora in matters of additions or omissions, we follow the Masora… even when the law is derived from them."
Rabbi Shelomo Ganzfried, redactor of the Kitzur Shulchan Arukh, offers a similar opinion in his work Kesset Ha-sofer (introduction to Part II, siman 28).
These discussions reflect a clear awareness by authorities and commentators through the ages of the many discrepancies between the Masoretic version of Tanakh and the version used and cited by Chazal.
Translated by Kaeren Fish
Today this fact is borne out by checking Torah scrolls using scanners. It turns out that that even well-known scribes make mistakes. According to data published by websites dedicated to this field, the vast majority of Torah scrolls checked by computer are found to contain errors.
 For a general overview of the subject, see Y. Yeivin's entry, "Masora,” in Encyclopedia Mikrait 5, Jerusalem 5728, columns 151-154.
 For a comprehensive review of the Babylonian tradition see Y. Ofer, Ha-Massoret ha-Bavlit la-Torah – Ekronoteiha u-Derakheiha, Jerusalem 5761.
 In the southern part of Eretz Yisrael there developed a third Masora, known as the "Masora of Eretz Yisrael,” but it is marginal in relation to the two main traditions.
Lists of these variants were noted in manuscripts, the best-known of these being Sefer ha-Chilufin (the Book of Variants), a scientific edition of which was published by A. Lipschitz, Jerusalem 5725.
Rabbi Mordechai Breuer published a number of editions of the Tanakh based on the Aleppo Codex. For his discussion of the Codex and its influences on the Tanakh accepted today, see: M. Breuer, Keter Aram Tzova ve-ha-Nussach ha-Mekubal shel ha-Mikra, Jerusalem 5737; A. Shamosh, Ha-Keter – Sippuro shel Keter Aram Tzova, Jerusalem 5747; Y. Ofer, "Keter Aram Tzova ve-ha-Tanakh shel R. Shalom Shakhna Yellin," in Sefer ha-Yovel le-Rav Mordechai Breuer, Jerusalem 5752, pp. 295-353; Keter Yerushalayim – Tanakh ha-Universita ha-Ivrit bi-Yerushalayim, Jerusalem 5760, and the accompanying volume He’arot.
At the time of writing, not all the volumes of this edition have been printed yet, but the entire Tanakh is available in this edition on DVD.
 This edition, too, is not yet complete; only a few volumes have been published.
 The Tosafot Yeshanim ad loc. question the count of the verses in Tehillim and the matter remains unresolved. Rav Hai Gaon, in a responsum, writes that the figures noted in the Gemara refer to "that Sefer Torah which they found in Jerusalem, which was unusual in its script and in its number of verses, and likewise the Sefer Tehillim and also the Divrei Ha-yamim, but now there is only one [accepted] version of the Torah, and only one [accepted] version of Tehillim, and one [accepted] version of Divrei Ha-yamim" (Teshuvot ha-Geonim, Harkaby, siman 3). He refers here to the Sefer Torah found in the Synagogue of Severus, which we mentioned in the previous shiur. This explanation raises its own difficulties, since a Sefer Torah contains no division into verses, and in any case it is difficult to see how Sefer Tehillim could possibly contain anywhere near as many verses as a Sefer Torah. For additional proposals see Segal, p. 886.
 Rashi, in contrast, maintains that the Masora version is correct: "This [opinion of Rav] is greatly mistaken, and we do not accept that the word should be written that way, for our manuscripts read 'ma'avirim.’"
 The approach of the Tosafot and other Rishonim contrasts with that of the Sages of Babylonia, who consistently maintain that it is impossible for there to be any contradiction or discrepancy between the version noted in the Gemara and the version of the Masora. If there are "verses that we find in the Talmud that do not appear [in that form] in the Tanakh,” they are not the result of Chazal using a text with errors. Rather, "our Sages would not make a mistake with a verse, for in their teachings they invest great effort so that the teaching of one Sage will not become muddled with the teaching of a different Sage, each taking care to cite [the law] as he heard it from his teacher; how much more so, then, [would they exercise caution] with the words of the Torah and Tanakh" (Teshuvat Rav Hai Gaon, Teshuvot Geonim Kadmonim, siman 78). Therefore, in this view, the phenomenon must be explained in one of three ways: "Either it is a scribal error, or it the teachings were disseminated by disciples who were not proficient [in the text], or the words were never uttered in the first place as the quotation of a verse."
 Rabbi Yishayahu Pick-Berlin, in his comments scattered throughout the Gemara (Vilna edition), notes this phenomenon and compiles a list of instances in his work entitled Hafla'ah she-be-Arakhin (on the Arukh), under the entry "meah.”
Including Tosafot Ha-Rosh on Berakhot ad loc.; and Rashba and Ritba on the parallel unit in Eruvin.
Maharshal, in his commentary Chokhmat Shlomo on Berakhot ad loc., writes: “The comments of Tosafot, who completely erase the (gemara’s) version of the text, are difficult in my opinion. Of course, textual variants can be expected from time to time in the Talmud, but to utterly dismiss a version of the text and to maintain that it was mistakenly written is hard to accept.”
Maharshal maintains that the quote in the Gemara is referring to the verse that appears in different form in the Masora text, and that "it is the way of the Gemara to alter verses.” As we find in the Hagahot ha-Shas there, Maharshal's explanation contradicts Rashi's approach.
See Rosenthal, pp. 400-401 and notes.
See Y. Maori, "Midrash Chazal ke-Edut le-Chilufei Nussach ba-Mikra,” M. Bar-Asher et al. (eds.), Iyyunei Mikra u-Parshanut 3, Ramat Gan 5753, p. 273 n. 32, and in the appendix, pp. 283-286.
 In the Responsa of the Rashba attributed to Ramban, siman 232.
The citations are from the Responsa of the Ridbaz noted above (part III siman 592). A similar view is expressed by the Meiri in his Kiryat Sefer (ma'amar sheni part II, Herschler edition p. 58).
Minchat Shai is the name of an important work by Rabbi Shelomo Yedidia Nortzi, published in 1626, that is a guide in matters relating to textual variants of the Tanakh. For more about the work and the author, see Z. Betzer, Minchat Shai al Chamisha Chumshei Torah, Jerusalem 5765, pp. 3-50.
 On Vayikra 4:34, Betzer edition p. 237.