Torat Emet - Textual Criticism Of The Torah

 Home    Introduction    Talmudic Issues    Essays  

 Pamphlets    Parsha    Original Essays    Contact Us 


On the Text of the Torah

by Gil Student

There is no text more sacred to Jews than the Torah. While it is not the sole source of our religion, it is a holy work whose every word is precious. It is therefore a critical task, and a religious imperative, to determine whether the text of the Torah that we have is correct and, if not, to fix it. However, because of the importance of this mission, there is no room for hasty judgements or speculative decisions. This is a very serious undertaking that requires the proper sense of humility.

In this essay, we will list the various evidence we have for the Torah - what are called the "witnesses" of the text. Additionally, we will discuss the usefulness of each witness and, in this, depart from the standard academic method. Emanuel Tov wrote in what quickly became the standard handbook on textual criticism of the Bible, "[M]any scholars, including the present author, believe that all readings which have been created in the course of the textual transmission ought to be evaluated " (Emanuel Tov, Textual Criticism of the Hebrew Bible, p. 295). We disagree. Almost anyone who attends synagogue regularly has witnessed the finding of a mistake in a Torah scroll. An average Torah has some mistakes and therefore the precise reading of any given word is suspect. There are, however, better than average scrolls and even excellent scrolls that have been reviewed carefully many times. Only those witnesses that are known to be excellent scrolls are valid evidence. Average scrolls, such as the one in our synagogue, can hardly be used as proof of the original Torah text.

In addition to this, we will also depart from academics in the following. We will assume that, absent any evidence to the contrary, the text that is agreed upon by the Jewish community - the textus receptus that is claimed to be the Masoretic Text - is correct. In halacha, this is called chazaka. We will abide by the status quo until it is categorically disproven and only in those areas that it is disproven. Thus, if it is demonstrated that the textus receptus is inaccurate in its differentiation between א and ה - sometimes switching them - we will draw no conclusion regarding the differentiation between masculine and feminine nouns. Those are very different areas. Just because - and if - scribes made mistakes between א and ה does not mean that the entire text is worthless. It only means that any given א might be a ה and vice versa. However, if we detect evidence that the scribes who wrote this text were consistently careless in many areas or had no compunction regarding intentionally changing the text, then we have no choice but to disregard the entire text.

The Samaritan Torah

The first witness we will consider is the Samaritan Torah. The Samaritans are an ancient group that was transplanted by the Assyrians from their native Kutah to the land of Israel.  They never fully assimilated into the Jewish people but adopted many of their practices (2 Kings 17:24-34). When the Jews returned from the Babylonian exile, the Samaritans strongly opposed them and particularly objected to their rebuilding of the Temple in Jerusalem (Ezra 4). One important theological change the Samaritans implemented was the relocation of the holy site from Jerusalem to Mount Grizim. The remnants of this group can still be observed sacrificing a lamb annually and offering it on Mount Grizim as a Passover sacrifice.

The Samaritans retain a Torah that is still written in the script used during the First Temple period (which we will discuss elsewhere). However, there are many differences between the text of the Samaritan Torah and ours. Most obvious are ideological changes that represent the different Samaritan beliefs such as the importance of Mount Grizim. Their Torah actually contains an extra commandment in the "Ten Commandments" that emphasizes the sanctity of this mountain. There are also a number of "harmonizing alterations" that appear to be obvious attempts to smooth differences between similar texts. Many commandments and narratives in the Torah are repeated in different terms. The Samaritan scribes changed words to make the different versions match. Overall, there seem to have been many intentional changes in the text of the Torah.

If so, the Samaritan Torah fails our test of being a reliable witness. Any given difference between the Samaritan text and ours can be due to the free hand the Samaritan scribes exhibited in developing their Torah. As Tov writes, "What characterizes the scribes of [the Samaritan Torah] and the Pre-Samaritan texts is the great freedom with which they approached the biblical text" (Tov, p. 85).

Furthermore, R' Chaim Heller has demonstrated that the Samaritan Torah has a consistent tendency of modifying the text as a form of interpretation. As he writes, "[I]t is evident that the Samaritan's deviations from the Massoretic text must be looked upon as explanations of this same text but not as proving the existence of a deviating original text" (R' Chaim Heller, The Samaritan Pentateuch: An Adaptation of the Massoretic Text, p. 3). We will look at examples later in this essay when we discuss this same phenomenon in ancient translations. However, this is very similar to the harmonizing of texts that has been previously mentioned. Additionally, we find slight changes of wording in the Samaritan Torah to fit the text to the Samaritan dialect of Hebrew. For example, the Samaritan scribes consistently removed verbs in the infinitive. Also, certain Aramaic words found their way into the text. Both of these types of changes reflect the translation of the biblical text into the Samaritan dialect, a phenomenon very consistent with the previously mentioned commentaries that were also inserted into the text.

Examples of Removal of Infinitive in the Samaritan Torah

Textus Receptus

Gen 8:3 וישובו המים על הארץ הלך ושוב

Gen 11:7 ויצא יצוא ושוב

Ex 13:3 זכור את היום הזה

Num 25:17 צרור את המדינים

Samaritan Text

הלכו ושבו . . .

יצא ושב . . .

. . . זכרו

. . . צררו

(Heller, pp. 186-187)

Examples of Aramaic in the Samaritan Torah

Textus Receptus

Gen 21:20 רבה קשת

Gen 22:2 יחידך

Ex 22:24 לא תהיה לו כנשה

Deut 4:6 חכם

Samaritan Text

רבי קשת

יחידאך

כנשיא . . .

חכים

(Heller, pp. 187-188)

Given all this, in our search for proof against the presumptive textus receptus the Samaritan Torah cannot offer sufficiently convincing evidence. At most, it can teach us about the Samaritans themselves.

Dead Sea Scrolls

In 1947, scrolls were discovered in the Qumran caves that changed the way we understand history. These voluminous texts teach us about a sectarian group who, beginning in the early second century BCE and continuing into the first century CE, retreated into seclusion to live lives of religious isolation. Among the library of this sect were over 200 manuscripts or fragments of biblical books. These texts have proven invaluable in understanding the Qumran sect and their attitudes towards the Bible (see Lawrence H. Schiffman, Reclaiming the Dead Sea Scrolls, ch. 10). However, our question is whether they can inform us about the true text of the Torah. While it is certainly tempting to rely on these documents of unparalleled antiquity, we must make sure that we do not succumb to the methodological pitfall of using them simply because they are available.

Qumran Manuscripts of Books of the Bible
Genesis                                       18 + 3?

Exodus                                        18

Leviticus                                     17

Numbers                                     12

Deuteronomy                              31 + 3?

Early Prophets                            12

Later Prophets                            45 + 1?

Writings                                      69 + 3?

(Schiffman, p. 163)

Of the many biblical texts found, approximately 100 of them being of the Torah, 20% have been identified as being of distinctly Qumran origin. This means that scribes in the community copied them from outside texts. Of these texts Tov writes, "From the great liberties which these scribes took it is evident that they do not reflect a tradition of precise and conservative copying, but rather a popular or vulgar one" (Tov, p. 108). However, the other 80% of texts were brought into the community from elsewhere and cannot be assumed to have these drawbacks. Three quarters of these manuscripts are almost identical to the Masoretic text. There are, though, minor differences. The other quarter of these texts show marked differences from the Masoretic Text that we have. Can these texts shed doubt on the textus receptus? Does the wide variety of different versions tell us that the Masoretic Text was only one of many versions of the Torah that were circulating at that time? To answer this, we must first understand the process of transmission of the Torah until that time.

Transmission of the Torah

At the end of his life, Moshe wrote thirteen Torahs - one for each of the twelve tribes and one for the priests to keep in the ark (Devarim Rabbah 9:9; Deut. 31:9, 24-26). The Torah only specifies the priestly scroll, but that might be due to the importance this text would play in history. The priests safeguarded this Torah and brought it to the Temple when it was built. Sefer Tagi (introduction in Torah Shelemah vol. 29 p. 87), an ancient work on the letters of the Torah, tells us that Eli the priest (of 1 Samuel) wrote a book with details of the Torah's script based on another of Moshe's writings. This, too, was kept in the Temple. Of what purpose were these books?

Professor Saul Lieberman explained based on Greek and Roman sources how books were published in the ancient world.

The authentic original copy would be deposited in a temple, a library or the archives. Such an act guarded the book against possible forgeries. In case of doubts or controversies regarding readings in the given book, the copy placed in the archives would be decisive.

(Hellenism in Jewish Palestine, p. 85)

Thus, Moshe's Torah scroll was kept in the Temple as a reference. Due to the antiquity and historical importance of this scroll, there were probably other copies of it that were kept in the Temple for everyday use. This would minimize the wear-and-tear on the important Torah of Moshe. Whenever a Torah scroll was written, it would be written based on these codices to guaranty the accuracy of its text. This ensured the integrity of the Torah through much of the Temple period.

But scrolls copied (indirectly) from Moshe's Torah were not the only texts written by the scribes. At that time in history, the only books of religious material that Jews were allowed to have were the biblical works (Gittin 60a). Therefore, there was a need to mass produce these books. Since scrolls written according to the demanding halachic requirements take a long time to write, they are impractical for mass production. However, an alternative to a scroll is what the Talmud called a megillah letinok. In the ancient times there was no printing press, of course, so Bibles were handwritten. However, in order to speed the writing to enable the mass production necessary to have teaching materials for schoolchildren, the normal stringencies required for a scroll were ignored. Scribes wrote from memory and without concern for exact spelling. Editorial insertions and even an occasional explanatory remark were not uncommon, similar to popular translation. These works were intended for the masses to use as study tools and were understood by all to be just that. In the terminology of the ancient world, these works were called vulgata (Lieberman, p. 25). It seems to reason that Jerusalem, home to the archive copies of the Torah and many learned priests intent on teaching G-d's word, was also the base of publishing for the vulgata.

Going back to Moshe's Torah, we find a possible reference to it in 2 Kings 22:8-11:

Then the high priest Hilkiah said to the scribe Shaphan, "I have found a scroll of the Torah in the House of the Lord." And Hilkiah gave the scroll to Shaphan, who read it. The scribe Shaphan then went to the king and reported to the king: "... The high priest Hilkiah has given me a scroll"; and Shaphan read it to the king. When the king heard the contents of the scroll of Torah, he rent his clothes.

R' Yitzchak Abarbanel asked how it is possible for Torah scrolls to become so scarce that the finding of one and its reading made such a deep impression on the king. With prophets like Yirmiyahu and scribes like Baruch ben Neriah alive, how was it possible that the righteous king had never seen a Torah? Abarbanel suggested that during the reigns of the wicked kings the Torah of Moshe was hidden by the priests in Jerusalem. When Yoshiyahu (Josiah) the righteous king came to power, the priests searched for and found this Torah and brought it to the king. One can only imagine the awe the king felt when he held and read from the Torah written by Moshe on the last day of his life.

What happened to this Torah of Moshe is unclear, although there are rumors of it resurfacing later in history. However, the history of the text continued. Crucial to this development of the text of the Torah was the Babylonian exile which was really much more than the word exile connotes. It was the total devastation of a civilization. Scholars and leaders were killed; homes and cities were destroyed. The survivors of this long and brutal struggle were forced to leave their homes with what they could carry and move to a foreign land. It was at this point, Radak in his introduction to Yehoshua tells us, that "the books were lost and dispersed and the sages who were skilled in Bible were dead." All the people had was what little they had brought with them and what they remembered. Many of the texts were damaged due to the battle and travel conditions. Of the texts available, some were actual Torah scrolls while others were vulgata which, as we discussed, were never intended to be exact replicas of the Torah. However, each community had to make do with what they had and they courageously tried to fix their Torahs based on the best information available. As Radak suggested, it was at this point that multiple spellings and even word differences were introduced into the biblical texts. The nation was roiling in shock from its terrifying losses and was not able to recreate the centralized Torah leadership it once had. But complete Torah scrolls were an immediate need and could not wait for a new generation of leaders to organize a unified text. Therefore, each community acted on their own to fix their scrolls of missing words and portions.

It was not until decades later, when the Persian government allowed the Jews to return to Israel, that talented leaders could once again assert authority and try to unify the people under a single text. Such a unifying gesture would surely have been viewed with suspicion by a hostile Babylonian overlord. However, the benevolent Persian kings encouraged Jewish unity and thereby gave the Men of the Great Assembly, as the leading rabbis of the time were called, the opportunity to gather the Jewish people together under the banner of Torah. We are told in Masechet Sofrim 6:4:

Three books were found in the [Temple] courtyard - the maon book, the za'atutei book, and the hi book... so they retained the [majority] reading of the two and abandoned the [minority] of one.

In each of the three Torahs that were found, there was a unique textual variant and the sages followed the majority. In other words, when Ezra returned to Jerusalem to rebuild the Temple, he was able to find three reliable scrolls with minor differences. The differences were as follows. In one, the word נערי was written in its Aramaic translation זאטוטי. In another, the word מעונה was written מעון without the final ה. And in the third, the word היא was written as הוא (but vocalized as 'hi') in eleven places. Other than these, the texts matched exactly which, frankly, is outstanding for texts produced by human hands.

Ezra and his colleagues used these three Temple scrolls to correct their texts and decided the very few controversies based on majority. This use of majority will serve as a basis for textual decisions later in history. (That this passage refers to the time of Ezra, see Torah Shelemah vol. 19 p. 254 n. 29, vol. 29 p. 102; Rashi, 1 Chronicles 8:29; Radak, Introduction to Yehoshua; Meiri, Introduction to Kiryat Sefer). This activity is spoken of in 4 Ezra which, although being full of later Christian interpolations, contains a core of ancient Jewish tradition. It reads (4 Ezra 14:44) "And it came to pass when the forty days were fulfilled that the Most High said to [Ezra] saying: Publish the twenty four books you have written so that the worthy and unworthy may read them."

This Torah that Ezra wrote based on three ancient scrolls with almost no differences has played a special role in history. Some claim that it remained until medieval times. However, in the Temple times this text was kept in Jerusalem and carefully copied by scribes for distribution. "We do not fix even one letter [on Chol Hamoed] even in Ezra's scroll" (Mishna Moed Katan 3:4). See Rashi on Moed Katan 18b that this might refer to the Torah scroll that Ezra wrote. (Fixing does not necessarily mean correcting an error. It can simply mean fixing a faded letter, which would be expected in an ancient scroll.). See also Mishna Kelim 15:6 and Tosefta Kelim, Bava Metzia 5:7. "The reviewers of the scrolls of the [Temple] courtyard receive payment from the communal trumat halishkah" (Yerushalmi Shekalim 4:2). These scribes served the public by disseminating accurate versions of the Bible. This was crucial because after the Babylonian exile there were many incorrectly "fixed" texts that were circulating. The accurate version that Ezra had prepared based on trustworthy manuscripts needed to be disseminated widely. These scribes, who were most likely priests, are mentioned by Josephus (Antiquities 12:3:3) as "the scribes of the Temple" and repeatedly throughout the Christian bible (e.g. Matthew 2:4, 5:20, 7:29, 9:3,...). Thus, the authorized version of the Bible was set by Ezra and the Men of the Great Assembly - the predecessors of the Pharisees (see Mishna Avot 1:1) - and was maintained by the priestly scribes in the Temple.

Back to Qumran

It is with all this mind that we can return to the manuscripts found at Qumran. Lawrence Schiffman characterized the Qumran sect as a group of renegade Sadducees who were opposed to the Pharisaic practices in the Temple. Schiffman wrote (p. xxii), "When the victorious Hasmonaean rulers adopted the rulings of the Pharisees regarding the conduct of the Temple in about 152 BCE, the loyal opposition - a band of pious Sadducees - retreated to the desert, taking up residence in Qumran." As an act of rejection, they secluded themselves from the general public and planned for the restoration of the Temple to its true (in their view) non-Pharisaic practices. As a group that rejected the Pharisee rabbis and the priests in the Temple, is it any surprise that they rejected the authority of the text authorized by Pharisees and maintained by scribes in the Temple? The surprise is that 80% of their outside texts were proto-Masoretic. What likely happened was that they recognized the damage that the exile did to the biblical text and accepted any corrected version. The preponderance of proto-Masoretic texts in Qumran demonstrates the success the scribes were having in replacing incorrectly "fixed" texts with Ezra's Temple-based version. However, even wrong traditions die hard and the existence of sects that rejected the authority of rabbis and the Temple made the total domination of Ezra's corrected text impossible. Since 20% of the texts brought to Qumran from outside were non-Masoretic, there was still a significant minority that were clinging to their old, battered exile versions.

This might also explain why scribes at Qumran did not exhibit "a tradition of precise and conservative copying." We have already noted that Emanuel Tov suggested that the scribes were writing vulgata, a concept with which we have become familiar. It could also be suggested that the Qumran sect, in their rejection of the Pharisaic leadership and the Temple scribes, accepted a fluidity of texts and therefore saw no reason to adhere carefully to previous texts. This might be contradicted by the midrash halacha that is evident in Qumran writings that implies a certain sanctity was given to the text. Regardless, while the Dead Sea Scrolls have opened up windows into history and have given us tremendous information about the Qumran sect and the world in which it lived, they cannot tell us about the precise text of the Bible. Because the sect accepted deficient texts, we cannot know whether any given manuscript represents Ezra's authorized version or another text. Therefore, alternate spellings and wordings that are suggested by the Dead Sea Scrolls can teach us much about history but little about the original biblical text.

Ancient Translations

Ancient translations are often used as "witnesses" of the biblical texts when these translations are retranslated back into Hebrew to demonstrate the vorlage - the text underlying the translation. This vorlage might differ from the textus receptus, offering us an ancient witness to a textual variant. For example, Genesis 2:19 reads ויבא אל האדם ([And the Lord G-d formed out of the earth all the wild beasts and all the birds of the skey,] and brought [them] to the man...) but retranslating the Septuagint into Hebrew indicates a vorlage of ויבא אותם אל האדם This method of reconstructing the vorlage shows us that the translator of the Septuagint had אותם in this verse and offers us a textual variant.

There are a number of ancient translations and we will briefly review the important ones. Unquestionably, the most important ancient translation is the Septuagint. The Gemara in Megilla 9a-b, which is supported by the Letter of Aristeas, says that the Egyptian king Ptolemy (II Philadelphus) had seventy two rabbis independently translate the Torah into Greek in the third century BCE. Miraculously, all of the rabbis' translations were identical. Whether that rabbinic translation is the Septuagint that we have today is unclear (see Lawrence Schiffman, From Text to Tradition, p. 92). After the translation of the Torah, other books of the Bible were translated into Greek over time.

The Peshitta is a translation of the Bible into Syriac, a dialect of Aramaiac, that was written in the second century CE. The oldest manuscript available dates from 460 CE. The Vulgate is a translation into Latin by the Christian church father Jerome in the late fourth century CE. Targum Onkelos is an Aramaic translation of the Bible written in the first century CE but composed, according to tradition, in the time of Ezra. Targum Yonatan as well as various Targum Yerushalmis are also translations of the Bible from the first century CE.

Using the method of reconstructing the vorlage of a translation has a number of pitfalls. A translation is by definition an interpretation of the original text. In order for someone to translate the Bible he must first understand it, which requires interpretation. This understanding will then be rendered in the translation. Furthermore, a good translation, which these ancient translations most certainly are, cannot simply render the Hebrew word for word, "a procedure which nearly always results in quaintness or awkwardness and not infrequently in obscurity. A translation which is stilted where the original is natural, heavy where the original is graceful, or obscure where the original is perfectly intelligible, is the opposite of faithful" (Preface to the JPS 1962 translation of the Torah). A translation that seeks to bring the Bible to an audience speaking a different language must adapt biblical Hebrew into the idioms and language structure of this new language. This sometimes requires changing word order, adding conjunctions (e.g. and), and adding pronouns. The Bible might be vague about the subject of a phrase but the new language does not allow similar vagueness. The translator must then interpret the verse and add an object to the text.

In the example given above from Genesis 2:19, the Septuagint seemed to add the word אותם into the verse, offering us a textual variant. But is that really a textual variant or just a function of translation? In the early tenth century, R' Saadia Gaon translated the Torah into Arabic. He undoubtedly used the Masoretic text, as can be seen from the criticism of his translation by colleagues living shortly after him. They only objected to some of his translations as interpretations of the Masoretic text but never questioned the text he used (see Heller, pp. 27-28). R' Saadia Gaon also inserted the Arabic equivalent of אותם into his translation. In modern translations, we can find the pronoun "them" in the JPS 1962 Torah translation and the Artscroll Stone Tanach translation. They both used the textus receptus but still implied a different underlying Hebrew text than was actually used. This is the nature of translation. The following chart will give a few of the many examples brought by R' Chaim Heller of this phenomenon. We include examples from the Samaritan Torah as explained above.

Examples of Interpolations into the Text by Translators
Textus Receptus
Gen 14:14
וירדף עד דן
"and he pursued until Dan"
 
 
Gen 18:14
למועד
אשוב אליך
כעת חיה
"I will return to you at the time that life is due"
 
Ex 21:36
או נודע
"Or it becomes known"
 
Ex 23:13
לא ישמע
על פיך
"They shall not be heard on your lips"
 
Ex 25:9
וכן תעשו
"And so shall you do"
Septuagint
 
וירדף אחריהם
עד דן
"and he pursued after them until Dan"
 
 
למועד הזה
אשוב אליך
כעת חיה
"I will return to you at this time when life is due"
 
 
ואם נודע
"But if it becomes known"
 
 
לא ישמע
מפיכם
"They shall not be heard from your lips"
 
 
כן תעשו
"So shall you do"
R' Saadia Gaon
 
וירדפם
עד דן
"and he pursued them until Dan"
 
 
למועד הזה
אשוב אליך
כעת חיה
"I will return to you at this time when life is due"
 
 
ואם נודע
"But if it becomes known"
 
 
לא ישמע
מפיך
"They shall not be heard from your lips"
 
כן תעשו
"So shall you do"
Artscroll
 
וירדף אחריהם
עד דן
"and he pursued them as far as Dan"
 
 
למועד הזה
אשוב אליך
כעת חיה
"At the appointed time I will return to you at this time next year"
 
 
ואם נודע
"But if it becomes known"
 
 
Inconclusive
"Nor shall your mouth cause it to be heard"
 
 
 
וכן תעשו
"And so shall you do"
JPS
 
וירדף עד דן
"and went in pursuit as far as Dan"
 
 
 
למועד
אשוב אליך
כעת חיה
"I will return to you at the time that life is due"
 
 
ואם נודע
"If, however, it is known"
 
 
לא ישמע
על פיך
"They shall not be heard on your lips"
 
 
כן תעשו
"So shall you do"

 
Textus Receptus
Deut 4:6
ועשיתם
"And you shall perform"
 
Deut 5:27
שובו לכם
לאהליכם
"Return, you, to your tents"
 
Deut 15:15
עבד היית
בארץ מצרים
"You were a slave in the land of Egypt"
Peshitta
 
ועשיתם אותם
"And you shall perform them"
 
 
שובו
לאהליכם
"Return to your tents"
 
 
עבד היית
במצרים
"You were a slave in Egypt"
R' Saadia Gaon
 
ועשיתם אותם
"And you shall perform them"
 
 
שובו
לאהליכם
"Return to your tents"
 
 
עבד היית
במצרים
"You were a slave in Egypt"
Artscroll
 
ועשיתם אותם
"And you shall perform them"
 
 
שובו
לאהליכם
"Return to your tents"
 
 
עבד היית
בארץ מצרים
"You were a slave in the land of Egypt"
JPS
 
Inconclusive
"Observe them faithfully"
 
 
שובו
לאהליכם
"Return to your tents"
 
 
עבד היית
בארץ מצרים
"You were slaves in the land of Egypt"

 
Textus Receptus
Gen 3:16
אל האשה אמר
"To the woman He said"
 
Gen 10:
בארצתם
בגויהם
"In their lands, in their nations"
 
Gen 17:1
בן תשעים
שנה ותשע
שנים
"Ninety years and nine years old"
Samaritan
 
ואל האשה אמר
"And to the woman He said"
 
 
בארצתם
לגויהם
"In their lands, by their nations"
 
 
בן תשעים
ותשע שנים
"Ninety-nine years old"
R' Saadia Gaon
 
ואל האשה אמר
"And to the woman He said"
 
 
בארצתם
לגויהם
"In their lands, by their nations"
 
 
בן תשעים
ותשע שנים
"Ninety-nine years old"
Artscroll
 
אל האשה אמר
"To the woman He said"
 
 
בארצתם
בגויהם
"In their lands, in their nations"
 
 
בן תשעים
ותשע שנים
"Ninety-nine years old"
JPS
 
ואל האשה אמר
"And to the woman He said"
 
 
לארצתם
ולגויהם
"By their lands and nations"
 
 
בן תשעים
ותשע שנים
"Ninety-nine years old"

The above examples are only a sample of the many textual variants from the Septuagint, Peshitta, and Samaritan Torah that are nothing more than natural adaptations of the text. This can be seen by comparing them with the translations of R' Saadia Gaon, Artscroll, and JPS that undoubtedly used the textus receptus.

Another source for implied textual variants is commentary inserted in the text. By comparing the "textual variant" with later commentaries that were unquestionably based on the textus receptus and never saw these ancient translations so could not have been influenced by them, we will see that these changes to the biblical text are legitimate commentaries of the original Masoretic Text. For example, Genesis 7:3 in the textus receptus reads גם מעוף השמים שבעה שבעה (Of the birds of the sky also, seven pairs). The Samaritan Torah has גם מעוף השמים הטהור שבעה שבעה (Of the pure birds of the sky also, seven pairs). This, of course, is the standard explanation that can be found in Rashi and many other commentaries.

Textus Receptus
 
Ex 21:13
ואשר לא צדה
"But for one who has not lain in ambush."
 
 
 
Num 12:8
ותמנת יקוק יביט
"At the image of G-d he gazes."
 
 
Num 24:17
ומחץ פאתי מואב
"And he shall pierce the nobles (lit.: corners) of Moav."
Septuagint
 
 
ואשר לא כוון
"But for one who has not intended."
 
 
 
 
 
וכבוד יקוק יביט
"At the glory of G-d he gazes."
 
 
 
ומחץ ראשי מואב
"And he shall pierce the leaders of Moav."
Commentary
R' Yonah Ibn Janach, Sefer Hashorashim, sv. צדה
... לענין ואשר לא צדה
הכונ המהדבר
"The meaning of 'But for one who has not lain in ambush' is intention."
 
Ibn Ezra, ad loc.
כטעם הראני נא
את כבודך
"In the spirit of 'Show me now Your glory' (Ex 33:18)".
 
Radak, Jer. 48:45
פאת מואב: דרך
משל על הגדולים
"The corner of Moav: An allegory for the leaders."

 
Textus Receptus
 
 
Ex 20:21
בכל המקום אשר
אזכיר את שמי
"Wherever I permit My name to be mentioned."
 
 
 
Num 32:7
ולמה תניאון את לב
בני ישראל
"Why do you dissuade the heart of the Children of Israel?"
 
Deut 15:11
לאחיך לעניך
ולאביונך בארצך
"To your brother, to your poor, to your destitute in your land."
Peshitta
 
 
 
בכל המקום אשר
תזכיר את שמי
"Wherever you mention My name."
 
 
 
 
ולמה תשברון את לב
בני ישראל
"Why do you break the heart of the Children of Israel?"
 
 
לאחיך העני
ולאביון שבארצך
"To your your poor brother and to the destitute in your land."
Commentary
R' Shlomo Parchon, Machberet Ha'aruch, chelek hadikduk, sha'ar hachiluf
- אשר אזכיר את שמי
משפטו אשר תזכיר את
שמי
"'Wherever I permit My name to be mentioned' - This means, 'Wherever you mention My name'."
 
Ibn Ezra, ad loc. based on Tanchuma, Shelach 9
תניאון כמו תשברון
"Dissuade - like break."
 
 
 
Ramban, ad loc.
כאומר לאחיך העני
ולכל אביוני ארצך
"As if it says, 'To your your poor brother and all the destitute in your land'."

 
Textus Receptus
 
Gen 37:35
כי ארד אל בני
אבל שאולה
"For I will go down to the grave mourning to my son."
 
 
 
Ex 15:25
ויורהו יקוק עץ
"And G-d showed him a tree."
 
 
 
Num 26:10
ויהיו לנס
"And they became a sign."
 
 
 
Samaritan
 
 
כי ארד על בני
אבל שאולה
"For I will go down to the grave mourning for my son."
 
 
 
 
ויראהו יקוק עץ
"And G-d showed him a tree."
 
 
 
 
ויהיו לנוס
"And they fled."
Commentator
R' Yonah Ibn Janach, Sefer Hashorashim, sv. אלל
ותהיה אל בענין על
כמו כי ארד אל בני
אבל שאולה
"'To' can mean 'for' as in 'For I will got down to the grave mourning for my son'."
 
Ramban, ad loc.
וטעם ויורהו יקוק עץ
שהראה אותו עץ
"The meaning of 'And G-d showed him a tree' is that He showed him a tree."
 
Midrash Aggadah, ad loc.
ויהיו לנס - על שם ובני
ישראל נסו לקזלם
"'And they became a sign' - like it says, 'And the Children of Israel fled at their sound' (based on Num 16:34)."

We have seen that textual variants in the ancient translations can be due to either the nature of translation from one language to another or interpretation by the tranlsator. These reasons explain the vast majority of variants, as Tov (p. 123) writes:

Although there are thousands of differences between [the Masoretic Text] and the translations, only a fraction of them was created by a divergence between [the Masoretic Text] and the Vorlage of the translation. Most of the differences were created by other factors that are not related to the Hebrew Vorlage.

However, there are a tiny minority of variants that cannot be explained this way. The questions we need to ask is whether the source of these few variants is a different text from the textus receptus and if that text is more accurate. Let us consider a much-celebrated example.

Children of Whom?

Deut 32:8

בהנחל עליון גויים              When the Most High (Elyon) gave nations their homes
בהפרידו בני אדם                And set the divisions of man,
יצב גבולות עמים               He fixed the boundaries of peoples
למספר בני ישראל              In relation to Israel's numbers.

The medieval commentators were divided on how to explain the last two phrases. Rashi understood it as meaning that G-d established seventy nations corresponding to the seventy members of Yaakov's (Yisrael's) family that went down to Egypt (see Gen. 46:27). Rashbam explained it as meaning that there were twelve nations in the land of Israel corresponding to Yaakov's twelve sons. Ibn Ezra suggested that the text does not imply a specific number but just means the Jewish population. Similarly, Sforno said that it refers to Israel's small numbers.

However, the Septuagint implies a variant that gives the verse a different meaning. The Septuagint has the last phrase as למספר בני אלים thus changing the meaning to the following: "He fixed the boundaries of nations in relation to the number of gods." Scholars in the 19th century debated which version was better but a discovery in the early 20th century greatly changed the debate. Thousands of Ugaritic (ancient Canaanite) tablets were found with a literature that enriched our understanding of their culture. We learned that their mythology had a god named El who, with his wife Ashera, had seventy children who were themselves considered gods. These seventy children of El - in Hebrew בני אל - seemed to be the referent in our verse. This appeared to confirm the Septuagint's version of the text. However, the case was clinched with the following fragment that was later found in Qumran:

[בהנחי[ל
] בני אל

While this does not seem like evidence of much, its placement in the manuscript makes it clear that this refers to our verse. This text definitely does not refer to בני ישראל but to בני אל or בני אלים, like the Septuagint's version. Should we change our Torah scrolls to reflect this well-attested variant?

Not so fast, wrote Professor Joshua Grintz in his Studies in Early Biblical Ethnology and History (Motzaei Dorot, pp. 242-257). For one, Ugaritic mythology has many more than seventy gods. With one wife, El had seventy children. But with his other wives he had hundreds more. There were more than seventy בני אל (children of El). Additionally, the number seventy appears nowhere in the text! It is simply a matter of commentary with which many commentators, such as Rashbam and Ibn Ezra, disagreed. The seventy children of El and Ashera correspond to Rashi's interpretation of the text and not the text itself.

Also, the beginning of the verse refers to Elyon, a not uncommon name for G-d (e.g. Tehillim 91:1,9; 92:2) but also the name in Ugaritic mythology for El's grandfather who was killed before El was born. If this verse is a reference to ancient pagan myths, as the Septuagint version implies, why are Elyon and El mentioned together when they lived in different generations?

Furthermore, the Septuagint refers to the children of elim - gods - and not the children of El. While the phrase "children of elim" can be found elsewhere in the Bible, nowhere is "children of El" found as a complete phrase. We therefore see that Ugaritic literature and the Qumran scrolls provide no support the Septuagint version.

We must also ask whether a benign reference to pagan mythology is in place in a poetic tribute to the one G-d of Israel. Why would an author - whether G-d, Moshe, or someone living at the revivalist time of Yeravam ben Yoash (as some scholars suggest) - insert such a reference? It lacks conformance to the literary theme of the section, to say the least.

What seems more likely is that the Septuagint version is based on a text that had been damaged and an exiled Babylonian community tried to fix it to the best of its ability. Familiar with the term "children of elim" (Tehillim 29:1) they thought that this was what the damaged text should read. After this text was "fixed" and temporarily accepted, it was copied and eventually made its way into the Qumran community and the final rescension of the Septuagint.

We asked before whether the few remaining variants from the Septuagint represent a different text from the textus receptus and whether that text was more accurate. What we have seen from the above example is that this difference originated in a variant text, but an inferior one. The textual variant comes from an early mistake that was never corrected. If so, we must be very skeptical of variants from the Septuagint. There is the possibility, if not the probability, that these different readings are born of an errant source.

Some modern scholars have maintained that the Hebrew text of the Bible should be revised on the basis of the textual differences in the Greek translation. It is indeed true that certain readings of the Septuagint have been confirmed by the discovery of ancient manuscripts. Yet care must be exercised in view of the complex history of the Greek biblical text and the fact that differing biblical texts are known to have existed in ancient times.

(Schiffman, From Text to Tradition, p. 94)

What we have seen is that the Babylonian exile brought an era of multiple texts in which there were many versions of the Bible. Ezra and the scribes found reliable manuscripts and tried to end this multiplicity. To some degree, they succeeded. Although, the existence of ancient vulgata and scrolls with inevitable scribal errors misleadingly imply only mitigated success. However, the uniformity of text that Ezra initiated did not spread to all sectors of society, particularly not to those who actively rejected the central authorities of Judaism. Thus, the Qumran sect accepted a multiplicity of texts that most respectable rabbis of the time would have rejected.

Because of the circulation of these unauthorized texts, non-rabbinic manuscripts and translations are suspect of - perhaps intentionally - using these discredited versions. Scholars note with surprise that following the destruction of the Second Temple the Masoretic Text almost immediately became the textus receptus. How could the era of multiple texts end so quickly? Scholars explain this with elaborate schemes of rabbis hunting down deviant texts and implementing regulations to raise up the selected text. However, with our understanding the matter is simple. The Masoretic Text had been the textus receptus since the time of Ezra and the scribes. The deviant texts were mainly the province of deviant sects that disappeared after the destruction of the Temple. The swift ascendance of the Masoretic Text was merely the survival of the dominant Pharisees and their ancient authorized version.

Talmud and Midrash

Another source for textual variants is the rich and voluminous literature of rabbinic exegesis. However, before we approach this topic we need to divide this category into three types of deviations from the textus receptus. The first is when the Talmud or Midrash cite a verse and quote it differently than the way it appears in the textus receptus. For example, in Pesachim 33a we find Daniel 4:14 quoted as ובמאמר קדישין instead of ומאמר קדישין, as it appears in our texts. Similarly, Berachot 17a tells us that Abaye was fond of quoting Proverbs 15:1 - "A gentle reply turns away (משיב) wrath". However, our texts of the verse read, "A gentle reply will turn away (ישיב) wrath". Do these and many other instances of this phenomenon imply a multiplicity of texts during the time of the Talmud?

Many of these "variants" can be attributed to copyist error. The written transmission of the Talmud is notoriously poor in terms of precise lettering and wording. Anyone who has studied Talmud or Midrash is familiar with the frequent textual emendations by the Gra, Bach, Radal, and others. In the medieval commentaries it is very common to see the text quoted differently, demonstrating that the text the commentators had was different than the one we have. Frequently, these commentators even discuss the merits of one version over another. This all is not very surprising because these texts represent what is, after all, an oral tradition. However, the contrast between the wide-ranging textual variants in the Talmud and the few variants in the bible is striking. The Talmud even varies greatly by printed edition. For example, the Warsaw Talmud is somewhat different from the Venice edition and both are different from the now-standard Vilna printing. This is something that commentators and halachists take into account. The masterly Dikdukei Sofrim, whose excellent introduction documents much of what we have said, was an attempt to bring together all the textual variants of the Talmud into one book. This is currently being continued and updated by the Machon Hatalmud Hayisraeli Hashalem. However, the Babylonian Talmud is in better condition than the Jerusalem Talmud, the Tosefta, and the Midrashim. See, for example, the comments of Raavad in Hilchot Maaser Sheni 1:10 and R' Tzvi Hirsch Chayes in Imrei Bina ch. 1. Given all this, it would not be surprising if a copyist accidentally changed some letters or even some words in a talmudic or midrashic citation of the Bible.

However, there is still need for further clarification. The Talmud in Zevachim 62b tells us of a time when R' Tarfon quoted Genesis 25:1 as "Avraham took another wife whose name was Yochni." He was immediately corrected that the verse says Ketura and not Yochni. This correction by the students tells us that the mistake was not a copyist's error but does that mean that R' Tarfon had such a different text than we do (and his students did)? Similarly, in Bava Kamma 81b it is said of someone, "About whom the Bible says 'In order to be good, do not be called bad'". The Gemara immediately pointed out that there is no such verse and suggests a similar verse (Proverbs 3:27) in its place. Did the author of the saying have a different biblical text?

The Yad Malachi 1:283 quoted a number of scholars who explained that many rabbis of the Talmud intentionally misquoted the bible as a strict interpretation of the dictum (Gittin 60b), "Things that are written [i.e. Bible] may not be quoted orally." They refused to quote the Bible when they could not read it from a written text. Therefore, they changed some words in the verses so as not to violate halacha as they understood it. For example, we quoted R' Tarfon as citing a verse, " Avraham took another wife whose name was Yochni." In a number of other places in rabbinic literature (e.g. Bereshit Rabba 25:3; Menachot 85a) we find names like Yochna and Yochni used as a generic name similar to the way John Doe is used today. R' Tarfon just inserted a standard name into the verse so he could avoid quoting it exactly thereby requiring reading it from a written text. See also R' Reuven Margoliyot's Hamikra Vehamesora ch. 13 on this issue. This explains why we sometimes find verses misquoted in the Talmud and midrashim. It is not that the sages did not know Bible or had a different version of it. Either the text of the talmudic or midrashic passage has been distorted by transmission, which we have shown is common for these works, or the original passage contains an intentionally misquoted verse. While it is still possible that the sages of the Talmud sometimes quoted a verse differently because their text was different, we have no clear evidence of this.

Before we arrive at the next category of talmudic variant, we need to discuss an important statement the Talmud makes on our topic. In Kiddushin 30a, Rav Yosef (early fourth century) says that we are not experts in chaserot veyeterot (defective and plene spellings). There are certain vowel sounds in Hebrew that can be spelled solely with vowelization (chaser=defective) or also with a letter vav or yud (yeter=plene). Both spellings are grammatically equivalent and are pronounced the same. At some point in time, doubts arose among the scribes regarding some words in the Bible, whether they are spelled chaser or yeter. While some have tried to minimize this statement to refer only to Rav Yosef who was blind (R' Reuven Margoliyot, Hamikra Vehamesora, ch. 4), the commentators and halachists did not understand it this way. Some have also tried to limit it to unusually large or small letters. However, R' David Metzger has demonstrated that this is insufficient to explain the entire talmudic passage (Torah Shelemah vol. 28 p. 288). Additionally, the rishonim clearly did not read the passage this way. Thus, the Rama ruled in Shulchan Aruch, Orach Chaim 143:4 that if a Torah scroll during a public reading in the synagogue is found to have a mistake then it must be closed and another Torah taken out. However, if the mistake was a chaser or yeter that differs from the textus receptus then we do not take out another Torah. Who can say, the Rama implied, that the new Torah that presumably matches the textus receptus is more correct than the current Torah being used? On this one issue ­- chaser and yeter - the textus receptus is considered imprecise (with one exception that we will discuss later). This does not, however, affect the meaning of the Bible, which is perhaps how these errors crept in. Never the less, the official Masorah, which we will discuss soon, has standardized all spellings, including chaser and yeter.

Aggadic Midrash

The second category of textual variant found in rabbinic literature is in an aggadic midrash (exposition). We have already discussed cases of verses that are misquoted in the Talmud and Midrashim. In a midrash, however, we sometimes find that the key word on which the exposition is based is different than in our texts. For example, in Shabbat 55b we find a dispute whether both of Eli's sons sinned or just one (See 1 Samuel 2:11-17, 22-26). Rav Nachman bar Yitzchak argued that both sons sinned because the word מעבירים is used (v. 24) which is plural. Rav Huna breih deRav Yehoshua answered that the word is actually spelled מעבירם which implies that only one sinned. However, the textus receptus has the word spelled מעבירים, not like Rav Huna breih deRav Yehoshua claimed. This cannot be a copyist's error because the entire argument is based on this spelling of the word. Similarly, Bereshit Rabbah 61:4 has the following:

"Avraham took another wife and her name was Keturah" (Gen. 25:1) - Rav Yehuda said: This is Hagar... He said to him: Does it not say, "And to the sons of the concubines (פלגשים)" (ibid. 6)? He answered: It says פלגשם.

Similarly, in Yalkut Shimoni, Beshalach 232:

From where do we know that even the sea was full of anger? Above it says, "And the water was a wall (חומה) for them" (Ex. 14:15). Anger (חמה) is written.

Neither of these two readings are matched in the textus receptus. From these and many other examples we seem to see that the rabbis of the Talmud and Midrash had a different text of the Bible. If their Bible did not read מעבירם, פלגשם, and חמה then how could they have made their drasha (exegesis)? The first thing to note is that almost every single drasha is on a chaser or yeter - a vowelization that can be written with or without a letter. As we already said, we are not experts in this area and should not be surprised if some rabbis had a different version of this. However, we do not need to raise this issue because a more nuanced understanding of aggadic drashot reveals that they are not based on the actual spelling of the words.

It is important to understand that aggadic drasha is a form of homiletics. Rabbinic sermons attached moral messages to the text of the Bible as a brilliant method of emphasizing a point. For example, Shir Hashirim Rabbah 2:1:4 exposits on the phrase ודגלו עלי אהבה ("and His banner upon me is love" - Song 2:4) that mumbling (ודלוגו) and childish musing (ולגלוגו) are beloved to G-d. Is there any question that this was merely a case of cleverly modifying a word for the sake of a sermon? Similarly, Bereshit Rabbah 45:9 takes the description of Yishmael of יד כל בו ("Everyone's hand against him" - Gen 16:12) and combines the last two words into כלבו (his dog) to compare Yishmael's wildness with that of a dog. If these two words were really combined in a text then the verse would not make sense. Rather, all would agree that the previous two examples are cases where the texts are admittedly maniplated for homiletic concerns.

R' Chaim Heller (The Septuagint References in Mandelkern's Concordance, appendix) has shown that a comparison of similar midrashim in different sources demonstrates that this method is very widespread and is used in seemingly simple drashot as well.

For example, Shmot Rabbah 41:5 has:

What does it mean, "When He finished (ככלותו)" (Ex 31:18)? R' Shimon ben Lakish said: Whoever says words of Torah that are not pleasing to his listeners like a bride (כלה) is to her groom would have been better not to say them.

Here, the midrash homiletically reads "When He finished (ככלותו)" as "Like his bride (ככלתו)". There is no question that this is merely a clever play on words to emphasize an idea like reading ודגלו as ודלוגו. The Tanchuma does this similarly:

"When He finished (ככלותו)" - It is written "Like his bride (ככלתו)". R' Shimon ben Lakish said: It is forbidden for a scholar to teach halacha to a group until the words of Torah are pleasing to their listeners like this bride who is pleasing to her groom.

The very same aggadic reading is said in a way that implies that it is based on a textual reading. However, the second case is the same as the first case. It is a homiletical expansion of a religious message that is attached to the Bible through a loose reading of the text. Placing too much weight on the phrase "It is written" is incorrectly applying rigorous analysis to a very loose method of homiletics.

Comparison of Midrashic Formulations
1. "It was on the day that Moshe finished (כלות) erecting the Tabernacle" (Num. 7:1) - It is written כלת. (Bamidbar Rabbah 12:8)

2. Why did they not count Achaz? Because he was embarrassed before Yishayahu. As it says, "At the road of Launderer's (כובס) Field" (Isa. 7:3). It is written כובש because he hid his face. (Sanhedrin 104 according to one version).

3. Rabbi Yehuda said: When he left his mother's womb he cut her uterus so she could not give birth [again]. As it says, "For his pursuing his brother with the sword and suppressing his mercy (רחמיו)" (Amos 1:11). It is written "his womb (רחמו)". (Tanchuma, Ki Tetze 4)

1. "It was on the day that Moshe finished (כלות)..."(Num. 7:1) - The rabbis considered it to have been written כלת (Zohar, Pinchas 226).

2. When the prophet would come to argue with him he would leave... As it says, "At the road of Launderer's (כובס) Field" (Isa. 7:3) that he hid his face. (Vayikra Rabbah 36:33)


3. Rabbi Yehuda said: When he left his mother's womb he cut her uterus. "Thus said the Lord: For three transgressions of Edom... and he destroyed his mercy (רחמיו)." (Pesikta Derav Kahana ch. 12)

Thus, the textual variants we find in aggadic midrashim do not necessarily reflect different versions of the Bible. They reflect the homiletic license that is typical of this literary genre. Some rishonim did not understand it this way. For example, Tosafot in Shabbat 55b said that the Gemara's version of the Bible was different than the textus receptus because the Gemara had מעבירים chaser while the textus receptus has it yeter. However, the Rashba (Responsa Attributed to Ramban, 232) wrote that we should not revise our biblical texts based on aggadic midrashim. Similarly, Rashbatz (Tashbetz 3:160) wrote, "We do not rely on aggadic midrashim to correct the texts." Even though we are not experts in chaser and yeter, we still cannot be sure that these midrashim were doing anything more than loosely reading the text.

Halachic Midrash

The third category of textual variants found in Talmud and Midrash comes from halachic midrash. There are countless cases of laws being derived from the Bible and in very few of the cases the text in our Bibles is different than in the midrash. For example, the Gemara in Sanhedrin 4b says:

It is taught: לטטפת (Deut 6:8), לטטפת (Deut 11:18), לטוטפות (Ex 13:16) - There are four. These are the words of R' Yishmael.

The Gemara seems to learn from the two chaser spellings of the word letotafot (which refers to tefillin) and the one yeter spelling that the two chaser imply singular and the one yeter implies plural (two). One plus one plus two results in four, the number of tiny scrolls that are in a tefillin box. The precise spelling of each word is crucial to the halachic midrash. However, in our texts the third word is also spelled chaser - לטוטפת. Similarly, that same Gemara on the previous side (4a) says:

קרנת (Lev 4:25), קרנת (Lev 4:30), קרנות (Lev 4:34) - There are four. Three to teach the mitzvah and one to teach that it is an absolute requirement.

From the fact that the first two mentions are written chaser and the last is yeter the Gemara learns that there are really four mentions in the Torah. This is used to learn an halacha. Yet, in our texts even the last קרנת is chaser.

However, it is very possible that, like aggadic midrashim, halachic midrashim are not based on the precise spelling of the text or are more complicated than they seem. For example, the drasha from לטוטפות we cited above is explained by Tosafot ad loc. as having no connection to the spellings of chaser and yeter. Rather, it is referring to the inner vav between the two tets that exists in only one of the three spellings. This extra letter, claim Tosafot, exegetically expands the word into two. Similarly, R' Menachem di Lonzano in his Or Torah suggests that the קרנת that the Gemara mentions are not those in Leviticus 4:25, 30, and 34 but those in Exodus 29:12, Leviticus 4:30, and Leviticus 8:16. In those places, the word is spelled twice chaser and once yeter, as implied by the Gemara.

This approach, however, is not universal. Many commentators takes these drashot at face value and acknowledge that in these matters the Talmud had a different reading of the Bible than we do. Once again, though, we are dealing with cases of chaser and yeter. The Talmud has לטוטפות and קרנות spelled full (yeter) while the textus receptus has them spelled defective (chaser).

We have already noted that the Talmud in Kiddushin 30a says that we are not experts in chaserot and yeterot. While on the plain level these differences do not change the meaning of the text, on a drash level they do. Drashot can sometimes be based on the spelling of words. If so, in the few cases in which there is a conflict between the Talmud's text and the textus receptus regarding chaser and yeter we have to ask whether the drasha and its resulting halacha still stand. Perhaps the comparative study of biblical texts should change halacha.

Of course, we cannot change halacha for this alone because, as we noted above, there are commentators who do not believe that these drashot are based on the different spelling of the texts. Additionally, perhaps the Talmud's text is correct and not the textus receptus (we will address this shortly). More importantly, however, is the dispute over the role of halachic midrash in the development of halacha. There are historians who claim that halachic midrash is a method in which rabbis created or derived new halachot by analyzing the biblical text with the exegetical rules of midrash. There are other historians who disagree with this assessment and claim that the halachot are ancient traditions that were transmitted orally. Halachic midrash was a relatively late development that attempted to unite the oral Torah with the written through midrashic exegesis (see Chanoch Albeck's Mavo Lamishna, ch. 3). According to the latter historians, even if the drashot were based on an incorrect text the halacha should not change. The halacha preceded the drasha and is not dependent on its accuracy.

What we have seen is that the only reliable criticism of our biblical texts that we can garner from all the ancient evidence is that the textus receptus may be inaccurate regarding chaser and yeter spellings. However, even a minor criticism such as this is serious in that it tells us that our Torah is not written exactly like the Torah that G-d gave Moshe. While it may be 99.99% accurate, even that tiny percentage of inaccuracy must be dealt with. For this, we turn to the development of the Masorah.

The Masorah

The term Masora generally refers to a tradition on how to write and punctuate the Bible. However, it also has more specific meanings. We shall only address the Masora in terms of orthography - the letters in the Bible. There are early talmudic sources regarding a tradition on the letters of the bible. For example, Kiddushin 30a says that the early scribes kept track of how many letters are in the Bible. In Avot 3:13, Rabbi Akiva praised the Masora as being a fence that protects the integrity of the Torah text (see the commentaries of Rabbeinu Yonah and Meiri).

Beginning in the sixth or seventh centuries, codices - Bibles with punctuation and accentuation - began to be written (Aharon Dotan in Encyclopedia Judaica, sv. Masora 2.2.2). In the eighth or ninth centuries there were Masoretes whose reputations last to this day - R' Pinchas Rosh Yeshiva and R' Asher ben Nechemiah (Ben-Asher). These scholars began putting into writing the accumulated knowledge that we now call the Masora. The Masora eventually contained somewhat cryptic notations indicating unusual spellings, identifying parallels, and noting many grammatical and orthographic phenomena. Exactly what should be written in the Masora was never clearly stated so that over centuries more and more was added. Not surprisingly, over such a long period of time differences emerged between the Masora texts. Scholars have identified three major systems of Masora - the Babylonian, the Palestinian, and the Tiberian systems. However, the differences between these systems lie in the area of punctuation and refer to issues that are obscure to most readers - phenomena like the gaaya and the chataf. Regarding orthography, however, there are no distinctions among the systems (Dotan, 5.; R' Mordechai Breuer, The Aleppo Codex and the Accepted Text of the Bible, introduction par. 16, 21). It is therefore possible to speak of a single Masora for the letters of the Bible.

There are still differences between the Masora texts. However, because they are not systematic differences we can use a powerful tool that we have to resolve these discrepancies. Earlier, we quoted the baraita that related the story of Ezra finding three Torah scrolls and resolving the few differences based on majority. If we use the majority rule to resolve the differences within the Masora and among the reliable codices that we have, we can reconstruct the original Bible and the original Masora.

Regarding the use of majority in our case, R' Yaakov Emden wrote, "Even if there is a dispute on a matter we must follow the majority like was done with the Torah scrolls found in the Temple courtyard" (She'elat Yaavetz 1:48). See also Binyan Tziyon 82 and Ginat Vradim 2:6 who agreed.

There are, however, those who disagreed on one point. As we discussed earlier, there are some who believe that halachic midrash may be taken as a legitimate source for textual variants. Therefore, even when the Masora has determined a spelling, if an halachic midrash has it differently we must follow the midrash. The Rashba (Responsa Attributed to Ramban, 232) ruled that we must fix our Torah scrolls so that they follow the halachic midrash when it differes (solely in regard to chaser and yeter). In this instance, Rashba claimed, we are not expert in chaser and yeter and do not follow the Masora.

However, as we noted above, there are those who disagree with that evaluation of halachic midrash. Those midrashim do not imply a textual variant when understood properly. It is presumably for this reason that R' Yaakov Ettlinger wrote, "What follows is that even in the case where an halacha is derived from [this word] the tradition is not certain enough to be relied upon more than another tradition" (Binyan Tziyon, 98). Indeed, anyone can check the textus receptus and will see that the practice is not to change the chaser and yeterot in our Torahs to match halachic midrash. We are therefore free to follow the Masora and majority of texts to determine the best and most authentic Torah version. As R' Meir Halevy Abulafia wrote in his introduction to Masoret Syag Latorah, "I felt the need to come forward and study and investigate the exact scrolls and precise Masorot, to study the disputes and ignore the more recent scrolls, and to follow the older reliable ones, and to decide among them following the majority rule."

This process that resulted in the textus receptus was recently repeated by R' Mordechai Breuer based on the oldest and most reliable manuscripts of the Bible and Masora. The following is his conclusion regarding the Torah:

There are only six instances in which it is difficult to say with certainty which text best fits the Masora. In all other cases there is a definite orthography which is the version in accordance with the Masora...

[I]n the version commonly accepted among the Ashkenazim, following the Ramah, there are 6 orthographies apparently contradicting the Masora... [T]he Yemenite community preserved a different version, which differs from the [Ashkenazi] text in nine places. Three of them are among the six doubtful places mentioned above, the six other instances being those same six places in which the [Ashkenazi] text differs, apparently, from the orthographic text of the Masora...

Thus we have three versions of the Torah: the Yemenite, the [Ashkenazi], and the version which is derived from [the majority of manuscripts]. The three versions differ from one another with respect to their source, and yet they are almost identical everywhere.

(Breuer, introduction par. 20)

Using the methodology of majority to determine the text of the Torah, we find that the three most authoritative versions differ in less than ten places. Less than 0.01% of the 304,805 letters of the Torah are under question.

Conclusion

We have traced the history of the Torah from its writing by Moshe, through its recreation based on manuscripts by Ezra, until its present form. We have seen questions arise from non-normative and non-Jewish sources like the Samaritan Torah and the Septuagint. And we have seen issues come up from the Talmud and Midrash. Through all this, the only issue that raised serious questions about the textus receptus of the Torah is that of chaser and yeter, vowelizations spelled with and without an assisting letter. This is acknowledged in halachic sources (see Yad Malachi, 1:283) but does not pose any serious limitation on our understanding of the Torah that G-d wrote through Moshe. While there are doubts whether a handful of letters (less than 0.01% of the Torah) were actually placed there by G-d or not, we have no compelling reason to doubt that the overwhelming majority of the Torah text (over 99.99%) has remained in its original form throughout the thousands of years of transmission.

We began with the task of searching for evidence that the Torah we have is incorrect and to find methods of fixing it. However, we have not found convincing evidence that anything has been changed or omitted in the Torah text throughout history. The “witnesses” we looked at were inconclusive at best and generally showed evidence of being inaccurate themselves. The Samaritan Torah was clearly changed by its scribes. The Septuagint’s differences from the textus receptus are largely translational issues but even those variants that are not are probably due to mistaken sources. The Dead Sea Scrolls emanate from a sect that rejected the authorized version of its time and knowingly accepted deficient texts. Finally, talmudic and midrashic variants can be due to many different reasons. Even those differences that are considered by some to be sources of variants, relate to cases of chaser and yeter. After tracing through the transmission of the Torah, we find that it has been preserved in an almost perfect form.


Back Home


Contributor(s): Gil Student
Last revised: 3/4/02
© Gil Student 2002