Do We Know the Original Text of the Torah?

The revelation at Sinai is a mystery and will likely always remain one. Notably, the Torah itself never says that Moses received the Torah at or from Sinai, only that God spoke the Decalogue to Moses and/or the people,^[1] and that he received there a series of laws which are quoted in the Torah at different points.^[2]

By the late Second Temple Period, however, Jews, Samaritans, and (eventually) Christians understood the Torah to be Moses’ book.^[3] As the Mishnah states:

^{משנה אבות א:א} משה קבל תורה מסיני, ומסרה ליהושע, ויהושע לזקנים, וזקנים לנביאים, ונביאים מסרוה לאנשי כנסת הגדולה.

^{m. Avot 1:1} Moses received the Torah from Sinai, and transmitted it to Joshua, and Joshua to the elders, and the elders to the prophets, and the prophets transmitted it to the Men of the Great Assembly.^[4]

Shavu’ot in its current form commemorates this receiving of the Torah at Sinai,^[5] and I therefore use this opportunity to apply some scholarly thinking about what we can or do know about the text of this ancient Torah.

Which Version Did Moses Receive at Sinai?

Samaritan tradition identifies Moses’ book with the Samaritan Pentateuch (SP), the textual version sacred to their community.^[6] In contrast, Jewish tradition identifies Moses’ book with the Masoretic Text (MT), known to us in many forms in medieval manuscripts and in modern printed editions, found in every synagogue and Jewish home. And these are only two of several textual forms of the Torah.^[7]

First, the MT and SP are both derivative of older text traditions. MT, with its vowels and cantillations marks (teʿamim), was only finalized in the early Middle Ages. But long before that, we find a purely consonantal version known as proto-MT, for example, in the fragmentary Leviticus scroll from Masada, MasLev^b (30 BCE–30 CE).^[8] Similarly, a Pre-Samaritan text is exemplified in the fragmentary Exodus scroll from Qumran, 4QpaleoExod^m. In addition to these two text traditions, other Hebrew versions of the Torah appearing among the Dead Sea Scrolls present us with scores of fragmentary Hebrew scrolls dating from the third century B.C.E. to the second century C.E., some of them preserving traditions that occasionally predate MT, such as 4QDeut^q.^[9]

Other textual variations appear in ancient translations of the Torah. The most important of these is ancient Greek Septuagint (LXX) translation. Other versions are the Syriac Peshitta, the various Aramaic targumim, the Latin Vulgate, and occasionally even the Old Latin translation of the LXX. All these ancient translations were produced directly from Hebrew texts. Some of these early Hebrew texts resembled the Masoretic Text, while others differed slightly or considerably from it.

This textual variety brings up the question of whether there is actually such a thing as “the” text of the Torah, i.e., an original text from which all the text forms we have stem—what scholars call (in German) an Urtext. The first step in answering this question is understanding the kinds of differences we find between these text forms of the Torah and how they came about.

The Relation Between the Torah Texts

How do we summarize the relations between the various Torah texts? The simple answer is that we are looking at a unified tradition, albeit one whose specific iterations differ from each other in a myriad of small ways.

The Gathering of the Waters: Day 2 of Creation

Take the creation story (Gen 1–2:4a). The content of the story is identical in the various traditions, but minor textual differences that do not affect the gist of the story. For example, let’s look at the passage in which the waters are gathered, starting with the familiar MT version:

^{בראשית א:ט} וַיֹּאמֶר אֱלֹהִים יִקָּווּ הַמַּיִם מִתַּחַת הַשָּׁמַיִם אֶל מָקוֹם אֶחָד וְתֵרָאֶה הַיַּבָּשָׁה וַיְהִי כֵן.

^{Gen 1:9} God said, “Let the water below the sky be gathered into one area, that the dry land may appear.” And it was so.

The LXX has this verse, but with two differences:

^{Gen 1:9 [LXX]} And God said, “Let the water that is under the sky be gathered into one gathering, and let the dry land appear.” And it became so. And the water that was under the sky was gathered into their gatherings, and the dry land appeared.[10] (NETS)

Both of these differences are also reflected in the Qumran fragments of 4Q8 (4QGen^h) and 4Q10 (4QGen^k). The first difference involves one letter: (final) mem (מקום) in MT and heh (מקוה) in the LXX Vorlage and 4QGen^h. The graphic similarity between the two words led to a scribal error in one direction or another. The second reading of the LXX—and probably also 4QGen^k—reflects a phenomenon named “harmonizing,” that is, the expanded text added several words to match the surrounding text—in this case, adding the execution of a command, which is typically noted elsewhere in Genesis 1.^[11]

Larger Harmonizations in SP

This particular harmonization is rather small, but the SP contains several much more extensive harmonizations that involve additions (duplications) of other Torah verses and a few rearrangements. The editors of the SP group were especially attentive to what they considered to be incongruences within and between Scripture stories. Special attention was paid to the presentation of the spoken word, especially that of God and Moses.

The most extensive example of harmonization in the SP is in the plague narrative of Exodus. In MT, before several plagues, God commands Moses to warn Pharaoh, and the MT text merely records that Moses “does so.” Instead, the SP reproduces the text of Moses’ speech to Pharaoh.^[12] Most scholars regard these editorial tendencies in SP as secondary. The same goes for the expansive texts that resemble SP that are found in the Dead Sea Scrolls.

Working With Translations

The evaluation of the details in the ancient translations requires intimate knowledge of the character of the translations, since the many differences between any given translation and the extant Hebrew can derive either from the translators’ interpretation of the Hebrew text or by their use of a text that differs from their source text.^[13] Therefore, in trying to reconstruct the early text of Scripture, when we are faced with a translation that does not reflect an extant Hebrew text, we first want to eliminate the option that it was caused by the translator’s exegesis. This is, for example, generally the case when comparing any of the Aramaic targumim with MT.

Indeed, when we look carefully at the various ancient translations with this procedure in mind, we find that only the LXX yields a good crop of Hebrew readings that differ from MT—differences that are based on a different text rather than on exegesis. Although many LXX readings in the Torah are, like in the SP, a result of a process of textual harmonization, the LXX also reflects a host of valuable ancient readings.

Relationship Between the Versions

Despite the differences between the Torah texts, an overarching unity in the totality of the texts is visible in the majority of the details. It seems to me that the various textual traditions can be reduced to one common text that presumably branched off in different directions in the course of transmission.^[14] But how do we sort through this?

When thinking about the unity or lack of unity of all the textual sources of the Torah, the enormity of the details is overwhelming. The first step is to collect these details, and to organize the texts into groups. It is then possible to at least search for the common text form which they all derived.

I find it helpful to organize the texts into three main groups:

1. The Masoretic Text and its congeners (=closely related texts),
2. The LXX together with the SP and their congeners,
3. A smaller group of mixed texts—namely those that are not exclusively close to MT, SP, or LXX—extant only in fragments from the Judean Desert.

The relation between the first and second groups can be defined. Furthermore, the three groups can be organized in a structure that is called a family tree, graphically depicting the relation between them.

Mixed texts—such as 11QpaleoLev^a (11Q1), contain a mixture of readings from the first (MT) and second (LXX/SP) groups, and in addition include independent readings.

MT—This group consists of the ancient forerunners of MT, the proto-Masoretic texts,^[15] such as the aforementioned Leviticus scroll from Masada, as well as the medieval full-fledged Masoretic Text. Other members of the group are several MT-like Hebrew texts from among the Dead Sea Scrolls, the (Aramaic) targumim, the (Syriac) Peshitta, and the (Latin) Vulgate.

LXX/SP—The LXX and SP seem to be strange bedfellows, since, in its present form, SP embodies the holy writ of the Samaritan community,^[16] while the LXX is the Greek translation of the Jewish Torah produced in Alexandria in the 3^rd century B.C.E. Yet, these two texts uniquely resemble each other in their many common harmonizing additions in the text.^[17]

How Much Did Abimelech Pay Abraham?

Remarkably, sometimes these harmonizations actually lead to a text whose meaning veers somewhat from the text it modifies. For example, Abimelech’s gift to Abraham in MT reads as follows:

^{בראשית כ:יד} וַיִּקַּח אֲבִימֶלֶךְ צֹאן וּבָקָר וַעֲבָדִים וּשְׁפָחֹת וַיִּתֵּן לְאַבְרָהָם וַיָּשֶׁב לוֹ אֵת שָׂרָה אִשְׁתּוֹ.

^{Gen 20:14} Abimelech took sheep and oxen, and male and female slaves, and gave them to Abraham; and he restored his wife Sarah to him.

In the Samaritan Pentateuch, Abimelech additionally gives her silver:

^{בראשית כ:יד [נ"ש]} ויקח אבימלך אלף כסף וצאן ובקר ועבדים ושפחות ויתן לאברהם וישב לו את שרה אשתו׃

^{Gen 20:14 [SP]} Abimelech took 1,000 pieces of silver, sheep and oxen, and male and female slaves, and gave them to Abraham; and he restored his wife Sarah to him.

The LXX has the same addition, χίλια δίδραχμα “a thousand didrachma’s.” Where does this addition come from? According to all traditions, Abimelech refers to this financial payment in his farewell to Sarah:

^{בראשית כ:טז} וּלְשָׂרָה אָמַר הִנֵּה נָתַתִּי אֶלֶף כֶּסֶף לְאָחִיךְ...

^{Gen 20:16} And to Sarah he said, “I herewith give your brother a thousand pieces of silver…”

In the presumably original text, reflected in the MT, these thousand pieces of silver probably represented the value of the flocks and slaves he had given Abraham. The version that eventually yielded LXX and SP, however, added this detail from v. 16 to v. 14, since it saw that detail as conspicuously absent, and wished to create harmony. According to the new version found in LXX and SP, Abraham receives twice as much value in reparation.^[18]

Other Harmonistic Texts

Other texts in this second group, that have much in common with SP and the LXX, are:

• The Qumran tefillin, such as 4QPhyl A, which reflects many variants and contains several Scripture passages in addition to the rabbinic prescriptions;^[19]
• 4QDeut^j (4Q37) and 4QDeut^k1 (4Q38), which may have served as master copies for the copying of tefillin;^[20]
• Liturgical scrolls, such as 4QDeutⁿ,^[21] which may have been used as some form of prayer at Qumran.

The main textual feature of these two groups is the presence of textual harmonizations, and therefore they are secondary when compared with MT, like the SP and LXX.

Edits and Errors in the MT

Given that the distinction between MT and the other traditions is based on the frequent presence of harmonization in the other traditions, it is hardly surprising that the Masoretic Text is generally the most trustworthy text in the case of the Torah. This is also the case when it comes to errors, since the textual transmission of MT was very precise after the first century B.C.E., and possibly earlier.^[22] Even so, MT is not always preferable, and the other texts sometimes provide important early readings that antedate parallel readings in MT.

For example, twice in the song of Moses, we find that the MT scribes engaged in theological censorship, removing references to other gods (Deut 32:8, 43) that can still be found in LXX and some Qumran manuscripts (4QDeut^j, 4QDeut^q).^[23] An example of a mistake is MT’s וַיֵּלֶךְ מֹשֶׁה “and Moses went” (Deut 31:1), which is clearly a metathesis (reversal of letters) for ויכל משה “and Moses finished” as preserved in the LXX translation and 1QDeut^b.^[24] In MT Jacob has 70 descendants, whereas LXX (Gen 46:20) and 4QExod^b (Exod 1:5) have the likely earlier reading 75 descendants.^[25]

One Archetype

All these cases where there are differences between MT and LXX or SP make up only about some 3–4% of the text; in the other 96–97% of the cases, the Torah text is identical across all text traditions. And these 3-4% of cases where there are differences reflect different text traditions, harmonizations, or errors.

To quote a recent study by Innocent Himbaza: “To state my position simply, yes, we should assume only one archetype behind the textual diversity of the Pentateuch.”^[26] By “archetype,” Himbaza means that all known texts, that is, all sources that you can see with your eyes, ultimately derive from a single source.^[27] Yet, even if we were to fully reconstruct this archetype, it was not the original text of the Torah because this presumed archetype is not earlier than the third century B.C.E., centuries after the Torah was composed or compiled.

The Original Text of the Torah?

If text critical reconstruction can get us only to a 3^rd century B.C.E. archetypal text, what about the original original? What can we say about the first text or texts ever of the Torah; when was it written and by whom? Here, text criticism offers very little direction.

Tradition states that the Torah was composed by Moses, during Israel’s wandering in the wilderness or at the end of that period (b. Gittin 60a), around the 13^th century B.C.E. According to Moses Maimonides’ (1138–1204) eighth principle of faith:

^{רמב"ם, פירוש המשנה, הקדמה לפרק חלק} והיסוד השמיני הוא תורה מן השמים. והוא, להאמין שכל התורה הזאת המצויה בידינו היום הזה היא התורה הנתונה למשה, ושהיא כֻלה מפי הגבורה, רצוני לומר, שהגיעה אליו כֻלה מאת ה'...^[28]

^{Maimonides on the Mishnah, Introduction to Helek} The eighth principle is that the Torah is from Heaven. This means to believe that this entire Torah, found is in our hands today, is the Torah that was given to Moses, and that it is entirely from the Powerful One, what I mean to say is, that that all of it came from God…

This belief, in one form or another, undergirds traditional Jewish, Samaritan, and Christian theology.^[29] While text criticism has no direct response to this belief, we can point to several elements in the text that would have certainly been different in any 13^th century B.C.E. text from the Torah we know today.

Technologically speaking, it would not have been possible to produce a scroll this size in that period.^[30] Also, the script or “font” would not have been the Aramaic script which we use in our Torahs today; indeed, even the ancient Hebrew script was not yet developed, and writing—and we have little evidence of substantial Hebrew writing in this period—would have been a form of the so-called pictographic Hebrew script.^[31]

Of course, critical biblical scholarship, even amongst its most conservative practitioners, does not consider a 13^th century date for the original Torah as plausible, and most prefer to look between the eighth and the fifth centuries as the most plausible period(s). Moreover, based on an analysis of duplicate versions of the same story, tensions within and between stories, contradictory laws, specific terminology, and other criteria, literary-critical scholars have pointed to the Torah as having been composed in stages, though there is much disagreement as to exactly how this took place: whether by combining different versions of the core narrative (documentary hypothesis), weaving together smaller units to create an overarching narrative (fragmentary hypothesis), periodically adding to a core text (supplementary hypothesis)—or any number of permutations of such processes.

This type of literary analysis, known as the historical-critical study,^[32] is the backbone of biblical studies. These processes precede by centuries the time of the archetypal text under discussion here, and it is unclear what the scholars who were and are involved in the so-called historical-critical study – literary critics – think about the issue of the original text. Textual and literary criticism move in different ways.

The Limitations of Text Criticism

Textual criticism, i.e., the comparison of extant text traditions, can be relevant in principle for recovering a text that preceded the putative archetype. For example, some scholars believe that the considerably deviating text of the LXX in the tabernacle chapters (Exodus 35–40) is based on an early Hebrew tradition differing much from MT.^[33] In practice, however, textual comparisons are rarely addressed in the literary-critical analysis of the Torah’s formation.^[34]

Textual criticism—and I say this as a textual critic—has to live within its natural limitations, namely, the boundaries of the textual data. Consequently, unlike the more speculative work of literary critical scholarship, text criticism cannot and should not hypothesize beyond the textual horizon, and thus the goal of this discipline is limited to reconstructing the 3^rd century archetype and tracing how the various versions of the Torah we now have, came about. The Torah, indeed, likely has a very complicated creation history, but different tools are needed to recover its earlier history.