In recent years there has been significant debate surrounding the ultimate goal of textual criticism. Traditionally the goal was simply to reconstruct the original text of the Old Testament. With the discovery of the Dead Sea Scrolls, it has become apparent that this goal is not as simple as it used to be. This post will explore the goal — or perhaps the goals — of textual criticism.

This is the seventh post in a series on the textual criticism of the Hebrew Bible/Old Testament. Other posts include:

• Textual Criticism of the Hebrew Bible – An Introduction (TCHB 1)
• Textual Criticism of the Hebrew Bible Resources (TCHB 2)
• Hebrew Witnesses to the Text of the Old Testament (TCHB 3)
• Early Versions of the Hebrew Bible (TCHB 4)
• Codex Sinaiticus: A Profile (TCHB 5)
• The History of the Biblical Text (TCHB 6)

All posts in this series may be viewed here.

The concept of an Urtext, the putative original text, depends partially on how we understand the origins of the five text types found at Qumran and their relationships to it (see my previous post on the Hebrew Witnesses here). There have been three primary models proposed to answer these questions. In spite of the importance attached to this issue, no conclusive answer is possible because of a lack of solid evidence from the time of their origins.

Lagarde’s Model: An Archetypical Urtext

Paul de Lagarde‘s model, historically embraced by the majority of text critics, presupposes one original text of a biblical book and that all textual witnesses derived from it. In practice, the majority of critics first collect the texts into text types, the MT, the LXX, and the SP, and from them reconstruct the eclectic Urtext.

F. M. Cross refined this process by his widely influential theory of “local texts.” In his view the texts developed in geographical isolation: Babylon for the proto-MT of the Torah, Egypt for the Septuagintal texts, and Palestine for the pre-Samaritan Torah and for the proto-MT in the Prophets and Hagiographa (see Frank Moore Cross and Shemaryahu Talmon, eds., Qumran and the History of the Biblical Text [Harvard University Press, 1975] 37; Buy from Amazon.ca | Buy from Amazon.com). Furthermore, in his view, while the proto-MT preserved the Torah in a superb, pristine state, elsewhere it conserved the expansionistic Palestinian text type (Ibid, 307-308).

Cross’s local text theory, however, does not adequately account for the network of agreements and disagreements among the texts and for the “non-aligned texts,” and no compelling evidence exists for the proposed provinces of the developing text types. For example, the paleo-Hebrew script, which Cross thought secured the pre-Samaritan text in Palestine, was later found in other text types. For example, 11QpaleoLev is written in paleo-Hebrew and sometimes aligns itself with all three text types and other times stands apart (see K. Mathews, “The Leviticus Scrolls (11QpaleoLev) and the Text of the Hebrew Bible,” CBQ 48 [1986] 171-207). Talmon modified Cross’s local text theory by pointing to three socio-political groups: Judaism and the proto-MT, the Samaritans and the SP, and the Christians and the LXX (Shemaryahu Talmon, “Aspects of the Textual Transmission of the Bible in Light of Qumran Manuscripts,” in The World of the Qumran from Within [Magnes, 1989] 71-116; Buy from Amazon.ca | Buy from Amazon.com).

Kahle’s Model: From Plurality into Unity

In contrast to Lagarde’s Urtext theory, Paul Kahle argued for a multiplicity of texts from which a standard text emerged (See his Cairo Geniza [Clarendon, 1951]). Basing himself on an analogy with the Aramaic Targums, he presupposed the same development from independent, vulgar texts to the final forms of the MT, the LXX of Ezekiel, the SP, and to certain extent of the biblical text as a whole.

Other scholars also hold to a number of pristine originals for certain biblical books. For instance, S. Talmon, makes his case for multiple “original” texts on the basis of synonymous pairs of parallel readings (“The OT Text,” The Cambridge History of the Bible I [ed. R. P. Ackroyd & C. F. Evans; Cambridge University Press, 1970] 1-41; Buy from Amazon.ca | Buy from Amazon.com). Similarly, M. Greenberg, makes his arguments for equally valid MT and LXX from an exegetical viewpoint (“The Use of Ancient Versions for Interpreting the Hebrew Text,” Congress Volume: Göttingen 1977 [VTSup 29; Brill, 1978] 131-148); while Peter Walters bases his arguments on parallel stories in 1 Samuel (The Text of the Septuagint: Its Corruptions and their Emendation (Cambridge University Press, 1973; Buy from Amazon.ca | Buy from Amazon.com). Eugene Ulrich, in a number of publications, has perhaps been the most recent scholar to champion this perspective.

According to this view, the varying text types of certain books, such as Samuel, Jeremiah, Ezekiel, Daniel, and Esther, call into question the notion of an original text and suggests instead a multiplicity of original, pristine texts. Text critics, so the argument runs, should aim to recreate these original texts, not one eclectic, archetypical text that may have never existed. This view may find support in the parallel synoptic texts in the Bible itself.

Tov, however, criticises this theory since it is so vague about the origin and relationship of these independent texts (Textual Criticism of the Hebrew Bible, 184-185; Buy from Amazon.ca | Buy from Amazon.com). It also underestimates the capability of detecting secondary readings within the textual witnesses. Moreover, because the text critic cannot decide the priority of one reading over another, it does not necessarily follow that both are original; one may still be secondary. In addition, Waltke has noted that when the theory of independent texts of equal textual status is extended to the view that they also enjoy equal canonical status, it is not satisfying from both a historian’s and theologian’s point of view. Most theologians will want to know whether the tenth commandment prescribes worship on Mount Gerizim and most historians would want to know whether the biblical historian recorded in Exod 12:40 that Israel spent 430 years before the exodus in just Egypt (MT) or in Egypt and Canaan (LXX, SP). Finally, the evidence of synoptic texts does not prove the existence of parallel texts. The differences between these texts may be due to a linear development within the texts where they are now embedded.

Tov’s Model: Original Editions

Tov more plausibly supposes that certain biblical books such as Jeremiah and Ezekiel experienced more than one stage in their literary development, at least one early short edition and a later expanded final edition. Before the later, final form was produced, the earlier forms were considered the original and copied. According to this argument some of the Qumran scrolls and the versions preserve these earlier literary stages as well as the final edition behind the proto-MT. Other biblical material, such as the different edition of the LXX versus the MT, the Targums, the Syriac Peshitta, the Vulgate of Proverbs and of Exodus 35-40 reflect different parallel editions. The date of the final stages differs from book to book and remains undetermined because it antedates the DSS. Tov explains his view: “Large-scale differences between the textual witnesses show that a few books and parts of books were once circulated in different formulations representing different literary stages, as a rule one after the other, but possibly also parallel to each other” (Textual Criticism). While Tov basically agrees with de Lagarde’s thesis that all texts ultimately go back to an original text, he nevertheless believes that it is “almost impossible to reconstruct the original form.”In Tov’s view the text critic ought not necessarily to reconstruct the earlier stages, such as the shorter Septuagintal text in the Prophets — that is the task of literary criticism — but the final edition, such as the fully developed proto-MT in Jeremiah and Ezekiel. Tov further explains: “This formulation thus gives a certain twist to the assumption of one original text as described in the scholarly literature. We do not refer to the original text in the usual sense of the word, since the copy with which our definition is concerned was actually preceded by written stages. Reconstructing elements of this copy (or tradition) is one of the aims of textual scholars, and usually they do not attempt to go beyond this stage” (Textual Criticism, 171).

All in all, Tov’s theory best fits the data. The final edited text is the end of the literary process and, at the same time, the starting point of the transmission of the text. Tov has put a new twist on the meaning of the “original” text. It now means “original edition,” a view that mediates between Lagarde and Kahle. This fits with evidence from the ancient Near East where texts developed by supplementing earlier sources with later material (see J. H. Tigay, ed., Empirical Models for Biblical Criticism [Fortress, 1985]). Scribes in the pre-Samaritan tradition added material from Deuteronomy into Exodus. While this is a much larger topic than I can cover here, it seems most plausible to me that the Pentateuch developed in a supplemental fashion, with earlier sources being expanded and edited over time. Likewise, in the exile the so-called Deuteronomist re-worked earlier books of the Former Prophets by supplementing them with a distinctive theology. This process of literary development can still be observed in the Qumran scrolls and in the ancient versions of certain biblical texts.

Tov wisely stops the process with the proto-MT for socio-religious and historical reasons. That text, he argues, became the authoritative text within Judaism. For that reason, he excludes the later midrashic literary compilations such as the Hebrew behind several sections in the LXX, namely, sections in 1-2 Kings, Esther and Daniel. In short, text critics should aim to recover the original edition behind the MT.

The church as well as the synagogue both accepted the edition behind the MT as authoritative. Both Origen and Jerome conformed the Septuagint and the Vulgate (respectively) to the proto-MT, so that the MT essentially became the standard text of the OT within the Church. Our modern English versions are based on the MT. That history should not be underestimated in deciding the question of “what is the original text?” The MT inherently commended itself to both the Synagogue and the Church as “the best text.” As the canon of the OT emerged in the historical process, so also the MT surfaced as “the best text” of that canon.

Conclusions

It should be noted however that when the canon was discussed, there were not discussions of which version of a biblical book should be considered canonical. This realization leads to one caveat. While I agree that from my community of faith (Protestant evangelical Christianity) the goal of textual criticism is best conceived of as recovering the original edition behind the MT, I still see immense value in exploring the different texts and versions of the Old Testament. This is especially the case for the Septuagint considering the historical and theological significance it has had for the Christian church. Thus, while I agree with Tov in regards to the goal of textual criticism, I think it is also valuable to balance his views with those of Eugene Ulrich who argues for multiple texts as the goal. He argues, “the goal of ‘textual criticism of the Hebrew Bible’ is not a single text. The purpose of function of textual criticism is to reconstruct the history of the texts that eventually become the biblical collection in both its literary growth and its scribal transmission; it is not just to judge individual variants in order to determine which were ‘superior’ or ‘original'”(“Multiple Literary Editions: Reflections Towards a Theory of the History of the Biblical Text,” in Current Research and Technological Developments on the Dead Sea Scrolls: Conference on the Texts from the Judean Desert, Jerusalem, 30 April 1995 [Donald W. Parry and Stephen D. Ricks, eds.; Brill, 1996] 98-99; Buy from Amazon.ca | Buy from Amazon.com). While Ulrich clearly takes his observations too far, he does remind us of the richness in the textual witnesses to the text of the Hebrew Bible/Old Testament.

The last few posts in this series discussed some of the major witnesses to the text of the Old Testament; this post will bring them together and describe a bit of the history of the transmission of the Hebrew Bible.

This is the sixth post in a series on the textual criticism of the Hebrew Bible/Old Testament. Other posts include:

• Textual Criticism of the Hebrew Bible – An Introduction (TCHB 1)
• Textual Criticism of the Hebrew Bible Resources (TCHB 2)
• Hebrew Witnesses to the Text of the Old Testament (TCHB 3)
• Early Versions of the Hebrew Bible (TCHB 4)
• Codex Sinaiticus: A Profile (TCHB 5)

All posts in this series may be viewed here.

The history of the text of the Old Testament/Hebrew Bible can be divided into five periods on the basis of the kinds evidence available:

1. from the time of composition to mid-third century BCE, for which no extant texts are available;
2. from mid-third century BCE to end of first century CE, when the variety of types of texts found at Qumran are attested;
3. from the end of the first century CE to the end of the tenth century, that is, from the survival of proto-MT alone to the work of Aaron Ben Asher;
4. from the end of the tenth century to the sixteenth century, attested by the hundreds of medieval Masoretic manuscripts; and
5. from sixteenth century to the present, the time of printed Hebrew editions of the Bible.

Arguably, the discovery of the DSS has so revolutionized our understanding of the text, that it ought to be marked off as a new era. Harold Scanlin, United Bible Society translation advisor, pretty much argued as much when he said: “These changes [in our understanding of the history of the text] are at least as significant as the nineteenth century revolution in New Testament textual criticism, culminating in the work of Westcott and Hort” (Harold P. Scanlin, “The Presuppositions of HOTTP and the Translator,” BT 43 [1992] 102).

For the purposes of these posts, I have treated the last three periods sufficiently in my discussion of the MT, therefore I will focus here exclusively on the first two periods.

From Composition to Mid-Third Century BCE

Discussion of this early period is necessarily conjectural since there are not extant manuscripts from this era. While a few scholars posit a very late date for the writing of many biblical books, based on a number of lines of evidence it is plausible that the majority of biblical books were composed by the mid- to late-third century BCE. (Some of the evidence includes early versions like the Septuagint, while other evidence is based more on the signs of development apparent — at least to me! — in the biblical text, as well as the intertextuality between many biblical books.) While perhaps this warrants a future post, suffice it to say that I am assuming most — but not necessarily all — biblical books were composed during this period.

During this period, it can be inferred both from extra-biblical and biblical sources a tendency both to preserve and to revise the text.

1. The Tendency to Preserve the Text.
There are three factors which demonstrate the early scribal tendency to preserve the text. First, the very fact that the biblical books persistently survived the most deleterious conditions throughout a more or less long history until the extant manuscripts demonstrates that indefatigable scribes insisted on its preservation. The books were written on highly perishable papyrus and animal skins in the relatively damp, hostile climate of Palestine. The prospects for their survival were most uncertain in a land that served as a bridge for armies in unceasing contention between the continents of Asia and Africa — a land whose people were the object of plunderers in their early history and of captors in their later history. That no other purported Israelite writings, such as the Book of Yashar (e.g., 2 Sam 1:18) or the Annals of the Kings (e.g., 2 Chron 16:11), survive from this period indirectly suggests the determination of the scribes to preserve the biblical books. Of course, I am assuming that there may be something behind many of these references to other writings, instead of seeing them as rhetorical devices only serving to give some verisimilitude to the writings.

Second, the OT itself (cf. Deut 4:2; 12:32; Josh 1:7; 24:25, 26; 1 Sam 10:25; Ps 18:30; Prov 30:6-7; Eccl 12:12) and relevant literature of the ancient Near East show that at the time of the OT’s composition a mindset favouring canonicity existed. For example, the famous Code of Hammurabi (ca. 1750 BCE) and the Hittite treaties of the Late Bronze Age (ca. 1400-1200 BCE), closely resembling Deuteronomy, call down imprecations on anyone who tampers with one word in them. This mindset must have fostered a concern for care and accuracy in transmitting the sacred writings.

2. The Tendency to Revise the Text.
On the other hand, both biblical and extra-biblical data show a tendency to revise the text during this period. This can be demonstrated by four strands of evidence. First, the post-exilic book, Ezra-Nehemiah, states that as Ezra read from the Book of the Law of God, he made it clear and gave the meaning so that the people could understand what was being read (Neh 8:8), implying he modernized and explained the earlier text.

Second, the many differences between synoptic portions of the Hebrew Bible strongly suggest that those entrusted with the responsibility of teaching the Bible felt free to revise the texts (Compare, for instance, 2 Sam 22 = Ps 18; 2 Kgs 18:13-20:19 = Isa 36-39; 2 Kgs 24:18-25:30 = Jer 52; Isa 2:2-4 = Mic 4:1-3; Ps 14 = 53; 40:14-18 = 70; 57:8-12 = 108:2-6; 60:7-14 = 107:14; Ps 96 = 1 Chr 16:23-33; Ps 106:1, 47-48). The differences between synoptic portions resemble the same sort of variations found in the Qumran scrolls, suggesting that scribes, before the extant texts, felt free to revise the text within the similar restraints attested in the Qumran scrolls as noted by scholars.

Third, this effort to clarify and update the text was entirely in keeping with textual practices in the ancient Near East. Albright said: “A principle which must never be lost sight of in dealing with documents of the ancient Near East is that instead of leaving obvious archaisms in spelling and grammar, the scribes generally revised ancient literary and other documents periodically.”

Finally, the book of Chronicles in its synoptic parallels with the pre-Samaritan Torah and with the MT’s Former Prophets exhibits the same kinds of revisions as found in the Qumran scrolls, reflecting the early revision of texts. In short, some biblical texts were being conserved and revised at the same time others were being composed.

3. Kinds of Revisions.
From ancient inscriptions and comparative Semitic grammar we can plausibly trace the development of the biblical text’s script and grammar. From epigraphic evidence it appears that in its earliest stage, the text was written in the Proto-Canaanite alphabet, a pictographic alphabet (see my post on Serabit el-Khadem for more about the origins of the alphabet). This script later developed into an angular, pre-exilic Hebrew script, sometimes called Phoenician. At about 1100 BCE short vowels, indicating case and tense, were dropped (See Waltke & O’Connor, Introduction to Biblical Hebrew Syntax [Eisenbrauns, 1990] § 8.1c, 29.4j). During the same period matres lectiones (“mothers of reading”, i.e., vowel letters) were gradually added to the text.

From Mid-Third Century BCE to Late First Century CE

From ca. 400 BCE until the destruction of Second Temple in 70 CE, there also was a tendency to preserve and revise the text, as attested by the DSS and by other Jewish literature from this period.

1. The Tendency to Preserve the Text.
Talmudic notices, calling for a careful preservation of the Hebrew Scriptures, are backed up by the discoveries in the Judean desert. The preservation of the proto-MT reflects its antiquity and preservation. In addition, the para-textual scribal elements attested in the MT about the uncertainty of a few readings, such as inverted nuns (thought to mark verses thought to have been transposed) and other extraordinary points, probably go back to this period and so show an early concern for the text’s preservation.

2. Tendency to Revise the Text.
On the other hand, the variant text types attested among the Qumran scrolls unambiguously show that the proto-MT, the pre-Samaritan, and the Septuagintal texts continued to be copied during this period and that those of the “Qumran practice,” and possibly of the non-aligned texts, arose at this time. These variants also find agreement in Jewish literature originating during the time in question, such as the Book of Jubilees (either late or early post-exilic) and the NT (ca. 50-90 CE).

As I have argued, at the end of this period the rabbis stabilized the text by preserving only the dominant proto-Masoretic text type. The fall of the Second Temple, the Jews debate with Christians, and Hillel’s rules of hermeneutics, all called for a stabilized text. And socio-political realities led to dominance of the early rabbis and their proto-Masoretic text.

3. Kinds of Revisions.
Sometime after the Exile the Jews switched from the pre-exilic, angular script to the post-exilic, Aramaic script, also called square script because most of the letters are written within an imaginary square frame (It should be noted, however, that some DSS of varying text types were still written in the angular paleo-Hebrew script). With the new script came five final letter forms, which helped the division of words. Here is a chart from GKC with the various semitic alphabets and scripts (click to enlarge):

In addition, matres lectiones continued to be added to the text and spellings were updated (orthography). These changes are seen (in varying degrees) in the variety of different texts types extant in the DSS. The pre-Samaritan text, for example, exhibits linguistic modernization, expansions, interpolations, and exegetical smoothing, as does the proto-MT.

Another significant revision from this period is the safeguarding of the tetragrammaton (YHWH) by occasionally substituting forms in the consonantal text (this is also seen in the translation of YHWH by “Lord” in the LXX).

Conclusions

As a result of this transmission history (only briefly sketched here), by the end of the first century CE, the biblical text had undergone a series of intentional and unintentional changes and a number of varying text types emerged. The relationship between these text types is rarely simple to discern, and some books appeared to have more than one final form (or at least they circulated in more than one version).