How the Bible Became a Book: A Review Article of Karel van der Toorn, Scribal Culture and the Making of the Hebrew Bible
A review article of Karel van der Toorn's, "Scribal Culture and the Making of the Hebrew Bible," by Dr. Richard S. Hess.
Van der Toorn, Karel. Scribal Culture and the Making of the Hebrew Bible. Cambridge, MA: Harvard University Press, 2007. x + 401 pp. Paperback, $18.95. ISBN 978-0-674-03254-5.
According to Van der Toorn, the Bible was produced between 500 and 200 B.C. However, before the Hellenistic Period (beginning c. 300 B.C.) there were no books and so the Bible did not exist as one. Van der Toorn accepts a limited level of literacy. He compares the literacy of Mesopotamia at less than 5 per cent, that of Egypt at less than 7 per cent, and that of Greece at less than 10 per cent. He thus suggests a similar percentage for ancient Israel. However, this is “high literacy.” Van der Toorn distinguishes it from the ability to write one’s name and to read a letter. This latter form of reading (and writing) was “quite common” in ancient Israel. He understands the books of the Bible to be anthologies in virtually every case. They were constructed by editors, not by the original authors. Therefore, the literary structure and the style of these books are a later imposition not reflective of the original authors’ intent.
This all begs the larger question of what constitutes a book. For this van der Toorn uses the definition of von Wilamowitz that it is “a text published by its author through the medium of an organized book trade for the benefit of an expectant public” (p. 25). As van der Toorn notes this would exclude the existence of virtually any book in the ancient Near East. However, it would also exclude the formation of a book by newspaper installments such as the books of Charles Dickens and other well known Victorian and later authors of the nineteenth and twentieth centuries. They did not publish “a text” by itself but a group of smaller texts, one after the other, similar to the construction suggested by van der Toorn for the biblical books. The definition would also exclude various e-books and other media that are created and “published” on the internet without the use of any sort of “organized book trade.” This is especially true of many written items that are as long as various biblical books in terms of number of words but are posted on blogs and other devices and thereby made available for free to anyone who wishes to access them (such as this review). Rather than reducing or limiting the number of written items, these contexts of “publication” have increased the number of volumes. They call into question what amounts to an overly simplistic definition of the production of books in the modern era and therefore to an overly restrictive definition regarding the production of books in the ancient world.
In discussing authorship, van der Toorn identifies compositions of the ancient world as largely anonymous and limited to occasional colophons that included the name of the scribe who copied the text and perhaps the name of the author. However, his assertion overlooks a wide range of genres that preserved the names of their authors in the opening lines of the work. Letters, prophecies, treaties, and other documents often named their authors at the beginning of the document. These were not pseudepigraphic compositions nor were they honorific authors, as van der Toorn identifies some texts. In fact, the composition of Deuteronomy that he describes as peseudepigraphic has been understood by some scholars as a treaty/covenant document. It, like other ancient Near Eastern treaties, identifies its author at the beginning of the document. This same confusion emerges when van der Toorn discusses the Mesopotamian Catalogue of Texts and Authors. This catalogue describes literary works other than those genres mentioned above. For example, it does not list prophecies. So it is not clear that van der Toorn compares similar items when he relates this catalogue to the Talmudic tradition of ascribing the compilation of Isaiah to Hezekiah. However, a wisdom text such as Proverbs is properly compared to the catalogue. Its “authorship” may indeed refer to editorial arrangement and activity as appears to be described in Proverbs 25:1.
Van der Toorn provides a helpful review of the scribal training and abilities in Mesopotamia and in Egypt. In both cases the scribes in the religious traditions achieved a middle class status somewhere between the menial workers and those of the religious elite. In the Mesopotamian tradition, the more advanced scribes rose beyond the levels of rote memory and copying to discussion and debate concerning the classical works of their profession. When he considers Israel, van der Toorn advances the text of Jeremiah 8:8-9 which is repeated many times in his book. He understands this text as implying that Jeremiah’s prophetic word stood in opposition with the Torah as created by the scribes of Jeremiah’s day. Thus Jeremiah argues that the Torah is a deceitful product of scribes. This may be one interpretation of this text. However, modern translations understand Jeremiah 8:8-9 as an indictment for falsely dealing with the Torah, not a charge of inventing a false torah. The problem is one of interpretation rather than van der Toorn’s concern with production.
One need not follow van der Toorn’s assumptions about the lack of authenticity for the unprovenanced Baruch seal in order to agree that many scribes were associated with the temple. Indeed, his observation of an absence of separation between the secular and the sacred, or more specifically between the palace and the temple, lead one to accept in large measure a sacred provenance for much of the scribal activity surrounding the production and preservation of the Bible. The comparison with Ilimalku of Ugarit is apt: He appears to have been a scribe connected with both the temple and the palace. The deposit of the description of kingship in the Shiloh sanctuary (1 Samuel 10:25), the holy background to the Torah (Hosea 8:12), and the connection of the book of Torah with the temple (2 Kings 22-23) all affirm the scribal connection with the temple. The Levites in Chronicles are involved in Torah instructions as well as in civil and other temple duties. Further, van der Toorn points to the sages as one group of explicitly religious professionals in Jeremiah 18:18. Deuteronomy 17:18-19 demonstrates how the king is to copy the Torah from a scroll “before the Levitical priests.”
It is in this context that van der Toorn makes an important observation concerning literacy (p. 97): “The epigraphic evidence suggests that training in rudimentary scribal skills was available throughout Palestine, but the formation of scribes who were ‘expert and wise’ required a program of study provided only in the temple school.” Indeed, this was my point in the note, “Writing about Writing: Abecedaries and Evidence for Literacy in Ancient Israel,” Vetus Testamentum 56 (2006) 342-46. Functional literacy was available widely in Israel. Whether or not the abecedaries from Palestine are to be understood as samples for engravers and potters is another matter. If the Izbet Sartah inscription is an abecedary, it seems difficult to explain the repetition of letter forms on the text as a sample text for an engraver. It seems more natural to understand here some “scribal scriblings,” however elementary the writer may have been. Modeling the Mesopotamian curriculum, van der Toorn suggests that Hebrew scribal training was divided into two parts. In the first phase, the twenty-two Hebrew letters were “easily mastered” and skill was then developed in the speed and clarity of style involved in the formation of each letter. Acrostics such as Psalms 25 and 119 formed the counterpart to the cuneiform acrostic, the Babylonian Theodicy. These were used as copy texts to instruct beginning students. Van der Toorn creatively identifies possible lists in the biblical texts that could have been used for the development of related vocabulary (p. 99): animals in Leviticus 11 and Deuteronomy 14, places in Numbers 33, jewelry in Isaiah 3:18-23, and “revealed things” in the later apocalyptic literature.
Reasonably it may be argued, as van der Toorn does, that the prologue of Ben Sira mentions the use of “the books of the fathers” as a curriculum for scribal instruction. The texts of Deuteronomy, Psalms, and Isaiah dominate at Qumran and in the New Testament’s use of the Old. Therefore, these are reasonably the ones most likely forming the essential books of the scribal curriculum in the Second Temple Period. Scribes would have committed them to memory. Advanced classes would have discussed these texts. Van der Toorn speculates that the additions to Isaiah lay behind this ongoing debate about the interpretation of the text, although it seems difficult to believe that some of the highest, most elegant style of Hebrew poetry should emerge from a committee.
Van der Toorn considers how many of the texts were transmitted orally and then written down and edited by scribes. The later did not merely reproduce a memorized form, but creatively integrated what they received. Thus the material of Genesis 12, 20, and 26, about the endangered ancestress, goes back to a single oral tradition. In both Jeremiah 7 and 26 the prophet compares the Temple with Shiloh. This and the Deuteronomistic phrasing lead van der Toorn to suggest another example of scribal editing. However, as he observes “there is no way for us to penetrate beneath the scribal interpretation of the event” (p. 114). Therefore, these suggested common origins remain speculative.
Building on the note of Proverbs 25:1 concerning the “men of Hezekiah” collecting the proverbs of Solomon, van der Toorn argues that there are many list forms in the Hebrew Bible that were gradually constructed by adding items to the list. For example, the Covenant Code (Exodus 20:22-23:33) and various laws in Leviticus with attached colophons bear witness to gradual accretion. In some cases, scribes invented new laws. The same is true of the two introductions of Deuteronomy 1:1 and 4:44. This indicates that the material before 4:44 was added later (p. 128). Such a conclusion controverts the view of Deuteronomy as a vassal treaty form where, as least in the second millennium B.C., the historical prologue was common. Thus two beginnings to the work would have been expected. This is clear in the treaty between Suppiluliuma and Niqmaddu II of Ugarit which begins “Thus says his majesty Suppiluliuma,” and then, after the historical prologue begins again, “Now Suppiluliuma, Great King, King of Hatti, has made the following treaty with Niqmaddu, king of the land of Ugarit, saying” (G. Beckman, Hittite Diplomatic Texts [SBLWAWS 7; Atlanta: Scholars Press, 1996], pp. 30-31). This calls into question the view that the only explanation for all such multiple introductions or colophon summaries must be the result of later scribal additions. Clearly such was not the case with the Hittite vassal treaties and documents that may have been modeled on them.
Van der Toorn seeks clues of literary construction in the notes that recur. For example both Exodus 40:33 and Numbers 7:1 mention how Moses set up the Tabernacle. This indicates a frame within which the book of Leviticus and the first six chapters of Numbers fit. This seems likely but does the observation say about this material. Is it necessary to conclude that these texts must have once had “an independent existence” (p. 131)? Perhaps this is so, but it is not necessarily the case. It may rather signal the end of the first major section of laws and events at Mt. Sinai (Exodus 19-40) and the end of the last major collection of laws and events there (Numbers 1-6).
Van der Toorn considers examples of adaptation in the Hebrew Bible, such as (1) the borrowing of Proverbs 22:17-24:22 from the Egyptian Teaching of Amenemope and (2) the common background of the first part of Psalm 20 and the Demotic papyrus with an Aramaic blessing virtually identical to the biblical text (but appealing to a different deity). He concludes this section with an overview of the laws governing the Israelite slave in Exodus 21:2-11, Leviticus 25:39-46, and Deuteronomy 15:12-18. These are seen as contradictory. However, it is not clear that they are, especially if the principle of debt servitude is recognized in all three as the only means by which an Israelite is allowed to “own” another Israelite. If that is the case, then it is not actually ownership of the person (as the meaning of slavery is traditionally understood) but a kind of indentured servitude, an ownership of the person’s labor productivity (with strict limits placed on what is permitted). While it is certainly possible that the scribes involved with various aspects of the production and transmission of the Hebrew Bible were able to edit two texts into one, it is not always so easy to recognize this feature. In particular, the flood story of Genesis 6-9 remains a classic example to which appeal is made for redundancies and contradictions (p. 140), despite the fact that alternative forms of literary analysis allow for a coherency (see e.g., G. J. Wenham, “The Coherence of the Flood Narrative,” Vetus Testamentum 28 (1978): 336-48; reprinted in R. S. Hess and D. T. Tsumura, “I Studied Inscriptions from before the Flood:” Ancient Near Eastern, Literary, and Linguistic Approaches to Genesis 1-11[SBTS 4; Winona Lake: Eisenbrauns, 1994], pp. 436-47).
Based on the view that Deuteronomy was composed as a single scroll c. 620 B.C., and that it would need to be replaced with a new copy every forty years, i.e., in 580, 540, and 500 B.C.; van der Toorn suggests that these are the times when innovations would be introduced in the scroll. Thus he sees the Covenant edition, the Torah edition, the History edition, and finally the Wisdom edition of the scroll of Deuteronomy. These are identified by the additions that appear especially at the beginning and the end of the scroll. While this reconstruction is a plausible one and has the benefit of some support from comparative studies, it is interesting to see how different it is from the recent study of Thomas Römer, The So-Called Deuteronomistic History: A Sociological, Historical and Literary Introduction (London and New York: T&T Clark, 2007). This latter volume proposes an initial work followed by two further phases of development, dividing all three parts into the Neo-Assyrian period, the Neo-Babylonian period, and the Persian period. However, the literary development envisioned is not primarily one of additions at the beginning and end of the scroll (so van der Toorn), but texts scattered throughout both Deuteronomy and the remainder of the Deuteronomistic history. Even if one were to set aside the concerns already expressed regarding van der Toorn’s approach, there is no sense in which an agreement has been reached on the subject.
Van der Toorn turns to consider the prophetic writings. He concludes, on the basis of the Balaam inscription from Tell Deir Alla and on what he identifies as citations of pre-exilic prophets in 1 Kings and elsewhere, that written prophets did exist in the pre-exilic period. In studying the Neo-Assyrian prophecies, he observes the manner in which multiple oracles from different prophets were collected on a single large tablet. This is different from the biblical approach where the oracles collected tended to be attributed to a single author. In fact, van der Toorn notes that this more closely parallels the Balaam oracles at Deir Alla than the Neo-Assyrian evidence. Both the story of Moses in Exodus 32-34 and that of Jeremiah in Jeremiah 36 suggest that the oracles were not dictated and written, but rather they were written from recollection by a scribe composing them sometime after the period of the prophet. Suggesting parallels with psalms of lament and with Job, van der Toorn argues that the so-called Confessions of Jeremiah in chapters 11-20 could not have been written by the prophet because “No one kept this kind of diary” (p. 189). Therefore a scribe wrote this to demonstrate a prophetic identity of what it means to be “a real prophet.” Much of this persona is borrowed from earlier prophets such as Samuel, Moses, Micah, and Elijah. Further, earlier prophetic oracles were mined for phrases used in the oracles against the nations as found in chapters 46-51. The oracles against Babylon in chapters 50 and 51 themselves are borrowed from earlier texts in Jeremiah. Thus the text is an anthology of oracles created by scribes who studied ancient texts and applied them to new situations.
However, is this a necessary conclusion? Could the prophets themselves have applied earlier oracles and phrases in this manner? We know, for example, that the Neo-Assyrian oracles alluded to Mesopotamian literary traditions (see Charles Halton, “Allusions to the Stream of Tradition in Neo-Assyrian Oracles,” Ancient Near Eastern Studies 46 : 50-61). Since van der Toorn would suggest that these oracles represent texts closer to the original oral pronouncements, as opposed to the larger collections of oracles in a prophetic scroll such as Jeremiah, then this sort of relecture appears to go back as far as we are able to reach to the original oracles. It would therefore not appear to provide prima facie evidence for scribal editing and reworking. For van der Toorn the sermons of Jeremiah represent Deuteronomistic works composed by scribes in that tradition. But it is not impossible that they were composed by the prophet Jeremiah. The demonstration that phrases and ideas were borrowed and reused in prophetic books does not preclude that these were borrowed and reused by the prophets themselves.
Van der Toorn then turns to consider the origins of the concept of revelation, arguing that it is a product of forces that came together in the late second millennium B.C. He defines revelation as “human knowledge from a culturally postulated superhuman source” (p. 206). In Mesopotamia the late second millennium B.C. was the time of the writing of Enuma Elish where it was ascribed to an Ancient. By this time authoritative and revelatory texts were understood as those that were the oldest and whose origins reached back to the gods and to the apkallu sages from before the Flood. A good example of this is the Gilgamesh epic. In the Old Babylonian edition, the hero of the story received wise counsel from a female innkeeper. In the first millennium edition, it comes from the deified flood hero (pp. 213-14). In the Babylonia Theodicy and in Ludlul, both Babylonian wisdom compositions from the end of the second millennium, the wisdom of the gods was described as remote and hidden.
Van der Toorn finds c. 1300 B.C. to be a transition from oral to written sources as centers for authority. Lexicographical texts became fixed at this time. This suggests that the written exemplar that would be copied by students replaced the oral teacher who had been the authority. The texts take on the power of revelation and access to them becomes privileged. Therefore they became shrouded in secrecy, as evidenced by the colophons that so identify them. Turning to the Torah and its putative origin in the discovery of Josiah’s servants, van der Toorn emphasizes Deuteronomy 29:28. He suggests that this reads like a Mesopotamian secrecy colophon. However, neither this text nor Deuteronomy 5:28-31 (which he also cites) resemble a secrecy colophon. Such colophons emphasize that only the initiated are to have access to the texts and to their understanding. This is simply not the case with Deuteronomy. Not only are the texts commanded to be read publicly to all Israel every seven years, according to Deuteronomy 31:10-12, but even the so-called secrecy texts that van der Toorn cites command that all the laws be proclaimed to Israel. Nothing here is to be kept secret. The same is true of the Levitical cult. The biblical texts suggest that no texts are to be recited in secret and the whole book is presented as given to all Israel.
Van der Toorn finds evidence for a temple library early in Jerusalem. He cites the texts of 1 Samuel 10:25; 2 Kings 22; and 2 Maccabees 2:13-15. He concludes that around 450 B.C. Ezra identified the Torah as the beginning of the canon (p. 248). Two centuries later the Prophets, Psalms, and Proverbs were added. By the time of Josephus, he and others identified most of the Writings as part of the Prophets. Thus Josephus and most of the New Testament describe the canon as consisting of the Law and the Prophets. The Persian authorities gave Ezra the power to codify the Torah and to enforce it on the community around Jerusalem. Van der Toorn concludes that c. 250 B.C. Malachi was added as an anonymous work to bring the scroll of the Minor Prophets to an ideal number of twelve and to close the canon. He feels that this is signaled when Malachi 3:22 (English 4:4) echoes Joshua 1:7 and forms an inclusio for the entire work of the Prophets; with its emphasis on following the Law of Moses. However, this echo is odd if it is intentional. Other than the general verb “to command,” not a single verbal phrase is repeated in the two verses. “Be strong and courageous” and “be careful to do everything” are key and repeated phrases in Joshua 1. However, they occur nowhere at the end of Malachi. There is no explicit inclusio here.
Van der Toorn would ascribe a second century B.C. date to Daniel but he suggests that the author of this scroll was able to convince the Jerusalem priestly authority that the work was of greater antiquity, dating back to the purported time of Daniel at the beginning of the Persian empire. For this reason it was included in the canon.
Van der Toorn’s work is an important contribution to scholarship. It brings an updated model of the development of the Old Testament as a Scriptural canon and bases this model on reasonable and detailed comparative analysis of the better known world of Mesopotamian scribal schools and religious (and other) texts. Van der Toorn joins with other scholars who reject a very late canon formation, posited by some as after the Roman destruction of the Second Temple. The use of comparative evidence from ancient Mesopotamian texts and scribal culture provides a much needed corrective to other approaches. Indeed, some of the concerns expressed in this review come as a result of not going far enough with the comparative evidence. Van der Toorn’s reconstruction must of necessity contain speculative and controversial elements, given our limited knowledge and the relatively few pieces of written materials directly relevant to the formation of the Hebrew Scriptures. However, any serious student of this subject will learn the essential evidence and issues through reading this important work.
Richard S. Hess, Ph.D.
Earl S. Kalland Professor of Old Testament and Semitic Languages