Biblia Hebraica Quinta and the Making of Critical Editions of the Hebrew Bible¹

Richard D. Weis

United Theological Seminary of the Twin Cities

Introduction

1. The field of textual criticism of the Hebrew Bible finds itself in a remarkable situation in that three scholarly editions of the text of the Hebrew Bible, each with its distinct character, are now under way at the same time. Moreover, a fourth has just been republished with an English translation of its front matter and comments (Dotan 2000). This is a rare and important opportunity for reflection on the making of text editions. Consequently, I aim to reflect on some issues concerning text editions of the Hebrew Bible through a comparative presentation of some of the essential decisions and principles that shape one of these text editions, Biblia Hebraica Quinta (BHQ).² The comparison will consider BHQ in relation to earlier editions in the Biblia Hebraica series published by the Deutsche Bibelgesellschaft and its predecessors (specifically, the third edition of the Biblia Hebraica edited by Rudolph Kittel and Paul Kahle [BH3], and Biblia Hebraica Stuttgartensia [BHS]), as well as the other two editions currently in progress, the Hebrew University Bible (HUB), and the recently announced Oxford Hebrew Bible (OHB).³ To the extent possible the comparison will be illustrated with reference to Jeremiah 23:1-9, for which a sample page and accompanying commentary according to the principles of BHQ are offered in the appendix.⁴

2. Before proceeding with the comparative presentation, I begin with two basic points by way of introduction. The first structures my presentation of the decisions and principles shaping BHQ. The second anticipates some of my concluding reflections. First, text editions are not simply statements of scholarly judgments about the text, but are first and foremost tools for others. Thus audience and purpose are, or perhaps should be, as important in determining the shape of editions as the text-critical philosophy of the editors. Indeed, an edition's audience and purpose set the overall scope of the edition and frame all the other decisions that shape it. Second, I will propose in conclusion that the nomenclature used up to now in the field of textual criticism to describe editions of the Hebrew Bible--i.e., the alternatives of critical or diplomatic--is too limited and simplistic a typology for describing and conceptualizing such editions.

A Comparative Presentation of Biblia Hebraica Quinta

3. The Biblia Hebraica Quinta, which is now in progress, will be a single volume containing text, Masorah, and apparatus, plus an accompanying volume containing translations and notes for the Masorah along with commentary on cases in the apparatus.⁵ It follows the pattern of Biblia Hebraica Stuttgartensia in using the Leningrad manuscript as its base text and in offering a single apparatus. However, it returns to giving the Masorah in an essentially diplomatic presentation, reporting what is actually in the manuscript, and it prints both small and large Masorah with the text. Since the Masorah is properly considered a part of the text of a Masoretic codex, we may say that for the first time we will have a complete edition of the text of the Leningrad manuscript. The small Masorah will be placed in the lateral margins, and the large Masorah will be in the lower margin between the text and apparatus. The apparatus of BHQ is organized on radically different principles from BHS and will contain many more substantive cases than either BH3 or BHS. Indeed, for certain types of textual cases it may approach the HUB in comprehensiveness. Its use of abbreviated characterizations of the readings in the apparatus, and its commentary volume, go well beyond the level of comment offered in the fifth register in the HUB.

Audience

4. The primary, and intended, audience of BHQ is persons who are not specialists in the textual criticism of the Hebrew Bible but who need a scholarly, not to say critical, edition of its text. This group includes scholars of various kinds, Bible translators, clergy, and students. This audience is worldwide and works in a variety of contexts with quite differing access to resources. That this is the edition's audience was a key factor in decisions that shape what BHQ will be and that distinguish it from the HUB and OHB, which, as far as I can tell, may have different primary audiences in mind.

5. The need of significant parts of BHQ's primary audience for a text edition that is portable and affordable probably was the single most important factor in the decision to project BHQ as a manual edition consisting of a single text volume with an accompanying volume of commentary. In turn, that decision has meant that the editors constantly have been faced, as were the editors of previous volumes in the Biblia Hebraica series, with questions of selectivity. Not everything that one would want in a complete critical edition of the Hebrew Bible fits in one volume, so selection becomes a necessity.

6. The nature of BHQ's primary audience also led the editors to make modest interventions in what otherwise is a diplomatic edition of the Leningrad manuscript presented within the overall edition. As may be seen on the sample page of BHQ for Jeremiah 23:1-9 in the appendix, these include: the addition of chapter and verse numbers to the text; the use of stichography in texts judged poetic that the manuscript itself treats as prose (e.g., vv. 5-6); the insertion of verse references at the head of notes in the Masorah magna to aid the reader in linking notes and text; and the insertion of dots between sîmanîm in those notes.

7. Because of the present state of our field, BHQ also has a secondary audience, that is, specialists in textual criticism. This is not the edition's intended audience, but we hope it will be of some service to them, and we have tried to keep their needs and concerns in mind in the preparation of BHQ.⁶

8. Keeping this secondary audience in mind has affected BHQ in many ways, of course. In one noteworthy instance it affects the edition's ability to serve portions of its primary audience. Originally, the editors experimented with a variety of ways to make the entire contents of the apparatus accessible to the many readers who use only Hebrew and Greek. In the end they abandoned these attempts in favor of giving every reading that will appear in the apparatus in its original language and a printed form of its own script. So, for example, the references in Latin found in apparatus entries for vv. 1 and 6; in Syriac in entries for vv. 4, 5 and 8; and in Aramaic in entries for vv. 4 and 5 would have been given in English translation had the general editors followed their original plan, whereas now the actual readings are presented. Unquestionably the current practice of BHQ makes the apparatus more transparent in relation to the textual evidence, and therefore a better tool. For scholars who control the requisite languages, this is an important gain. For those who do not read some of the relevant languages, it makes some of the information inaccessible to large parts of its primary audience, making the tool perhaps less useful for them. This decision, correct as it is, illustrates the kind of compromises necessarily involved in making text editions.

9. Both audiences of BHQ, as they have been defined, include adherents to a variety of schools of thought about both the history of the text of the Hebrew Bible and what constitutes textual criticism, who nevertheless need to find BHQ a useful tool. This fact led the editors to discard forms for the critical apparatus, or (at one stage) apparatuses, that organized the evidence of the witnesses according to a particular theory of the text's development, an arrangement not dissimilar from that in the HUB. This also led, I think, to the admission into the apparatus, in very limited and clearly marked ways, of conjectures and cases that have previously been regarded as text critical in nature even though there is no arguable variation in the Hebrew text.

Purpose

10. The purpose of BHQ in relation to its primary audience is to provide them with a clear statement of what the BHQ editor judges to be the earliest attainable form of the Hebrew/Aramaic text that can be discerned on the basis of the surviving manuscript evidence, and that is useful for translation and exegesis, and also to provide them with a basis for criticizing the editor's judgment.

11. This purpose with this audience is the basis for an important set of characteristics of BHQ. The need of non-specialists to be able to determine the preferred text of the Hebrew Bible for purposes of translation and exegesis led to the decision to indicate the editor's recommendation for the reading that is to be preferred in each case, although in practice this only needs to be marked when the preferred reading is not found in the base text (e.g., in apparatus entries for vv. 3, 6, and 7 of the Jeremiah sample in the appendix).

12. Of course, this distinguishes BHQ from the HUB, which does not explicitly indicate such preferences, although I would argue that the HUB's comments on the contents of its apparatuses often clearly imply preferences. The comment in HUB on the Old Greek reading of <grc>tou= laou= mou</grc> for the Masoretic Text's <heb>co)niy</heb> in Jer 23:3 provides an example of this (Rabin, Talmon, Tov 1997: 118). To comment "hardly <heb>(my</heb>" unambiguously implies a preference for the reading in the Masoretic Text, in spite of the claim by the HUB editors that they "desire to present nothing but facts" (Goshen-Gottstein 1965: 7). This is not materially different from the apparatus entry in BHQ, which characterizes the Old Greek reading as assimilation to the usual expression in the Bible, and thus implies a preference for the reading in the Masoretic Text.

13. On the other hand, BHQ's systematic practice of indicating preferred readings aligns it with the OHB, which necessarily will indicate preferences systematically through its construction of an eclectic text. If one may take Ronald Hendel's edition of Genesis as indicative of the future OHB, OHB will insert its preferred readings into the text. In BHQ these will not be inserted in the running text, but instead will be presented in the apparatus. However, the result is essentially the same, the reader has a recommendation from the editor about the best text for use in translation and exegesis, over against the HUB, where such preferences intentionally are not systematically presented in spite of the exceptions just pointed out.

14. The fact that BHQ will systematically indicate the editor's recommendation for the preferred reading among the surviving witnesses in a case has implications for the edition's practical use. This practice on the part of the editors turns the running text on the page into a reference point around which the reader constructs the text that is preferable. Thus the proper use of the edition does not consist of simply reading this base text as if that were the best text, but rather consists of using base text and apparatus together.

15. For an editor to indicate a preference among the surviving readings in a case introduces an unavoidable degree of subjectivity into the edition. For that reason, BHQ strives to make its subjectivity explicit and to equip the reader to make his or her own judgment. So the edition provides for either abbreviated characterizations for non-preferred readings (e.g., "via root <heb>pcc</heb>"; "assim-Ezek 34:31"; "lit"; "homarc"; "assim-usu"--see the apparatus to Jer 23:1-9), or a more extensive comment on the entire case. By making explicit the reasoning implied by the preference of one reading among others, these devices position the reader to consider other options. In this way BHQ is to be distinguished from earlier editions in the BH series, which famously indicate preferred readings through the use of various imperatives such as "read" or "delete," but rarely explain the origin of the other readings that are adjudged to be derivative of the preferred reading. The OHB apparently will follow a practice similar to that in BHQ by giving abbreviated characterizations that explain the origin of readings not preferred for reconstructing the text.⁷

16. Yet another BHQ decision, to include the evidence of all constantly cited witnesses in each case, except purely linguistic cases, provides the reader basic data with which to form her or his own judgment should she or he find that of the editor lacking. This decision, a further consequence of BHQ's purpose in relation to its primary audience, also distinguishes it from its predecessors and aligns it with the HUB, which also provides a fuller report of the primary evidence for the text, although it ought to be noted that the HUB does not always give the reading, but sometimes only describes the class of variation in ways that do not allow the reader to discover or infer what the actual variance is (e.g., as in the HUB apparatus entries for Jer 23:3 <heb>(l</heb> and <heb>nwhn</heb>, and 23:4 <heb>)lhym</heb> [Rabin, Talmon, Tov 1997: 118]). The reader can see the combined effect of these decisions to give all the evidence, and to more fully sketch the editor's argument about the evidence, by comparing BH3 and BHQ in the following cases: BH3 23:7<grc>a</grc> = BHQ <heb>b.:n"y</heb>; BH3 23:8<grc>e</grc> = BHQ <heb>hid.ax:t.iym</heb>; and BH3 23:9<grc>a</grc> = BHQ <heb>lan.:bi)iym</heb>.⁸

17. This shift in the approach to presenting cases in the apparatus of Biblia Hebraica points to a corresponding shift in the way the edition should be used. Previously the reader of Biblia Hebraica had one choice, whether to follow the editor's instruction or suggestion if that seemed helpful exegetically. Now the reader best uses the forthcoming edition by critically assessing the editor's argument and proposal for a given case in light of the full presentation of the evidence and then adopting the text that emerges from that critical dialogue as preferable. In other words, BHQ will enable its readers deploy the full range of text critical argument, referring to external evidence as well as internal, whereas its predecessors only equipped the reader to use internal evidence, and indeed, implied that that was all that mattered.

18. The edition's stated purpose in relation to its primary audience, namely, to provide the earliest attainable text of the Hebrew Bible that is useful for translation and exegesis, was a crucial factor in defining the criteria governing the selectivity that was necessitated by the size of the edition. Specifically, this purpose led to the adoption of the two primary criteria for including a case in the edition's apparatus: that the variation be significant for translation and/or exegesis, and that at least arguably a Hebrew/Aramaic variation must lie behind it.

19. The results of following these two criteria are quite striking when compared to the selection of cases in the apparatuses of BH3 and BHS. By comparison with BH3 a number of very minor cases are dropped. For example, the cases in BH3 at 23:2<grc>a</grc> and 2<grc>b</grc> are simply plene spellings that have no impact on translation and exegesis and so are excluded from BHQ.⁹ On the other hand, when we compare the BHS and BHQ apparatuses the rules just mentioned result in the inclusion in the BHQ apparatus for vv. 1-5 of seven(!) potentially substantive cases--including one preferred reading!--that do not appear at all in the BHS apparatus (<heb>Um:piciym</heb>, <heb>)Eloh"y yi&rf)"l</heb>, <heb>hfro(iym</heb>, <heb>co)niy</heb>, <heb>wahA$ibotiy</heb>, <heb>y"xat.U</heb>, <heb>melek:</heb>).¹⁰

20. The result, interestingly, is something very close to the core of substantive cases in the first two apparatuses of the HUB for Jer 23:1-9a<grc>a</grc> (Rabin, Talmon, Tov 1997: 117-119). The reader of the HUB primarily gains access to four additional categories of readings: variants in OG, V, S, and T that are not significant for translation or exegesis; variants from sub-groups of witnesses for OG; variants from Rabbinic literature; variants from a selection of medieval manuscripts. The first and fourth categories of variants are, of course, part of the total picture of the text's transmission, but seldom impact translation or exegesis. The variants in the second category, while potentially affecting translation or exegesis, more likely belong to the tradition of the OG, rather than attest a variant Hebrew Vorlage. The variants from Rabbinic literature also are often not relevant for translation or exegesis. When they are, they either are an insufficient basis for recommending a preferred reading, or second the evidence of other witnesses. For text-critical specialists seeking an overview of the total picture of the text's transmission these are all essential data. That the HUB should include them is entirely appropriate, given its purpose and intended audience. That BHQ should not include them is equally appropriate in light of its purpose for its intended audience.

Philosophical and Pragmatic Choices

21. A number of other decisions that are important in defining the character of Biblia Hebraica Quinta were made on philosophical and pragmatic grounds unrelated to audience and purpose. Thus, in implementing the selectivity necessitated by restricting the edition to a single volume, the editors adopted a third criterion beyond the two just mentioned (see above, par. 18): the inclusion of all non-orthographic variants from early Hebrew textual witnesses, for example, the biblical manuscripts among the Dead Sea Scrolls or the Samaritan Pentateuch. The decision to include this material comes from the editors' conviction that, due to the age and relative scarcity of such witnesses, they are of high enough importance that their non-orthographic variants ought to be reported without the filter of an editor's judgment about their significance for translation or exegesis.

22. Another such decision distinguishes BHQ markedly from the earlier editions in the BH sequence. A unit of variation will be noted in the apparatus only when one or more variants among the surviving textual witnesses exist, not--as Rudolph Kittel indicated in the preface to the first edition of the Biblia Hebraica--when there is a perceived exegetical difficulty regardless of the presence or absence of variation among the witnesses (Kittel 1905: iii).

23. This difference in approach is exemplified in the respective editions' treatment of Jer 23:3. In the apparatuses of BH3 and BHS, one finds the note that verse 3 has been added.¹¹ There is no reference to testimony of textual witnesses. The background for this case actually has nothing to do with the extant textual witnesses, among which there is no evidence at all for a form of the text without verse 3. The background is to be found only in the exegetical difficulty that arises when the text is evaluated in the light of certain assumptions about literary consistency.¹² This exegetical difficulty is the reason behind the apparatus references in BH3 and BHS. The total lack of evidence for the omission of verse 3 is the reason why there is no such entry in BHQ.

24. This change in the basis for including cases in the apparatus also changes the status of the readings and witnesses in the apparatus. They become the evidence for what the reading of the text should be, rather than a fund of potential solutions to an exegetical problem. This difference in the status of readings and witnesses is exemplified in the lack of a full report of the evidence of the witnesses in the apparatuses of the BH3 and BHS, even to the extent that some variant readings are omitted from cases that the editors of these editions included (e.g., BH3 v. 8<grc>e</grc> which refers only to the Old Greek variant, omitting the Syriac variant [a plural passive verb with no suffix]). This contrasts with the full inclusion in the BHQ apparatus of the testimony of all regularly cited witnesses.

25. This view, that textual criticism is to interpret the surviving evidence for the text and its transmission, is behind the BHQ editors' aim to reconstruct the earliest available form of the text, whether that can be said to be the original form of the text or not. At the same time it is the historian's interest in getting as close to the beginning of the transmissional process as possible that is behind the editors' interest in the earliest available form of the text.

26. Certainly one of the key philosophical perspectives that shapes BHQ is the recognition that the process of the transmission of the text was not only a mechanical matter, but that the text was in some degree transmitted in terms of its meaning for the ancient copyists' and translators' communities. Of course, this is not such a strange assertion concerning translators, but it is a possibility less commonly asserted about copyists, as BHQ does. Thus editors of the individual biblical books in BHQ are expected to be alert not only to the kinds of changes that arise from the mechanics of manuscript production but also to the kinds of changes that are made in consequence of meanings assigned to the text by its readers. The editors have elaborated correspondingly the repertoire of terms that can be used to describe the origin of variants. This extensive, and typologically structured, list of characterizations is indicative of the complexity of the process of text transmission in the editors' view. Individual book editors within the BHQ team will undoubtedly employ this repertoire with differing emphases, according to their own perspectives and the reality of the specific witnesses with which they have to deal.

27. These characterizations, however, are not simply an undifferentiated accumulation of terms, but conform to an underlying typology. This typology arises from several key conceptions about the process of transmitting texts. The underlying distinctions are organized in the following chart (Figure 1).

THE TYPOLOGY OF TERMS USED TO CHARACTERIZE READINGS IN THE APPARATUS OF BIBLIA HEBRAICA QUINTA

Not Bearing on the Case Bearing on the Case

Marking a Difference Treating Difference as a Change

Describing Change Describing Motivation

Scribe as Mechanic Scribe as Reader

Reaction Intention

Form Meaning

Figure 1.

THE TYPOLOGY OF TERMS USED TO CHARACTERIZE READINGS IN THE APPARATUS OF BIBLIA HEBRAICA QUINTA
Not Bearing on the Case	Bearing on the Case
	Marking a Difference	Treating Difference as a Change
	Describing Change	Describing Motivation
	Scribe as Mechanic	Scribe as Reader
	Reaction	Intention
	Form	Meaning
Figure 1.

28. First, not every surviving reading bears on a case. In some cases this has been due to physical damage to manuscripts, a factor long recognized in textual criticism. In other cases this is because the reading arises from redactional change that intended to produce a new text, rather than changes deriving from the process of transmitting an accepted text, a factor that has particularly claimed the attention of text critics with the publication of the Dead Sea Scrolls. Second, not every textual difference can be successfully interpreted as a change. In some cases all we can do is to mark the nature of the difference. Third, we cannot always discern the motivation behind a change. In some cases the best we can do is to describe the mechanics of the change.

29. Fourth, when we describe the motivations behind text-transmissional change, the picture we have of the scribes who transmitted the text is crucial. Do we see them simply as mechanics, copying without particular reference to the meaning of what they copy? This has certainly been a view of scribes that has informed text-critical work for a long time, and it is the view behind BH3 and BHS.

30. Alternatively, do we also see the scribes as readers, who may introduce changes as they react in "readerly" ways to their exemplar, or as they pursue their own intentions about the proper form and meaning of what they read? By the end of the twentieth century this view had become firmly rooted in textual criticism of the Hebrew Bible, as can be seen from the serious attention given to the corresponding categories of textual change in a standard discussion of the field (Tov 1992: 258-275). A particularly nuanced elaboration of the nomenclature for transmissional changes arising from the identity of scribes as readers is a distinctive feature of the work of the Hebrew Old Testament Text Project (Barthélemy 1982: XVII-XX),¹³ and also especially BHQ.

31. This view of scribes, and of the nature of the process of transmission, is partly at the root of the difference between BHQ and its predecessors in the BH series in the matter of selecting cases for the apparatus. Earlier BH editions often assumed that the process of transmission was essentially a process of corruption through mechanical error. The effects of transmission would appear as "rough" spots in the text, and so exegetical or philological difficulty was an effective marker for a case needing text-critical judgment. As a result of BHQ's more complex view of the process of transmission, we must assume that the process of transmission does not necessarily "roughen" a text but may also "smooth" it. Therefore, exegetical difficulty is not a reliable marker of cases needing text-critical judgment. Only variation among the surviving witnesses is a reliable marker.

32. As noted above, BHQ also takes seriously the survival of diverse literary forms of the text into the transmissional history of some books of the Hebrew Bible, for example, Jeremiah. This appears in the characterization of variant readings stemming from such diverse forms as "literary" (abbreviated as "lit" in the apparatus), and thus not relevant to establishing the text at hand. The editors' philosophical commitment to keeping that distinction clear is expressed in this particular fashion, however, because it is the only practical option within the limits of a one-volume edition (as opposed to printing two different texts of Jeremiah, for example).

33. Both the decision to follow a form of edition employing a base text from a single reference manuscript accompanied by a critical apparatus, and the decision to retain the Leningrad manuscript as the source of that base text have both philosophical and pragmatic grounds. When the general editors decided these matters, they were convinced that the time was not yet opportune for an edition built around an eclectic text. It seems to me that two issues were of particular importance in this decision.

34. One was the view that an eclectic text needs to be an editio critica maior, that is, a comprehensive text edition, rather than an edition that presents only a selection of cases of textual variation. This is an issue that the editors of the recently announced OHB project will need to take seriously. If the eclectic text and apparatus for Genesis 1-11 published by Ronald Hendel (Hendel 1998: 109-149) is taken as a test case for the OHB, then the edition will have to face the question of how it can claim to present an original text. Hendel excludes from the apparatus, and thus from the reconstructed text, variations he classifies as "minor variants," whereas he includes those he classifies as "major variants." However, any original text would have included readings that fall into both classifications (Hendel 1998: 115).

35. The other consideration influencing the BHQ decision not to offer an eclectic text was that the lack of early Hebrew witnesses extending over the entire Hebrew Bible forces an editor to present at least some retroverted forms for preferred readings (e.g., the preferred readings in Jer 23:3 and 7 [see Appendix, Figure 3]). The BHQ editors judged that in the early 1990s the field did not have the full range of critical editions and studies of translation techniques for the versions needed to put such an edition on the firmest possible foundation. This is still the case. By putting the retroverted form in the apparatus, the reader still has a choice to read the text of L instead, whereas putting the retroverted form into the text would take that option away.

36. Some would also point out, as a third issue, that the need to rely on reconstructed Vorlagen for some readings and the Masoretic Text for others will, in an eclectic text, either produce a questionable mixing of unvocalized text (the versional Vorlagen) with vocalized and accented text (the MT), or will lead to the creation of anachronistic readings when, for example, an Old Greek Vorlage is reconstructed in vocalized form.¹⁴

37. Alongside these philosophical considerations was the simple pragmatic consideration that, at a stage where the use of conventional production methods was envisioned, the Deutsche Bibelgesellschaft had a thoroughly proofed text of L already set in type. Then, when electronic methods came to the fore, the availability of an electronic text of L weighed heavily. These practical considerations, of course, contributed to the editors' decision to retain L as the edition's base text. At the same time, they were mindful that the HUB was presenting a diplomatic edition of the Aleppo Codex, and the Madrid school had published a diplomatic edition of the Cairo Codex of the Prophets (Perez Castro 1979-1992), so it seemed of value to the field to maintain an edition of the Leningrad Codex in print through BHQ. Finally, of course, since BHQ uses its base text as the reference point against which to collate the evidence for the text's transmission, that L is the earliest complete manuscript of the Hebrew Bible makes it especially useful since the edition then has a consistent reference point throughout.

Concluding Reflections

38. In conclusion, I offer some reflections about the making of editions (based on experience so far with BHQ). Before returning to my initial point about the field's classification of editions, I lift up four other issues:

There are multiple audiences that need scholarly editions of the text of the Hebrew Bible. These audiences need different things from an edition. Thus the field needs multiple editions of differing types. It is therefore a good thing that we have these three distinct text editions, BHQ, HUB, and OHB, in progress.
All text editions of the Hebrew Bible will be shaped to some extent by compromises. So in evaluating editions, we need to ask how well the strengths and weaknesses resulting from a particular set of compromises serve the edition's purpose for its intended audience. To a certain extent, it is not a question of some competition for the one "best" edition. Among BHQ and BHS, BH3, and earlier editions of BH, because they have the same audience in mind, and aim at the same or similar purposes, I think there can be a comparative assessment of efficacy. On the other hand, among HUB, BHQ, and OHB I do not think there can be such a clear comparative assessment of efficacy since they serve different audiences and aim at different purposes with those audiences.
As a field, and as makers of editions, we always need to keep in mind the question of accessibility (i.e., size, portability, and cost). In the academies of the developed world accessibility means one thing. For many users of editions outside this particular context it will mean something else. This has specifically been a deciding factor in some deliberations over the design of BHQ.
Since text editions are tools that will be used by readers holding to a variety of schools of thought, and since such tools are most serviceable when they most nearly approach a clear and even-handed presentation of their contents, each edition needs specifically to address the issue of how to cope with the unavoidable subjectivity introduced by indicating a preference among the various readings attested in any particular case. Earlier editions in the BH series entirely overlooked this question insofar as they expressed preferences, often quite strongly, but without making the perspective explicit. HUB addresses the question by not expressing preferences among the readings. For its audience of text-critical specialists that seems an appropriate solution. BHQ addresses the question by making choices, but by making their terms as fully explicit as possible so that the reader is aware of the subjectivity. The question of how OHB will address this is especially acute since it will propose an eclectic text, so that by its very nature it must make choices among the readings. However, the early indications are that it will follow the path of BHQ by making clear the terms of its choices.

39. I now return to my second original point, namely, that the typology critical-diplomatic is too simplistic for describing or conceptualizing text editions. What do we mean when we say an edition is "critical"? Ronald Hendel offers a helpful definition in saying that textual criticism involves the exercise of "an educated judgment concerning the textual data" (Hendel 1998: 109). This is a field-specific version of a standard dictionary definition of "critical," namely, "exercising or involving careful judgment or judicious evaluation" (Merriam-Webster 1999: s.v. "critical," 2c). I contend that by Hendel's own definition the entire Biblia Hebraica series of editions is unquestionably critical. Although I would disagree with many of the judgments presented in BH editions before BHQ, I would argue that to refer to the Biblia Hebraica series as "diplomatic," as has been done, makes little sense (Hendel 1998: 109; Tov 1992: 289; but see now Tov 2001: 373-374). Its judgments are expressed in its apparatus, rather than through a composite text, but the editions of the BH series unquestionably exercise "an educated judgment concerning the textual data" (Hendel 1998: 109). When one remembers the explanatory notes in the HUB, I think one must say that this edition too meets the definition of "critical," although that classification might be more easily supported if one employs another dictionary definition of "critical," namely, "including variant readings and scholarly emendations" (Merriam-Webster 1999: s.v. "critical," 2d).

40. All of this suggests that we need more complex ways of categorizing editions than the old, simplistic dichotomy, "diplomatic-critical." In particular, we need a set of sub-categories within the category "critical," if such differing editions all fall within that one classification.

41. My proposal is this (see Figure 2):

First, we should distinguish between critical editions (e.g., HUB, BHQ, OHB) and diplomatic editions (e.g., Biblia Hebraica Leningradensia [BHL], the Madrid edition of the Cairo Codex, the DJD series).
Second, among critical editions we should distinguish between descriptive editions (e.g., HUB) and prescriptive editions (e.g., BHQ, OHB). Descriptive critical editions present the testimony of the witnesses for the text's transmission. Prescriptive critical editions also express a preference for specific readings among the possibilities arising from that transmission history.
Third, among prescriptive critical editions we should distinguish between a collative type (e.g., BHQ) and an eclectic type (e.g., OHB). Between the collative type and the eclectic type, the differences are at least two. The first is in the nature of the principal text printed on the page. In collative editions, the text on the page is essentially a reference point against which the information in the apparatus is organized, as can be seen with BHQ. In eclectic editions, the text is the end result of a decision-making process recorded in the apparatus. The second difference is in the relation of the apparatus to that text. In collative editions the apparatus is both descriptive of the textual evidence in relation to the reference point and prescriptive of possible reconstructions of the most preferable text among the options collated. In eclectic editions the apparatus presents the evidence and arguments that justify the text printed.

A TYPOLOGY OF SCHOLARLY TEXT EDITIONS

Diplomatic editions
(e.g., BHL, DJD, Madrid ed. of Cairo Codex) Critical editions
(e.g., HUB, BHQ, OHB)

Descriptive critical editions
(e.g., HUB) Prescriptive critical editions
(e.g., BHQ, OHB)

Collative type
(e.g., BHQ) Eclectic type
(e.g., OHB)

Figure 2.

A TYPOLOGY OF SCHOLARLY TEXT EDITIONS
Diplomatic editions (e.g., BHL, DJD, Madrid ed. of Cairo Codex)	Critical editions (e.g., HUB, BHQ, OHB)
	Descriptive critical editions (e.g., HUB)	Prescriptive critical editions (e.g., BHQ, OHB)
	Collative type (e.g., BHQ)	Eclectic type (e.g., OHB)
Figure 2.

42. Of these options, I would argue that the collative prescriptive type of critical edition (illustrated by BHQ) combines the best of both descriptive and prescriptive types of critical edition. By collating the testimony of the witnesses in each case against a reference manuscript, it presents the evidence for the text's transmission in as neutral a way as possible. By proposing preferred readings, it makes clear to its readers that all readings are not of equal value. By putting those preferences in the apparatus, collated against a base text that is an extant witness, it makes clear the subjective nature of the process of reconstructing a text. By requiring the reader to put preferred reading together with base text, it also makes clear that any reconstructed text is a virtual text, not an actual text. In the current situation in the study of the text of the Hebrew Bible where many preferred texts necessarily will be based on subjective reconstructions of versional Vorlagen, this seems to me the clearest and most balanced approach, at least for the audience BHQ aims to reach, which is primarily composed of persons who are not specialists in the textual criticism of the Hebrew Bible.

Appendix

Figure 3. Sample page from Biblia Hebraica Quinta


Figure 3. Sample page from Biblia Hebraica Quinta

Figure 4. Sample pages from Textual Commentary to Biblia Hebraica Quinta



Figure 4. Sample pages from Textual Commentary to Biblia Hebraica Quinta

© TC: A Journal of Biblical Textual Criticism, 2002.

Endnotes

1I wish to express my appreciation to my colleagues Adrian Schenker and Arie van der Kooij for reading and commenting on an earlier version of this paper. Other versions were presented in a panel on critical editions of the Hebrew Bible at the annual meeting of the Society of Biblical Literature in November, 2000, and in a colloquium at Luther Seminary in St. Paul, Minnesota in February, 2001. I thank participants in those contexts for their reactions and comments, especially Ronald Hendel, James A. Sanders, and James Limburg.

²I write as a member of the Editorial Committee of Biblia Hebraica Quinta, but I do not claim to speak on the committee's behalf.

³The Hebrew University Bible is known in editions of Isaiah and Jeremiah. Volumes for Ezekiel and the Twelve Prophets are in varying stages of preparation and production. The Oxford Hebrew Bible was announced at the annual meeting of the Society of Biblical Literature in November, 2000, by one of its general editors, Ronald Hendel. Its editors have designed it to offer not a base text drawn from a single manuscript but an eclectic text which reconstructs the editor's judgment of the earliest attainable form of the Hebrew text. This will be supported by an apparatus giving the evidence for the reconstructed text. The OHB will also offer, in a parallel column, a diplomatic text of the main text of the Leningrad manuscript, but not its Masorah. Where more than one edition of a book existed in antiquity, parallel texts and apparatuses for each ancient edition will be included.

⁴No explicit comparison with Dotan's valuable edition (BHL) will be made since that is a diplomatic edition only of the running text of the Leningrad Codex without Masorah or apparatus. It really belongs to a different type of edition (see the Concluding Reflections to this article) and thus is difficult to compare to HUB, BHQ, and OHB.

⁵The first official fascicle of BHQ, containing a general introduction and the editions of the Megillot, will appear from the Deutsche Bibelgesellschaft in 2002. It is hoped that the entire edition will be complete in 2010.

⁶However, it is clear that something on the order of HUB is much more to the point for the needs of specialists in textual criticism of the Hebrew Bible, if anything can be said to be satisfactory for our needs short of recourse to the raw materials behind any of these editions.

⁷Based on the sample for Genesis 1-11 published by Ronald Hendel (Hendel 1998: 120-149).

⁸For the BH3 references see Kittel 1937: 746. For the BHQ references see the sample page in the Appendix.

⁹For the BH3 references see Kittel 1937: 746. For the BHQ references see the sample page in the Appendix.

¹⁰For the BHQ references see the sample page in the Appendix. For the appropriate point of comparison in BHS see Elliger and Rudolph 1983: 826.

¹¹For the BH3 reference see Kittel 1937: 746. For the BHS reference see Elliger and Rudolph 1983: 826.

¹²See, for example, Rudolph 1968: 145: "Nachdem 2b den schlechten Hirten in einem kurzen allgemeinen Wort die Strafe angesagt hat, folgt in 3f. ein Heilswort für das Volk. Fällt schon auf, daß das "Verstoßen" von 2 hier plötzlich nicht mehr bildlich, sondern wörtlich verstanden wird vom Zerstreuen Israels unter die Völker, so ist noch auffallender, daß hier eine Diaspora in der ganzen Welt vorausgesetzt wird (zu s. o. die Note). Da das für Jer's Zeit nicht zutrifft, kann der Vers nicht von Jer stammen. Ist dann 4 die Fortsetzung von 2? Daß der Anschluß gut ist, läßt sich nicht leugnen: . . . So ist also kein Grund, 4 für Zusatz zu halten (gegen Cornill); nur 3 ist als verständliche Glosse aus der nachexilischen Zeit zu streichen (vgl. 29, 14)."