Journal of Hebrew Scriptures - Volume 12 (2012) - Review
This study is a slightly revised version of a doctoral dissertation completed in 2008 under the direction of Robert P. Gordon, Regius Professor of Hebrew at Cambridge University. Harvey's initial purpose seems deceptively simple: to account for variations or inconsistencies in the way that the Septuagint translates some 887 instances of יהוה אלהים found in the MT. The LXX typically renders יהוה as κυριος, אלהים as θεος, and the combined title as κυριος ὁ θεος, although there are numerous exceptions to this rule. Harvey organizes the history of debate in methodological terms, reviewing previous lexical, grammatical, and theological/narrative approaches, several of which concern a particular passage or book rather than the corpus as a whole. His own proposal is to offer a comprehensive survey of the relevant data, although he acknowledges that detailed examination of every occurrence is not possible within the confines of a single book. As a basis for comparison, Harvey takes the Leningrad Codex of BHS (which he designates MST, or Masoretic Survey Text, as representative of the Masoretic text type) and the best available critical texts for the LXXwhether the Göttingen editions, Margolis, or Rahlfs. The latter are collectively referred to as the Septuagint Survey Texts (GST) (p. 12; not, as might be expected from the previous example, SST).
After describing in detail his method of data collection, subset selection, and analysis (chapter 2), Harvey provides exhaustive tabulations of occurrences of יהוה אלהים in each book of the MT, whether in relation to total occurrences of יהוה or in terms of types of appositive (articles, constructs, suffixes, nomina recta), syntactic function (subject, object, construct, complement, direct address), identity of the speaker, and variants in synoptic parallels (chapter 3). Also listed are variant readings from the Samaritan Pentateuch, Dead Sea Scrolls, and the collection of medieval manuscripts collated by Kennicott. To choose examples at random, we learn for instance that the use of יהוה אלהים as an address is rare. Psalms accounts for 30.4% (17/56) of the occurrences, 12 Chronicles for 28.6% (16/56) and 1 Kings for 12.5% (7/56) (p. 45). Similarly, undetermined and construct forms followed by צבאות also occur more frequently as subjects. Fifty-eight per cent (22/38) of the undetermined cases and 61.1% (11/18) of the construct forms followed by צבאות are subjects. These figures are noticeably higher than the 40.8% (362/887) overall average (p. 48).
Patience is required as the reader waits to discover the relevance of such meticulous detail to the main thesis of the study. Comparative data is plotted on bar graphs, with fuller information recorded in separate appendices (as is also the case in subsequent chapters). The graphs, incidentally, prove difficult to decipher due to their use of different black and white patterns (light/dark print, dots, vertical/diagonal bars, etc.) to designate different sets of data. At a list price of US $120.00, the volume is already prohibitively expensive, but this material would nonetheless have been much clearer in colour (as it appears to have been originally). Moreover, whereas for the majority of bar graphs differences in line height indicate proportional variations of numerical frequency or percentage of occurrences, in Figure 23 (p. 123) differences in height correspond rather (and without explanation of the fact) to instances of אלהים ,יהוה, and יהוה אלהים, respectively. Notwithstanding the author's claim that different bar heights were used to aid legibility (p. 122 n. 9), they instead prove confusing. This is equally the case with Figure 24 (p. 126), which endeavours within a single graph to represent translation variants between 195 occurrences of אלהים/יהוה and κυριος/θεος in the MT and LXX (cf. Figure 28, p. 166). Yet a third style of graph (a single bar for each book, subdivided proportionally into different patterns) appears in Figures 26, 29, 32, and 33. Finally, one cannot help but notice that rotating a bar graph 90° counterclockwise to make it fit onto a standard 6 x 9 book page renders explanatory text that has already been rotated 90° within the graph upside down on the printed page (Figures 2, 8, 1012, 1419, 2223, 2526, etc.). This is, at the very least, visually infelicitous.
A fourth chapter addresses questions of context in the target language. Harvey enumerates a wide range of variations in Greek translation: instances of יהוה אלהים that the LXX renders simply as κυριος; instances of κυριος θεος without a corresponding יהוה אלהים in the MT (he reports 189 of these); various renderings of צבאות and other additional elements; the presence or absence of definite articles or pronominal suffixes. The data he provides is again highly specific: we learn, for example (p. 81), that among 59 cases in which the Greek translation does not correspond precisely to a pronominal suffix in the MT, there is one instance each of יהוה אלהיך ,יהוה אלהי, and יהוה אלהינו that lack a corresponding possessive pronoun in the LXX (Num 22:18; 1 Kgdm 25:29; and 4 Kgdm 18:22, respectively, although Harvey does not provide the references). Further complicating the picture is evidence from non-Septuagintal Greek traditions, more specifically where the readings in question differ from the LXX in regard to one or both of κυριος and θεος, with or without the presence of additional elements in either language (pp. 8899). Points of note include instances in which the Old Greek, against the LXX, lacks possessive pronouns anticipated by the MT; the transposed use of θεος for יהוה and κυριος for אלהים; or the observation that Old Greek not infrequently renders MT אדני יהוה as κυριος θεος where the LXX has κυριος alone. Accounting for degrees of correlation between צבאות and παντοκράτωρ, δυνάμεων, or σαβαωθ, etc., in the three traditions is especially complex, but thankfully there are somewhat fewer instances of these. Harvey speculates that since in seven specific cases the LXX renders MT אדני יהוה as κυριος ὁ θεος, perhaps elsewhere as well (in Isaiah, for instance) the same Greek sequence reflects an original אדני יהוה against MT יהוה אלהים (pp. 83, 85). He makes a similar suggestion with regard to אדני יהוה and אדני יהוה צבאות in relation to κυριος θεος in non-LXX traditions (pp. 95, 99). Once more, readers may be forgiven if they find it difficult navigating through the overwhelming amount of detail contained in these chapters.
The author's ultimate purpose emerges more clearly in his fifth and sixth chapters, which examine 34 instances of יהוה אלהים where אלהים is undetermined and not followed by צבאות (p. 100) and 22 occurrences of יהוה אלהי[ם] [ה]צבאות. The former task occupies some 32 pages, the latter a full 59. Here it becomes clear that this study intends not to survey translation variants so much as to propose alternative original readings in Hebrew on the basis of the LXX. In Exod 9:30, for example, it seems likely (p. 106) that the original reading was יהוה rather than יהוה אלהים (following LXX τον κυριον). In 2 Sam 7:25 Harvey proposes replacing יהוה אלהים with אדני יהוה on the basis of LXX evidence; he draws the same conclusion for several texts in 1 and 2 Chronicles, Psalms, and Jonah, but in these cases on the basis of stylistic considerations, or else as a result of apparent harmonization and conflation in the process of textual transmission. Genesis offers a particularly challenging example: between 2:4 and 10:9 there is no consistency in the LXX translation of יהוה ,אלהים, or יהוה אלהים other than the fact that אלהים and יהוה אלהים are never rendered as κυριος. Accordingly, a particular Hebrew Vorlage cannot be inferred from specific instances of LXX κυριος θεος. Yet assuming the validity of the Graf-Wellhausen model, Since texts attributed to the Yahwist are generally characterized by the use of יהוה, it is more likely that all 20 occurrences of יהוה אלהים in Gen. 2.4b3.24 read יהוה prior to being combined with the narrative in 1.12.4a, where the name אלהים predominates (p. 131). This is an important methodological departure. Having initially proposed to review translation variants on the basis of a single diplomatic exemplar (Leningradensis), the author next proposes improvements to an eclectic text on the basis of both Septuagintal and internal or contextual evidence. Here, however, he moves behind the canonical text (in whatever form) to reconstruct a pre-redactional source. How these different projects relate is not explained.
Analysis of יהוה אלהי[ם] [ה]צבאות is, of necessity, no less complex. Harvey recounts the history of debate regarding the grammatical function and meaning of צבאות, then lists all variants of the divine name that employ צבאות and their Greek translations (with further breakdown according to syntactic function, varieties of attendant expression, and textual variants; pp. 13350). He finds evidence for transposition of elements, recensional additions, and the progressive expansion of translation formulas over time. As to conclusions, Harvey proposes that all five instances of אלהי that precede צבאות in Jeremiah could be secondary (p. 165); this could be the case likewise in 2 Samuel and 1 Kings. After an especially detailed review, he concludes, it is possible that every occurrence of אלהי that is followed by ה]צבאות] in Amos is secondary (p. 186); similar conclusions for Hosea and the book of Psalms either could be or are possible. In contrast to previous examples, however, none of the last three proposals appeal to Septuagintal evidence.
A brief final chapter summarizes the preceding argument, then observes (p. 201), The main conclusion arising from this study is that יהוה אלהים is an artificial compound designation. This assessment is based on the irregular distribution of the 38 occurrences, the questionable usefulness of undetermined אלהים as an appositive after יהוה and the possibility of providing a viable explanation of how each of these designations could have been introduced into the Hebrew Bible. However confidently asserted, this conclusion is undermined to some extent by the overwhelming complexity of the evidence: given a) not inconsiderable variations in both Hebrew and Greek textual traditions, b) evidence for stylistic variations, transpositions, and translation variants even within individual books and specific text families, and c) the difficulty of establishing reliable developmental trajectories, at what point is the probability of a proposed reconstruction reduced beyond statistical usefulness?
Notwithstanding the occasionally diffuse aims of this study, it will be of interest primarily to specialists in Septuagintal and Hebrew textual traditions. The published text reflects the same remarkable attention to detail that is evident throughout the study itself: I was able to find only a single typographical error in more than 250 pages.