Abstract
This article deals with the question of the nature of and scholarly approaches to studying Greek syntax in the Septuagint. The concrete point of departure is the publication of A Syntax of Septuagint Greek by T. Muraoka (Leuven: Peeters, 2016). The author discusses Muraoka’s work, while touching upon general trends in Septuagint scholarship, and reviews the book in a detailed manner. The author’s theoretical considerations are illustrated by two case studies that demonstrate the problems associated with Muraoka’s approach to syntax in the Septuagint. By way of conclusion, the author reflects on future directions in research on the Septuagint and its language usage.
The first volume of Henry St. John Thackeray’s landmark Grammar of the Old Testament in Greek was published in 1909. The introduction includes the following statement:
It is true that no final grammar of the LXX can be written at present. But the grammarian cannot wait for the final verdict of textual criticism. Grammar and criticism must proceed concurrently, and in some ways the former may contribute towards a solution of the problems which the latter has to face.1
Takamitsu Muraoka’s monumental Syntax of Septuagint Greek of 2016, seen by its author as completing Thackeray’s work,2 presents a superficially similar sentiment in its own introduction:
Whilst aware of valuable contributions which a syntactic research on [Septuagint Greek] conducted along the lines of translation technique are capable of delivering, I am not only a scholar, but also a human being. As such I am conscious of some limitations. When I started in earnest on the present work, I was already seventy-four years old. To investigate every question from the perspective of translation technique would have been sheer madness. Nor could I wait for, or was willing to wait for, completion of such a gigantic project somewhere, sometime. In the course of this research, as the reader can see at many turns, I did try to establish if this or that feature shows significant signs of influence of the linguistic structure of the source text. But I had to be realistic. Those who believe that a syntax of [Septuagint Greek] can be only complete after a thorough and systematic investigation of the Septuagint from the perspective of translation technique, I could only wish that they undertake such a research and hopefully revise or supplement what is presented here.3
Both authors state their unwillingness to wait for the completion of work in a relevant research sphere before attempting grammatical analysis of the Septuagint. For Thackeray, who focuses on orthography and morphology, that relevant sphere is textual criticism. For Muraoka, writing on syntax, it is translation technique.
Similar the sentiments may seem, but there is a key difference in motivation and practice. Thackeray, writing at a time when no critical edition existed and as a member of the editorial team that was working on the Larger Cambridge Septuagint, seeks to bring out the mutual dependence of grammar and textual criticism. He consciously aims to complement the work of the latter discipline in the Grammar of the Old Testament in Greek and usefully addresses textual issues throughout.
Muraoka clearly understands that Septuagint translators consciously (and as far as the respective linguistic coding-systems of Greek and Hebrew allowed) imitated the structure of their source texts—this practice produced abnormal patterns in their syntax that can seem highly distinctive in relation to contemporary Greek usage. He also writes with the benefit of a significant body of translation-technical research (especially that emanating from the famous Helsinki School) at his disposal. Nevertheless, he tends to ignore relevant questions in his book. He does consider various features in relation to matching Hebrew and Aramaic phenomena, but makes no attempt at a systematic treatment.
The result of this practice and also of some others considered below is a seriously flawed work. In the following treatment I will seek to indicate the major issues. In addition, I will argue that the time is not ripe for attempting Muraoka’s specific objective, a comprehensive syntax of the Septuagint.
That does not mean Septuagint grammarians ought to wait. In fact there is urgent need for activity and perhaps at last we have enough scholars available with sufficient training and interest to undertake it. Many more people are now studying the language of the Septuagint than was the case twenty years ago. But we need to choose objectives that are currently feasible. There is much work to do. The final part of my paper will address the challenges before us. My aim there is to identify some key objectives and to offer a viable approach for achieving them.4
1 Preliminary Observations on Muraoka’s Syntax
The publication of Muraoka’s Syntax was a momentous occasion for students of the Septuagint, who had long been starved of any extended treatment of the topic. The first volume of Thackeray’s Grammar was subtitled “Introduction, Orthography and Accidence.” A second volume on syntax was planned, but never written. Robert Helbing, whose concise Grammatik addressing “Laut- und Wortlehre” had appeared in 1907, had similar intentions, similarly unfulfilled.5 For a long period the discipline had to make do for syntax with an even older work. The first decade of the twentieth century was a busy one for Septuagint grammarians, and already in 1905 Frederick Conybeare and St George Stock had produced their valuable Selections from the Septuagint, which contains a modest grammatical introduction dealing partly with syntactic phenomena. This book was retitled Grammar of Septuagint Greek many decades later. The change seems to have granted it an oddly authoritative status on syntax (at least up until the appearance of Muraoka’s study), but was essentially a marketing exercise. The work’s grammatical section is contained within some 73 (and the portion addressing syntax only 28) of its 382 pages and is unchanged from the original and now seriously dated form of the 1905 publication.6
Muraoka’s contribution follows his important Greek-English Lexicon of the Septuagint (henceforth GELS; completed in 2009 and now the standard work of its type for the corpus)7 and is another formidable output. The Syntax runs to nearly a thousand pages and contains analyses of a vast number of syntactic and morphosyntactic features. The book presents a rich collection of data and many observations and insights that should prove of enduring value. Scholars will be trawling its findings for the foreseeable future. Simply to have completed the work is a great achievement, a testament to Muraoka’s seemingly indefatigable appetite and capacity for accomplishing mammoth projects. I want to stress this because the assessment that follows, though offered in a spirit of engagement, is highly critical. What is more, he apparently wrote the whole work in less than four years.8
That revelation, however, ought to introduce a note of unease for the reader. Impressive as the speed of production is, it also raises questions. To write a syntax of the entire Septuagint in such a short period of time, some desiderata must inevitably have been sacrificed. A careful examination reveals that this is indeed the case. The presentation and analysis of the data involve significant problems. It is not my intention here to present a conventional review.9 My aim is rather to investigate what I see as key conceptual and methodological issues. I will first note in passing, however, some other characteristics of the book that ought to be borne in mind.
The title seems to me to invite the assumption that this is a descriptive grammar, presenting a synthesis of current scholarship on each feature of the syntax treated. For the Septuagint, however, such a treatment is at present impossible, since the study of its syntax is really still in its infancy (see further §5 below). What we have here is essentially a work of argumentative scholarship. This can be seen from a glance through the footnotes, sometimes deployed for the purpose of conducting debates with other writers. The book thus has a strangely combative cast for what is clearly intended to be a reference work.
Meanwhile, there are numerous minor mistakes, including faulty citation of secondary sources, unsatisfactory or inaccurate referencing in both the text and the indices, and basic errors of fact. Examples are not hard to find and some will turn up in the following discussion (see for instance n. 10 below). They are perhaps one result of the extraordinary pace with which Muraoka works, though that ought not to be advanced as an excuse.
In addition, the Syntax (although beautifully produced) can in certain respects be hard to navigate. This difficulty stems from a number of factors. Muraoka’s style is dense and highly discursive. There is a lack of clarity in the layout (despite his habit of exhaustive subcategorization) and an at times bewildering tendency towards terminological obscurantism and conflation or confusion of grammatical categories. For example, it would be a mistake to assume the contents pages offer a straightforward guide to discussions of the categories indirect speech and indirect command and when one does find them, introduced under functions of the infinitive and consecutio temporum, the content is not entirely accurate.10 And then there is the large number of footnotes, which often contain material that might have been summarized in the main text.
2 Three Fundamental Issues for the Analysis of Septuagint Syntax
Let us now turn to the larger problems to which I have alluded. By way of approaching them we should remind ourselves of some basic facts. The Septuagint is a body of texts mostly translated from original Hebrew and in some places Aramaic works. These are supplemented by some original Greek compositions. Taken all together the corpus amounts to one of our most important surviving specimens of early Koine Greek. We have reached a point where a series of studies has demonstrated that the Septuagint’s linguistic evidence can no longer be hived off and ignored by students of the history of Greek, as it long was, on the grounds that its translation Greek has been influenced by the source languages. On the contrary, once the nature and extent of the Hebrew or Aramaic influence on each book has been isolated these texts can reveal a large amount of valuable information about the Greek of their period.11
Anyone analyzing the syntax of this material is obliged to consider three fundamental issues: (a) the relationship of the Greek in translated books to its source text—this involves both the nature and degree of any genuine bilingual interference and more importantly (at least in terms of the scale of impact) the translators’ characteristic methods of rendering their Hebrew or Aramaic texts into Greek; (b) the relationship of the Greek, whether translated or not, to contemporary non-biblical Greek, both literary and documentary; (c) the relationship between the different parts of the corpus itself, produced by different translators or original composers over a period probably of more than two centuries.
Muraoka’s introduction addresses all three issues in a revealing discussion that immediately raises concerns. On the first topic we have already been able to observe (in the extended quotation in my introduction above) his views regarding translation-technical analysis, both his recognition of its importance and defence of his own lack of systematic effort in that direction. He essentially states that he was determined to write a comprehensive syntax, but felt that at the age of 74 he did not have the time available as an active scholar to attempt the translation-technical component himself.
Muraoka also observes that a number of syntactic features do not necessarily depend on specific features in the source text and implies that a translation-technical analysis is not necessary to underpin their assessment.12 The fact, however, that for some features it is inappropriate to begin from specific source-text features does not grant us liberty to ignore the relationship of the Greek features in question to whatever linguistic structures of the source text more or less closely correspond to them. Muraoka’s method interferes with his capacity to assess adequately a very large number of the syntactic features he analyses in his book.13
On the second issue it is encouraging to note that Muraoka sees the language of the Septuagint as essentially the Greek of its time.14 This fits with a growing scholarly consensus that is securely grounded in the evidence. It rapidly becomes apparent, however, that his understanding of the Greek of the time depends heavily on standard grammars, almost all of which describe either classical or New Testament Greek. The one exception is an excellent choice as far as it goes, Mayser’s classic study of the Ptolemaic papyri, but it is unclear to what extent Muraoka understands that that great work of the pioneering age of papyrology is now becoming outmoded.15 He also ignores some other secondary literature, such as Geoffrey Horrocks’s highly relevant book on the history of Greek. On the whole Muraoka seems uncomfortable outside biblical material and more generally with recent research. A specific example is the choice to rest his analysis of verbal aspect on the theoretical position reflected in the seminal but on this topic long superseded grammar of Kühner and Gerth, having noted but retreated from the large body of more recent work.16 Another is his treatment of the genitive absolute construction, discussed by Aitken and Dhont in the introduction to this issue.
This impression of reliance on outdated research and the biblical environment is reinforced by a glance at the “Index of Passages” at the end of the book. Muraoka cites many ancient works from outside the Septuagint, but these citations are mostly drawn from the New Testament or classical literature. When it comes to documentary Koine sources, crucially important for Septuagint studies, he lists only limited references to papyri and almost none to inscriptions.17 As for Koine literature, there are almost 330 citations from the New Testament, but approximately 31 from sources such as Polybius, Plutarch, Josephus, Philo, Pausanius, the Palatine Anthology, Achilles Tatius, and so on. All of this has to limit our confidence in the picture of early Koine Greek that Muraoka presents in the Syntax. One sees at once that it is likely to be a somewhat inaccurate portrait. It should emerge clearly enough from what follows that this is indeed the case.
As for the third issue, Muraoka is of course fully aware that difference between parts of the Septuagint is a significant feature, but he does not believe that a separate treatment of the different parts is “absolutely necessary.”18 By way of example he rejects the idea of writing “two separate grammars for translated books and original compositions,” pointing out that the original compositions amount to only 5 percent of the corpus. It is sufficient in his view simply to point out features particular to individual books. But the distinction between translations and original compositions is, as Muraoka also appreciates,19 far from the only kind of heterogeneity to be found within the Septuagint. In different parts of the corpus we find different translation techniques and this can have a significant effect on the Greek syntax. The level of education and literary pretension manifested by each translator or group of translators is also important. Stylistic tendencies are another factor to consider, though Muraoka is cautious about taking these into account.20 The result of his approach on this front, treating the evidence from different portions all together and observing differences only occasionally and inconsistently, is to produce a levelling effect in the presentation of the material. His interpretations have considerable potential to confuse the unwary reader and in places seem to have confused the author as well (see §4 below).
Muraoka’s approach to these three issues has far-reaching consequences for the reliability of his findings in the Syntax. Extensive sampling reveals that it often results in unsatisfactory treatment of specific features. The case studies that follow illustrate the sorts of problems that frequently recur. It is important to emphasize that similar difficulties turn up throughout the work.21
3 Case Study 1: The Optative Mood
Consider Muraoka’s treatment of the optative mood. The well-known decline of the optative towards eventual desystematization, though not so swift as is sometimes suggested,22 was already underway in the early Koine period and the Septuagint provides important evidence for the nature of the process.23 According to Nigel Turner’s 1963 analysis, on which Muraoka seems to rely heavily, there are 539 optative forms in the corpus. Turner identifies seven functions of the mood in Septuagint books.24 These are the volitive (termed desiderative by Muraoka) and potential functions in independent clauses and in subordinate clauses the conditional, oblique (namely indirect-question and indirect-statement uses), comparative,25 and final functions. His figures are:
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
Muraoka is interested in the distribution of the optative within the Septuagint and at first sight does appear to be getting to grips here with the issue of relationships between different parts of the corpus. He states that it is “only sparingly used” in Septuagint books “the literary ambition of whose translators is modest,” but that its “overall frequency … can hardly be said to be negligible.” This, he writes, is true not only of books with a generally recognized Atticistic tendency. He asserts that “use and retention of the optative is thus one of the important indices of stylistic and literary level of various [Septuagint] books.”27
The claim is supported by statistics. There are 99 instances in Psalms and 123 in Job. In Ruth, “a short novelette,” there are 14, while in Lamentations, “only slightly longer … and … a piece of poetry” there are only 2. For Muraoka this “suggests that whether to use or not to use the optative and how often is to a significant degree a feature of the author’s or translator’s idiolect.” He also stresses the significance of genre, observing that nearly 40 percent of all Septuagint optatives are found in Psalms and Job.28
This may seem like an effective analysis, but something crucial is missing. There is absolutely no investigation of possible impacts from methods of translation on the use of the optative, or indeed of the influence of the linguistic contexts found in the source text on distribution of the form in the translation. Turner cannot be blamed for overlooking these issues, especially the first. When he wrote the Helsinki School’s studies in translation technique had yet to start appearing and his focus was in any case different; he was interested mainly in the language of the New Testament. Muraoka has no such excuse. He has also ignored the findings of the present writer’s 2001 study of the optative (cited above), which demonstrates the significance of translation methods for the use of this mood in at least the Greek Pentateuch.
If Muraoka had taken translation technical issues into account, his interpretation would surely have been modified. As we saw above, 434 of Turner’s 475 optatives in independent clauses are volitives. That amounts to 91 percent of them, and 80 percent of all optatives in the corpus. This is an unusually high frequency for Greek of the period. As it happens, the volitive is the one function of the optative that in translated books has clear motivation from the source text.
That motivation involves both broad contextual factors (as already noted by Conybeare and Stock)29 and specific Hebrew forms, most typically the volitive jussive, but also certain particles. For instance, in Pentateuchal books I find 65 volitive optatives out of a total of 80 occurrences of the mood (Turner’s figure is 70) and 43 (or 66 percent) of those 65 volitives are matched by Hebrew volitive jussives or related forms. If we add another 15 examples matching particles, mostly
It seems probable this link between frequency of volitive optatives and the source text—an example of what John Lee suggests terming stylistic interference31—could be extended from the Pentateuch to embrace translation Greek in general.32 To test the idea I have now examined those 14 examples of the optative in Ruth that Muraoka notes. It turns out that they are all volitive in function. And at least 13 of them match volitive jussives or related forms in the Hebrew. In addition, note that there are no examples in Ruth 3, a chapter where motivating features are lacking in the Hebrew text. The fact that the pattern in this book aligns with that observed in the Pentateuch does not necessarily mean it will everywhere else as well. What it does tell us is that wherever we find optatives in translated books of the Septuagint, we would improve our understanding by testing the relationship of these forms to the source text.
Muraoka is right to stress genre as a factor motivating use of the optative. Style, literary level, and idiolect preference are surely also influences on choice of optatives in the corpus. But he has completely overlooked another highly significant aspect of the usage. This is a case where the lack of any translation technical analysis (or even reference to work of this sort done already) seriously weakens his treatment.
There are other points to make regarding the distribution of the optative in the Septuagint. These concern its appearance in non-translated books. Muraoka states “[t]hat the final optative is confined to 4M[accabees] is significant.”33 This is certainly true, given that that late book is not only an original Greek composition, but also manifests Atticistic tendencies. Here, however, he could have gone somewhat further. As I have remarked elsewhere, 4 Maccabees occupies less than 28 of the 2,125 pages of Rahlfs’s edition of the Septuagint, but it contains 15 of the 41 potential optatives, half the conditional instances, 2 of the 7 oblique instances, and the 9 secure instances of the final optative.34 Several of the remaining examples of subordinate-clause functions are to be found in other original Greek compositions. So it is important to be aware that these functions are mostly very rare, if found at all, in translated books.
Meanwhile, Muraoka observes in a footnote an argument of the present writer regarding the comparative optative (for instance
Muraoka’s handling of the optative, then, provides clear examples of the weaknesses I outlined in §2 above. The main problem is the compromising lack of translation-technical analysis (which has previously been shown to be significant for an understanding of this mood). But we also find the other issues: problematic reliance on outdated scholarship, limited control in some respects over the nature of the Greek in question, and unsatisfactory analysis of internal differences within the corpus. The result is a treatment that fails to capture adequately the functional dynamics of this important Septuagint feature.
4 Case Study 2: A Phantom Syntagm
It is also worth testing the way Muraoka’s approach can impact on analysis of rare phenomena. I examine here his treatment of the isolated periphrasis
In GELS Muraoka asserts that this syntagm, meaning “be obliged to do” something, is attested for the first time in Greek in the Septuagint.39 So it ought to engage our attention at once. He gives there three examples, from 2 Kingdoms, 2 Chronicles, and Tobit (the instance of particular interest here). A fourth, from Numbers, is added in the extended treatment in the Syntax.40 In all of these
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
Additional examples are introduced in the Syntax where
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
The impersonal use deserves close analysis, and Muraoka’s interpretations are questionable.41 For the present purpose, however, apart from observing in passing that any idea of obligation in these cases arises from contextual effects, I will set this group aside and focus on the four instances of the personal construction among which he places the Tob 5:15 (GI) example.
Do these really form a unity? Even in purely formal terms they manifest three different types. In Num 8:11 the infinitive is introduced by
Let us examine each passage in the light of these other factors to see how they can inform our understanding. I begin with the example from Numbers, which was probably translated in the third century BCE:43
Num. 8:11
καὶ ἔσονται ὥστε ἐργάζεσθαι τὰ ἔργα κυρίου
והיו לעבד את עבדת יהוה
This is a typically mimetic rendering, which neatly matches the components of the underlying Hebrew to the extent possible (Greek has no feature corresponding naturally to the Hebrew direct-object marker
Muraoka translates “they shall be doing the works of the Lord” and states that “
The remaining examples all belong to books of the Septuagint which are later productions of uncertain date. Those from 2 Kingdoms and 2 Chronicles should be considered together:
2 Kgdms 10:11 (matching MT 2 Sam 10:11)
καὶ εἶπεν Ἐὰν κραταιωθῇ Συρία ὑπὲρ ἐµέ ,καὶ ἔσεσθέ µοι εἰς σωτηρίαν ,καὶ ἐὰν υἱοὶ Αµµων κραταιωθῶσιν ὑπὲρ σέ ,καὶ ἐσόµεθα τοῦ σῶσαί σε·
ויאמר אם תחזק ארם ממני והיתה לי לישועה ואם בני עמון יחזקו ממך והלכתי להושיע לך
2 Chr 30:17
καὶ οἱ Λευῖται ἦσαν τοῦ θύειν τὸ φασεκ παντὶ τῷ µὴ δυναµένῳ ἁγνισθῆναι τῷ κυρίῳ
והלוים על שחיטת הפסחים לכל לא טהור להקדיש ליהוה
2 Kingdoms 10:11 belongs to a portion of the books of Kingdoms (1 Kgdms plus 2 Kgdms 1:1–11:1) that is “probably one of the most literal of the Septuagint.”47 The Greek that results is often syntactically artificial. In the passage in question the translator replicates a stylistic pattern in the source text. But, as perhaps always, there is an element of independence too. This explains the choice of the
LSJ glosses this instance of
The 2 Chronicles translation is also highly literal.50 In 2 Chr 30:17, however, the expression
It has challenged its interpreters. LSJ glosses
The articular infinitives in both these instances are in themselves natural Greek. They are deployed as loose expressions of purpose, essentially the “exegetical adjunct” use described by Geoffrey Horrocks.53 Their awkwardness within their respective contexts reflects the translation techniques employed. These constructions do not require the special category Muraoka has proposed.
That leaves the Tobit example. This book presents us with fascinatingly complex textual problems. In Greek it is preserved in three versions, a short one known as GI, in which the example in question appears, a long one known as GII, and a third known as GIII (not preserved in a form distinct from GI in our manuscripts for the verse relevant here).54 The relationship of these versions to one another, to Aramaic and Hebrew fragments discovered at Qumran, and to the original Greek translation of the book is a matter of ongoing discussion, but it can now be stated with confidence that “Tobit was certainly translated into Greek from a Semitic language, and from a text which was close to the version attested in [the Qumran fragments].”55 The evidence of the different Greek versions ought always to be compared when investigating linguistic issues.
There is no surviving Hebrew or Aramaic version of Tob 5:15. The GI version reads:
The GII version’s corresponding passage as preserved in Sinaiticus has the present tense form
In MS 319 (1021 CE), which also has the GII version for this part of the book, we find, however, the periphrastic future
The periphrasis
All the core evidence for the syntagm expressing the sense “be obliged to do” something has now disappeared. Muraoka has constructed a completely artificial syntactic category from a group of disparate examples. It was only possible to do so by failing to address the fundamental issues outlined in §2 above. Once all relevant data are taken into consideration the category crumbles.
5 Future Directions
Work on the language of the LXX is by no means complete … It is a field that is still immature, despite all that has been done. The LXX falls in a period when our knowledge of the Greek language keeps growing, both because it is being better studied and because new data are being found. Much of the present approach to the language of the LXX goes back to foundations laid in another era. A full appreciation of the language of the LXX translators is still to be achieved.62
The problems I have sought to demonstrate in Muraoka’s treatments of the optative mood and the (in my view imaginary) “be obliged to do” syntagm are not isolated cases in the Syntax. There is a high likelihood that a very large proportion of its analyses will eventually prove to be either seriously weakened or totally compromised by the method that informs them. The book cannot be relied upon as a reference work and its findings have considerable potential to mislead scholars inexperienced in navigating the special challenges of Septuagint syntax. They will all need to be tested through careful analysis along the lines demonstrated above.
In his 2004 review of the second instalment of Muraoka’s Lexicon John Lee observed that “the possibility exists of making a good lexicon better.”63 That comment does not apply so comfortably mutatis mutandis to the Syntax. In this case my impression is that the whole work will in due course need to be rewritten. That does not mean it is a useless book. We would be much poorer without it since, however problematically, Muraoka has blazed a trail. My hope is that his extraordinary efforts will act as a powerful stimulus for further research. Be that as it may, how might we work towards the “full appreciation” of Septuagint language which the epigraph to this section reminds us we currently lack? The following remarks offer some suggestions towards facilitating this process.
What we should not yet be trying to do is to write another syntax of the entire Septuagint. That is a goal for the future. If done properly, such a work would obviously be very valuable. As Muraoka asserts, “Few would seriously doubt the need and importance of a comprehensive study of Septuagint Greek.”64 But we do not know enough yet to produce it.65 What could successfully be attempted at this point is a grammar of the Greek Pentateuch, the portion of the corpus we know most about and on which most work to date has been done. This is especially true now that John Lee’s magisterial Greek of the Pentateuch has appeared. Notice too that I use the term grammar, not syntax. It is important to acknowledge that Thackeray’s admirable 1909 book is inevitably out of date. It is not only an effective treatment of syntax that would be required in this sort of undertaking. Such a work ought to have as its goal the elucidation of the complex nature of the material, dealing both with its natural Greek characteristics and those that arise from methods of translation and genuine Hebrew interference.
Besides working towards a grammar of the Pentateuch, we need to keep building our knowledge of the other parts of this heterogeneous corpus. We need to move beyond what is currently a largely theoretical understanding and gain a more secure grasp of details. Most would probably accept that the Pentateuch has a definite influence on later translation practice. But what can we identify precisely about the nature and extent of that influence? New technologies (especially searchable databases) provide excellent opportunities for advances here, for instance in refining our appreciation of the use of distinctively Pentateuchal vocabulary in later books. And there is a likelihood that we can build a much better understanding of ways in which those other books might have influenced one another as well.
Admittedly, working out as clearly as possible the relationships between the various books will be no easy task. We have few firm dates to use as internal anchor points in assessing the material. The documentary evidence from the second and first centuries BCE is less extensive and less easy to use than that from the third century. We still lack full agreement on how many books are original Greek compositions.66 And the hypothesis of a “Septuagint style” imitating translation language in original Greek compositions needs thorough testing. Even when these sorts of problems and inhibiting factors have been sorted out as effectively as possible, there will probably always remain more than one plausible way to relate individual books and larger portions to one another. On many issues we will have to be willing to settle for balancing probabilities.
Nevertheless, there is plenty of room for optimism. Such scholars as James Aitken, for instance in his work on the unpromising Ecclesiastes,67 have indicated the rewards that can be gained from close analysis of the material. We now have GELS, NETS, various handbooks,68 and for most books a critical edition to support our efforts. And despite the issues I have raised, we have the rich data contained in Muraoka’s Syntax as well.
The most feasible way forward is to continue developing the burgeoning linguistic and stylistic focus on all the different parts of the corpus, extending our increasing level of control over the Greek Pentateuch to other books. For most of them fresh work needs to be underpinned by systematic translation-technical analysis. And there needs to be full engagement with contemporary Greek, especially the vital evidence of the documents.
Such efforts can be encouraged and supported by revisiting Thackeray’s famous and highly influential classification of books “from the point of view of style.”69 This impressionistic grouping of the material reflects Thackeray’s fine feeling for Septuagint language and accordingly has an enduring value. It is still normally used as the basis for assessments. But as others have observed the classification is unsystematic and muddles categories.70 Careful assessment of individual books has already shown that some fit their assigned places rather awkwardly.
For instance, Ruth is placed by Thackeray in a class of “literal or unintelligent” translations,71 presumably because it is a kaige book. Close reading indicates, however, that it is neither literal nor unintelligent. For a single example consider the pronominal use of the article in the expression
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
Citation: Journal for the Study of Judaism 2023; 10.1163/15700631-bja10056
This pronominal use in combination with
Even one feature like this in a supposedly low-level work ought to alert us to the need for reassessment of Thackeray’s categories.75 Lee has suggested a framework for replacing them in his 2014 study of the language of Isaiah. He proposes “three main categories or criteria of classification,” namely translation method, style, and level of Greek.76 This is a model worth testing and potentially refining. Lee’s own work on Isaiah demonstrates its value.
It is only when we have achieved a better understanding of the linguistic and stylistic character of the various components of the Septuagint that we will be able to contemplate writing an effective description of the language of the entire corpus, including its syntax. Investigating all books systematically according to Lee’s criteria ought to lead to genuine and exciting progress. When we have made that progress we will be in a position to replace both Thackeray’s Grammar and Muraoka’s Syntax.
Bibliography
Aitken, J.K. “Rhetoric and Poetry in Greek Ecclesiastes.” Bulletin of the International Organization for Septuagint and Cognate Studies 38 (2005), 55–77.
Aitken, J.K., ed. T&T Clark Companion to the Septuagint (London: Bloomsbury, 2015).
Barr, J. The Typology of Literalism in Ancient Biblical Translations (Göttingen: Vandenhoeck & Ruprecht, 1979).
Conybeare, F.C., and St.G. Stock. Selections from the Septuagint (Boston: Ginn and Company, 1905). Expanded edition published as Grammar of Septuagint Greek with Selected Readings, Vocabularies, and Updated Indexes (Peabody: Hendrickson, 1995).
Corley, J. “Judith.” In The T&T Clark Companion to the Septuagint, ed. J.K. Aitken (London: Bloomsbury, 2015), 222–236.
Dhont, M. Style and Context of Old Greek Job (Leiden: Brill, 2018).
Evans, T.V. “The Comparative Optative: A Homeric Reminiscence in the Greek Pentateuch?” Vetus Testamentum 49 (1999), 487–504.
Evans, T.V. Verbal Syntax in the Greek Pentateuch: Natural Greek Usage and Hebrew Interference (Oxford: Oxford University Press, 2001).
Evans, T.V. “Approaches to the Language of the Septuagint.” Journal of Jewish Studies (2005), 25–33.
Evans, T.V. “Periphrastic Tense Forms in the Greek Tobit.” In Interpreting Translation: Studies on the LXX and Ezekiel in Honour of Johan Lust, ed. F. García Martínez and M. Vervenne (Leuven: Peeters, 2005), 109–119.
Evans, T.V. “The Optative at PMichZen 36.3.” Zeitschrift für Papyrologie und Epigraphik 179 (2011), 218–220.
Evans, T.V. “The Optative at PMichZen 36.3: Addendum.” Zeitschrift für Papyrologie und Epigraphik 180 (2012), 278.
Evans, T.V. “Numbers.” In The T&T Clark Companion to the Septuagint, ed. J.K. Aitken (London: Bloomsbury, 2015), 58–67.
Evans, T.V. “Verbs of Sexual Intercourse in the Greek Pentateuch: A Lexical Analysis.” Journal of Hellenic Studies (forthcoming 2023).
Gheorghita, R. Review of T. Muraoka, A Syntax of Septuagint Greek. Journal of the Evangelical Theological Society 61 (2018), 638–641.
Good, R. “1–2 Chronicles (Paraleipomena).” In The T&T Clark Companion to the Septuagint, ed. J.K. Aitken (London: Bloomsbury, 2015), 167–177.
Haddad, C.J. A Linguistic Analysis of Roman Republican Official Letters in Greek, Dissertation (Department of Ancient History, Macquarie University, 2018).
Helbing, R. Grammatik der LXX: Laut- und Wortlehre (Göttingen: Vandenhoeck & Ruprecht, 1907).
Hiebert, R.J.V. “4 Maccabees.” In The T&T Clark Companion to the Septuagint, ed. J.K. Aitken (London: Bloomsbury, 2015), 306–319.
Horrocks, G.C. Greek: A History of the Language and Its Speakers, 2nd ed. (Chichester: Wiley-Blackwell, 2010).
Hugo, P. “1–2 Kingdoms (1–2 Samuel).” In The T&T Clark Companion to the Septuagint, ed. J.K. Aitken (London: Bloomsbury, 2015), 127–146.
Jones, S.A. Syntax in the Septuagint: With Special Reference to Relative Clauses in Greek Numbers, Dissertation (Faculty of Graduate Studies, Trinity Western University, 2015).
Joosten, J. “Elaborate Similes: Hebrew and Greek. A Study in Septuagint Translation Technique.” Biblica 77 (1996), 227–236.
Joosten, J. Review of T. Muraoka, A Syntax of Septuagint Greek. Biblica 98 (2017), 631–633.
Kreuzer, S., et al. Handbuch zur Septuaginta, 5 vols. (Gütersloh: Gütersloher Verlagshaus, 2016–).
Lee, J.A.L. Review of T. Muraoka, A Greek-English Lexicon of the Septuagint: Chiefly of the Pentateuch and the Twelve Prophets. Bulletin of the International Organization for Septuagint and Cognate Studies 37 (2004), 127–139.
Lee, J.A.L. Review of T. Muraoka, A Greek-English Lexicon of the Septuagint. Bulletin of the International Organization for Septuagint and Cognate Studies 43 (2010), 115–125.
Lee, J.A.L. “The Literary Greek of Septuagint Isaiah.” Semitica et Classica 7 (2014), 135–146.
Lee, J.A.L. The Greek of the Pentateuch: Grinfield Lectures on the Septuagint 2011–2012 (Oxford: Oxford University Press, 2018).
Lee, J.A.L. “Back to the Question of Greek Idiom.” In The Legacy of Soisalon-Soininen: Towards a Syntax of Septuagint Greek, ed. T. Kauhanen and H. Vanonen (Göttingen: Vandenhoeck & Ruprecht, 2020), 13–25.
Ljungvik, H. Beiträge zur Syntax der spätgriechischen Volkssprache (Uppsala: Almqvist & Wiksell; Leipzig: Harrassowitz, 1932).
Mayser, E. Grammatik der griechischen Papyri aus der Ptolemäerzeit, 2 pts. in 6 vols. (Berlin: De Gruyter, 1906–1938; 2nd ed. of vol. 1 by H. Schmoll, Berlin: De Gruyter, 1970).
Muraoka, T. “How to Analyse and Translate the Idiomatic Phrase מִי יִתֵּן.” Bulletin of the International Organization for Septuagint and Cognate Studies 33 (2000), 47–52.
Muraoka, T. A Greek-English Lexicon of the Septuagint (Leuven: Peeters, 2009).
Muraoka, T. A Syntax of Septuagint Greek (Leuven: Peeters, 2016).
Piwowar, A. Review of T. Muraoka, A Syntax of Septuagint Greek. The Biblical Annals 7 (2017), 291–295.
Salvesen, A.G., and T.M. Law, eds. Oxford Handbook of the Septuagint (Oxford: Oxford University Press, 2021).
Silva, M. Review of T. Muraoka, A Syntax of Septuagint Greek. Bulletin of the International Organization for Septuagint and Cognate Studies 51 (2018), 216–224.
Soisalon-Soininen, I. Die Infinitive in der Septuaginta (Helsinki: Suomalainen Tiedeakatemia, 1965).
Sollamo, R. “Some ‘Improper’ Prepositions, such as ἐνώπιον, ἐναντίον, ἔναντι, etc., in the Septuagint and Early Koine Greek.” Vetus Testamentum 25 (1975), 773–782.
Stuckenbruck, L., and S. Weeks. “Tobit.” In The T&T Clark Companion to the Septuagint, ed. J.K. Aitken (London: Bloomsbury, 2015), 237–260.
Thackeray, H.St.J. Grammar of the Old Testament in Greek, vol. 1 Introduction, Orthography, and Accidence (Cambridge: Cambridge University Press, 1909).
Turner, N. Syntax, vol. 3 of J.H. Moulton, A Grammar of New Testament Greek (Edinburgh: T&T Clark, 1963).
Usener, K. “Zur Sprache der Septuaginta.” In Septuaginta Deutsch: Erläuterungen und Kommentare zum griechischen Alten Testament, ed. M. Karrer and W. Kraus, 2 vols. (Stuttgart: Deutsche Bibelgesellschaft, 2011), 40–52.
Thackeray, Grammar, 1–2.
Muraoka, Syntax, xxxv: “It is an immense pleasure to be able to present herewith what H. St. J. Thackeray proposed, but God did not dispose.”
Muraoka, Syntax, xli.
Versions of this article were presented at the Koine Greek Colloquium at Macquarie University (19 May 2017) and at the Septuagint within the History of Greek Seminar at the University of Cambridge (20 April 2018). It is a pleasure to thank members of the audiences at these events for valuable responses.
On the intention of both Thackeray and Helbing to produce a syntax volume cf. Evans, Verbal Syntax, 5 and n. 15.
Conybeare and Stock, Selections from Septuagint, 25–97.
This is not meant to suggest the Greek-English Lexicon of the Septuagint is free of faults; its errors are numerous and in need of correction (for some examples see Evans, “Verbs of Sexual Intercourse”). It represents, however, a major advance on all other lexica of the Septuagint produced to date; for detailed evaluations see Lee, Review 2004, 127–39; Lee, Review 2010, 115–25.
Muraoka turned 78 around the time the Syntax appeared and states that he was 74 years old when he began “in earnest” (see again the quotation in my introduction above).
For laudatory early reviews see Piwowar, Review 2017; and Gheorghita, Review 2018; more balanced are Joosten, Review 2017; and Silva, Review 2018.
See Muraoka, Syntax, 355–56, 577–84, 695–97 (note that the category direct speech does have its own section, §79 at pp. 690–94). The relevant discussions contain various oddities and confusions; e.g., at p. 355 and n. 5 Num. 9:4
For a brilliant recent example of this approach see Lee, Greek of Pentateuch.
Muraoka, Syntax, xl. This assertion is influenced by his “reader-centred” approach to interpretation of Septuagint language, deployed in both GELS and the Syntax. It seeks to capture the way an ancient reader without knowledge of Hebrew or Aramaic would have understood a Septuagint construction simply as a piece of Greek. Muraoka opposes it to what he terms the “translator-centred” approach, where the original translator’s understanding of the source text is a guiding factor (GELS, viii–ix; Muraoka, Syntax, xl–xli; on the two approaches applied to lexicographic practice, as manifested in GELS and NETS, see the critique at Lee, Review 2010, 117–24), but asserts that “the two approaches … do not have to be mutually contradictory” (Syntax, xl; cf. Lee, Review 2010, 121). A fundamental problem for Muraoka’s approach is the inherent vagueness of his notion of the “reader.” Are we to imagine that all ancient readers (or listeners for that matter) would have interpreted the material in the same way? See also Lee, Review 2010, 121: “Muraoka faces the prospect of not one but multiple later readers spread over centuries, and uncertainty about how any of them would have understood the LXX.”.
My impression is that although Muraoka defends his approach, he implicitly acknowledges its inadequacy—note the words “I am conscious of some limitations” and the invitation to others to “revise or supplement what is presented here” in the quotation in my introduction above.
Muraoka, Syntax, xxxvii.
Mayser’s Grammatik remains a valuable resource, but is an increasingly blunt research tool in terms of both data and interpretation and has to be used with care. For a specific case in point see Evans, “Optative at PMichZen 36.3,” 218–19; Evans, “Optative at PMichZen 36.3: Addendum.”
Muraoka, Syntax, 250–54, esp. 252; on Muraoka’s treatment of verbal aspect see also Silva, Review 2018, 221–23.
Only ten items are listed in the Index of Passages (Muraoka, Syntax, 889) under the heading “Epigraphical data” (where, since all but one are drawn from papyri, “epigraphical” must mean “documentary”). We are, however, directed in a footnote to Mayser’s Grammatik for additional references “very often” made (on Mayser see above n. 15). Peculiar habits of citation obscure some others. For instance, a classic example of topicalization in the well-known Zenon papyrus PMichZen 29, transcribed by Muraoka as
Muraoka, Syntax, xliii.
Muraoka, xliii.
Muraoka, xliii.
Failure to address one or more of the three fundamental issues raised here is pervasive and affects the reliability of the majority of entries. For the reader’s convenience I list here illustrative examples cited in this article additional to the following case studies: verbal aspect, the genitive absolute construction (and one might extend this to the functions of the participle in general), the pronominal use of the article. Other obvious examples are the treatments of anacolouthon and of renderings of the Hebrew infinitive absolute. A full listing would be very long. I invite researchers to apply the points raised in the section above to any entry in the Syntax. The experiment is likely to produce revealing outcomes.
Evans, Verbal Syntax, 178; on the persistence of the optative see also Haddad, Linguistic Analysis, 312–16, demonstrating the health of this mood in official Roman documents of the second and first centuries BCE (Haddad finds 14 certain examples in his corpus, only 4 of them in formulaic constructions (and so arguably fossils); they include 10 potential, 2 conditional, and 2 final optatives).
Evans, Verbal Syntax, 175–97; Evans, “Approaches,” esp. at 29–32; Lee, Greek of Pentateuch, 84–86, 159–64, solving the puzzle of Num 11:29
Turner, Syntax, 119.
On this function see Joosten, “Elaborate Similes”; Evans, “Comparative Optative”; Evans, Verbal Syntax, 190–97; Lee, Greek of Pentateuch, 84–85.
These figures cannot be entirely trusted and should be treated as rough approximates (see Evans, Verbal Syntax, 179, 268).
Muraoka, Syntax, 320.
Muraoka, 320–21.
Conybeare and Stock, Selections from Septuagint, 72; cf. Evans, Verbal Syntax, 179.
Evans, Verbal Syntax, 181, 183–84. My figures are based on the Göttingen edition (but could be slightly modified in light of Joosten, “Elaborate Similes,” 231–32, who argues persuasively for the identification of two more comparative optatives in Deut 22:26, where the Göttingen edition has subjunctives in accordance with the majority reading of the manuscripts).
Lee, “Back to Question,” 18.
Evans, Verbal Syntax, 260–61.
Muraoka, Syntax, 326.
Evans, Verbal Syntax, 179. Turner’s (Syntax, 120–31) figures for 4 Maccabees are 4 volitive and 15 potential optatives in independent clauses and 13 conditional, 2 oblique, and 9 final optatives in subordinate clauses, a total of 43. Turner’s total of 13 instances of the final optative in the LXX (as shown above) includes 4 variants from other books. Muraoka has probably chosen to ignore them for that reason. Note that Hiebert (“4 Maccabees,” 310) finds “some 46 [examples] in total” and calls the optative “an indication of the author’s syntactic sophistication.” (Hiebert is editor of the forthcoming Göttingen edition of 4 Maccabees.)
Evans, “Comparative Optative,” esp. at 497–503; Evans, Verbal Syntax, 190–97. For some responses (all implicitly or explicitly positive) see Usener, “Zur Sprache der Septuaginta,” 47; Jones, Syntax in Septuagint, 22 n. 51, 67–68; Lee, Greek of Pentateuch, 84–86.
Muraoka, Syntax, 325 n. 4.
Muraoka, Syntax, 349 n. 4.
Muraoka, 349.
GELS s.v.
Muraoka, Syntax, 349–50.
On the Isa 48:22 example, for instance, see Lee, Review 2010, 123.
Soisalon-Soininen, Infinitive in der Septuaginta might have supported such work.
Evans, “Numbers,” 59–60.
Cf. NETS (Flint): “and they shall be so as to perform the works of the Lord.”
Muraoka, Syntax, 349.
Muraoka, 340.
Hugo, “1–2 Kingdoms (1–2 Samuel),” 130.
LSJ s.v.
Muraoka, Syntax, 349, mistakenly citing the passage as 2K 10.11KG (where KG = kaige recension); at GELS s.v.
Cf. Good, “1–2 Chronicles (Paraleipomena),” 170–71: “The Greek Chronicles is a very literal translation…. The translation of Chronicles follows the Hebrew text closely with a word-for-word representation of the morphology and syntax of the Hebrew source text or Vorlage … Sometimes the kind of stereotyping translation employed produced structures that diverge from the norms of Greek syntax.”
LSJ s.v.
Muraoka, Syntax, 349–50.
Horrocks, Greek, 94: “the substantivized infinitive functioning as a gerund …, typically governed by a preposition to impart a determinate sense to the expression, but also used alone in the genitive to express purpose … This latter construction … was then employed simply as a ‘strengthened’ infinitive, used loosely as an exegetical adjunct or even as a complement after control-type verbs … the nominalized infinitive quickly became a stock feature of the Koine and, though based on a classical construction, soon acquired a frequency and range of usage that went well beyond the practice of classical prose writers.”
For detail on the nature of these different versions see Stuckenbruck and Weeks, “Tobit,” 237–39, 246–51 (on GII), 252–54 (on GI), 256–58 (on GIII).
Stuckenbruck and Weeks, “Tobit,” 239.
Evans, “Periphrastic Tense Forms,” 114; cf. NETS (Di Lella): “But say, what wages shall I give you—a drachma a day and what is needed for you as well as for my son?”
Muraoka, Syntax, 349.
Evans, “Periphrastic Tense Forms,” 114; cf. NETS (Di Lella): “I will give you a drachma a day as wages, and what is needed for you as well as for my son.”
Cf. Stuckenbruck and Weeks, “Tobit,” 248–49.
Thackeray, Grammar, 24.
Evans, “Periphrastic Tense Forms,” 112–17.
Lee, Greek of Pentateuch, 2.
Lee, Review 2004, 139.
Muraoka, Syntax, xxxvii.
For interesting soundings and analysis of prior scholarship see Jones, Syntax in Septuagint.
See, e.g., Corley, “Judith,” 227–29 for the ongoing debate on whether the book of Judith is a translation from a lost Hebrew source (this, for what it is worth, is the present writer’s view) or an original Greek composition.
Aitken, “Rhetoric and Poetry,” 55–77.
E.g., Aitken, T&T Clark Companion; Kreuzer et al., Handbuch zur Septuaginta; Salvesen and Law, Oxford Handbook.
Thackeray, Grammar, 12–16 (quotation at p. 12).
See, e.g., Sollamo, “Some ‘Improper’ Prepositions,” 775; Barr, Typology of Literalism, 283; Lee, “Literary Greek,” 136.
Thackeray, Grammar, 13.
Muraoka, Syntax, 3; Lee, Greek of Pentateuch, 34–36. It is instructive to compare the two analyses.
Lee, Greek of Pentateuch, 34–35; Mayser, Grammatik II.1, 57.
Lee, Greek of Pentateuch, 35. There are three examples in Ruth involving change of subject where the construction might have been used, but is not employed: Ruth 2:2
Muraoka (Syntax, 321 n. 2) observes the Ruth 3:14, 16 instances of
Lee, “Literary Greek,” 136.