Bryn Mawr Classical Review 2005.07.57
James D. Ernest, The Bible in Athanasius of Alexandria. The Bible in Ancient Christianity 2. Leiden: Brill, 2004. Pp. xiv, 482. ISBN 0-391-04176-2. €133.00.
Reviewed by Hugh Houghton, Institute for Textual Scholarship and Electronic Editing, University of Birmingham (H.A.G.Houghton@bham.ac.uk)
Word count: 1955 words
Following Charles Kannengiesser's weighty "Handbook of Patristic Exegesis" comes this second volume in the series The Bible in Ancient Christianity, a similarly substantial offering from James D. Ernest detailing the presence and function of Scripture in the writings of Athanasius. This is a revised version of Ernest's doctoral dissertation (under Pheme Perkins at Boston College) and, whilst retaining all the characteristic features of the genre (including a critical literature review as the opening chapter and over a hundred pages of appendices, bibliographies and indices), constitutes an impressive monograph on a subject which has previously only been treated in articles and isolated chapters. Ernest used the text of the TLG in combination with his own mark-up scheme to create an electronic "textbase", which he then classified and interpreted according to a range of criteria. The result is an extensive study which is particularly strong on the mechanics of Athanasius's use of the Bible and the distribution of citations in the selected writings and covers key topics in a thorough and detailed manner.
The writings in question consist of Athanasius's works which survive in Greek and are generally considered authentic, treated under four headings: Apologetic writings (Against the Pagans and On the Incarnation), Dogmatic-Polemical writings (Against the Arians 1-3), Historical-Polemical writings (Three Apologies and the History of the Arians) and Pastoral writings (Personal letters and the Life of Antony). These are considered in chapters two to five respectively. Although a chronological approach might have shed some light on the development of Athanasius's thought and exegesis, the dating of many of these writings is contested. Instead, the decision to treat the works by genre enables Ernest to highlight the different aspects of Athanasius's use of Scripture, notably the principal distinction between his unitary reading of the Bible, illustrated by recourse to selected key verses and theological "paradeigmata", and his use of scriptural types as moral "exempla", persuasive weapons in his rhetorical armoury .
Each genre is used to exemplify a different facet of biblical reference together with one or more wider issues: chapter two focuses on the formulae used to introduce citations (included in Table 2-5) alongside the rhetorical structure of the writings; chapter three concentrates on the polemical employment of "touchstone" texts (Tables 3-4 and 3-6), and discusses Athanasius's exegetical principles and vocabulary as well as the history of the Arian controversy; chapter four describes the use of imitative "exempla" (Tables 4-6 and 4-9) with historical parallels; chapter five considers more "exempla" with, for the first time, an account of the actual text of the biblical references (Tables 5-3, 5-8 and 5-9), to which is attached a discussion of Athanasius' canon of scripture. Despite its appealing variety, this structure is slightly disorientating until revealed in retrospect: until that point, the impression that like will be compared with like has been reinforced by the presence early on in each chapter of identical tables summarising the types of scriptural references and their canonical distribution for that particular genre. Indeed, the wealth of background information offered for these broader subjects and the occasional necessity of incorporating material from works outside the nominated genre for each chapter (such as the discussion of the Festal Letters in chapter three [134-5] or the animal exempla from Against the Arians in chapter four [196-8]) prompts the suggestion that these topics might have been better served by an individual treatment in separate chapters, free from generic constraints. Nonetheless, the book is commendably free from repetition and well-indexed.
A wealth of material is adduced in footnotes, especially in chapter three, where occurrences elsewhere of the technical terms under discussion are presented in full. In the discussion of Athanasius's exegetical principles, despite suggesting that the distinction "anthropinos" (or "sarkikos") and "theikos" may have originated with Athanasius , Ernest dismisses this technical vocabulary as "mere window-dressing" . Instead he affirms that Athanasius' exegesis is controlled by a metanarrative rather than a systematic method . While he makes a strong case for this, one wonders how far this surprisingly modern position has been influenced by Athanasius's own refusal to portray himself as a biblical scholar or consider alternative readings of scripture. Despite the irony that "the founder of the dogmatic exegesis of the great councils did not want to be seen as an exegete" , Ernest is still able to describe Athanasian characteristics in handling Scripture even if Athanasius was largely independent of the "traditional trappings of Alexandrian biblical interpretation" . The inclusion of biblical usage is a useful contribution to the debate concerning the authorship of Against the Arians 3 [143-9] and at least part of the Encyclical Letter of Alexander (Henos somatos) , as is the observation that Antony reflects on Scripture in a non-Athanasian way in the Life of Antony , although Ernest supports the ascription to Athanasius .
An interesting, and potentially transferable, feature of this study is Ernest's identification and categorisation of scriptural references, which generated his initial corpus of material. Text-critical studies of patristic citations, such as the SBL New Testament in the Greek Fathers series, normally have three categories of biblical references: citations, adaptations and allusions. These are defined according to their significance for textual reconstruction and as such have limited application outside the discipline. Furthermore, the scope and significance of each category can vary from Father to Father, and it is sometimes necessary to propose refinements. Seeking "to approximate the text critics' attention to detail"  (!), Ernest suggests five categories, based on an initial distinction between whether the biblical reference is explicitly indicated as such by the use of a citation formula or not. This information is not currently featured in the text-critical definitions and is a valuable and welcome inclusion, despite the pessimism of some textual scholars.1 There are two types of "marked" reference: Citations are defined by Ernest as scriptural text preceded by a marker of direct discourse, while Allusion covers explicit references to the Bible which are not accompanied by quoted material . "Unmarked" uses are defined according to their correspondence to the scriptural text: Quotations indicate exact or nearly-identical text, Reminiscences looser but identifiable correspondences and Locutions more general biblical language . So far, so good, although in practice, due to the pervasive influence of the Bible on Athanasius (and most Church Fathers) distinct Locutions are hard to isolate and rarely function as scriptural parallels because their reference is so vague: they are not included in the tables of citation distribution and are all but abandoned .
The method begins to look less convincing when, once the data has been entered and classified automatically on the basis of a sample (which is not further described), the references are reduced to numerical values. This is not so much of an issue for citations and quotations, even though their percentage correspondence with the biblical text (described as "quotation fidelity") must be taken with a pinch of salt given that we cannot know the exact reading of Athanasius' Bible.2 However, such quantitative descriptions are ill-suited to allusions and reminiscences, despite their identical treatment in the summary tables. The verbal correspondence of these categories with the Bible is often minimal (almost half of the reminiscences in Table 2-5 are three words or fewer) although their extent may be far greater. For instance ἐπειδὴ δὲ ποικίλα καὶ πολύτροπα τὰ τοῦ διαβόλου βέλη is described in Table 5-3 as a one-word reminiscence of Ephesians 6:16, ἐν ὧι δυνήσεσθε πάντα τὰ βέλη τοῦ πονηροῦ τὰ πεπυρωμένα σβέσαι. Granted that this is the sole occurrence of the word βέλη in the New Testament, it is only the wider context which permits the identification of the reminiscence, covering, at the least, the four-word phrase including the two definite articles. When dealing with a handful of words, the distinctions between reminiscences, locutions and even quotations become hard to define. For example, although μονογενὴς θεός originally stems from John 1:18, it is debatable whether the phrase taken out of context constitutes a quotation . Such references should certainly be listed, but the reductive numerical value (which also forms the basis for the total words of Scripture in the work, the "fidelity quotient" and other calculations) may give a misleading impression. A more satisfactory treatment of biblical reminiscences is the extended illustration of the imitative relationship with Paul in the Defence before Constantius [224-233] and other occasions when Ernest considers Athanasius and the Bible side by side, since the impact of such correspondences reaches beyond the extent of direct verbal correlation.
It is a hard task to assemble an exhaustive collection of biblical references, especially when including looser correspondences whose identification is often more subjective: the temptation to see biblical references in every collocation must be resisted if the list is going to be meaningful. Ernest appears to have undertaken his collection from scratch, and offers in Appendix B a 39-page scriptural index to the writings covered in this study. As a measure of accuracy, I compared the scriptural references in Against the Pagans and On the Incarnation in Table 2-5 with Thomson's edition of both works. It is sometimes difficult to identify the suggested reminiscences, and alternatives could be proposed, but this is only to be expected. More concerning are a number of occasions when Thomson identifies scriptural parallels which are lacking from Ernest's list, e.g. 2 Cor 4:18 in CG8, Romans 1:25 in CG13, Matthew 13:13 in CG23 (a three word quotation), Matthew 21:33-41 in DI 13, 1 Cor 1:24 in DI 19, quotations of Hebrews 11:35 and Acts 2:31 in DI 21, Acts 2:24 in DI 27, Acts 2:37 in DI 51, Acts 19:19 in DI 53 and Romans 9:5 in DI 55. There are even two introductory formulas which Ernest misses: in DI 56, Matthew 24:42 is not a reminiscence but a citation, preceded by "the saying is salutary which prepares us for that day and says...", while the clear reference to "expelling demons by the prince of demons" in DI 48 (Mark 3:22 and parallels) is omitted altogether, despite the indication "just as the Jews said to him when wishing to abuse him". There are also inconsistencies between the summary figures: in Table 2-3, CG has 9 reminiscences and DI has 59 allusions, but in Table 2-6, CG goes up to 12 reminiscences and DI gains another allusion when Tables 2-7 and 2-8 are combined; again, although Tables 4-2 and 4-4 agree on figures of 182 citations, 69 quotations, 152 allusions and 119 reminiscences in the Historical-Polemical writings, the text on page 201 only has 180 citations, 69 quotations, 134 allusions and 116 reminiscences.3 Of course, such inaccuracies are only of minor importance in the overall assessment of the study, but they do present a salutary reminder of the danger of placing excessive reliance on numerical information and claims of exhaustiveness.
The final result is therefore what might be expected from Ernest's combination of technology and traditional scholarship. The tables, statistics and results of key-word searches present the broad outlines of a picture, for which the detail still has to be supplied by close attention to Athanasius' text and its context. The strength of the synthetic approach means that the biblical information can be brought to bear on wider issues, although it is slightly frustrating that, thanks to the varied focus of each chapter, direct comparison between citations of different genres is not possible because the same textual or introductory information is not supplied. Broader topics are handled thoroughly, although the material is sometimes dispersed and the treatment can become discursive. Chapter six provides a fluent and lucid summary of the occasionally dense argumentation of earlier chapters. The book is well-presented (as one would hope at the price)4 and deserves to feature prominently in the bibliography on Athanasius and patristic exegesis more generally.
1. The criterion of intention is inherent in Fee's definition of a citation (which Ernest refers to on page 29) and the use of a formula is the clearest indication of this. Although Ehrman in Didymus the Blind and the Text of the Gospels (1986:12) laments that a citation formula is no guarantee of the verbal accuracy of the following quotation, such information is bound to vary from author to author. Both Brooks and Mullen in their respective volumes on Gregory of Nyssa and Cyril of Jerusalem in the same series use the proportion of citations preceded by an introductory formula as a guide to intentionality.
2. Ernest summarises a recent dissertation by John Brogan, The Text of the Gospels in the Writings of Athanasius [26-27], although he does not appear to use its textual data: despite the closeness of Athanasius to the "secondary Alexandrian" Gospel witnesses, this does not permit the reconstruction of a single, invariant biblical text.
3. There are also inconsistencies in the total words of Scripture in Table 2-3 (1163 for CG, including 32 in locutions, and 1665 in DI including 16 in locutions) and Tables 2-6,7,8 (where the totals of unchanged and adapted words are 1075 for CG and 1551 in DI, not including locutions). Similarly Table 3-2 lists 11990 words of Scripture (including 108 in locutions), but the total of unchanged and adapted words in Table 3-3 comes to 11203 (not including locutions). However, these discrepancies are presumably to be explained by the "interpolated" and "substituted" words of Scripture which are included in the earlier tables, but not in the canonical breakdowns.
4. There are a few minor typographical errors, and there appears to be some text omitted between pages 435 and 436.