2010.10.04 | H. A. G. Houghton. Augustine’s Text of John. Patristic Citations and Latin Gospel Manuscripts. Oxford: OUP, 2008. 424 pp. ISBN 978-0-19-954592-6. Hardback.

Reviewed by Dan Batovici, University of St Andrews.

This is one of the two reviews presented in the review-session dedicated to Hugh Houghton’s Augustine’s Text of John at the International Medieval Congress, Leeds, July 2010, (session 1630). The second review is available here, and the author’s response here.

Thanks to Oxford University Press for the review copies.

To start on a personal note, little before I came across Hugh Houghton’s book I was looking at Eriugena’s text of John, namely at the Johannine Greek readings Eriugena mentions in his commentary on the Fourth Gospel. Some of these readings, which are explicitly presented as coming from codicibus grecorum, simply didn’t survive in any of our manuscripts. Interestingly enough, the only manuscript we have today of this commentary is thought to contain, in a number of corrections and additions, Eriugena’s autograph. Since this means that the evidence the manuscript provides was not affected by textual transmission, it seemed a good starting point for assessing the relevance of these readings for the Greek text of John.

And then I found this book, which brought me to significantly reconsider my approach (and to postpone indefinitely the completing of this particular project). Moving away from myself for now, it should be said that the following are just thoughts provoked by the reading of this very stimulating book. As such, they are not meant in any way to diminish Hugh’s contribution; it is a very solid work, which will most likely prove significant in all related fields.

Indeed there are many good things to say about Hugh’s book; previous reviews, however, would have had that covered. To be specific from the outset, the present review will focus on the text-critical issues the volume raises. I would then only mention that, even for the hardcore Johannine textual criticism reader, this book offers a number of very interesting side-analyses such as the insight on the administration and use of biblical manuscripts in the churches of the early centuries (22-30). Particularly interesting is the presentation of Augustine’s ‘biblical exposition and citation technique’ presented as a background for assessing textual issues, which also offers an excellent insight on Augustine’s attitude towards the Bible, and, furthermore, on his method of biblical exegesis (44-56).

Turning now to the text-critical aspect of the book, of primary importance is authors’ understanding of the relevance of the various sorts of Johannine citations in Augustine’s works.

In order to assess the reliability of the diverse types of citations for the textual history of the Fourth Gospel in Latin, Hugh starts from presenting the evidence that suggests Augustine is using physical biblical codices in his sermons for quotations (32-6; 58). This kind of rather reliable use of the biblical codex is then opposed to quotations ‘likely to have been produced from memory’ (37): the former group contains passages which are the subject of Augustine’s exegesis, while the latter contains ‘illustrative quotations from other parts of the Bible’ (37).

To quote the author, ‘the initial citation of each verse of a lectionary passage seems to be the most likely to represent the text of Augustine’s manuscript, after which he reverted to memory’ (38). Hugh also notices that Augustine’s sermons that are known to have been dictated and not delivered publicly contain more biblical citations, and that this ‘may indicate that Augustine consulted volumes from his library during the process of composition’ (39). Furthermore, with another quote: ‘Even in sermons, with the manuscript in front of him, it is unlikely that he used it for anything other than the lectionary passage’ (67).

Moreover, shorter biblical references are deemed likely to be written from memory, while in quoting larger passages Augustine would have used a codex, ‘as demonstrated not just by the length and accuracy of the extracts, but also by the sequence in which they appear’ (39). In case of longer quotations sometimes the impression is given that the preacher is looking for the reference’s place in the gospel (58).

In a similar manner, quotations which are introduced by the title of their source (57) are also considered to be coming from a manuscript; yet at p. 68 is said that an introduced quotation points more to the fact that Augustine is invoking the biblical authority than to the use of an exemplar.

On pages 60-1 there is presented Augustine’s awareness of the importance of exact biblical quotation, with examples in which he quotes a whole verse ‘but with one crucial alteration, in order to emphasize the correct reading’; it is not clear whether this is really relevant for Augustine’s care for transmitting a correct text; one wonders whether this is more than a rhetorical ploy.

Hugh then defines the category of ‘mental text’, which is discussed by Cornelia. As far as text-critical considerations are concerned, it is remarkable that the author offers patterns to describe Augustine’s specific alterations of the biblical text (68), namely flattening (68-9) and conflation (70); these are accompanied, however, by alterations that are not easily to ascribe to Augustine and which may be taken as such from already altered testimonia (72). Other places allow for even bigger ambiguity: since Augustine’s readings are found in some biblical Latin manuscripts, an eventual case arguing that the variant are due to Augustine is dissolved (74-5).

Another category is further introduced, that of citations proper: to quote, ‘scriptural texts usually introduced by a quotation formula or pragmatically distinguished in some way,’ end quote. Within this category Hugh distinguishes – quote – ‘between primary citations, which are most likely to have been drawn from a codex, and secondary citations, which probably derive from memory’ (73-4), a distinction Hugh argued for in a different publication. The latter citations are said further on to ‘cannot be ascribed to biblical manuscripts with the same degree of confidence’ (78).

A trustworthy sub-set of primary citations is that formed by instances which Augustine explicitly presents as coming from a different biblical manuscript (79).  The other sub-set comprises primary citations: as Hugh will argue in case of Tractatus in Iohannem and De consensu euangelistarum, the initial citations and the ‘sequential treatment’ respectively are ‘most likely to have been drawn from manuscripts’ (93). These two form together the ‘most significant evidence, as there is a high probability that these transmit the text of manuscripts known to him’ (99).

Yet I am not sure at all why the author then says that ‘alterations of these observations by later editors is almost inconceivable’, with regard to the first subset (79). Nor why is it necessary, in the case of the second subset, that – quote again – ‘the form of primary citations indicates that most of the biblical references in Augustine’s writings have not undergone later revision or adaptation’ and that ‘we can be fairly confident that we have the original scriptural text’ (86).

The author mentions that ‘Augustine’s citations follow a general pattern of Old Latin readings in his early works, and Vulgate forms in those composed after 404’ (86), and that some variants that are not found anywhere (86). Since it is expected that a possible redactor to have corrected these to the text of Vulgate, if anything, these primary citations are considered to be unaltered (86).

While it is true that this is the natural expectation from a redactor, isn’t it still possible that the preserved text be nonetheless ‘corrected’ in its history of textual transmission, as any other text, and that the initial one had simply more Old Latin (keeping the pattern) and nowhere-else-to-be-found readings?

Turning now to secondary citations, they are shorter quotations and display ‘flattening’ alterations, seemingly drawn from memory (87). They also display the mentioned pattern – earlier works more Old Latin than the later ones who display a Vulgate text. Hugh notes, I quote: ‘However, there is a slight delay in the influence of the Vulgate on this material, which further suggests that Augustine was quoting from memory’ – end quote (87). Furthermore, ‘his mental text is primarily valuable for the rendering of key words and phrases’ (88).

I will turn now to Hugh’s presentation of Tractatus in Iohannem. This is included in the second part of the book, which analyses Augustine’s quotations of John in his different works. The introduction says, concerning the migration of Augustine’s biblical quotations towards the Vulgate, that ‘correspondences with Old Latin witnesses in earlier works give way to readings from Jerome’s revision to the results are immediate in citations made with reference to a codex, and more gradual in Augustine’s mental text’ (103; same 120). Yet couldn’t this simply describe the resistance to change of the citations that are commentary proper, as opposed to initial quotations (which would then just be more altered towards the text of Vulgate) as the NA 27 editors suggest?

Augustine’s commentary on John is assessed by the means of a threefold classification of references: the continuous text (which is the initial quotation of the commented text), sequential variants (which are within the commentary of the passage) and non-sequential citations (which are ‘illustrative material cited from memory, either out of the order of the Gospel or from elsewhere in the bible’ (108-9)).

Hence, the initial quotation is the text commented on, while the sequential variants are those repeated in the commentary and considered in the book reviewed here ‘a result of relying on memory’ (108).

Hugh mentions then that previous authors noted that the continuous text follows the text-type of Jerome’s revision, the Vulgate (109). However, he notices there are also many readings paralleled in surviving Old Latin manuscripts (111); these, accompanied by a number of readings which are did not survive at all elsewhere, are considered to ‘tell against suggestions that the biblical citations in this work have been changed by latter copyists to accord with the text known to them’. With the risk of repeating myself, is it really impossible the initial text written by Augustine from a codex just had slightly more Old Latin readings?

To contextualize Hugh’s following exposition, it is useful to recall that the larger bibliographical context for this book is that which reevaluates the importance of the evidence found in the texts of the Church Fathers for the text of the New Testament. It is in this particular domain the here reviewed book brings its contribution. The author thus mentions to the importance for New Testament textual criticism of Patristic evidence, both because such readings are datable and localizable and because some readings have not been preserved in surviving manuscripts (79; 91).

I have to say I missed the discussion of this bibliography, which is present in part at least in the References section, but is reflected rather scarce in the bulk of the volume. Within this bibliography, there is one quite opposing methodological stance with which Hugh does not interact in this volume, well not explicitly. The importance of Patristic evidence is an issue also addressed by the editing committee for the Greek text of the New Testament. And the editors of NA 27 have a quite different take on what kind of quotation bears more value fur such purposes. As it is described in the introduction ‘the distinction is drawn between the text commented on (txt), called the lemma, and the text of the Father ascertained from the commentary proper (com)’ (30*-1*) In distinguishing this, the preference is manifested for the latter, which would largely correspond to Hugh’s sequential variants: ‘the text ascertained from the commentary proper reflects the text of the Father more accurately,’ since ‘especially in later manuscripts the running biblical text has often been replaced by a different text type’ (n. 9, 31*). This is quite the opposite take on the subject: the initial citations are prone to alteration, while the other citations are less likely to be changed.

Now back to our book. Hugh does engages – indirectly – with this possible position on pages 113-4 on Tractatus in Iohannem where he acknowledge the possibility that sequential variants, by displaying more Old Latin readings, ‘at a first glance […] may appear to support’ the likelihood of a later revision for the initial quotations; then mentions that it has been suggested that – I quote – ‘the citations within the body of the commentary are more resistant to change than the lemma’ (113). Yet, given that Vulgate readings are ‘grammatically incorporated’ along with the Old Latin readings ‘would require that the hypothetical editor to have been zealously interventionist in one sentence while remarkably careless in the next’ (113).

He then argues that ‘this pattern, however, is better explained as Augustine lapsing into the Old Latin forms of his mental text after citing the initial passage from a codex’ (113).

Hugh proposes a model in which what he calls Augustine’s mental text (mainly based on the latter’s memory of Old Latin readings) forms a convenient category for assessing ambiguous citations. This is an ingenious model because it seems to explain in a coherent manner the differences between the initial quotation and the rest of citations. Also really interesting is the case he builds against the sequential variants being drawn from a codex, by offering a pattern of alteration specific to Augustine’s working method: flattening (68-9) and conflation (70).

The model is then applied to most Johannine variants which, in the works of Augustine, differ from the text of the Vulgate. Their analysis forms the impressive textual commentary comprised in the third part of the book, covering really half of the book, very insightful on Augustine’s treatment of the bible. As such, the proposed model is indeed seductive and a highly probable one. Yet I’m not sure exactly why an opposite model, similar to the one used by the committee of NA 27, would not be equally possible.

Rather I am intrigued by the insistence on Augustine’s memory; throughout the book this seems to act as an umbrella covering most ambiguities and inconsistencies of non-initial biblical citations. Is it impossible that Augustine, when citing in the commentary proper to use the same codex in a freer manner, which includes flattening and conflation? Do we really need employ to such extent a volatile category such as that of memory?

On page 114, the Vulgate form terrena is used, after the initial quotation on eight lines from 3 to 22; then, in line 23, there is the terrestria variant which is found in some Old Latin manuscripts. Is it really more plausible that – quote – ‘Augustine himself reverted to terrestria at the end of his exposition when citing from memory’? Do we need to imagine an abrupt switch in Augustine’s mind – the different reading is only one line away – for what could simply be a freer use of the text after repeating the same term eight times, perhaps, as Hugh rightly suggests ‘influenced by the similarity to caelestia later the verse’?

I am hence not convinced the initial citations were spared of alterations otherwise expected in manuscript transmission; the other model still seems possible, in which Augustine would have used manuscripts closer to Old Latin – reflected mainly in sequential quotations of the commentary proper, but less now in the initial quotations of the text commented on, where the process of copying would have altered them closer to Vulgate. Augustine’s gradual move over the years from Old Latin to Vulgate would still be possible; only the pace of this change would be then slower, as it would be more accurately reflected in sequential variants than in initial quotations. Altered as they may be, they might be construed – in this other scenario – to reflect not only memories of Old Latin text, but also the use of Old Latin manuscripts; in fact they are at times discussed in relation with manuscripts (that have survived or not) (161; 183).

All things considered, I want to thank Hugh Houghton for his work, which already is a positive influence my own work. I hope it will do that to others too. I am confident, however, that future studies either on the Latin text of John or on Augustine’s biblical text will find it rather hard to avoid it.

Dan Batovici
University of St Andrews
db47 [ at ] st-andrews.ac.uk