2011.01.04 | H. A. G. Houghton. Augustine’s Text of John. Patristic Citations and Latin Gospel Manuscripts. Oxford: OUP, 2008. (13.8×21.6), 424 p. ISBN 978-0-19-954592-6. Hardback.

This is Hugh Houghton’s response in the review-session dedicated to his Augustine’s Text of John, at the International Medieval Congress, Leeds, July 2010, (session 1630). The two reviews, signed by Cornelia Linde and Dan Batovici are available here and here.

First of all, I’d like to thank Cornelia and Dan for the time they’ve spent preparing for this session and for their very detailed and careful reviews: it has been a real pleasure to hear such constructive engagement with the text. I’d also like to pay tribute to their courage in inviting me to hear and respond to their critiques in person. As the book had its origins in a PhD, I’ve already had one viva on this material: it’s been much nicer this time to have had some idea in advance of what my examiners would be saying!

First of all, I’m glad that the structure of the book has been vindicated in the way Cornelia and Dan divided up their reviews. PhD theses can be notoriously unreadable, so when I came to rework it for publication I arranged the three sections on a decreasing scale of readership. The first part brings together general information and summarises the overall conclusions of the study. So if you’ve only got time to read 100 pages it should all be there. (By coincidence, this is also the amount one can access for free on Google Books, although that probably won’t help sales!) The second part, going through Augustine’s works and examining their biblical quotations, will, I hope, be of interest to a range of people working on exegesis or editing Augustine. Finally, the third – and least readable – part is the textual commentary for what Dan called the “hardcore Johannine textual criticism reader”.  We should be grateful that he hasn’t raised too many points from this section!

Time is far too limited to address all Cornelia and Dan’s questions, and it would be good to have time left at the end for discussion. So I’m going to try to respond to as many of Cornelia’s points as I can under the broad headings of “The Nature of the Old Latin Bible” and “The Authority of Scripture”, and then answer Dan’s observations about Augustine’s use of the Vulgate and the analysis of biblical commentaries.

The nature of the Old Latin Bible

Cornelia asked several times about the versions of the Bible Augustine used and which texts had official sanction. Part of this is to do with the nature of the earliest Latin translations: Augustine claims in De doctrina christiana that, as I translate, anyone “who believed himself to have a modicum of ability in both [Latin and Greek] … hazarded his own translation”. In reality, surviving Old Latin versions are a lot more consistent than is often claimed, and, what we seem to have are successive revisions of a limited number of early translations.

Yet the concept of versions is bound up in the wider issue of using early manuscripts for the text of the Bible. Every manuscript is different, and therefore every Bible is different. For example, a Christian community might have been using scriptural codices which did not include the story of the woman taken in adultery, or claimed that Barnabas, not Barabbas, was set free instead of Jesus, or even had that famous alternative reading in the Ten Commandments “Thou shalt commit adultery”. This would not have made any difference to the way the manuscript was used in the liturgy, or revered as sacred scripture. Clearly, more obvious copying errors would usually have been corrected pretty quickly, but plenty have been left to stand! Augustine’s comparison of manuscripts was a standard preliminary stage in working with a text [see page 18 on reading practices in antiquity], just like modern scholars should ascertain whether they are using a current edition of a work, or whether there is a corrected second edition or important subsequent material which needs to be taken into consideration.

In fact, I suspect that modern attitudes to “the Bible” even after the invention of printing are not so different to those in the time of Augustine. What do you mean when you refer to “the Bible”? Do you mean the King James Version, the Revised Standard Version, the New Revised Standard Version, the Good News Bible, the Contemporary English Version, the New Living Translation with the words of Jesus in red? The list is endless. Even if you take a printed Latin Vulgate or an edition of the Greek text as your archetypical “Bible”, you still have to specify which edition. None are entirely identical, and among modern translations – as with the Old Latin tradition – it can be very difficult to specify when a text is sufficiently different that it constitutes a new version rather than a variant form or light revision of an earlier version.

In the light of this, I would like to suggest that authority is co-located in an overall concept of Scripture and the individual exemplars which one uses for one’s text. When Augustine says “it is written in the Bible”, how would he prove it? In the same way as he challenges his opponents in the dialogues, by inviting them to bring him a manuscript with that reading [e.g. Contra Faustum (p.18-19)]. And if they could, he then goes on to outline his text-critical criteria for evaluating that witness: how old is the manuscript? Is it accurate elsewhere? How does it correspond to Greek manuscripts?

So I think that Cornelia’s question about “which was the standard version” can only be answered by the picture I’ve reconstructed from his works of Augustine with codex in hand. Augustine’s Bibles had the authority of Augustine the bishop, and if that was challenged, then it was reinforced by the authority of Augustine the scholar, the library he had built up at Hippo and his textual researches over his lifetime [cf. the retraction over Sirach 34:30 (p.20)]. Yes, Augustine does prefer the bibles of “more learned and responsible churches” [AU do 2.15.22 (p.7)] and apparently expresses a preference for North Italian versions [AU do 2.15.22 (p.7)], but each individual codex still had to be weighed on its own merits.

I hope this resolves the tension which Cornelia rightly notes between Augustine’s emphasis in some places on an exact text, and the authority he accords elsewhere to a much looser form of words which is so far from known versions that it’s hard to believe it ever occurred in a biblical codex. When a modern-day preacher claims that “The Bible tells us that ‘God so loved the world that he sent his Son to redeem the human race'” most of us wouldn’t bat an eyelid. Only the hardcore textual critics would object that those words are a heavily ‘flattened’ form of John 3:16 which won’t be found in any printed Bible and bear only a limited relationship to the text of the Gospel in Greek.

Biblical commentaries

On that note, let’s turn to the issues Dan raises about my analysis of Augustine’s sermon-commentary on John. Does this really transmit the biblical text used by Augustine, or has it been altered? When I began my research, I had exactly the same presuppositions as Dan that later copyists or editors would have interfered in such a way as to make it impossible to retrieve Augustine’s text. It was only after thorough investigation of the material that I was led to make the claims that I do in the book, that the biblical text has been transmitted accurately.

First, shorthand dictation is an entirely mechanical activity: stenographers – particularly those trained for work in the law courts – do not leave gaps to be filled in later, even for biblical quotations, or rewrite using their own words. The only way to keep up with the speaker is to follow them word for word, including any deviations in the form of their biblical citations.

Second, the picture I paint of Augustine only using a biblical codex for his first citation, and thereafter repeating the text from memory, introducing minor variations, is not confined to this commentary: it’s also there in other sermons [cf. the example of gradual flattening in AU Jo 38 (pp.69-70), but also AU s 342 (p.133), 126 (p.125), 129 (p.127), 143 and 145 (p.129), (and AUs 122, p.124)]. We know from the Retractationes that Augustine never got round to revising his Sermons for publication: if he had, might he have made the biblical text more consistent?

The big question is whether Augustine himself used the Vulgate in his commentary, or whether it was introduced by a later editor. For this, we really need a better edition than the 1953 Corpus Christianorum text. During my research, I looked both at manuscripts produced in the time of Augustine, such as the St Petersburg Codex, and the earliest surviving witnesses to the Tractatus in Iohannem to see if the biblical text was indicated in any way which would make it easy to identify and replace. It seems that the consistent indication of citations and sources only really takes off in the scriptorium of Wearmouth-Jarrow in the eighth century. Before this, there are occasional diplai in the margin, marking the odd citation, but nothing consistent or comprehensive.

Besides, as I’ve taken pains to show, terms which are characteristic of the Vulgate – such as ostium rather than ianua in John 20 – are embedded into Augustine’s exposition, both in the Tractatus and On the Agreement of the Evangelists. It would be a peculiar editor who altered single words grammatically incorporated into Augustine’s text, yet failed to ensure that longer verbatim citations matched the Vulgate. Yes, it’s just about possible. But the picture of these commentaries gradually being brought into conformity with the Vulgate over a long period is unnecessarily complicated. We know from Letter 71 to Jerome, written in 403, that Augustine had a copy of the Vulgate Gospels, and had analysed them in some detail. That was around the time he wrote On the Agreement of the Evangelists, and three years before he began preaching his commentary on John in Hippo. Why should he not use for this the most recent revised version which he had so highly praised? The problem then is to explain the inconsistencies, and I think that the model of the preacher gradually reverting to the forms he knows from memory is both textually and psychologically plausible.

Does this explanation conflict with the editorial policy in the Nestle–Aland Greek New Testament, as Dan suggested? I don’t think so. I carefully tried to avoid using the word lemma in my discussion of Augustine’s commentary on John, because it doesn’t have lemmata in the sense this word is used for other commentaries [The only time I do use it, on page 113, it is repeating an observation of Bogaert.]. In Origen, and even the Latin Ambrosiaster, lengthy portions of the biblical text are copied out verbatim at the head of each part of commentary. These are physically distinct from the commentary, and as such are particularly susceptible to updating to bring them into conformity with whatever biblical text is in use at a later time or place. The result of this is that we have manuscripts of other commentaries where the biblical text in the lemma is not the same as the one in the discussion. In Augustine, however, the closest we have to a lemma is the title of each sermon, which does seem to have been taken from a different source (although, again, this is something which a new edition must clarify). Instead, as Dan explained, I have taken the continuous text out of the commentary, where Augustine quotes it one sentence or one phrase at a time. The initial citation is simply the first occurrence of the biblical text within Augustine’s exposition [See p.108: this occurs within the continuous text.]. As such, this corresponds to what the Nestle–Aland editors call the “commentary proper”, and the apparent conflict is, I hope resolved. (Incidentally, if you want my methodological survey of the text-critical use of patristic material, it is Chapter One of my PhD, which never made it into this book, although it may yet become a separate article.)

I’m very pleased that both Dan and Cornelia mentioned primary and secondary citations, since I think that this is a valuable distinction. In the article where I introduced these categories, I was slightly more nuanced than in this book: primary citations are those “which are most likely to have been drawn from a biblical manuscript” [p. 74] (for the reasons Dan explained), whereas for secondary citations we simply can’t tell, and must be more agnostic about their value in applying them to a biblical text [“Secondary citations, in contrast, may feature variants and abbreviation characteristic of citations from memory. They are normally shorter. While it is possible that they were made with reference to a codex, there is no explicit or implicit indication of this. The fact that the majority of secondary citations were probably quoted from memory, in accordance with ancient custom, does not mean that they are textually insignificant. Memory must be memory of something, even if the accuracy of someone’s recall may vary. Nonetheless, they do not demonstrate the same direct connection with biblical manuscript tradition that characterises primary citations. ” NTS 54 (2008) p. 451]. In the case of Augustine I do think that most of his secondary citations were made from memory, but I wouldn’t want to identify all secondary citations with Augustine’s mental text. The identification of primary citations, however, tells against Dan’s suggestion that the pace of change was slower than I present. If you are reading a text from a codex, the pace of change is immediate: you read exactly what’s in front of you. If you are quoting from memory, however, it takes longer for the new version to become the dominant form which you cite, and that’s what I think I have shown by dividing the material in this way.

Conclusion
I have not answered every question, but I must draw to a close. What the last two points have made clear, and what I have come to realise as I have been working on the edition of the Old Latin versions of John following the publication of this book, is that every author – every work, in fact – is different. A lemma in Ambrosiaster is not necessarily the same as a lemma in Augustine. The identifying characteristics of a primary citation in, say, Cyprian, may not be the same as a primary citation in Gregory the Great. The sort of analysis which I have done in this book on Augustine’s use of biblical codices, his methods of composition, the transmission of his works and so on really has to be done afresh for every author if their quotations are to be used as evidence for the text of the New Testament. We can’t just take what editors put in italics or inverted commas and include it uncritically in a biblical apparatus. We need to know something about its context and how it relates to the rest of the author’s references to that verse. In this age of hyperlinks and electronic corpora, it is becoming easier and easier to gain access to this information, but it still needs careful human analysis. I’m delighted that this book has already been a positive influence on Dan’s research, and I hope that it will inspire others to similar studies. My thanks again to Dan and Cornelia for teasing out some of the implications of my work, and I hope that others will extend it yet further. Thank you.

Panel Review of Augustine’s Text of John

Hugh Houghton

Leeds IMC Session 1630

H.A.G.Houghton@bham.ac.uk