The problem with finishing is that you never really do finish. You produce your text, replete with footnotes — and you think it is done. You feel that you should receive advice from your peers and betters, and so you importune others to read it, some of who do, and you revise (probably not as much as you should) in light of their feedback and your own re-reading — and you think it is done. You submit it, you receive further comments, you have it accepted — and you think it is done. You receive queries from the copy-editor and you are grateful for being saved from several slips and refine it accordingly — and you think it is done. You see the proofs and realise that there is more to be corrected and you work by the midnight oil to improve it at that late stage — and you think it is done. Of course, it is not. It remains imperfect and provisional. Your last word is only part of the ongoing conversation.
I have very recent experience of this, with the forthcoming catalogue of the manuscripts of Christ Church, Oxford. This is the work mainly of Ralph Hanna, but I helped at a late stage, updating the descriptions and adding some more (of sixteenth-century manuscripts), as well as expanding the introduction. In that introduction, we survey what little is known of books of the previous institution, whose Norman buildings provide now the college chapel which doubles as Oxford’s cathedral. Until their dissolution in 1524 by Cardinal Wolsey, making way for his new foundation of Cardinal College, these were the buildings of the Augustinian priory of St Frideswide’s. As we say in the introduction, it was not known for being a place of learning, and only a few manuscripts are associated with it. We also say that ‘only a single literary manuscript has been identified as being owned by’ it, and technically that is true: the bible of English medieval institutional provenances, Neil Ker’s Medieval Libraries of Great Britain, now available on-line as MLGB3 (thanks to James Willoughby and Richard Sharpe), mentions only that codex as the one literary survival. I have now, however, convinced myself that another volume should really be added to that list and so should have appeared in our introduction.
The manuscript is hardly unknown: it sits in the Bodleian with the shelfmark MS. Digby 177. It is an obvious candidate for coming from the priory, as it provides a unique copy of a description of the miracles attributed to St Frideswide, said to have been compiled in the 1180s by Prior Philip of the Oxford house. In revising W. D. Macray’s nineteenth-century catalogue of the manuscripts given to the Bodleian in 1634 by Sir Kenelm Digby, Andrew Watson, working with the materials of the late Richard Hunt, addressed the issue of this manuscript’s provenance and expressed unresolved ambivalence: ‘it is possible that [it] comes from St Frideswide’s Priory, Oxford, but … it may be no more than a section with an Oxford interest which has been detached from a larger book with no Oxford connection’. It was, of course, Andrew Watson who provided the Supplement to Ker’s MLGB and he saw no reason there even to hazard the suggestion that it is expressed so tentatively in the revision of the Digby catalogue. What, then, persuades me that the issue should be reviewed?
First, against the suggestion that this manuscript was part of a larger book, Watson’s own comment can be quoted: ‘the last page looks as though it had been the final page of a unit on its own’. The last recto is, indeed, rubbed, and so is the first recto, suggesting that this fascicule travelled alone for some of its life. Morever, as Watson also notes, it reached Digby from the Oxford antiquary, Thomas Allen and it appears in his catalogue, listed alone as an item (‘fo. 7’), in contrast to the volumes entered immediately before and after it where multiple contents are listed. In other words, it is likely that Allen came by it in its present state, unencumbered with other material, and this may well have continued its prior existence, as a discrete codex.
The codicology of the manuscript is strongly suggestive of its Oxford provenance. The main part is written in an elegant bookhand on the cusp between so-called protogothic and a textura rotunda. The final columns (fol. 28vb– 30rb) are in a darker ink and by different hand, spikier and yet closer to being fully gothic. That addition provides the tale of an extra miracle which, it says, happened ‘in ciuitate oxoneforde eciam nostris temporibus’ — it appears, in other words, to be updating the collection with a recent occurrence. Even if the main text was not produced in Oxford, it would seem likely that this addition was made there.
In addition, the title added at top left of fol. 1 may be notable in its phrasing: ‘Incipit prologus domini philippi prioris de miraculis sancte fridwide’. That the author is known but it is felt unnecessary to state of where Philip was prior hints that this was written within the community. Moreover, there are signs of later use of the volume, not just notes in plummet the bottom margin of fol. 15v-16, showing that there was continuing interest in the text, but also at the top right of the final verso where an acrostic is added, in a thirteenth-century anglicana hand, on the name ‘Fridesuuida’. Wherever this was, there was a continuing devotion to a saint whose cult was localised to Oxford and centred on the priory named after her.
The clinching evidence would, of course, be an ex libris. It seems to me that there was once one, near the top left of the first folio, just right of the later shelfmark, ‘A 14’. I have tried checking it under UV but to little avail. Its secret remains, for the moment, just beyond our grasp, as frustrating as any branch of fruit with which Tantalus was tormented.
Even without that, though, I feel there is enough to merit at least proposing an association with St Frideswide’s as probable, though by no means certain. With, however, the proofs of the introduction of the Catalogue now back with the type-setter, it is too late to add a footnote, and so that work is out-of-date before its off the press. I have half a mind to beg them to stop and not complete the publication process: we all have a duty only to publish when we can place our hand on our heart and promise we believe a work is as polished as it could possibly be. As I have said before, if a work is half-decent, then that is not good enough. But assuming for a second that the publishers would even countenance a delay, it would not be a momentary pause: this one hypothesis creates several ramifications which deserve to be pursued. Pitted against that, our society piles on the pressure to see texts in print — it prefers something to be available than to be perfect. The result, of course, is that the threads woven together to form the text begin unravelling as soon as the fabric is complete. If we are to be finishers, we are to be the heirs not to Tantalus but to Sisyphus.
Addendum: the delight of the online is that one can, of course, update. Having completed this draft, I came across this talk by Andrew Dunning which I was not able to attend but which, using different evidence, makes a persuasive case for the manuscript I discuss here being Prior Philip’s fair copy of his collection of the saint’s miracles. I am pleased that there will be someone to point out the oversight in the Christ Church catalogue.
Yes, yes, I know. I have been silent for too long, leaving my audience shuffling in their seats uncomfortable at this Cagey performance. I did warn you when I started this site that I have neither the character nor the time to blog incessantly. But, by any standards, the hiatus since the previous post has developed from being a pregnant pause to a laboured silence. It is not, I would like to insist, because I have had nothing to say. Au contraire: several posts have been crafted in my mind, only to fail to be downloaded from brain to laptop. Perhaps their time will come or, more likely, they have been sent to the recycling bin of forgetfulness.
There is, however, one topic that has refused to be ignored and insists that I write something. My research in the last few days, in Oxford, London and Rome, has made me think about the uses and the limitations of the lectio probatoria: those intriguing and infuriating records in some medieval library catalogues of a terse extract, usually just the first word – if you are lucky, two or three – of a volume’s second folio (and so sometimes called the secundo folio). The practice of providing this information appears to have begun in Paris in the thirteenth century and became fairly common in France and in England (but, in contrast, very rare in Germany). The purpose of it is clear: while the opening of a text should always be the same, by the time a scribe and his pen come to the start of the next folio, it is unlikely that he will have reached exactly the same place as his exemplar or another copy to hand, and so the first words of that page can act as a diagnostic, identifying the specific volume where the title alone may not.
The use of such evidence in identifying extant manuscripts and so reconstruct their provenance is well recognised. Daniel Williman and Karen Corsano, in particular, have published studies using the lectio probatoria as a tool for manuscript provenance studies. More recently, James Willoughby has shown that the practice of recording the second folio continued into early print culture: this may seem, at first, nonsensical as what precisely marks out the (relatively) mass-produced printed book is that individual copies are not unique and all of one printing will begin each page with the same words. Yet, in defence of cataloguers who continued the practice, this did not, in the earliest decades of the new technology at least, have to worry them: several copies with identical layout may exist but only one was in their library and so the lectio probatoria remained a useful finding aid. And, as Willoughby explains, for the latter-day bibliographer, it can also be useful in helping identify precisely which printed edition the library owned: while multiple copies were identical, they were highly unlikely in their layout and thus their secundo folio to be exactly the same as the multiple copies of another printing.
But the historian’s use of the lectio probatoria has its limitations. There are obvious duds: the list-writer who records ‘et’ or ‘quod’ as the first word of the relevant page is providing the minimum of information which may not have been of much use then and is certainly not for our purposes. In other cases, the wording may still be fairly mundane but more revealing when combined with the knowledge of the text in the volume, which is usually the first piece of information listed. Yet, I have encountered cases where a search for the relevant phrase in the text cited suggests not it or anything like it occurs at an appropriate point early in the work. Either the cataloguer made a mistake or, as seems to be the case in some inventories, the text listed is not necessarily the first. This is the case with the indenture drawn up for the third and largest gift made by Humfrey, duke of Gloucester to Oxford University in 1444 (and which we know only from the copy recorded of it in the University’s Archives). My reconstruction of what happened is that the inventory-maker, working quickly, picked up the volume and (as was customary, because of the way clasps on bindings closed a book) opened it from the back, flicked through to find a title, and then moved to the beginning to record the lectio probatoria. What, then, he and other list-makers like him were providing was not a record of first work and second folio but a note stating that a manuscript included the cited work somewhere between its covers and, in addition, had the word or words recorded at its second folio.
The lesson from this is that to understand the evidential possibilities of the lectio probatoria, we first have to appreciate the particular modus operandi of that specific cataloguer. There are a couple of other rules of thumb that we need to follow and each can be introduced by a cautionary tale.
The first involves a manuscript of Juvenal, now in the Bodleian, that I was consulting last week. I was interested in it because it has been attributed to the collection of Robert Flemyng, dean of Lincoln (d. 1483), in whom I am interested as he was an early aficionado of humanist manuscripts, and himself a scribe competent in humanist cursive. The provenance of the manuscript – MS. Lat. class. e 30 – has been reconstructed on the basis that the secundo folio agrees with the lectio probatoria of an entry in the 1474 catalogue of Lincoln College, Oxford, where it is stated that the book was given by Flemyng. The manuscript, a small mid-fifteenth-century Italian humanist product is the sort of volume that Flemyng could have picked up on his travels – but it has at the foot of its opening page two coats of arms, one of the Loredan family, the other suggested to be that of the Malipiero family (though the tinctures are wrong). This helps localise the manuscript to the Veneto; it is not impossible that Flemyng bought it second-hand but there is no other evidence to relate it to him – he does not, as he does in a good number of the other books he owned, annotate this volume. Nor does the present Oxford location help localise its earlier existence: we only know it was in England at the start of the twentieth century and may have been elsewhere before that. What is more, there is a problem with relying on the lectio probatoria as deciding evidence. Remember that this is a work of poetry, where the scribe has to respect line divisions. This obviously makes it much more likely than with a prose text that two manuscripts could have the same opening of the second folio. In such a case, the lectio is much less probatoria than we should like.
The second case I present to you (if any of the audience are still in the house) involves a Vatican manuscript for which Williman and Corsano have proposed a provenance. It is a copy of Vegetius the secundo folio of which accords with the lectio probatoria of a volume of that work as recorded in the 1389 catalogue of Dover Priory. This would be interesting evidence of the migration of manuscripts to Italy from England, except that the manuscript itself – MS. Vat. lat. 4492 – disproves the reconstruction by the fact that it was written in Rome in 1408. The book simply came into the world too late to be that recorded in the Dover listing. By the fifteenth century, then, at least two copies of the text Vegetius existed – at opposite ends of Europe – with the same secundo folio. An unlikely occurrence in a prose text, certainly, but one which, given the laws of probability, is going to happen in some cases.
What I want to emphasise in this case – and which is relevant to the Flemyng example as well – is a fact so obvious it should not need stating: the evidence of the first words of the second folio should not be taken in isolation. It cannot be used as a trump card, rendering all other known facts redundant. If it does not accord with the other information available, alarm bells should ring and another explanation must be sought.
This, though, is not to suggest that we should ignore the possibilities the lectio probatoria provide – we too often work with so little evidence, we cannot lightly dispense with this precious piece of proof. Indeed, we can probably make more use of them than we are accustomed to do. They can sometimes help us provide details about manuscripts that are no longer extant. I have enjoyed, on some occasions, identifying more precisely the opening contents of a volume from the record of the lectio probatoria. So, for instance, the first of Humfrey’s gifts to Oxford (made in 1439), one of the entries reads:
Item oraciones Tullii 2o fo. aut quelibet
Checking those two words, I realised that they came at the appropriate point in just one of Cicero’s orations, the Pro Sexto Roscio Amerino, a work which was re-found in 1415 by Poggio Bracciolini (a friend of ours from other posts on this site). The brief entry, then, demonstrates that the duke was providing classical texts which had been circulating for less than a quarter of a century. This much has subsequently been noted in print in an important article by Rod Thomson, but we can add further comment. First, the opening text was an unusual one, most of the copies of the ‘new’ orations choosing other speeches with which to begin. That nugget of information may help us reconstruct the origin of the copy that Humfrey owned. Second, the specific location of the words in the relevant text can give us some general sense of the size of the volume. It can only be general – the changes in possible shape combined with the varieties of scribal practice would not allow anything more exact – but, in this case, with the phrase falling just over 500 words into the oration, we are probably looking at a large quarto volume, perhaps something with between 20 and 25 lines to a page.
I provide that final example to suggest some of the possibilities of the lectio probatoria which are, as yet, underused. It may not be able to match as often as we would like a record with an extant manuscript – and we should use caution, respectful of other evidence and conscious of the type of text that we are studying – but it may well be able to give some insight into a volume that no longer exists. To put it another way: we are, perhaps, too keen to imagine it as a key to unlocking the secrets of what we do have, rather than recognising it as a peep-hole onto what we can no longer touch.
Saturday saw me in the stunning setting of Durham’s Castle, for a conference on the Medieval Library. It was organised under the aegis of the Society for the Study of Medieval Languages and Literature, the publishers of Medium Ævum. The papers took us from the classical precedents (an excellent paper by Matthew Nicholls) to the arrival of print (James Willoughby on characteristically learned form), but through them I sensed some persistent questions.
Later modern societies might conceptualise ‘the library’ as an independent building, a specific pin-point on the map. But for centuries up to, perhaps, the eighteenth, the library was defined rather by its physical or conceptual proximity to other rooms. As Matthew Nicholls mentioned, a classical library might stand next to a mousieon where learned conversations could occur. In the medieval monastery, a library would take upper floor space; below might be the refectory in which the books themselves came alive by being read (as they did in Medingen, as described by Henrike Lähnemann). Similarly, academic libraries – like those in Cambridge about which Peter Clarke talked lucidly – would hold collections which may be useful for study, the focus of which was the lecture hall. For princes (a subject in which the conference’s speaker, Hanno Wijsman, is such an expert), there may be a place in their palace where their books were kept, as in the tower of the Louvre for the codices of the French kings, but the manuscripts would also be seen in the great hall or chamber, where acts of presentation are usually depicted as happening. In other words, we associate books pre-eminently with libraries but their lives were not confined to that specific space. To take this further, it could be said that the library was the place where the book went to rest, the busy-ness of its life occurring elsewhere in the building.
So, the papers at the conference made me think about the limits of libraries, their particular purpose and place in the odyssey of a book. The pre-eminent intention of a library was – as was clear from the discussions like Richard Gameson’s bravura review of images of libraries and their furniture – the safeguarding of knowledge through the protection of books. Yet, as Matthew Nicholls pointed out, this could be self-defeating: a library could itself succumb to fire, flood or other disaster, leaving us with only the titles of its books, not their contents. As Matthew put it ‘libraries can be bottlenecks rather than thoroughfares in the circulation of knowledge’. Presenting your work to a library-owner might gain you prestige and patronage, but not posterity. Thomas Bodley, famously, boasts in the motto of his Library quarta perennis – the fourth will last forever – and libraries now have an institutional certainty that is alien to their predecessors. Yet, that of the earlier Libraries of Oxford University, two died and one (that of Alfred) never existed, might give us pause for thought and remember that even libraries should have a memento mori perennially before them.
But if libraries are designed, however much they fail to do so, to safeguard knowledge – what knowledge? There seems to have been a long association of three concepts: the bibliotheca, religio and sapientia. The libraries are repositories for particular sorts of wisdom and what is interesting is what is excluded from the definition. Ovid complained that his books were banned from Rome’s libraries (which was to their advantage, as they now survive). The collecting of medieval libraries was – as the Cambridge examples discussed by Peter demonstrated – necessarily haphazard: even if there was an original rationale, that could be undermined by the addition of new gifts, and if a donation itself had a special focus, it would often join a collection that worked by different rules. There were also practical limits to a library – a physical space can only take so many books. In my experience, a large library in the later medieval England would include 500 volumes, a very large collection perhaps 800 – 900. The great challenge – as James Willoughby showed – came with print and the exponential increase in the number of books available at a cheap price. That, of course, made the limits of the library an all the more insistent issue. And so began the quixotic early-modern project to reverse Babel and gather together universal knowledge in one place. But, even then, the basic truth remained: whatever the quasi-religious status of learning with the library its temple, the bibliotheca was never the repository of knowledge, but of some knowledge. In that sense, at least, the medieval library may have the advantage over its latter-day successors: it was conscious of its own limits.