Today deserves a little fanfare: the latest instalment of Lost Manuscripts has been made available this morning. There are only 32 new fragments, coming from 19 host volumes, so not a full drum-roll and brass band, please — but some piano trumpetting will not go amiss.
This is the third batch drawn from the collection of Samuel Harsnett (1561-1631), a Colchester boy who rose to be Archbishop of York. We must be grateful that he left his library to his hometown, and perhaps equally grateful that the authorities did not look after his books too well: that is, they refrained from hyperactive campaigns of ‘restoration’ and so most of the books are in their early modern bindings, with some usually sensitive repair work done in the 1970s. With those early bindings often come flyleaves and pastedowns from manuscripts. The procedure on the Lost Manuscripts’ website has been the same as before: to give brief descriptions of the bindings, to catalogue the fragments and also, where possible, to bring together separate fragments from the same manuscript and to record that. For reasons I have explained elsewhere, the imaginary location where those physically divorced fragments share an existence as a partially reconstructed manuscript is the city of Babel. The Babel numbers now run up to forty, with twelve new inhabitants appearing today. They include some elegant arrivals — my personal favourite is a full leaf from a fine copy of Moerbeke’s translation of Aristotle’s Metaphysics — but others, it must be admitted, turn up looking rather shabbier.
The intention of this project has never been simply to make freely available, in virtual format, information and images of these books, though that is a benefit in itself: if the result of the website is that someone appreciates more fully what can be found in an early modern book, then it has done a service. My aspiration, however, is that over time we can ask much larger historical questions about the death of manuscript culture. We are far away from being able to do that, but I do have a couple of tentative comments which I think deserve investigating further.
1. The long life of tearing up manuscripts
It is already known that, while the heyday of book destruction was the Reformation period, it began long before that and did not end in the mid-sixteenth century. The Harsnett collection itself has examples of this and it goes back much further: it was an element inherent in manuscript culture, new codices sometimes cannibalising older ones, not just through the re-use of parchment, creating palimpsests, but also through the recycling of discarded pages in bindings. The sixteenth century inherited this practice, which was certainly most widespread in the first three quarters of the century. It did, though, continue and not just in Oxford (as is sometimes said). The material used did, in some cases, change, with an increase in recent paper waste, including booksellers’ accounts, being taken out of the rubbish (or the privy) and employed to serve the purpose that parchment manuscript leaves had previously provided. There are, however, other occasions, at the very end of the sixteenth and into the seventeenth century, when medieval codices were available and were deployed in bindings. In some cases, these were from what must have been quite impressive volumes, being used in known a book-collector’s library: the books of Richard Bancroft (1544-1610) provide notable evidence of this from his time as archbishop of Canterbury (1604 until his death); I will discuss this in more detail another day. Perhaps the manuscripts involved were already too damaged to consider preserving them, but it remains striking that, in this generation of the early antiquaries, destruction could be thought an appropriate process.
2. Not one process but many
What also has intrigued me from the findings we have made so far is the variety of practice. It is obvious that there are differences in use which define the types of fragment that survive: some binders included pastedowns, others had large strips as flyleaves, yet others small reinforcing pieces. There were differences as well in the quality of manuscript, and the parts of it they would employ. In some cases, this must have been governed by issues of availability, a dynamic that must lie behind the increasing use of recent documentary material which I have just mentioned. That, though, would not explain, for instance, the habits of some binders who cut up the manuscript and kept only those pieces with no or minimal text. As some clearly did this consciously, it raises questions about others’ practices: did the provision of an elegant leaf as a pastedown, replete with text and sometimes illumination, say something about the interests or even the loyalties of the binder or of their client? There is a basic opposition in the practices between those which were intended to hide the origin of the material used and those that celebrated it, but even here these describe two extremes of a range of practices. Our intention here must be to get into the mind of the binder when they had the manuscript before them and set about re-using it: we want to recover the impulses that moved their hand as the knife came down upon the parchment.
There are, of course, further issues raised by the increasing amount of evidence that is being gathered, and any comments at this stage must be provisional. But at least we are beginning to know what questions we can ask – and I am confident they will be thought-provoking questions.
In between the various commitments of a teaching term, I have been working on the release of a second batch of entries for the Lost Manuscripts project – and today is the day they are launched upon the world.
As you may remember, the pilot project involves cataloguing and digitizing the manuscript fragments found in bindings of books once owned by Samuel Harsnett and now resident in the library of the University of Essex.The rationale has been to treat each fragment both as an object worth describing in its own right and as witness to a manuscript which once existed but is now lost to us. In doing this, then, it engages with the developing and exciting practices of ‘virtual reunification’ – the process of bringing together on the internet elements of a work of art which are physically dispersed. Its manuscript variant is sometimes called ‘fragmentology’, an undeniably unlovely term (but I have complained about that before now). What the project hopes to bring to the metaphorical table, beyond a set of new examples, is thinking about the standards of cataloguing we might require for fragments and how they may differ from those for complete codices.
The intention of the Lost Manuscripts site is to make the images and the descriptions freely available. The bulk of the work for the pilot project was done in the summer of this now-closing year, but the pages are being released in small batches to allow for checking and the addition of further information. So, today is one step on a journey, and a small one at that. In the first batch, there were twenty ‘lost manuscripts’; in this, only eight. That is partly because this group includes several fragments which do not meet the rules we have set for creating a lost manuscript: if only one remnant remains from which it is impossible to extrapolate what might have been, then it is not allowed into that virtual realm of reborn codices we like to call Babel. So, a stray strip of music, or a single scrap of a copy of the Digest stands beyond that city’s gates. Of course, in those instances, our hope is that more work and more discoveries will allow us to link up that lone fragment with others – and then the doors will be opened to them.
All the same, there are some interesting finds in the fragments now there to view. Some of these are discussed on the Highlights page of the site. They include an example I have mentioned before of how some binders chose to save not the text from a manuscript but precisely those parts which provided virgin parchment. There is also a useful reminder that, while the process of dismantling manuscripts is, in English history, particularly associated with the disruption of the Dissolution of the Monasteries, the reformers by no means invented the practice. The fragments involved in that case are a personal favourite at the moment because several come from one medical manuscript for which we have been able to identify precisely from what text they come: it is painstaking and, indeed, thankless work which, in this instance, involved learning more about the varieties of urine than any layperson could ever really want to know.
As the project develops, what is also coming into sharper focus is the range of questions we should be (funding permitting) asking at the next stage. We can detect a variety of practices that took place in different binderies – what were the reasons for those differences and how did they develop from the fifteenth to the early seventeenth century? We also know that a broad sweep of medieval texts were ripe for cutting up, but did the variety of manuscripts and the balance between them shift over the sixteenth century? What, fundamentally, was the logic of the destruction of manuscript culture in the early modern period? These are big issues which will need ‘big data’ to begin to answer them – but they are ones that are surely worth asking.
Erik Kwakkel, one of our leading palaeographers (and by ‘our’, I mean us Europeans), has just posted on his blog a stimulating article on the size of margins in medieval manuscripts. As he rightly points out, they tend to dominate the page – that is to say, the text block forms the minority of the space. He remarks that this is astonishing, by which he means in a culture where parchment was rare and thus, we might add, there existed a culture of the palimpsest. The focus of his article is pre-Gothic; for me, working mainly with humanist manuscripts, such proportions are an expected part of what I describe as the ‘eloquent page’ which is, simultaneously, a work of conspicuous consumption: the text is isolated in a cream-white sea of margins. With scribes working after the invention of print, this could be taken to extremes: I spent a little time working out the proportions of text to margin in the manuscripts of the prolific Netherlandish scribe of the early sixteenth century, Pieter Meghen: the text could form under a third of the page. Here, the intention seemed to be to mark out the superiority of manuscript over printed text by the ostentatious waste of space on fine parchment, rather than the packed page on cheap paper that made print a commercial, if not often an aesthetic, success.
I have also been reminded of the importance of blank sections of manuscript by one of my present projects, creating an on-line database of fragments in the bindings of the books of Samuel Harsnett, archbishop of York (d. 1631), now held in the University of Essex. Many of these sixteenth-century bindings include pastedowns, flyleaves binding strips, reinforcing pieces or tabs taken from manuscripts. Here was a culture in which creating the finished product of the bound book involved destroying older books. You might baulk at that phrasing: a manuscript might have entered the bookseller’s already too damaged to be valuable as anything other than scrap, but the practice still suggests at the least a connivance with dismantling. What is more, what has struck me is that for some sorts of uses in bindings – particularly trapezoidal binding strips – some binders clearly had a distinct preference for blank parchment. They would not, we can assume, have had it made to order and instead they reserved for use margins from codices: we can know for certain that this is happening when a flick of pen-flourishing or a rubbed plummet note is still partially visible on the cutting (as in the examples I posted on Twitter). That is to say, those parts without text may have had a better chance of surviving through being recycled than those which the scribe had short-sightedly blemished by inflicting ink and letter-forms upon it.
All this discussion, though, misses a larger, very obvious point: the margins form only part of the blank space of a page. The textblock itself works by having space between the words and between the lines. Even at the most compressed, the height of minims may only be half the distance between lines and, while we must take into account ascenders and descenders, we are likely to have to conclude that at most about half of what we call the textblock is occupied with letters. That is to say, in the less-than-50% reserved for words, little more than 50% might be taken up with them. That is because (as I have put this in a lecture I gave last year), script needs space like the candle needs shadow – except that the script is the shadow casting darkness on the light of the page.
We could say that this makes the words all the more precious; they are in dialogue with their own absence, and that dialogue is surrounded by what might look like silence. I like to think of this as a metaphor for the social context in which these manuscripts were produced: the words being the minority, just as the literate who could engage with the words through reading were the few. I am not saying that scribes actually thought of the mise-en-page in these terms, though, as I will suggest in print soon, I do think we can see some being very conscious of themselves working within a majority-illiterate society and conscious of the limits to literacy, including their own. More obviously, though, the inhabting of the margins of gothic manuscripts by grotesques often so at odds with the text, at some level speaks to this paradox. Decoration and text are not in conversation; they are speaking different languages.
The presence of that decoration (another blemish on the virgin parchment) reminds us also that, just as oral culture was by no means uncreative, so these margins are not mute. An implication of what I have just said is that, if we look to all space on the page, we can recognise that it is not all blank in the same way. These spaces, too, are places, with their own articulation. How well can we hear their voices?
Let me entice you into the half-light, into the region inhabited by manuscripts which are no longer fully alive but which have not disappeared entirely – the world of undead books. I have become convinced that working here could have the power to be transformative of our understanding of manuscript culture. I want to encourage you to travel with me on this adventure.
All of us who work with manuscripts – or, indeed, with the earliest printed books – are conscious that we deal with only the minority that survives and can only dream of what once might have been. Early on in our researches, we come across those shards of evidence that exist between the two states, often collected in guardbooks or in boxes, though sometimes loose or hidden within other books – fragments. I first engaged with them when working on my doctorate and I could not resist their lure. On the emotional level, there is something tantalising about this evidence of what we have lost; on an intellectual level, it is hard not to relish the challenge of identification.
Perhaps the fascination of them is why I have allowed myself to be seduced back into studying them time and again. When, in 2003, I became an Editor for Oxford Bibliographical Society, my first task was to oversee the reprint of Neil Ker’s classic Pastedowns in Oxford Bindings, first published in 1954 and the classic study of a corpus of fragments whose place of dismemberment is localisable, since, in sixteenth-century Oxford, binders (more often, it is said, than anywhere else) strengthened and prettified the bindings they put on books by gluing a section from a discarded manuscript to each of the book’s boards. The reprint was not simply a reproduction. It involved providing some light updating, based on Ker’s own notes, those of another hero of Oxford manuscript studies, Richard Hunt, and the work of David Pearson, who had already supplemented Ker’s work in his own Oxford Book-binding 1500-1640 (Oxford, 2000). It resulted in the thirty-page addenda and corrigenda, work which made me conscious of how much more could be done with these broken survivors of an era of destruction.
Even at that point just over a decade ago, the potential of an on-line database of fragments was already imaginable. (Indeed, a review of Codicologica from 1983 threw out the suggestion of a ‘computerized information bank’). In the years since working on Ker, I have mused with friends and colleagues about the opportunities there might be for doing just that. Now, thanks to the University of Essex, ‘seed-corn funding’ has been made available for a pilot project which, in the coming months, will see created a digital catalogue, with images, of a particular set of fragments, in situ as strips, flyleaves and pastedowns in bindings of books once owned by Samuel Harsnett (1561-1631), Archbishop of York and son of Colchester, who left his library to the town; those books are now in the safe keeping of the University of Essex. The intention, after this pilot, is to move on and to build up a larger database of fragments in the British Isles. How that will be done is not what I intend to discuss here. Instead, I want to consider the intellectual requirements and possibilities of such an undertaking.
Of course, the digitising of fragments now has a plethora of precedents. Juergen Berger has helpfully pointed that a listing of some of these has recently been provided by Kaspar Kolk, who is himself working on manuscript fragments in his native Estonia. His survey suggests the range of endeavours occurring across Europe and in North America. The El Dorado for many of these ventures is the aspiration of bringing together elements from one manuscript which are now dispersed. How attractive this possibility can be is suggested by the interim result for a small poll related to my own project: I invited viewers to help name the enterprise and, to date, the preference is for my jocular suggestion of Fragments Reunited (that will teach me to try a joke).
To achieve any reconstruction, however, requires some painstaking research and the scholar needs all the help both the Internet and hard-copy sources can provide. The ability to identify a text is unimaginably greater – I mean simpler and quicker – than in the mid-twentieth century when Neil Ker was at work. But so many fragments are little more than scraps and thus defy identification by text alone. And when the words are susceptible to being pinned down to a particular work, there remains the issue for every piece of parchment – even if it represents the most uncommon text – of ascertaining whether it does come from the same codex as any other fragment of the same composition.
I say this not to arouse your sympathy for the hard-pressed archival archaeologist but, instead, to raise an issue of how we catalogue fragments. One of the most important websites being built at present is the Inventory of Medieval Manuscript Fragments in Norway. They have done sterling dectetive work in organising extant sections of leaves by their original manuscripts. In doing this, they have used the rules for cataloguing developed by J. P. Gumbert. His guidance provides the noble principle that each fragment is a manuscript in its own right, worthy of being given the same treatment as a full survival. At the same time, another guiding principle of his inventory approach is the need to work at speed, with each entry being as pared down as possible. If, though, we are going to maximise our chances of making accurate identifications, I would urge that we need to include details which are not always necessary for a description of a complete manuscript, while also ensuring other data are fully searchable. So, it is not usual to measure the space between lines in a conventional description (though this can be deduced by dividing the height of the written space by the number of lines) nor the height of minims but, as a cutting is likely not to provide the full text block, these have the potential to be important diagnostics. Similarly, if the cutting is from the centre of a bicolumnar folio, we will not know the dimensions of the columns but we can measure the width of the central reservation, which might well help us make an identification. This last is a datum that should always appear in a description but if, in a database it is recorded only as one part of the dimensions, its ability to act as a comparator is all but lost. In other words, if it is going to be fully searchable, the information recorded needs to be broken down to a level of detail not usually considered necessary. My own experience is that entering these data does not slow down the process of cataloguing by more than a few seconds – and can reduce substantially the time needed later for compiling the incomplete jigsaw that are the related fragments.
It will already be clear that I am not certain that we have fully realised what we need if we are to make the most from fragments. That is likely to be because we have not yet appreciated the entirety of their potential. To give these battered remnants the attention they deserve, we needs must adopt the mantra that a fragment is a manuscript but, in an obvious and fundamental way, that is untrue and, on my submission, can even undermine our recognition of each scrap’s significance. A fragment is not an island entire of itself, nor is a cluster of them simply an archipelago. Or, rather, if it is, we are like marine geologists looking for the submerged mass which connects the elements together. That is to say, each fragment (however tiny) is a witness to a whole manuscript and should be taken as an invitation to envisage how that codex would have looked. Faced with an insignificant and scruffy survival, it may seem hubristic to think we can move from that to conceptualising the pristine object, but, if we use all the information available, and work both by extrapolation and analogy, it is not impossible to glimpse, at the very least, the original codex. This is why I would urge that, when we record fragments, we should in effect provide a double catalogue, once as the individual piece, respecting its present condition and location, and once as a testimony to a recovered manuscript – the section that, in the database I am developing, will be called ‘Babel’.
You might ask whether it is worth the effort spending the extra time on that process of reconstruction. If the potential stopped with the completion of the catalogue entry, perhaps – in many but not all instances – it would not be. That is not, though, where our work should end, for the greatest gains are to be had by analyzing the gathered weight of data that a sustained project can provide. If we continue for the moment thinking about the fragment as witness to the lost manuscript, a question that will press itself on us is how the volume came to be dismantled and half-discarded. We may think we know the answer: we can explain that, in various societies, there have been moments of destruction in their history, and we might cite as an example England in the mid-sixteenth century. That, though, is not precise enough: we should ask ourselves what we can learn about the details of the individual journey each single manuscript took from wholeness to dismemberment. If a fragment sits in a binding, we can often tell in what book-shop it must have been torn apart and we might then ask how it could have reached this bookseller (who so loved books that he broke up old ones to strengthen the new). Were all books used, for instance, in Oxford bindings in the sixteenth century pulled from local resources? We cannot know that – yet. Our goal, I am suggesting, is that we should see an endpoint of working with fragments to be about gaining a sharper understanding of the processes and levels of loss that have occurred. This matters because, as we all know, the surviving complete manuscripts that we have are a minority of those that were produced in medieval Christendom. We are faintly aware that what survives is probably unusual; what we surely need to do is to have a stronger sense of how unrepresentative the extant full codices are of the manuscript culture that existed. The study of fragments invites us to move beyond the comfort of what we have and to develop a history that more fully recognises what we have lost.