The Digital Archive, Scholarly Enquiry, and the Study of Medieval English Manuscripts
A. S. G. Edwards
¶ 1 Leave a comment on paragraph 1 0 The potentialities of digital technology, for those of us concerned with the study of medieval manuscripts, invite some consideration. Our archive is the world: our materials are scattered across libraries and continents by more than five centuries of accession, de-accession, and re-accession, a process that still goes on. The possibility of virtual access to such geographically dispersed materials has now become a reality. Digital technology offers the opportunity to make accessible unimaginably greater amounts of material than hitherto, and to make them available to us in the comfort of our own homes. And not just there. One scholar has recently urged the benefits of digitizing medieval manuscripts by pointing to the fact that they enabled him to study them in such a form while he was on his boat.1 We may feel that this may not be the most compelling argument for the benefits of the expanding digital culture. But it does point to its increasingly pervasive ubiquity.
¶ 2 Leave a comment on paragraph 2 0 How does such ubiquity affect the scholarly study of manuscripts? Going digital, as libraries are increasingly aiming to do, is generally seen as a good thing. Ease of access, as the man in his boat asserts, is often seen as the trump card. The ability to examine what we want without stirring very far is seen as an irrefutable benefit. To suggest that there are drawbacks to such digital developments means aligning oneself with the Luddites, adopting stances akin to talking nostalgically of typewriters, microfilms, and slides to invoke the inferior technologies of the past.
¶ 3 Leave a comment on paragraph 3 0 Ken Pennington, in the letter just cited (note 1), tells us that “there is virtually no downside to digitization.” But matters may not be as clear-cut as he claims. Let me offer two simple examples from my own experience that suggest the benefits and limitations of the use of digital evidence in manuscript research.
¶ 4 Leave a comment on paragraph 4 0 Bodleian Library MS Douce 104 is a manuscript of Piers Plowman with a unique illustrative cycle comprising a series of seventy-two miniatures, all of them (with one exception) marginal. Since no other manuscript has more than a single miniature, this cycle has been the subject of considerable study in recent times, particularly in efforts to use them to clarify the “meaning” of Langland’s poem.
¶ 5 Leave a comment on paragraph 5 0 Some years ago, when I looked at images of the manuscript, I formed the opinion that at least some of the miniatures must have been inserted before the text was copied. Such a view inevitably problematizes a postulated purposive relationship between text and image. But so counterintuitive an opinion needs buttressing by clear-cut evidence. Computer technology should now be able to conclusively establish the stratification of ink and pigment and provide a basis for informed assessment of this aspect of the manuscript’s production.2 The issue here demonstrates the potential of what might be termed the “Gradgrind school of digital technology”: it offers just the facts through its potential to examine aspects of detail at a particular point or points in a manuscript, and enables the differentiation of speculation from certainty that can serve to securely advance scholarly discussion.
¶ 6 Leave a comment on paragraph 6 0 If this example offers a clear instance where digital technology has the possibility of definitive clarification of a problem, this must be in large measure because it does not involve the insertion of human agency into an interpretative process. Let me offer an example of a rather different kind. For some time I have been interested in the activities of Otto Ege (1888-1951), an American collector and entrepreneur, who, in the first half of the twentieth century, found notoriety and a great deal of money by buying manuscripts, cutting them up, and selling them leaf by leaf, thus making problematic any attempt to study what were in some instances important codices. Anyone attempting to reconstruct the original manuscripts rent asunder by his depredations must now confront enormous problems of location and identification of these separate fragments Ege created. But as more library holdings are digitized, the possibility of such identification and the digital reconstruction of the original becomes proportionately greater. Having images from various libraries, of individual leaves in their collections, is the most effective way of identifying the constituent parts of the various manuscripts Ege disassembled.3
¶ 7 Leave a comment on paragraph 7 0 But digitization of manuscripts is concerned with image, not necessarily with either fact or thought. Such images are only useful in appropriate contexts: in the case of Ege’s fragments, key supplementary information is needful, particularly, of course the size of the individual leaf, which obviously will not be apparent on a digital image on screen: the size of the text space, and also the date and circumstances of its acquisition; color register may also be relevant on occasions. An image itself is not enough. Not all libraries are able to provide the necessary information, since not all those who create such projects understand why such data can matter. The result is (for me) a fair number of either ill-tempered or bemused conversations with American librarians in which I either ask them for information, the relevance of which they find baffling, or in which I offer them information about the history and material of a manuscript leaf that they find irrelevant. Digitization is such respects runs the risk of operating in a vacuum, one in which the act of archive creation exists separately from necessary material and historical context.
¶ 8 Leave a comment on paragraph 8 0 These examples may suggest some of the more obvious potentialities and pitfalls of the creation of digital archives as these relate to medieval manuscript study. I raise them since I am not convinced that the issues that underlie them are always clearly grasped by at least some of those who initiate digital projects, as I will try to suggest in what follows. At least some undertakings have suggested that digitization has become an end in itself, existing in a conceptual void, creating data that is essentially inert.
¶ 9 Leave a comment on paragraph 9 0 We get a glimpse of the dangers that can underlie archive creation in the recent project to digitize all the manuscripts in Corpus Christi College, Cambridge. This is a project that cost millions of dollars in the costs of digitization. But the interpretation of the images so produced is underpinned by metadata derived from a tool created more than a century ago and never commended as a model of comprehensiveness or accuracy, M. R. James’s catalogue of the Corpus manuscripts, which provides the basis for the physical descriptions and details of content for these manuscripts.4 The digital hare is uneasily yoked to a print tortoise.
¶ 10 Leave a comment on paragraph 10 0 A couple of issues follow from the examples of Ege’s fragments and the Corpus collection. Even though they differ in scale and kind, they are linked by the fact that what ultimately makes the digital archive meaningful is actual expertise in matters to do with the nature of the original materials, in this case codicology and content, at the very least. Technology is not enough in such circumstances. Hard-won knowledge creates expertise that in turn forms the basis for appropriate scholarly contextualization of the digital image. And without such contextualization, the digital archive is no more than expensive testimony to the absence of thought. The launching of digital archives requires such contextualization that must be drawn from expertise based on the first-hand examination of the original materials, from an understanding of their material forms and histories. Without this information, the digital archive is data without meaning.
¶ 11 Leave a comment on paragraph 11 0 The point may seem obvious enough. But behind it lies a larger question: to what extent is the digital archive an impediment to the study of manuscripts? A while ago, I received an announcement of a lecture by David Zeidberg, Avery Director of the Library at the Huntington Library. The lecture was titled “A Library of Last Resort: The Huntington in the Twenty-First Century,” and the accompanying blurb reported that it would discuss the Huntington’s research holdings and talk about current trends in public and academic libraries. It goes on: “In the general move to digital resources, the Huntington is now distinguished as a ‘library of last resort’ where researchers have access to rare original materials as well as to digital resources.”
¶ 12 Leave a comment on paragraph 12 0 The claim is, of course, rhetorical. Other libraries do still offer access to original materials even when such materials have been digitized. In some libraries, like the British Library and the Bodleian, there can be wearisome protocols attached to such actual access to a Real Manuscript, with the proffering of microfilm as an alternative as generally the first step in what can be a protracted negotiation. We are not yet, however, at a point where the virtual is regarded as synonymous with the actual by most institutions. But the Avery Director at the Huntington does point to the growing reality that research libraries see the provision of surrogates, of whatever kind, as the appropriate fulfillment of their responsibilities.
¶ 13 Leave a comment on paragraph 13 0 Some research libraries seem increasingly less interested in either the credentials of the researcher or his or her motives in seeking access to originals than in making it as difficult as possible to do so. Some libraries require detailed explanation for seeking an original rather than using a surrogate. The reason most often advanced for restricting access is conservation: the need to protect delicate materials from uncouth hands. The reality may be shaped by other factors, often the most significant of which are likely to be security and staff time. Accessing the archive is one aspect of digital politics in the library: it is cheaper not to permit people to look at what has to fetched, policed, and carefully re-examined on its return. The deployment of such resources of labor and expertise can be reduced by digitization.5
¶ 14 Leave a comment on paragraph 14 0 Such realities are rarely explicitly acknowledged by libraries, perhaps because the denial of access to originals can easily be shown to be inherently dubious, if not wholly wrong. In spite of my nautical medievalist, happily accessing images on his boat, there is a great deal that digitization cannot tell you about a manuscript that you need to know, starting with the obvious facts of real size, physical structure (quiring and cancellation), distinction of materials (vellum or paper), differences in ink, and color registers. In later periods, punctuation may be an issue not resolvable just by digital images.6
¶ 15 Leave a comment on paragraph 15 0 No competent scholar is likely to place extensive trust in any form of surrogate, whether it is digital, photographic, or microfilm. Undoubtedly, digital images are superior in quality to other forms of reproduction, often in color and sometimes manipulable. But in general they are used in the same way as other less expensively produced images, to provide a text or texts; in such respects they are a more expensively produced form of photograph or microfilm. And they remain open to most of the same cautions about their use as apply to other such surrogates.
¶ 16 Leave a comment on paragraph 16 0 A countervailing argument might stress that digitization gives the benefit of extensive free accessibility of images. But the truth is rather more complex. Freedom of access does not come free, but at an unquantifiable price. That price is already seen in the reluctance, I have already noted, of libraries to allow access to originals where a surrogate is available, to save expense. But is digitization a money-saving activity for those institutions that employ it extensively? To put matters at their simplest, digitization is not a one-off expense. Behind it lurks the whole question of site maintenance, of the need for regular updating to preserve accessibility, of conserving the creating technology to preserve what has been created. No one can say what such ongoing site maintenance is likely to cost over time. Digital Realpolitik seems often to tend to substitute short-term pragmatism for any clearer sense of the future. But it seems clear that the more that is digitized, the greater the expenses of site maintenance will be on a continuing basis. Such costs could quite quickly force libraries into a situation where efforts have to be devoted to holding back the tide of technological decay rather than doing anything else, such as expenditure on staff and books. This also means that forms of digital research that could be of great benefit to scholarship are avoided, perhaps because of rising costs of other kinds digitization. Such research now has the potential, for example, to make advances in the recovery of damaged parts of a text. Although many of the surviving Cotton manuscripts were damaged in the 1731 fire, the British Library does not seem to have made any effort to explore the of digital tools to recover damaged parts of these manuscripts.
¶ 17 Leave a comment on paragraph 17 0 If the costs of digital archive development are, at best, worrisome, what about the digitals archives that are being created? Do they justify the potential expenditure? For medievalists, a particular benefit of digital activity has been held to be the creation of such archives for the editing of texts, particularly those texts that survive in multiple, complexly variant forms. The potential of the digital edition to create new forms of representation and conceptualization is often urged by its exponents. But the problems of such forms need to be stressed.
¶ 18 Leave a comment on paragraph 18 0 One reservation has to do with what we mean by “editing.” Take, for example, what seems to me a very useful electronic resource, the electronic form of Auchinleck Manuscript (National Library of Scotland MS 19.2.1) prepared by the late David Burnley and Alison Wiggins, now fully available online.7 This makes accessible a full transcription of the manuscript, with links to images, manuscript description, and a variety of other resources. Given the length, codicological complexity, and range of contents of this manuscript, this is a sensible way of presenting this material. We must be grateful to those who prepared it for giving us a scholarly tool of lasting value. But this is not an edition (and never claims to be such), and its merits are those of a meticulous, well-annotated transcript, the various components of which can be employed for a variety of purposes, pedagogic and academic.
¶ 19 Leave a comment on paragraph 19 0 Not all such electronic editorial projects are so clearly focused in their sense of what is possible. Consider, for example, the Electronic Canterbury Tales project. This was begun more than a quarter of a century ago by the late Norman Blake, and Peter Robinson. The Electronic Canterbury Tales is a project that received somewhere in excess of a million pounds of funding over a fifteen-year period in England and is now receiving more funding in Canada. Its professed aim was to produce a series of edited fascicles (originally on CD) of each of the Canterbury Tales, with full variants (prepared according to a collation system of Robinson’s devising) and with accompanying images.8
¶ 20 Leave a comment on paragraph 20 0 This endeavor has been linked to large promises of what it would achieve. Peter Robinson has claimed that “the computer techniques we have pioneered [in the Canterbury Tales Project] may (at last) permit us to recover the text that Chaucer actually left behind.”9 It is surprising that such a claim can be made on the basis of “computer techniques.” Ralph Hanna has correctly observed of this enterprise that “one does not  . . . from a machine or rule, but on the basis of what Bentley called ‘ratio’ or Bentley called ‘discriminating faculty.’”10
¶ 21 Leave a comment on paragraph 21 0 But if such a large claim is not likely to be supportable, what does the digital archive here do to contribute to the editing of Chaucer’s poem? How much is actually new here? The answer appears to be: not a lot. The transcriptions of all manuscripts certainly are. Not all the Canterbury Tales manuscripts had been fully transcribed hitherto, although the most textually important largely had been, with a fairly high level of accuracy.11 And it is hard to feel that there is value in transcribing all the manuscripts. Some manuscripts are manifestly copied from others that survive, so have no claim to authority. And the substantive variants for all the manuscripts were collated, very accurately and comprehensively, by teams of graduate students at the University of Chicago in the 1920 and 1930s in the preparation of J. Manly and E. Rickert’s study, The Text of the Canterbury Tales (1940).12 It is not clear therefore how electronic transcriptions will help in themselves “to reconstruct the text Chaucer actually left behind.”13 What is needed is not redundant creation of electronic transcripts, but actual analysis of the corpus of variants that has existed for more than seventy-five years.14 That is, the application of thought to textual criticism. Whatever thought may achieve, given what is available, digital assistance is unlikely to be helpful. What can be achieved will depend on the quality of an intellect to analyze the existing evidence. This is not a question of inputting variant forms of text into an electronic archive. The full online transcription of large numbers of manuscripts will serve very little editorial purpose. Textual editing, particularly in Middle English works, entails analysis of variation through collation, which involves examination of the original for aspects not securely discernible through digital images, such as changes of ink or evidence of palimpsest.
¶ 22 Leave a comment on paragraph 22 0 Such uncertainties about digital archiving as an aspect of editorial method can be extended to include another ambitious project, the Piers Plowman Electronic Archive, begun at the University of Virginia, which has so far been the recipient of over 1.5 million dollars in direct grant funding. The current on-line preamble to the project announces that
¶ 23 Leave a comment on paragraph 23 0 The Piers Plowman Electronic Archive [is] a collaborative open-access project, presents the rich textual tradition of Piers Plowman, a fourteenth-century allegorical dream vision attributed to William Langland. Three distinct versions of the poem (A, B, and C) survive . . . The long-term goal of the project is the creation of a complete archive of the medieval and early modern textual tradition of Langland’s poem. . ..15
¶ 24 Leave a comment on paragraph 24 0 The archive will be formed by the production of transcriptions of individual manuscripts, some of which have already been completed. But how do they contribute to our understanding of the textual tradition of Piers Plowman in a way that is different from the very extensive corpus of variants already exists in the various volumes of the Athlone Press edition? What is a “rich” textual tradition? And how is it possible to announce the project’s conclusions in its early stages, in its expressed belief that there are “three distinct versions of the poem” at a time when four or five versions of Piers Plowman, or perhaps more, seem possibilities that need to be seriously assessed? The text of Piers Plowman appears to have been far more fluid than that of the Canterbury Tales, with the combination in some manuscripts of different parts (A+C), the problematic Z text, held by some to be an early draft of the work, and other more local fluctuations (the RF manuscripts of the B-Text provide one obvious example).16 These variations have long been common knowledge. It is not easy to understand how an “archive” of all textual forms of the poem will extend knowledge in useful ways.
¶ 25 Leave a comment on paragraph 25 0 There seems a lack of clarity about both controlling assumptions and method in this project, a belief that the creation of a full text electronic archive is meaningful in itself. The Piers Plowman project no longer announces, as it once did, an intention to use such transcripts to “propose a set of solutions to editorial problems without suggesting that they will have final authority.”17 This is probably a wise decision since it was never clear how multiple transcriptions might contribute to such “solutions.” But the desire to create an electronic archive seems to have preceded any clear sense of why or how it might be a contribution to any understanding of the textual tradition of the work. Are all the manuscripts so variant? Or only some? Should the nature and extent of variance determine which manuscripts should be transcribed in full? If so, what purpose would they serve?
¶ 26 Leave a comment on paragraph 26 0 The relationship of the edition to the archive, at least in Middle English terms, seems to me at times a curiously problematic one, creating perhaps pressures to feed the archive rather than actually engage in editing. If this suggests a certain absence of thought it can perhaps be set alongside other forms of current Middle English archive creation that show related tendencies. An obvious example of this is the Middle English Prose Brut project, an undertaking funded by the Arts and Humanities Research Council (AHRC) that aimed to produce (at the very least) detailed online descriptions of the manuscripts of this work. This it has signally failed to do: about a third of the manuscripts are undescribed at all, and the descriptions that are there are not consistent in the data they include. And having run out of funding, it stopped work, leaving us with this drifting digital hulk. The inadequacies of the undertaking extend beyond incompleteness. Descriptions of individual manuscripts vary markedly in their levels of detail and generally fail to provide categories of information the user may seek. For example, there are no consistent records of the histories of ownership and sale, about which much information was retrievable. 18
¶ 27 Leave a comment on paragraph 27 0 To repeat: what is an electronic archive for? If it is to be meaningful it must involve the assembling of data on a different scale and (potentially) with a different level of accessibility and searchability than can exist within print culture. But accessibility and utility are not synonymous, and access to such archives can, at times, only reveal the inadequacy of the vision and of the executing will that underlays their creation.
¶ 28 Leave a comment on paragraph 28 0 On various grounds, technical and conceptual, the Brut project testifies to the capacity of archive creation to become a monument to intellectual inadequacy, to the inability to apply thought to the fruitful application of digital methodologies. One might contrast this project with another much more longstanding one created by experienced manuscript scholars who understand what such an archive might properly do. I am thinking of the Digital Scriptorium now based at UC-Berkeley. This is a long-established online site, one that has remained resolute in its limitations. Its explicit aim is to provide minimal data: a brief physical description, references to provenance and further descriptions, and some images (frequently restricted to a single leaf) capable of being enlarged. The firm limits it places on its scope make it extremely useful. It does not pretend to be a substitute for the real thing, but provides basic accurate information about the real thing that may provide a firm toehold for further research. As an example of method it can scarcely be faulted: it understands clearly what it is doing and why.
¶ 29 Leave a comment on paragraph 29 0 In contrast, the modern academic world encourages people to apply for large amounts of money for digital enterprises without any demonstrable sense that they know how to use it to any productive purpose and who are not prevented from getting such access by effective screening procedures. As anyone with experience as a reviewer for funding bodies will know, this means that funds that could be deployed usefully in other research activities (such as visiting libraries to look at original materials) are expended on “projects” with costly digital components and very limited plausibility. The pressure to appear to be digital in grant applications is a pressure to think expensively, not necessarily to think expansively.
¶ 30 Leave a comment on paragraph 30 0 It is made more so by the particular historical moment. In Great Britain we live in an academic world where the ability to obtain large grants is more valued by our managers than our ability to do good research, where funding strategies seem increasingly to be more important than writing good books. In such a world, the criteria for success are opportunism and plausibility, since getting large amounts of grant funding depends on the creation of digital (hence expensive) strategies. If we as academics are to be seen as only as good as our last successful big grant, while at the same time the difficulty in obtaining such grants is to steadily increase, creating digital archives becomes not so much a holy grail as a blatant beast, the quest for whose shimmering elusiveness will deflect us from fruitful forms of enquiry. The constant pressure on the humanities to go digital runs the risk of making technology an end in itself at the expense of more valid, less expensive, forms of scholarly enquiry.
- ¶ 31 Leave a comment on paragraph 31 0
- This is the claim of Ken Pennington; see his letter “Better on the Continent,” Times Literary Supplement, 12 March 2010, 6. This is a response to a brief article of mine, “Medieval Bytes,” Times Literary Supplement, 29 January 2010, 15. [↩]
- Research to establish the sequencing of text and image is currently being undertaken by Professor Stephen Shepherd of Loyola Merrymount University. [↩]
- For the most detailed and authoritative account of Ege and his activities see Scott Gwara, Otto Ege’s Manuscripts (Cayce, SC: De Brailes Publishing, 2013). [↩]
- M. R. James, A Descriptive Catalogue of the Manuscripts in the Library of Corpus Christi College, Cambridge, 2 vols. (Cambridge: Cambridge University Press, 1912). [↩]
- The other argument for reduced access to originals is conservation, the need to prevent physical deterioration by excessive handling of originals. This is a complex question that does raise issues of responsible access. But the suspicion remains that it used with some frequency as an excuse rather than a reason for denying access to an original. [↩]
- See, for example, the report of Robert Louis Abrahamson, “Rummaging through a Beinecke Treasure Chest: An Appreciation by a Visiting Fellow,” Beinecke Illuminated no. 1 (Fall 2014), who reports (p. 11) on the request of a colleague for an “on-the-spot-check” of punctuation in one of Robert Louis Stevenson’s unpublished stories since it was unclear on the pdf she had of the manuscript. [↩]
- Through the National Library of Scotland website: auchinleck.nls.uk, accessed August 14, 2016. [↩]
- For some account of this project see Peter Robinson, “The History, Discoveries, and Aims of the Canterbury Tales Project,” Chaucer Review 38, no. 2 (2003): 126-39. The passage of time has made a number of the claims made there open to doubt. [↩]
- Peter Robinson, letter to the editor, The Times [London], July 14, 1998, p. 19. [↩]
- Ralph Hanna, “The Application of Thought to Textual Criticism in all Modes—with Apologies to A. E. Housman,” Studies in Bibliography 53 (2000): 163-72 (165). [↩]
- See, most recently, The Norman Blake Editions of the Canterbury Tales: The Multitext Edition, edited by Estelle Stubbs, Michael Pidd, Orietta da Rold, Simon Horobin, and Claire Thomson, with Linda Cross, an online edition, https://www.chaucermss.org/. [↩]
- J. Manly and E. Rickert, The Text of the Canterbury Tales, 8 vols., (Chicago: University of Chicago Press, 1940). [↩]
- Robinson, letter to editor. [↩]
- A further problem is the form in which the completed parts of the “Electronic Chaucer” have been marketed: as CDs, which are no longer readable, as are other publications promulgated in this way, like those initially issued by the Society for Early English and Norse Electronic Texts (SEENET). [↩]
- Piers Plowman Electronic Archive, Society for Early English and Norse Electronic Texts, last updated March 18, 2018, http://piers.chass.ncsu.edu/texts.html. [↩]
- The discussion by E. Talbot Donaldson, “MSS R and F in the B-Tradition of Piers Plowman,” Transactions of the Connecticut Academy of Arts and Sciences 39, no. 2 (1955): 181-212 still invites consideration as does the later work on the so-called “Z Version” by A. G. Rigg and Charlotte Brewer. See their edition of this version: A.G. Rigg and Charlotte Brewer, eds., Piers Plowman: The Z Version, Studies and Texts 59 (Toronto: Pontifical Institute of Mediaeval Studies, 1983). [↩]
- I quote from the earlier online preamble: Piers Plowman Electronic Archive, January 10, 2015, http://www3.iath.virginia.edu/seenet/piers/archivegoals.htm. [↩]
- See A. S. G. Edwards, “Bruts for Sale,” in The Prose Brut and other Late Medieval Chronicles, eds. Jaclyn Rajsic, Eric Kooper, and Dominique Hoche (Cambridge: Boydell & Brewer), 218-31. [↩]
A. S. G. Edwards
Honorary Professor of English – University of Kent