Version classiqueVersion mobile

Text and Genre in Reconstruction

Willard McCarty

6. Electronic Editions for Everyone

Peter Robinson

Texte intégral

1. Books Defy the Digital Revolution

  • 1 A version of this paper was given by myself with Barbara Bordalejo in the Digital Texts seminar or (...)

1In January 2004, I gave a lecture on electronic scholarly editing at the University of Virginia.1 At the beginning of the lecture I asked the audience, of around 60 people, three questions. The first question was: who among them had bought a movie on DVD in the last year; who had bought a piece of music on CD-ROM or by download in the last year; who had taken digital photographs? Almost everyone in the audience had done all three. The second question: who in the last year had bought an electronic book? Only three people – around 5% of those present – had done this. The third question was: how many people had bought a conventional, print book? Everyone had done this.

2One could guess the answers to further questions. We may doubt that as many as three people in the group had bought a film on videotape, or music on vinyl record, or taken a photograph using film. In film, music and photograph, the digital triumph is near complete. Yet books – despite the frequent proclamation of the ’death of the book’ – remain stubbornly locked into non-digital, old-style, analogue printed and bound paper format. Gutenberg would recognize a modern book; Monteverdi would be puzzled by a CD-ROM (let alone an iPod). Even among a highly-select audience at the heartland of digital texts, the University of Virginia, just five percent of the audience bought electronic books, while everyone bought print books, and everyone bought digital films and music.

3Why have books defied the digital revolution? What implications does this have for those of us who are engaged in the making of scholarly editions in digital form?

2. No Problem with the Digital

4One response is to deny the premise: one could argue that it is misleading to base this argument on the relative numbers of people buying films, music and books. It could be argued that the correct measure is readership, not ownership. By this measure, it is certainly true that we daily absorb massive quantities of textual information, of the sort that used to be exclusively printed, via computer screens: think of emails, news sites (many set up by traditional print organizations, such as newspapers), academic journals, online resources of all kinds (encyclopaedias, search engines), and the daily working deluge of memos, notices, circulars, papers, which arrive in our electronic inbox each day – not to mention SPAM mail. But this argument only points out, even more sharply, the contrast between the relative failure of electronic books and the success of every other digital medium. It appears we have no problem with digital text: we read digital texts all the time, just as we listen to digital music and watch digital films. So why is it so few of us buy digital books? Nor is it that digital books do not exist: publishers and computer companies have spent fortunes on trying to persuade us to invest in specialized machines to read electronic books, and on converting printed books into the forms required by these machines (see the Appendix). Yet, none of these machines have achieved anything like acceptance, few of us buy electronic books, and airport bookstalls continue their thriving trade. The determination of millions of people to go on buying print books is even more striking when one considers that the text of very many of these books (indeed, of almost every book of any popularity published before 1900) is available free on the internet. So why do we – hundreds of thousands of us a year – go on buying copies of Pride and Prejudice when we could get the same text, very likely in identical wording, from the Web (indeed, from the University of Virginia) – or even, just find the copy we bought a few years ago, which is probably lying somewhere around the house. Compare, again, the situation with films and music. Film and music publishers each year have to spend a fortune on stopping people illegally downloading digital films and music, in order to force us to buy what they sell. But print publishers (who, nowadays, are likely to be film and music publishers also) have to spend very little on stopping illegal book downloads, while the public continues to flock to bookstores.

3. Media of Distribution and Media of Performance

5We can see, too, that there is a difference between books and both films and music. To enjoy films and music we need a player: the electronic file (or, in the pre-digital age, the film reel or the vinyl record) is firstly a distribution medium, requiring a separate performance medium. The performance medium might be a computer, an I-Pod, a Discman, or portable media player, or a home hi-fi or cinema system. The point is that distribution and performance media are separated in films and music, as they are not in print books. For digital films and music to succeed, they need only to provide a more efficient distribution medium – which, of course, the digital form does easily. One could phrase this in terms of ’compelling advantage.’ Simply as distribution medium, digital media in film and music have compelling advantages in terms of cost, fidelity, portability and convenience. Add to these advantages in terms of content (’add-ons’ such as deleted scenes, director’s commentaries, interviews, alternate endings), and add too advantages in the ultimate performance, in terms of accuracy and detail of sound and image, speedy movement to any part of the work, replay, freeze-frame and magnification, and it is no wonder that in film and music digital media have, in just a few years, driven their vinyl and tape predecessors to the edge of extinction. We note too that, in essence, the film or music on DVD or CD is the same film or music as on the outdated analogue. There are a few additions, and a few improvements, but these are not decisive. Even without these, the digital distribution medium itself constitutes a compelling advantage. There is no need for the film or music itself to be fundamentally rethought. It is sufficient just to present the same film or music: just better, cheaper.

  • 2 I here betray that I am writing as a reasonably comfortably-off citizen of the developed world. Wh (...)

6Let us now apply this distinction between distribution medium and performance medium to books. Firstly, let us consider distribution medium. For the great majority of books, digital methods carry little advantage in terms of distribution medium. Yes, digital methods might compress the book onto a computer chip, and might allow this to be sold for a few pennies, and might allow it to be instantly available, for download to your computer. But most books are already very small and convenient, and one could argue that a book on a computer chip is actually too small: how would you shelve it? (and, reader, consider too how many flashsticks – portable computer storage on a chip – you have lost). Further, books are already cheap: the difference between the few pounds most books cost and a few pence is not a compelling advantage.2 Finally, books are already widely available: not just at bookstalls, or newsagents, but increasingly at supermarkets, pharmacists, anywhere selling anything – not to mention the immense success of online booksellers such as Amazon. There is an exception to this in the case of large, multi-volume publications: dictionaries, encyclopaedias, catalogues, scholarly journals etc. In these cases, the printed volumes are so large, and usually so expensive, that digital distribution constitutes a compelling advantage. There is a rule of thumb here: if the computer on which you want to read the book is smaller and more convenient than the printed book, then you will very likely read the book on the computer. Now, even quite small encyclopaedias are larger than laptops, and hence their disappearance from the bookstalls, while most scholars with appropriate access will routinely read scholarly journals online, not in print (or, will print from the online version and read that). However, these are exceptions: for most books, digital distribution is not a compelling advantage.

7Secondly, as a performance medium, most printed books are far superior to any digital form of those books yet in existence. It is a cliché commonly repeated that one cannot read a computer in bed (or the bath, or on a beach). As with most clichés, this statement is much more than halftrue, much more than half the time. Inescapably, we must conclude that for many kinds of book – literary and popular fiction, biographies, selfimprovement or religious tracts, all kinds of specialist magazine – people prefer print books to electronic books for most forms of recreational reading. Indeed, we should not be surprised by this. The modern print book builds on five hundred years of refinement since Gutenberg, and indeed on more than a thousand years of the codex before that, and on many centuries of the roll before that. The major effect of the technological revolution on book publishing has been to bring this development to a new perfection, so that books are better printed, more attractive, and cheaper, than ever before (see the Appendix). Further, advances in paper, font, print and binding technology mean that publishers can tune the physical book to its content as never before. Reading a book is more pleasant (and cheaper) than ever before. Compared to such rich diversity, books on computer screens are bland and monotonous.

8In summary: books are both distribution medium and performance medium. They both carry the text and allow it to be read. For electronic books to succeed they must have an overwhelming advantage in either or both of the distribution medium and the performance medium. Electronic books have a marginal (if any) advantage only as a distribution medium, and they are at a considerable disadvantage as a performance medium. Considered in this light, it is not at all surprising that print books are prospering in the digital age. Indeed, it is only surprising (and, a measure of the blindness of technological hyper-enthusiasm) that so many otherwise sensible and capable people should have thought that electronic books could completely replace print books (see the Appendix).

4. Scholarly Editions, Then and Now

9So far I have been considering the general case of all kinds of text-based publication. Our primary concern, as textual scholars, is with scholarly editions. Given the context I describe above, how might scholarly editions fare? In the early days of the academic discovery of digital text (a whole decade ago, around 1995) the answer seemed clear. For scholarly editions, the future – then – was digital. Articles by myself, Jerome McGann, Susan Hockey, and others, laid out what seemed to us – then – to be the compelling advantages of digital editions (McGann 1995, Hockey 1996, Robinson and Deegan 1994). We could include all the material – all the images, all the transcripts, all the collations – which print editions have to leave out. By the magic of hypertext, we could do away with all the cumbersome mechanics of footnotes, appendices, front and back matter: everything the reader could possibly want could leap out at a mouse movement. In this analysis, digital scholarly editions would be superior both as distribution medium (all that information on a single disc!) and as performance medium (hypertext!). How could they not succeed?

10Looking back, we failed to see some critical factors which might have undermined our confidence. Firstly: it appears that rather few readers (indeed, rather often, only the editors) actually want to see all the images, all the transcripts, all the collations. Traditional print editions acted as filters, straining out all this information so that readers did not have to see it: if readers do not want to see it, then including all this is no advantage at all. Secondly: as with other kinds of print book, scholarly editions in print form have developed highly sophisticated means of both compressing and expressing information. A superlative example of this is the Nestle-Aland edition of the Greek New Testament (Nestle et al. 1993). The 27th edition of this packs into a single volume, small enough to fit into a largish pocket, all the information anyone working to a high level of detail with the tradition of the Greek New Testament is likely to need: a full record of all significant variants in some thirty key manuscripts, including even marginal and other corrections found in these manuscripts, and a conspectus of key variants across the versions of the New Testament in other languages and in patristic and other citation (Nestle et al. 1993). Of course, one has to learn to understand the highly-abbreviated forms used by the Nestle-Aland, but each year thousands of students do just that. On this analysis, the advantages of digital editions as distribution media (people do not want what digital editions can distribute) and as performance media (print editions are really rather good) are less compelling than we supposed.

  • 3 Up to August 2007 Scholarly Digital Editions had sold 38 internet licenses of the Parliament Rolls (...)

11Further, we failed to reckon with some other points. There is the demon of copyright. Some of the most exciting digital edition projects focussed on modern authors. It can be difficult enough gaining permission for print editions for these; for digital editions, in some notorious cases, it has proved impossible. But even for older texts, where there should be no copyright issues, there have been problems. Arranging for digital photography and reproduction rights is, with very rare exceptions, arduous and too often forbiddingly expensive. Further, co-operative projects involving different scholars at different institutions can founder on copyright, as individual scholars or institutions (for whatever reason) attempt to enforce control over what they see as ’their’ work. There is also the sheer expense and difficulty of making digital editions. For years, the audience for scholarly editions in any form has been shrinking. In this context, the commitment of resources necessary for making an electronic edition is difficult to justify – and indeed, is likely to be justifiable only in a very few cases, usually of flagship texts by flagship authors. Finally, there is the conservatism of our community, which might long continue to use print forms even where the advantages of digital forms are decisive. There is a rather spectacular instance of this conservatism in the case of the Parliament Rolls of Medieval England, on which I advised. This is now available in two forms for institutional use. It is available in print, in sixteen rather large volumes, from Boydell and Brewer, at a little under £2000 a set (Given-Wilson 2005a). It is also available in electronic form, from Scholarly Digital Editions, by internet license to institutions for approximately half the price (Given-Wilson 2005b). By any reasonable standard, the internet version is superior to the print version: it is easier to use, it consumes no valuable shelf space, it can be used by any number of students and scholars, anywhere in the university, anytime. Yet up to August 2007 around three times as many institutions have opted for the print version as have subscribed to the electronic version.3

12This appears to be leading us to a rather depressing conclusion: that the effort of making scholarly editions in digital form is not worth the candle. Is there a future for this work? Is there a place for digital methods, in the making of scholarly editions?

5. Imagination, Powerful Tools and Familiar Objects

13Here, we offer two answers to these questions. The first, brief, answer is: there is certainly a place for digital methods in the making of scholarly editions, even though the final mode of publication is print, not electronic. Scholars may use computers to gather, order and explore data, while making an edition. In particular, they may gather information on variation between witnesses and submit this to different analytic programs to gain a rapid and clear sense of the development of the textual tradition. We have ourselves used computers very effectively in this way, and we can expect that computers may continue to be so used. In this respect, scholarly editors may be little different to many other scholars, who use a wide variety of computer tools (including standard email, word processing, internet and e-journal applications), but whose output goes through the normal channels of print. In this view, we would be using some new tools, and our scholarly editions will be better for them, but the editions themselves will not be very different from those scholars have made for centuries.

  • 4 ’Cuando escribo: ”El pájaro anida en el arco iris”, os presento un hecho nuevo, algo que jamás hab (...)

14This falls far short of the kind of revolution we imagined ten years ago, and far short of what we thought scholarly editions in electronic form might do. However, I think that the potential of scholarly editions in electronic form remains as great as scholars first imagined it, though I and others were naïve about the difficulties of realizing this potential. The potential remains, and there may be ways this potential can be unlocked. But I think it cannot be in any form which looks and behaves substantially like any printed book, or any scholarly edition we have seen so far. I am reminded of the Chilean poet Vincente Huidobro. He initiated a literary movement called ’creationism’ (completely unrelated to anti-Darwinian babble) about which he said ’when I say ’The bird nests in a rainbow’ I present you with a new fact, something that you have never seen, that you will never see, and that however you would really like to see.’4

15We could say that scholarly editions in digital form are rather like Huidobro’s bird nesting in a rainbow: a wonderful idea that we would all like to see. Yet, anyone who has seen a Dreamworks film knows that in Hollywood – and so in every movie cinema in the world and in every home with a computer or a television – birds can nest in rainbows. All you need is imagination, and some powerful tools. So here is our second answer: we can make scholarly editions which do all that we have dreamed. All we need is imagination, and some powerful tools.

  • 5 Shirley’s remarks are in the ’The prologe of the knyghtes tale’ prefaced to the copy of the Tales (...)

16What can we imagine, then? Consider, further, Huidobro’s bird nesting in a rainbow. The point of this is that we can imagine a bird nesting in a rainbow, because we know what birds are, what nests are, what rainbows are. These are familiar objects: it is the combination which is startling. One can extend this further, more generally, to the virtual world of the Web. Where the Web is at its best, transforming and enriching lives, it takes familiar objects and puts them in a new and liberating context. The world we call ’Web 2.0’ is particularly rich in examples. For generations people have written diaries, people have sent messages to each others, people have collected albums of photos and other personal memorabilia. These are familiar activities. But in the world of the internet, these familiar activities have been transformed into blogs, MySpace and FaceBook; all examples of ’social networks’ on the Web. At the very least, the countless hours spent by millions now on these activities have the virtue John Shirley ascribed to reading Chaucer: they prevent people doing worse things.5 But of course, they have far more virtue than this, as anyone who has spent more than a few minutes in these intoxicating environments can testify: they offer the sense of belonging, of being sustained by a myriad of contacts, built on near-instant communication in many forms. To take another example: for centuries, we have had all kinds of compendia of useful knowledge, usually in the form of encyclopaedias. In the Web, these have been turned into a new kind of compendium, the Wikipedia, written and endlessly rewritten by its readers. Here, the ordinary object has indeed been transmuted into something very new, an encyclopaedia with some of the characteristics of a social network, as its readers are also responsible for making and sustaining it. Recently, we have become familiar with scholarly web communities, such as the Digital Medievalist and Digital Classicist groups, which incorporate aspects of wikis, blogs, newsfeeds and other elements familiar from social networks into invaluable new environments for collaborative work. If we can create digital editions on the internet which have even a fraction of the impact of these extraordinary objects, then indeed we may have made a new bird nesting in a new rainbow. How, then, can we imagine these? We have some clues. Firstly, we should identify the fundamental familiar objects which will constitute our bird in a rainbow. Secondly, we should imagine how these might be combined in a new way.

17For we textual scholars, the first fundamental familiar object is this: knowledge about texts. We know where texts come from, how they were made and distributed; we know how their meaning is shaped by the circumstances of their creation and dissemination. Show us a book, or a manuscript: we can tell you about how they were made and the texts they contain, and about other books and manuscripts containing those texts. Show us a text, and we can tell you about the books and manuscripts which contain it, and how they differ, and what these differences mean. The second fundamental familiar object is the methods we use to discover what we know; the painstaking work of finding and analyzing the forms of the text. So we can offer anyone interested in texts two things. We can offer what we already know; and we can offer means to find what we do not know.

6. Mediating Textual Scholarship

  • 6 This was on 15 June 2001. The following day (Bloomsday), Gabler and Rose together examined the Eum (...)

18Most of the time, this knowledge, this method, comes to us in a second hand way through books, articles, notes, catalogue entries etc. Indeed, we have become so used to receiving knowledge of texts in this way that it seems to us that textual scholarship is, actually, printed textual editions. But of course it is not: the print editions are simply the way in which most of us receive textual scholarship, most of the time. Textual scholarship precedes print editions, as what textual scholars know and do. Usually, we know textual scholarship only in severe print. But it does not have to be this way. In 2001 I had the extraordinary experience of being in the National Library of Ireland with Hans Gabler and Danis Rose as they examined a page of the Joyce’s draft of the Circe episode of Ulysses, then on display there.6 This is a very difficult page to interpret, and the two scholars fell into an intense discussion on exactly what Joyce wrote, in what order, how he changed it, why, and how these writings connected to the rest of the novel. Not only was I fascinated: in a few moments everyone in the gallery had gathered around Gabler and Rose, to hear what they had to say. One can suppose that few of these onlookers would think of reading (say) any of the Joyce editions made by Gabler or Rose, but that day they were rapt (Gabler 1984; Rose 1997).

  • 7 The James Joyce Archive has, in 63 volumes, facsimiles of almost all Joyce’s notes and drafts for (...)

19This incident suggested that not only are there other ways of communicating textual scholarship beside print, but that print may not even be the best way to communicate textual scholarship. Consider how one might read (say) the first page of Joyce’s Ulysses. You might have your reading text; then you might (if you are fortunate) have a copy of Gabler’s and Rose’s editions close by; you might also have the relevant volumes of the Joyce Archive, with copies of the key Rosenbach (1975) manuscript, and of various typescripts and galleys.7 These could be spread out on a large desk, open at various pages, with an apparatus of bookmarks pointing at other pages. It has been a commonplace of commentary on hypertext editions that the advantages of the digital medium could allow any reader to have access to all this material, with a rapidity and convenience never before possible. But I think our early attempts to do this have foundered, on exactly the lack of imagination which cannot see a bird nesting in a rainbow. What we have done in the digital form is mimic the way all the materials on this first page might be accumulated in print form. For example – to change the example from Joyce to my speciality, the Canterbury Tales – we have made digital copies of single manuscripts, with transcriptions, collations, descriptions and analyses (Canterbury Tales Project, Robinson 2003b). This is a rather comprehensive collection, but it is still only one collection, of only one aspect of the Tales, presented from only one point of view. There are many other materials one might want to consult which we have not included. Particularly, one might look for a tale-by-tale, line-by-line commentary, such as is found in the Variorum volumes. One might also wish to examine other editions, beside our own, and hunt down discussions of particular lines by various scholars.

20So far, digitization (where it has happened, and there are many areas where it has not) has presented all these disparate resources as distinct from each other. Just as the scholar in the library must find the various volumes he or she wants and lay them out on the desk, ready to consult, so the digital scholar has to locate the various resources he or she wants, work out how each one works, and spread them across the digital desktop. There are advantages, in speed and convenience, but there are also significant disad- vantages. The greatest is that while all books work in much the same way, this is emphatically not true of digital resources. We have all experienced the sinking feeling, when opening up the latest grand digital resource, that there are wonders here which cannot be found: or, even more frustrating, having found them once, being unable a few weeks later to find them again. Further, the cataloguing of digital resources is uneven, at best: a search engine is not a catalogue, but that is what we have to use most of the time. So, we are looking at the first line of the Canterbury Tales, in the Hengwrt manuscript. We have the splendid Estelle Stubbs edition of this, which includes a collation of every line of Hengwrt with the Ellesmere manuscript (Stubbs 2001). We ask ourselves: is there somewhere, out there on the net, an alternative transcription to this line, as given by Stubbs? There is: in the University of Michigan Corpus of Middle English Prose and Verse. We ask ourselves: is there somewhere an image of the first page of the Ellesmere manuscript? There is: at the Huntingdon Library website, in San Marino, California. We know that Norman Blake edited the Tales from the Hengwrt manuscript. Is there an electronic version of this first line, in that edition, somewhere? There is: at the Oxford Text Archive (though, unfortunately, with restricted access). We could ask: are there electronic versions of other editions, including this first line? There are: for instance, at the University of Virginia, and in the Michigan Corpus of Middle English Prose and Verse. It is up to the reader to find all these.

21From this sketch, one can see how far digital scholarship remains bound by the model of print: each scholar (or group of scholars) makes a separate digital object. It is up to the individual reader to locate all these separate objects, and make sense of them. In the worst cases, this is like one of those wonderfully circular games: you can only find what you want if you know what it is you want and where it is. As most readers know only the text in front of them, they are unlikely to find the riches lying about them, and are unlikely to be able to make sensible use of them, if they do find them.

7. Imagining Otherwise

22It is time to imagine a bird nesting in a rainbow. The way we have it now, all these transcripts, descriptions, facsimiles, commentaries, translations wait passively on the Web for someone to find them, just as print books wait in a library for someone to read them. Suppose instead that they did not wait, but they went out over the Web to find someone who was reading something closely related to them. Imagine you are reading the first line of the Canterbury Tales. In Michigan, a computer server thinks: that person is reading the first line of the Tales. I have a transcription of the first line, and it sends a message to the computer on which you are reading to that effect. Simultaneously, in Oxford, in Virginia, in San Marino, other computer servers realize they have something relevant to what you are reading, right now, and they too send messages to your browser: we have something for you. Instantly, on your browser, a message appears, indicating that there are relevant materials waiting for you, at all these sites. With a little intelligence, the materials could be sorted, into transcripts, images, commentaries, etc, and the message on your browser tuned (or removed altogether) to reflect just what information you are interested in. This is the equivalent, if you like, of books in the library shelves realizing that you are reading Chaucer, and unobtrusively shuffling off the shelves onto your desk.

23Does this sound like science fiction? It should not. Something like this already works every time you do a Google search: the links in the margins of your screen are generated by Google itself, calculating from your search just what you are interested in and creating links accordingly. Every time you log into Amazon, you will see a list of books which Amazon has chosen for you, based on your past purchases. There are many other instances. All these are examples of a key aspect of ’Web 2.0’: expressions of the principle that web applications should not simply wait for the user to do something, but should try to anticipate what the user might desire and should set to drawing together resources to satisfy that desire.8

24This vision has many implications for readers. In our study, or our library, we take books from the shelves and spread them on the desk. In the digital world, the reader may choose from the resources supplied by servers around the world, and spread them out, as he or she chooses, on the digital desktop. Again, we can see something like this in personalized pages such as that offered by Gmail, where you can populate your live desktop with newsfeeds, reports on the weather, stock market updates, clocks, and more. In the examples I have given so far, I have spoken of access via the text: through the first page of Ulysses, or the first line of the Canterbury Tales. Equally, access could be by the source object. I am looking at a digital image of the first page of the Hengwrt manuscript. Around the world servers gather material relating to that manuscript (descriptions, discussions, commentaries), to that page (digital images, observations), and to the text on that page, line by line and word by word (including, other versions of that text on that page). All this is then available for the reader, to explore as he or she wishes.

25There are implications for scholars, too. Up to now, if scholars wished to make a digital edition, they had to do everything, from initial encoding to final publication. In particular, they had to create a publication interface, so readers around the world could read what the scholars had made on their computer. This required access to considerable resources, usually only available to a few well-funded projects, and in turn meant that each of the editions so made were digital islands, entire in themselves and separate from others. In the vision here offered, scholars have only to make the individual components of the edition: the transcription of this line in this manuscript; the collation of the various versions of these words in this line; the commentary on this line etc. These are then placed on servers, from where they are harvested into each reader’s browser. This removes entirely the need for the scholar to provide any kind of interface (though they still might, if they wish). The effect is to open up scholarship: no longer will scholars need to be part of an elaborate and expensive project to contribute to digital scholarship. This raises further questions: if anyone can make a transcript of the Hengwrt manuscript and put it on the Web, how will the reader choose which of these to read? The restriction of scholarship to the academy provided forms of control and quality assurance; do we need these; how could they be provided if we do, and if we do not need them, how will we fare? And where will publishers be in this world?

26We said above that to make the bird nesting in the rainbow, we would need imagination and powerful tools. Here is what we have imagined. What tools will we need to make what we have imagined? In fact, most of the tools already exist. There are many protocols, in the form of ’web services’, which allow computers to talk to computers, and then send information back and forth between them, exactly as described here. It is now straightforward to embed commands into browser interfaces to access these web services. Thus: you start reading a text, any text, of the Canterbury Tales. As each line of the text is loaded into the browser, a command is triggered: find materials related to this line. The browser broadcasts a message: what do you have for me? Around the world, servers are waiting for such a message. They send replies; the browser gathers the responses, and adjusts the display accordingly. The technical infrastructure for this is widespread. However, we do lack a key component. We need agreed referencing schemes, so that the browser knows exactly what it has, and servers recognize that they have relevant materials. Again, the fundamental encoding infrastructure for such agreed referencing schemes has been laid down by the Text Encoding Initiative in general, and more particularly by the Canonical Text Services project.9 These are not yet worked out in sufficient detail to support applications of the sophistication here described, and this is work to be done.10

8. Realizations

27In a previous article, I characterized this model of editions as ’fluid, cooperative and distributed’ (Robinson 2004). Some key elements of this model can be seen in the ongoing work on the digital edition of the 28th Nestle-Aland Greek New Testament, on which I am advising the Deutsche Bibelgesellschaft and the Münster Institute for New Testament Research.11 This is described in more detail in an article I have written for Ecdotica (Robinson 2007b). In summary, this is arranged so that for any verse of the Greek New Testament, the reader may choose what elements should appear on the screen: the Nestle-Aland text; the conventional apparatus, as printed; a new digital apparatus, providing more detail for key manuscripts; full transcripts of the key manuscripts; a German-Greek dictionary; and an ’explanation’ window that gives a brief summary of the symbols and conventions used by the edition. In the Nestle-Aland work, all the elements are being provided by a single editorial group. However, the design of this work is such that it could be opened out by allowing other elements to be added to the desktop which are not provided by the editors (for example, different dictionaries for different languages, different commentaries). In the reverse of this, elements from the Nestle-Aland could be exported to other interfaces. This deconstruction of an edition into separable parts, possibly held on different servers, each capable of being made by separate scholars, and each capable of being linked with others in an infinity of ways, takes us a considerable way towards our vision.

28I believe that this model of editions provides the compelling advantage which will (at last) lead to readers routinely consulting digital editions online, and to large-scale creation of digital editions for readers. For readers this will provide immediate and convenient access to information about the text they are reading, with the ability to order, filter and arrange this information onscreen as they wish. For scholars this will remove the need to build complex publication interfaces for the editions they make. Rather, they can concentrate on the intellectual work of making transcripts, collations, analyses, commentaries (indeed, just as they would have done for print publication), leaving it to the well-organized magic of the Web to bring these to the readers. Such editions will be our bird nesting in the rainbow: familiar things, in a remarkable combination. They will be electronic editions for everyone.


29The first part of this article advances the premise that books that we read (such as novels, biographies, recreational literature), as opposed to books that we reference (encyclopaedias, dictionaries), have failed to cross the digital divide. Publishers produce books in print; we buy them and read them in print, much as we have for six hundred years.

30There have been determined, and very expensive, efforts to break this pattern, so that printed books would go the way of the vinyl record. There was a particular flurry of activity around the turn of the millennium. In a 1999 article written for The Economist, with the somewhat menacing title ’Beyond Gutenberg’, Bill Gates promoted the advantages of eBooks, without going quite so far as to suggest that they would drive print books out of existence (article available on​presspass/​ofnote/​11-19billg.mspx, accessed 14 August 2007). (Gates is rather fond of declaring obsolescence: besides books, a Google search for ’Bill Gates obsolete’ shows that he has declared television, DVDs and US High Schools obsolete.) On 6 January 2000 Barnes and Noble and Microsoft released a press release proclaiming that ’the sale of electronic books will reach US $1 billion in 3-4 years, and could overtake the sale of print books in ten years.’ (Reported in Wired, 6 January 2000;​techbiz/​ media/news/2000/01/33488, accessed 14 August 2007). A 2001 review article on eBooks and eBook readers by Sarah Ormes in the online journal Ariadne, 10 January 2001 (​issue26/​e-book/​, accessed 14 August 2007) lists a range of e-book providers and technologies.

31In retrospect, Ormes’ article marks a high-point of what now appears as misplaced confidence (though even in this, she strikes a note of caution: ’But finally these electronic dreams seem to be about to come true’: note the ’seem’). What did happen to the Barnes and Noble/Microsoft prediction? Industry sources (​notices/​ebooks.htm, accessed 14 August 2007) suggested that in 2002 500,000 eBooks were sold in the US. While this number is not negligible, it should be compared to the 1.6 billion print books estimated as sold in that year. The same site estimates eBook sales in February 2004 as $900,000, and cites a figure of $2,591,469 for eBook sales for the third quarter of 2003. This would give an annual figure of around $10 million for 2004. According to the Barnes and Noble/ Microsoft prediction, electronic book sales in 2003-4 should have been $1 billion, not $10 million. Barnes and Noble ceased electronic book sales in September 2003 (​2003/​10/​electronic-bookweb-, accessed 14 August 2007).

Electronic book readers have fared no better. The Rocket reader, in 2001 the market leader according to the Ormes article, was taken over by Gemstar, to become the Planet eBook reader, which appears to be no longer available: the Planet eBook site now does not mention the reader at all, and the company appears to have adopted Adobe pdf technology (//www., accessed 14 August 2007). Franklin Electronic Publishing (, accessed 14 August 2007), who created the eBookman, still exists, though the device does not.

32One cannot say that the failure of electronic books is due to a lack of supply. There are a very large number of books available in electronic form, to be read using software on general computer devices (some of the software is tuned for particular platforms, such as Franklin’s Mobireader, intended for mobile devices). Franklin lists over 25000 fiction titles and nearly 12000 nonfiction titles on its website, available for individual download at prices generally somewhat lower than the equivalent print book (e.g.: the Oxford Advanced Learner’s Dictionary is $29.95 on Franklin, $34 on Amazon, against an Oxford list price of $46; The Best American Erotica 2006 is $9.99 on Franklin, against $11.20 on Amazon and a list price of $14. It is revealing, however, that from the Franklin web homepage it appears that this company concentrates on ’niche markets’ (particularly of language students, for whom they offer a range of handheld dictionary readers) rather than on their general eBook offerings. Many more titles are available in Netlibrary (over 100,000 titles, according to, accessed 14 August 2007), but Netlibrary appears to concentrate on supplying the library market rather than retailing direct to consumers.

33The experience of Microsoft appears particularly relevant. Microsoft continue to offer their eReader software, able to run on any handheld using the Microsoft Pocket PC standard. The Microsoft Reader website (http://​reader, accessed 14 August 2007) offers outlines of popular titles and an extensive catalogue of eBooks (over 29000 fiction titles – The Da Vinci Code is third on their list, at $5.94). Again, prices for books listed on the Microsoft site are only slightly lower for eBooks than for print books: Nancy Gibbs’ The Preacher and the Presidents: Billy Graham in the White House is listed as an eBook for $16.14, against a new price listed of $26; Amazon had the title at $17.81. For both Franklin and Microsoft, however, one has to dig rather deep on their websites to find the electronic book catalogues. It appears that electronic books are ancillary to the main business of both companies. One could draw a similar conclusion from the way in which Amazon treats electronic books (at least, for on 10 September 2007): where electronic books are available, their presence is signalled as a note against the listing for the print book, rather than as saleable items in their own right. There is a section ’Digital downloads’ on but this leads only to downloads of films and music, not books (10 September 2007).

34One could summarize this state of affairs as follows: there are a very large number of eBooks available, able to be read on the most-widely used computer platforms. However, there appears no imminent likelihood that these eBooks will completely supersede print books. They appear to coexist with print, targeting a niche market of particular users, rather than the general reader. (Since first drafting this Appendix in August 2007, the launch of the Amazon Kindle reader has dramatically increased the profile of eBook readers. However, the determinedly print-like presentation of the books on the reader, and the marketing of print and eBook versions together, suggest co-existence rather than replacement.)

35On the other side of the ledger, there is plentiful evidence to document the thriving state of printed books. Statistics gathered by the International Publishers Association (IPA) suggest that in the period from 1990 to 2000, the number of individual titles published across the world increased by around 50%. The figures offered by the IPA are incomplete, with some countries not providing data for all years. However, there are comparable statistics for a few typical countries: in the UK, individual titles published grew from 63756 to 110155; in Spain, from 42207 to 60426; in the USA from 46743 to 64711 (see​statistics/​annual_book_prod. html, accessed 13/8/07). For the period after 2000, Bowker reported that book production in 2004 showed a 14.1% annual increase in titles published, with a 43% increase in fiction titles. Even university press output – often cited as a problem area for academic publication – was higher, at a 12.3% increase (Press release May 24, 2005; see​press/bowker/2005_0524_bowker.htm); the Bowker figures, available on their website, for 2006 and 2007 suggest a slight decline from this all-time high; see also the cautionary notes concerning the difficulty of counting titles at​2007/​08/​09/​british-book-production/​, accessed 14 August 2007. There is an extremely useful summary of statistics relating to world-wide publishing, with links to websites and comments, at​sites/​para/​resources/​statistics.cfm (accessed 14 August 2007). Overall statistics on the number of publishing houses (for example, on the ’parapublishing’ site) give a similar picture to those on individual titles published. Statistics on print runs and on actual books sold (let alone read) are harder to come by, and do not appear to be available in any systemized form. It is possible the overall increase in titles published in the last decades has been accompanied by smaller printruns, with the rise of ’print-on-demand’, so that fewer books overall are sold but a larger number of titles. The ’parapublishing’ site gives somewhat anecdotal information on actual numbers sold (e.g. 1.6 billion books sold in the US in 2001, around 7 for every person in the US). Overall, these figures hardly suggest an industry in decline.

36Useful articles on this subject are Robert Darnton’s New York Review of Books article ’The New Age of the Book’ (46.5, March 18, 1999;​articles/​546, accessed 14 August 2007) and Mark Moss’ The Future of the Book’, College Quarterly, 7.3 (2004), at​quarterly/​2004-vol07-num03-summer/​moss.html, accessed 14 August 2007.


1 A version of this paper was given by myself with Barbara Bordalejo in the Digital Texts seminar organized by Willard McCarty at the University of London in March 2007. Although I am credited as the only author of this paper, much of Dr Bordalejo’s contribution remains: particularly, the ’bird nesting in a rainbow’ trope. As so often, acknowledgement is scant justice to the debt I owe (and owe to many others, some mentioned in the citations in this article, with whom I have discussed digital editions over the years).

2 I here betray that I am writing as a reasonably comfortably-off citizen of the developed world. Where there is a desire for the book, but very little money, even a small difference in price might be critical: so that it is not surprising that a high proportion of downloads from the University of Virginia electronic text centre were, in 2001, to IP addresses in the less-developed world (personal communication, David Seaman).

3 Up to August 2007 Scholarly Digital Editions had sold 38 internet licenses of the Parliament Rolls of Medieval England. Boydell and Brewer printed 100 copies of the 16 volume edition, retailing at £1950, and had sold out by May 2007, and were printing additional copies to meet the demand (SDE company records; personal communication from Boydell and Brewer staff).

4 ’Cuando escribo: ”El pájaro anida en el arco iris”, os presento un hecho nuevo, algo que jamás habéis visto, que jamás veréis, y que sin embargo os gustaría mucho ver.’ Huidebro 1925; I owe the idea and the reference to Barbara Bordalejo (who, however, is not responsible for the development of the trope in this paper).

5 Shirley’s remarks are in the ’The prologe of the knyghtes tale’ prefaced to the copy of the Tales preserved in British Library Harley MS 7333. His words are: ’O. yee so noble and worthi pryncis and. princesse other estatis or degrees. what euer yee beo th[a]t haue disposicione, or plesaunce. to rede or here the stories of olde tymis Passed to kepe yow frome ydelnesse and slowthe. mescheuing other folies that might be cause of more harome filowyng’ [O ye so noble and worthy princes and princesses or estates or degrees: whatever ye be that have disposicion or pleasure to read or hear the stories of old times past, to keep you from idleness and sloth, mischief or follies that might be cause of more harm following] (Published in Solopova 2000).

6 This was on 15 June 2001. The following day (Bloomsday), Gabler and Rose together examined the Eumaeus notebook of Ulysses, on display in the Sotheby’s showroom in Dublin prior to its auction, again before a large and impromptu audience. The notebook has since gone into private hands, giving extra significance to this public examination (Private communications, Hans Gabler 7-10 September 2007).

7 The James Joyce Archive has, in 63 volumes, facsimiles of almost all Joyce’s notes and drafts for the novels and other creative works, and was published by Garland in 1977 and 1979 (Groden, 1977-79). Only 250 copies of this were published and these are now rare. There is an excellent account of the making of the archive (and other matters) in Michael Groden’s ’Perplex in the Pen – and the Pixels: Reflections on the James Joyce Archive, Hans Walter Gabler’s Ulysses, and ’James Joyce in Hypermedia’ (Groden 1998). The major exclusion from the James Joyce Archive is the Rosenbach manuscript of Ulysses, published in three volumes of facsmile in 1975 (Rosenbach 1975).

8 See the Wikipedia definition of ’Web 2.0’, [accessed 15/8/2007]. The term was coined and given currency by O’Reilly Media: see [accessed 15/8/2007].

9 See and respectively [accessed 12/3/2010].

10 A draft paper prepared by the author on the encoding and technical infrastructure needed to realize what is discussed in this article was placed at [accessed 12/3/2010].

11 The ongoing work, and the final edition, may be seen at [accessed 12/3/2010].


Peter Robinson is co-director of the Institute for Textual Scholarship and Electronic Editing, University of Birmingham, textual editor specializing in Chaucer, software developer and founder of Scholarly Digital Editions. Since the 1980s he has pioneered digital techniques for editing, most notably the use of philogenetic software applied to the study of large textual traditions. Currently he is working with the Institute for New Testament Textual Research, Birmingham, on both the Nestle-Aland 28 and the Digital Nestle-Aland editions. His Canterbury Tales Project is a normative starting point for scholars interested in digital editing.


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search