Torna all'indice

Ecdotica n. 1 (2004)

Rassegne (1)

JOHN LAVAGNINO
ON HYPERTEXTS

Paolo D’Iorio and Daniel Ferrer, editors, Bibliothèques d’écrivains, Paris, CNRS, 2001, pp. 256

María José Vega, editor, Literatura hipertextual y teoría literaria, Madrid, Mare Nostrum, 2003, pp. 288

Jerome McGann, Radiant Textuality: Literature After the World Wide Web, London, Palgrave, 2001, pp. 272

«I have been up to London to get the book I am writing, out of the British Museum. I have got a lot of it out, and I shall go again presently to get some more; and when I have got it all, there will be another book». She slung a strap of notebooks off her arm, and advanced to the fire with the smooth, unswaying motion of a figure drawn on wheels. «So many people were there, getting out their books. It doesn’t seem to matter everything’s being in books already: I don’t mind it at all. There are attendants there on purpose to bring it to you. That is how books are made, and it is difficult to think of any other way. I mean the kind called serious: light books are different.Mine ought to be quite a success. It will be just like the ones I am getting it out of, and they are standard books. I put things from several into another, and then it is called a biography. What have you done to-day?»

This working method described by Charity Marcon, in Ivy Compton- Burnett’s novel Daughters and Sons, has its literal counterpart in reality: Johann Joachim Winckelmann wrote an autobiography this way, a sequence of quotations from classical literature strung together to describe his life. In Bibliothèques d’écrivains, studies of ten writers and their libraries edited by Paolo D’Iorio and Daniel Ferrer, we learn in detail about Winckelmann’s approach to making books out of other books, and about other, quite different ways of using books. The focus is on these working processes as much as on the actual collections, and most of the essays have a perspective informed by critique génétique. But the studies are valuable to anyone with any sort of interest in the origins of books; there are discussions of sources for individual works that editors and critics of those works would need to know about, but together the studies also demonstrate the great diversity of possible relationships between books. Editors and critics know that they can never really study any single work in isolation; what we always need to be reminded of is that there are more ways for a text to derive from other texts than we can imagine.

In an excellent introduction, Ferrer points out a distinction that becomes very evident as you read these essays, between two kinds of writers: the extractors and the marginalists. Extractors, such as Winckelmann and Flaubert in their very different ways, pull chunks of material from books and store them away for later use in constructing their own writings; in both of those cases the structure of the intermediate notebooks containing these materials has an influence on the structure of the final book that’s produced. The extractor might assemble a substantial personal library, as Montesquieu did, but the notebooks still have a more important place: indeed, Winckelmann owned very few books and showed little sign of wanting to acquire more.

Marginalists carry out their interaction within their books, rather than extracting material: and their marginal notes are not limited to annotations, but may include any kind of writing, not necessarily closely linked to the text. Stendhal wrote what amount to journal entries in the margins of his books. Marginalists are much more likely to care about their libraries, because they’ve written some of the material in them. Where the extractor may be led to destroy books (by extracting pages to paste into notebooks), the marginalist merely defaces them.

Of course some writers interact with their books in both ways:Valéry, for example.Ownership of substantial libraries and posthumous preservation of annotated books grow more common in recent periods, too, so the picture becomes more complicated. And you can also note that writers are involved with other books to differing degrees: perhaps few writers will have nothing to do with other books at all, but of those considered here Robert Pinget seems to have had the weakest relationship with them. There are significant sources and books that mattered a lot to him, but on the evidence here you would also judge that he could probably have written a lot without any library at all.He did a bit of annotation (and also added his own illustrations and decorations to a few books), but apparently no extensive extraction; and it does seem to be the extractors who are most heavily dependent on books, to the extent that their writing can at times involve more selection and arrangement than composition. Although there are marginalists whose notes have been raised to the status of works by being collected and published (Stendhal, Schopenhauer), those are still not the works for which they’re principally known; these studies offer no example of marginal composition of a major work, or even of marginalia developing into one. The marginalist’s activity may be important to his or her thinking and writing, but it may also represent an urge to respond that’s peripheral to real work: it could just be the equivalent of talking back to the TV. But though it may be less consequential, it often gives us more personal insight: the notebooks of the extractor can seem very mundane and practical, sometimes with little individual comment, whereas the marginalist’s stance is very likely to be personal and responsive (as in Schopenhauer’s various expressions of disagreement and spleen recorded here).

The extractors discussed in Bibliothèques d’écrivains are: Winckelmann (by Élisabeth Décultot), Montesquieu (Catherine Volpilhac- Auger), Flaubert (Anne Herschberg Pierrot, Claude Mouchard, and Jacques Neefs),Woolf (Daniel Ferrer), and Joyce (also Ferrer). The marginalists are: Stendhal (Hélène de Jacquelot), Schopenhauer (Sandro Barbera), Nietzsche (Paolo D’Iorio and Frank Simon-Ritz), Valéry (Judith Robinson-Valéry and Brian Stimpson), and Pinget (Jean-Claude Liéber and Madeleine Renouard).When you look at the subjects in detail, they aren’t quite so neatly clustered, though; within broadly similar approaches there are of course many individual differences among these authors. But there is a further set of similarities in the analytic methods that the contributors adopt.

One problem all must address is that of understanding the surviving library (or the collection of extracts derived from it).Many of the essays make efforts to convey an impression of the overall nature of the library: by talking about its scope, listing subjects and authors included, and pointing out unusual features. That sort of characterization generally works well; it’s at the level of detail that the scholars are really pushed to great effort.How much does the ownership of a particular book signify? It doesn’t mean that the author read it or even chose to acquire it. Do marginalia give us an index of importance? Maybe, but it’s still only what happened to get recorded at particular moments, and the marginalist could have thought differently at another time; and a book marked up extensively for practical ends might matter to the author much less than a book read over and over again but unmarked. Samuel R. Delany observed (in The Motion of Light in Water) that he felt more influenced by some writers he hadn’t read through than by the books he knew well: partial knowledge left the imagination more free to expand. Of Valéry we read here that he frequently sampled a book rather than reading it through: but skipping itself doesn’t have a fixed significance. The argument here is that his skipping doesn’t reflect disagreement or dissatisfaction, but rather his feeling that he had seen what the book was getting at. (Infuriatingly for the scholar, he also sometimes peered into pages not yet cut open, so the pattern of opened pages isn’t completely dependable evidence of what he looked at.)

So a body of evidence like a library collection may look like a set of individual bits of information that can be examined separately; and sometimes a single surviving book offers very clear material evidence that it was a source (there are significant instances of that in Nietzsche and Pinget described here, for example).But in practice it frequently requires care and judgment to weigh the evidence, and the way things matter to different authors will vary. One reason that libraries are more valuable as evidence than individual books is that there is the possibility of seeing notes and marginalia in the context of a writer’s typical practice: so we learn it’s significant that Schopenhauer did not put a bookplate into his copy of Hegel’s Phenomenology of Mind.

These difficulties are instances of a broader problem for scholarship, that all information is insufficient. No matter how copious our material, we will always find some questions unanswered and will wish for further documents and data. And nearly every contributor here has a story about sloppy record-keeping or the needless loss of materials. One of the few observations I can make that doesn’t seem to have been quite specifically anticipated here is that these libraries show a pattern that occurs with books generally, that the bigger,more serious, more expensive book survives better. Nobody discussed here seems to have had very much of a collection of trivial or frivolous literature; the closest you get is the fascinating information that Nietzsche owned a book of stories by Bret Harte, the American writer of frontier tales. (Sadly, his sister lost it in Paraguay.) We don’t hear about any cookbooks or travel guides or stacks of old magazines. And a completely intact library would still be incomplete, as the essay on Valéry makes clear. It has an extended case study of Valéry and Dante, in which some crucial evidence of Valéry’s interest comes from his marginalia in someone else’s copy of Dante – marginalia he only added at the owner’s request. Visiting Anne Quellennec he several times asked to see her copy of Dante so he could read particular passages he had in mind; those he marked in her copy aren’t singled out in the same way in his own.

Every one of these studies moves from descriptions of the substantial evidence offered by writers’ libraries to the further stage of making arguments about the connections between pieces of evidence and what they signify. And every contributor could say far more than will fit into an essay or a book. We hear of only one project to use nontraditional means for analyzing libraries and publishing arguments about them: the HyperNietzsche project plans to combine the scholarly publication of primary documents with the creation of a space on the World Wide Web for linking them together with critical commentary. Broader discussions of this hypertextual approach to writing and what it means for readers may be found in the collection Literatura hipertextual y teoría literaria, edited by María José Vega. This is the most recent survey of the field of hypertext theory and practice; it includes translations of essays dating back as far as 1990 and not previously available in Spanish, as well as a group of recent and new essays written in Spanish. The focus is firmly on the intellectual problems of understanding hypertext rather than on any technical questions of making it work; the key concepts are linking and nonsequentiality. The contributors of the earlier essays are: Stuart Moulthrop in two essays, on the importance of pushing for openness to diversity in hypertext systems, and on how close to breakdown and failure we typically are in such systems; John Tolva on the standard objections to hypertext,with sensible responses countering them;Marie-Laure Ryan on the way that virtual-reality systems can provide both immersion and interactivity, properties that are in conflict in textual works; Jürgen Fauth on the inferiority as involving narratives of works not wholly subject to authorial control; Anja Rau and J. Yellowlees Douglas in two different essays on what the actual reading experience of some specific hypertextual works is like; Robert Kendall on various temporal aspects of hypertext reading; and myself on hypertext editions. The contributors of the more recent essays are: María José Vega with an introduction that surveys the topic; Emilio Blanco on the canon and hypertext, noting among other things the tendency towards the formation of canons of hypertext fiction and criticism, despite the opposition to such hierarchies of most people in those fields; Enrique Santos Unamuno on the pre-digital tradition of nonsequential writing; Gonzalo Pontón on the genre of hyperdrama, like hypertext with optional branches but acted out before spectators; Laura Borràs Castanyer on hermeneutics and reader-response theory as contributions to an account of the experience of reading hypertext; Neus Rotger on a deconstructive perspective on texts that are open-ended, changing, and decentered; and María Morrás on the theoretical affinity between hypertext and scholarly editions, and the practical lack of progress in the field.

In its necessarily sequential presentation, the collection moves from considerations of how hypertext is related to the literary tradition and the canon, to hypertext in relation to traditional reading practices and to virtual reality, and then to more detailed critical studies and reflections on its use in scholarly editions.The principal points of contact with Bibliothèques d’écrivains are in the strong focus on individual authors and on modes of reading and study that move around within a body of materials related to one author.And both books justify their existence as collections by adding up to more than the mere sum of their parts. But where the study of an author’s library tends to bring you back to a consideration of the author’s activity, the essays here frequently look at the activity of readers in working with texts – whether by considering the history of nonsequential writing that required readers to play a more self-consciously constructive role, or by recording the reflections and initiatives of readers at work, or by thinking about perspectives on reading from literary theory and how they might apply to hypertextual literature. Books like Bibliothèques d’écrivains can say a lot about the variety of ways authors work and draw on the existing body of literature; one thing Literatura hipertextual y teoría literaria shows us, like much work on reader response, is the astonishing variety of ways in which people read.

Literary theory has never quite figured out how to make sense of the enormous diversity of reception, of the range of ways readers find to take things; it can make the case for diversity of reception as necessary and valuable rather than merely mistaken, but managing that diversity itself remains a problem. But it has been making efforts, and one value of the newer essays in the book is that they broaden the range of theoretical approaches that are brought to bear. North American hypertext criticism around 1990 looked most often to Barthes and Derrida; the European contributions here broaden the range to include Gadamer, Jauss, and Iser, and it’s a welcome development. There is also a broader literary background: hypertext criticism has always discussed print precursors, and that happens here too, but there is reference at several points to the experiments of the Oulipo, which are rarely mentioned in the North American tradition. In that tradition, a work of hypertext fiction may be split into many small chunks of text which the reader may see in many different sequences, but the chunks of text are fixed: no random process is going to interfere with these carefully-crafted sentences. The criticism associated with that tradition has always stressed the reader’s freedom, but of course the authors never envisaged an unlimited freedom; the perspective of this collection helps clarify the distinctive nature of that body of writing.

The focus of the book is very much on hypertext fiction, and not on other literary genres: the book is very unusual in having an essay on hyperdrama, but there is little reference to poetry in digital forms. The focus is also very literary: there is little consideration of nonfiction. Or of the World Wide Web, the huge hypertextual system that is largely nonfictional and nonliterary: the book’s index has more citations for Gutenberg than for the Web. Of course, many of the contributions were written before the Web achieved its dominance, and it is a technically less advanced system than some used even in the 1980s by hyperfiction writers. But although Literatura hipertextual y teoría literaria is not primarily about hypertextual presentation for nonfictional purposes such as the presentation of scholarly editions, it does have importance for editors. The essay by María Morrás, «Informática y crítica textual: realidades y deseos», is a valuable meditation on why we haven’t seen more editions in hypertextual form, and on how we might achieve some progress. The essays as a group frequently consider the question of medium: of how the digital medium is different from print, and what the consequences are for us; and editors thinking of creating digital editions need to reflect on this. And the attention to the activities of readers is important too: should we be trying to model the diverse interconnections within a body of writing, or just leave that to readers to figure out? A further question for digital editions is whether we should be trying to provide more ways of interacting or of recording reactions; what can we do to support the work of extractors and marginalists? Extractors have it comparatively easy in the digital world; the marginalists do not.

Jerome McGann’s Radiant Textuality: Literature After the World Wide Web presents a marginalist’s views on literature and computing and his ideas about where we should be going. This is another collection of essays from diverse points of view: although the chapters are all by one person (except one that’s a collaboration with Lisa Samuels), they were written over the span of nearly a decade and are presented so as to describe an intellectual journey rather than to cover it up. And one chapter is itself in the form of a dialogue, a form McGann has used for critical writings now and then for decades. Though the book is not designed to present a tidy picture, it’s easy to identify two recurring subjects: an editorial one (the presentation of literary works in digital form), and a literarycritical one (the possibilities for using computers to further our exploration of literary works). The balance shifts from the first to the second as the book progresses; or, in another sense, the book starts out feeling more like Bibliothèques d’écrivains and ends up feeling like Literatura hipertextual y teoría literaria, as the focus of interest shifts: from the migration of forms of traditional scholarship into the digital world and their transformation by it, to digital means of supporting critical reflection and commentary.

The chapter most devoted to editorial matters is «The Rationale of Hypertext», which began to circulate on the Internet in early 1994 but wasn’t published on paper until 1996. It makes strong arguments for the creation of scholarly editions in hypertext form that would include all states of the texts in question, not as apparatus but as page images or transcriptions, and in whatever forms were relevant: so that, for example, songs wouldn’t just be notated but would be hearable. The arguments exist in fuller form in his books The Textual Condition (1991) and Black Riders: The Visible Language ofModernism (1993), but are complete in themselves here; they are the underpinning of the digital edition of Dante Gabriel Rossetti’s writings and pictures that he’s been working on since that time. McGann and others had been circulating ideas of this sort earlier, but this essay was still very influential, and in 2004 it also takes some effort to remember that digital image and audio publishing on this scale looked much more difficult in that distant era.

This strand continues, but the subject shifts from the opportunities offered by digital publishing to the inadequacies of digital representation. The argument is mostly conducted at a very high level, without much reference to the specifics of digital technology; it stresses the difficulty of representing the multiple, interacting, and shifting features of language, difficulties that by the logic of the argument apply to any attempt to represent or reproduce artworks by whatever means. The argument has the advantage of indicating the existence of real intellectual problems with digital representation that need attention, and not just technical issues; but the conclusions are so sweeping and general that they can have no effect on practice.There is too little here about the form of the problem that editors actually face: the work of choosing and arguing for your compromises given that the naïve goal of perfect reproduction is impossible.

Behind the second strand of the book, the literary-critical strand, is an intellectual disposition against settling on fixed and conclusive interpretations. Some critics have sought to establish the meaning of texts in an objective, impersonal way; McGann sees criticism as an activity of individual response and engagement. He’s not simply advocating some sort of random speculation: in this book as in many others by him, there’s a strong interest in historical context. But the context itself needs interpretation; there isn’t a set of nice hard facts about texts that we can uncover and use as an unchanging point of reference. McGann sees criticism not as an end (and therefore as something that might actually be conclusive) but rather as part of a process that should send you back to the work for further reflection. This is one reason for his interest in the dialogue form for criticism; it also means that readers seeking a body of dogma about criticism or editing from his books have tended to find themselves surprised by the directions his work subsequently takes. In this book, the literary-critical strand ends up with a description of the «Ivanhoe Game», an approach to criticism of works that’s based on rewriting or supplementing them: on direct intervention rather than on separate commentary. The details of the game don’t matter: the essential feature of it is that you knowingly and openly modify a work and make your changes known to others for their scrutiny and response.

The Ivanhoe Game as described here is played on computers, but it isn’t really tied to the studies of digital representation that the book’s other strand talks about; the game uses computers for convenience and not to do something that’s impossible in the paper world. The two strands of the book are almost entirely independent: they are united by an intellectual disposition but not by their content. In this disposition, a primary value is reflection. What McGann wants from computers is a spur to thought. Hence the examples that excite him are always of displays generated by the machine that make him think, rather than any conclusion that’s computed: he’s aware of things like word searches in large-scale full-text collections or mechanical techniques for automated collation, and he acknowledges their value, but they aren’t what inspire him and he only mentions them in passing. He’s much more interested in the use of image-processing software to distort images in ways that make previously-unrecognized features leap out at you. There’s recurrent reference to the “analysis” that is facilitated by computers, but what he means by this is that computers can generate a new display that a person looks at and thinks about; it’s not the computer that is taking things apart in some significant way. He’s as interested in a nonsensical or wrong display as in a considered one, because there can be just as much material for thought in it. That’s not to say that he wants to deny the distinction between historically documented texts of Rossetti and randomly garbled ones: but the deformed version can prompt thought in a different way, and he’s on the side of anything that makes him think. Seeing the many forms in which Rossetti published his works is one way of getting more material for thought, and it is important that they’re ways Rossetti chose; seeing mechanically distorted images of his paintings can also prompt thought, without involving any delusion that this is how Rossetti meant them to be seen.

McGann shares the interest in Oulipo that appears in the more recent contributions to Literatura hipertextual y teoría literaria. And he mentions his own Oulipo-ish game, dating back to before he used computers for anything: one player would scramble the words of a passage of verse and then see if others could reconstruct it from them. They stopped playing this game because it was too easy. This Oulipo connection is typical of the impressive and distinctive range of literary interests that lie behind Radiant Textuality but are not strenuously advertised by it. It still remains one person’s set of interests: anyone wanting a complete survey of «literature and the World Wide Web» will need to supplement this in many ways, perhaps by reading Literatura hipertextual y teoría literaria and looking at the HyperNietzsche web site to begin with. The advantage and disadvantage of the book is that McGann does not spend time on things beyond his own intellectual sphere.

Radiant Textuality, like Literatura hipertextual y teoría literaria, is a book that takes the digital world seriously and develops a way of seeing it that’s focused on a much higher level than the merely informational. Both books have the disadvantages ofwork that’s at the edges of familiar terrain: they don’t give us well-tested models for work that we can just apply without further reflection, as Charity Marcon seems to have done; they generate ideas in abundance and require readers to think and discriminate.