6.3 Hypertext as rhizome, collage, and montage
This section will be relatively light on images, despite being about visual images for "hypertext"--that is to say, hypertext in general. Two such images have enjoyed wide currency in the last decade, namely hypertext as rhizome and hypertext as collage, even though the visual particulars of either term have never been spelled out (or shown). In a certain way, they are instances of the elusive generalized image mentioned in the introduction, and both have been pressed into service to characterize what is different about hypertext as a way of organizing information. A third, somewhat more abstract metaphor for textual and hypertextual structure, namely linkage as cinematic montage, is nearly as old as montage itself but has had a revival with the advent of hypertext.
"The Rhizome" opens one of the pivotal texts of Post Modernism, Gilles Deleuze and Félix Guattari's A Thousand Plateaus. The rhizome is contrasted to the tree and tree-like structures as master images of the "books" specifying the relation and connectedness of things. A tree has both root and shape, but the rhizome has no center and ramifies (potentially) in various directions without limit. First published in 1980, "The Rhizome" was not offered as a figure for hypertext or the Web, but the traits they highlight are salient ones of webs and the Web and were quickly applied to hypertext by Kathleen Burnett among others. It gave its name to rhizome.org, a major site and archive of net.art from 1996 on, and a section "Hypertext as Rhizome" was added to George Landow's Hypertext in edition 2.0. It is now very established as a way of thinking about the Web, although it is also recognised that it is a much more comprehensive term as used by Deleuze and Guattari. (See for example the logo-and-explanation page for the CHID program at the University of Washington.)
Strangely enough, although A Thousand Plateaus has illustrations, it does not include one of a rhizome, and the CHID page is to be congratulated for having a go at it, since as has been noted by Darryl Laferte among others, unbounded things cannot be imaged. (They also say rats and ants are rhizomes.) The image at the left illustrates the point, since it depicts just a little fragment of the World Wide Quackgrass rhizome. The particular botanical properties of rhizomes do not seem of much interest to Deleuze and Guattari--as for example that they are not roots but stem-like bodies growing underground with scales and nodes that can sprout roots--but they do say enough to indicate that it is an impossible object or, as Landow says, "a counter paradigm, not something realizable in any time or culture" (42). There are a number of discussions of the adequacy of the analogy of Web to rhizome (and to its sister concept, smooth space) (e.g. by Stuart Moulthrop (1994) and Martin E. Rosenberg (1994), but this discussion has nothing to do with the visual representation of the rhizome, of which there is little or none, but with the concept and growth habit. The rhizome hypertext metaphor appears to be what is sometimes called a verbal image (as opposed to a mental image and to a visual one--this last being what we mean by image here). W.J.T. Mitchell discusses verbal image as a very broad and loose category, suggesting that we would all be better off without it (1994: 19-21). We seem very close to tree in the graph theory sense that has severed its connection to the plant world—i.e., it is a term for a rather abstract morphological form.
Deleuze and Guattari also describe a third figure, the radicle-system or fascicular root, which is intermediate between the tree/root and the rhizome. Its representation is multiple and fragmented--in short, radicle-system is found in collage (and in Hegel, Joyce, and Nietzsche). It is not all the way to rhizome, they say, for it frequently reveals a higher unity behind the apparent multiplicity of fragments, just as, we might say, a collage is in fact a bounded work and a composition and one may talk about the whole, although the whole is no longer one based on representation. Not surprisingly then, if one thinks of hypertext not in terms of the totality but in terms of individual works, the analogy with collage springs to mind. In the Electronic Word, speaking of digital replication and repetition, Richard Lanham declares:
The same aesthetic operates at the heart of electronic text, though we seldom notice it for what it is--an aesthetic of collage, the central technique of twentieth-century visual art. (40)
(Visual art here is apparently restricted to the static kind.) As we have seen, a case can be made for at least sharing the center with montage, or generalizing them to the notion of multiple, usually mixed-mode images. Lanham is speaking immediately of Andy Warhol's Thirty Are Better Than One (silk-screened images of Marilyn Monroe) which is certainly multiple, though not exactly your classic collage. Replication, juxtaposition, and differences of scale are the main traits he ascribes to collage, as well as an oscillation between looking at and looking through the fragments (i.e., Bolter and Grusin's hypermediation—38-41).
Landow develops the point by deriving the traits of "textual collage" from the hypertext link. Because it joins different things, the link "inevitably produces juxtaposition, concatenation, and assemblage." And then a step farther:
If part of the pleasure of juxtaposition inevitably tends towards catachresis and difference for their own ends and for the effect of surprise, sometimes surprised pleasure, that they produce. (171)
Since catachresis is the yoking together of disparate things, an "extravagant, unexpected, far-fetched metaphor" (Lanham, Handlist, 31), and these are the hallmarks of metaphysical poetry, all hypertext links aspire to the condition of the metaphysical conceit.
This passage catches a certain possibility for witty play with the hypertext link, but as a deductive argument, it has problems. Hypertext links don't always produce juxtaposition, which I take to be an abrupt placing together with no connection or transition. A link can be quite explicit about the connection it makes--in fact, Mark Bernstein argues, too explicit. There seems to be a quiet equivocation with the senses of different, moving from "nonidentical" ("Try a different card.") to "disparat want to follow" (talk at Hypertext 2000); www.markbernstein.org/talks/ HT00.html" (the Monty Python pseudo-transition "And now for something completely different"). Hypertext links can join one section of an argument to the next one: no great leap or witty yoking together there. One suspects that Landow is trying to pass off a kind of hypertext linking that he likes as the inner telos of links. To be fair to Landow, however, we should note that he elsewhere in the book maintains that:
Hypermedia as a medium conveys the strong impression that its links signify coherent, purposeful, and above all useful relationships, from which it follows that the very existence of links conditions the reader to expect purposeful, important relationships between linked materials. (1997: 126)
pointing out that users who cannot bridge or integrate the material at the target of the link will experience the text as incoherent and confusing. We might resolve the matter this way: a page appears juxtaposed to the first if the viewer cannot predict what the relation of the page will be to what she has been viewing; if after viewing the second page, she is still unable to see the relationship of the second page to the first, she may treat the link as an unresolved enigma, a Deleuzian nonsignifying discontinuity, or an annoying bit of user unfriendliness or authorial flakiness.
Implicit in the comparison of hypertext to collage is the notion that we could become better readers or writers of hypertext by studying collage, although it is rarely spelled out which collages or by whom. A notable exception is a recent article by Joseph Janangelo. Janangelo has suggested that a salutary model for student writers of hypertext is one of Joseph Cornell's "homage" collections (kept in a valise), namely the one dedicated to Ludwig II of Bavaria ("Toller Ludwig"). Though not a collage strictly speaking, the piece is certainly a multiple. Many of the things in the valise seem only obscurely related to the Mad Prince, but the principles of selection become apparent with some study of the book, a copy of which is included in the valise. (We are not able to study this book ourselves, of course, since the piece is always shown under glass at its home museum in Philadelphia.) This is a surprising piece to offer as a model for how to write hypertext, especially "persuasive" hypertext, since it is mute according to the best Greenbergian strictures (though there is text involved between the covers of a biography of Ludwig, a copy of which is included in the valise). Janangelo reports the work of Dickram Tashjian on the piece, in which various bookmarks that Cornell placed in the book are used to provide background on why some of the particular items were included. Interestingly, having unpacked the significance of some of the valise and celebrated its subtlety and richness, Janangelo turns on it and judges it as in need of revision if it is to function as a hypertext essay, at least with him as a reader and grader. There are a couple of common failings of student hypertexts that Janangelo hope to mitigate or correct, including the non-selective inclusion of everything the student finds on a subject, and for that purpose the analysis of "Homage to Ludwig II" makes some sense, not as an example of design or visual form but of selecting materials with relevance to a particular aim.
At this point we may say the tables have completely turned round on the relations of the visual and textual. We began with the Modernist ban on mixing visual and textual modes, in part to prevent the corruption of painting by literature and narrative. By the 1980s, the influence of Modernism had declined, and there have been numerous applications of language-based categories to visual works. But now we have explicit urgings to apply visual ideas to the new quasi-textual realm of hypertext. Articles like Janangelo's are not yet common, however, and it remains to be seen whether we have here an interesting experiment or the beginning of a new approach to teaching (hypertext) reading and writing.
Landow credits his student Lars Hubrich with the observation that hypertext, since it is experienced as a succession of screens might be thought of in therms of cinematic montage—i.e., the principle guiding the editing of shots and sequences in a film. The analogy to hypertext is imperfect, of course, since viewers are given no choices in realizing the sequence (no "Choose your own movie" as yet), but it does remind us that we do have a great deal of experience bridging and interpreting visual shifts, jumps, discontinuities and juxtapositions.
Cinematic montage is not uniform, however; it has its Modernism and PostModernism as well. Modernist practice is most often discussed in relation to Eisenstein's theory and practice. Lev Manovich takes the core of classic cinematic montage to be dissonance: "Montage aims to create visual stylistic, semantic, and emotion dissonance between different elements," and this he contrasts to compositing, which is the blending of elements into a single, seamless whole (144). It is obvious that Manovich's "montage" is the temporal equivalent of what we have been calling collage. 4
Juxtaposition and interruption are also the traits that Walter Benjamin welcomed in montage, seeing them as undermining the illusionism of realism and the camera and thus as tools of a critique of culture. It is not entirely clear to me from the quotations assembled by Susan Buck-Morss what kind of montage Benjamin has in mind, but it is clear that he was trying to develop a mode of critical discourse that reduced theorizing and authorial interpretation to a minimum in favor of series of "images" (both verbal descriptions and graphic images including photographs) plus some commentary that would evoke the very dense world of nineteenth-century Paris and in particular the transformations in material and social relations brought on by "high capitalism." This project was unfinished at the time of Benjamin's death in 1940; the notes for it and some sketches and proposals have been published and translated into English as The Arcades Project (referring to the central image of the glass-over blocks of streets and shops that were a center of nineteenth-century Parisian life). Benjamin speaks of trying to carry to perfection "the art of citing without citation marks" (Buck-Morss, 67) by which he does not mean plagiarism but simple presentation without embedding in an authorial discourse. There is a distinctly modern quality to this attempt to present "dialectical images" one after the other without the connective and integrative speaking of an author, but it greatly disconcerted Theodore Adorno, at that time his friend and supporter, who felt that the "images" did not speak for themselves but read as wide-eyed, uncritical descriptions--or, one might say, jottings on the "Paris street scene" for inclusion in one of the new feuilletons of the day. Buck-Morss mounts a defense of Benjamin's practice (interestingly using closeup images of the Eiffel tower to show its construction) and in a larger sense, her entire book is an attempt to make evident the guiding themes and purposes of the vast mass of "images."
This case makes it clear that the issue of "montage as a structuring principle" extends well beyond visual images and beyond and before hypertext. It did occur to Giles Peaker of the School of Art and Design at the University of Derby, however, that Benjamin's intention might be better realized as hypertext. 5 Working mainly from two pieces that attempted to weave into essays bits of the clippings in the many folders ("convolutes") of the project, Peaker sliced pieces out of them and recombined them (usually two or three) onto eleven pages. 6 He then cross-linked the pages on the basis it appears of semantics and themes, so that each page has about four links to it from other pages and with a variety of anchor words and phrases. He scanned and added some of the images that Benjamin's editor Rolf Tiedemann and Susan Buck-Morss had recovered or collected, but with no captions or identifiers. Unfortunately, this is one of the places where the images noticeably do not speak for themselves and where their placement on certain pages can mislead. For example, Peaker puts a picture of a woman adjusting her garter on the "Prostitute" page, but the woman in question is a wax figure which fascinated André Breton and perhaps Benjamin as a freezing of the ephemeral. (Buck-Morss, 369; the rights to the photograph are hers.) We cannot know what Benjamin would have done with the labelling and identifying of his images, but it would seem to be a good idea to give them some words to help viewers align them with the themes and topics of the work.
Peaker's "Fragments" are almost a pilot project; mounting and linking all the fragments of the Arcades Project would be life work. Peaker's head note focuses on the power of the links to place pieces in new aspects—i.e., by making the page with a source link the context for the target page of the link. In fact, it also presents pages in different lights because the links are different, and sometimes there are two or even three links on a page to the same page, though with different phrases and words as the triggering sources. This "different aspects" purpose is a different—and more integrative—one than that of disruption of context initially noted by Benjamin as the essence of montage.
These matters are discussed more fully and in relation to recent Web writing in Dillon (2004), where in addition can be found a reworking and amplification of Peaker's Arcades web.