In the July, 1945 issue of The Atlantic Monthly, Vannevar Bush, who had served as the first director of the Office of Scientific Research and Development, the agency established by Roosevelt to coordinate federally funded defense research, published an article entitled "As We May Think." In it, he pointed out the increasing gap between the growing mountain of research and the inadequacies of methods for transmitting and reviewing its results, which he blamed in part on the artificiality of systems of indexing. He suggested that the human mind operates by association. "With one item in its grasp, it snaps instantly to the next that is suggested by the association of thoughts, in accordance with some intricate web of trails carried by the cells of the brain." He proposed "a mechanized, enlarged, and intimate supplement to an individual's memory, a future device" which he called a "memex" using electro-mechanical technology as a device for associative indexing, a reading and writing machine that would allow "wholly new forms of encyclopedias to appear, with a mesh of mesh of associative trails running through them." Users would create "endless trails" of links...exactly as if the physical items had been gathered together from widely separated sources and bound together to form a new book." 

In a corrective to Bush, Norbert Weiner pointed out that the use of mechanical aids would still be limited by classifications. "In the case where two subjects have the same techniques and intellectual content but belong to widely separated fields, this (connection) still requires some individual with an almost Leibnizian catholicity of interest." ( Cybernetics, p. 158) 

Douglas Englebart, who had served as a naval radar technician during the war, took a degree in electrical engineering after reading Bush's 1945 article. Putting together his experience of trying to discern the real threats represented by blips on the radar screen and Bush's call for the augmentation of human capacities, it was Englebart who laid the foundations of personal computing in the early1950's, by proposing to put the computer together with a television-like screen as a medium for representing symbols. The orignal inventor of word processing, Englebart also invented the first actual working hypertext environment, calling his system Augment.

The term "hypertext" was coined by Ted Nelson, a cranky and apocalyptic visionary who described himself as "a rogue intellectual, social critic, and designer of interactive computer systems for our world of tomorrow". Working with mainframe computers in the 1960's Nelson had come to realize the machine's capacity to create and manage textual networks for all kinds of writing. During the next twenty years he worked on what he called project Xanadu, "a computer program intended to make possible a new unified electronic literature." (Nelson somewhat ruefully acknowledged that his use of the name of the never-finished palace of the protagonist of Orson Welle's "Citzen Kane" carries "both a magic and a curse.") Nelson described his project in his book Literary Machines, first published in 1981 and later republished as Literary Machines 91.1 in 1991. Its cover carries the admonition on its cover "DO NOT CONFUSE IT WITH ANY OTHER COMPUTER BOOK. 


Nelson considers literature (including belles lettres and scientific literature) as interconnected writings. By 'hypertext,'Nelson explained,"I mean nonsequential writing,text that branches and allows choices to the reader, best read at an interactive screen. As popularly conceived, this is a series of text chunks connected by links which offer the reader different pathways." (0/2) In the project Xanadu, Nelson proposed to make all published work electronically available, to make each reader able to make and publish links and comments, and to make royalty payments automatically upon quotation, which would in fact be a form of purchase. Despite Nelson's millenarian tone, ("Xanadu's intended networkmight just revitalize human life")the project might be a successful strategy for preserving copyright. In1988, Autodesk, the highly successful software company which developed Autocad, the most widely used computer-aided design program, aquired Nelson's project Xanadu, although it soon isolated Nelson from the programming, giving him the title of distinguished fellow, and abandoned the project after four years of sponsorship. 

More recently, some literary critics have used the concept of hypertext as a basis for a theory of literature. Writing Space, by Jay David Bolter (1991, and Hypertext, The Convergence of Technology and Contemporary Theory, by George P. Landow. are both attempts to theorize hypertext as literature. 

Both writers adopt the position that writing is the general term which describes not only hypertext but other hypermedia, or multimedia, as well. Bolter: "A hypermedia display is still a text, a weaving together of elements treated symbolically. Hypermedia simply extends the principles of electronic writing into the domain of sound and image." (p. 27) In a later discussion of the structure of a future hypertext encyclopedia, Bolter draws the line between a hypertext with sounds and images and "a simulated environment which offers the user the illusion of perceptual experience" (p.98) Bolter rejects the latter as an "anti-book", believing that without the distancing and abstracting quality of text, "the reader becomes a mere viewer, and the encyclopedia becomes interactive television." 

For Jay David Bolter, we live in the late age of print, and, like many recent writers on architecture, he starts with Victor Hugo's oft-quoted passage from Notre-Dame de Paris, 1482 in which the archdeacon says of the book in relation to the cathedral, "Alas. This will destroy that." (how have architectural writers interpreted this passage, what relation to current concern?)

Bolter describes a writing space, first as "the physical and visual field defined by a particular technology of writing." For him, "Each technology gives us a different space." (p.11), which can be as a literal, conceptual, and metaphoricwriting space. The concept of a writing space is further elaborated as "the interplay of writing materials and writing techniques used." (p.41) These structures are also described as hard structures, "the tangible qualities of the materials of writing" and soft structures, "those visually determined units and relationships that are written on or in the hard structures." (p. 41). In addition, the writing space is "literate culture's root metaphor for the human mind." (p.11) 

The history of writing spaces can be traced from ancient scrolls through medieval writing to modern printing and into the prospects of electronic writing. For Bolter, "The computer rewrites the history of writing by sending us back to reconsider nearly every aspect of earlier technologies." (p.46) Because of the malleability and complexity of the computer's space, consisting of the videoscreen as well as the electronic memory, aspects of previous technologies which had been excluded from book technology, may return in Hypertext. For Bolter the two-dimensional writing surface has often been the site of negotiations between pictorial space and textual space

The spatial experience of hypertext shares characteristic features of the printed page, as mediated by contemporary computer interfaces, the monitor and "desktop" icons. The surface of the screen is transformed, metaphorically at least, into a "writing space." Most importantly, the spatial experience is defined by the electronic links between texts. 

As a surface, the printed page has been a site of renegotiated relationships between textual space and pictorial space, and the history of these changing relationships, from pictographic writing to filmic superimpositions of text and image, provides a rich basis for imagining the hypertextual surface. The transformation of the page as it is detached from a material surface and can include sound, film clips, and other channels that go beyond the "bandwidth" of the book, provide an opportunity to renegotiate some of the relations which the book tended to fix. When the computer is used as a word processor, as a " virtual typewriter", the page is "realized" through the printer. Hypertext, on the other hand, loses its distinctive features when printed on a page. But at present, because we understand the reading and writing as marks on a surface, the space of a "chunk" of hypertext seems primarily two-dimensional.

Hypertextual links, on the other hand, seem to add dimension to the textual surface. It is here that the spatiality of hypertext starts to come into its own, for the linked text suggests a complex spatiality, perhaps ressembling the fractional dimensions of of fractal geometries which are more than two-dimensional but less than three dimensional, or the space of "wormholes" which shortcut through time and space, or the reconfigured spaces of telemedia. The space of hypertext is like a constantly changable constellation. Rather than being a stable territory, it changes according to its mappings, for its features are primarily relational, and those relations can determined anew through every use. It is this capacity for reconfiguration that situates hypertext at the intersection between the urban and the textual experience. Movement in hypertextual space effects the production of associations, of narratives, of "cognitive mappings." This movement traces its lineage from the Aristotelian concept of topos through Giordano Bruno's walks through London, to Walter Benjamin's descriptions of Baudelaire in the crowd, to the Surrealists' dérives , and Michel de Certeau's affirmations of the "spatial practices" of Everyday Life. 


see hypertext city

The contemporary network promises new forms of discursive production and sociality. "Tech-noir " fiction like William Gibson, or Phillip Dick provide some of the scenarios for the future on which to project our current preoccupations. 

It is in this increasing congruency between representations of space and representational space (to use Henri Lefebre's terms) that Hypertext overlaps with other problematics of the relationship between cyberspace and the city. Whatever one's position on this problematic, hypertext would seem the (techno)logical extension of writing into the hyperspaces of postmodernism

George landow argues that Hypertext suggests the possibility of reducing, if not doing away with, the distinction between the reader and the writer. Hypertext has been described as nonlinear because various choices are possible at any time. (This meaning is quite different from the mathematical / scientific usage) Bits of texts are linked in a variety of ways. Instead of following a line of thought, whether in reading or writing, one navigates a conceptual space which undermines the dominance of linear narrative. Book technology had not only affirmed the dominance of linear narrative, but had made the writer the author-ity on reading. In the Hypertext environment, acts of reading are individual and idiosyncratic, depending on the person, the context and their particular preferences. Although they can be retraced, each act of reading, according to the proponents of Hypertext, encourages a new itinerary of reading. It is also very easy to add to the text, to manipulate the text itself, which again tends to dissolve the distinction between reading and writing. This can result in all kinds of strange problems, such as the erosion of notions about copyright, which is the legal guardian of authorship and textual stability. Some writers are claiming that these technologies are enabling precisely what post-structuralist writers have been theorizing: destabilizing the text, destabilizing the relationship between reader and writer. George Landow has written about the convergence of critical theory and technology but has also revealed that while for some this has meant the unhappy death of the author and the end of the book, he and other proponents of Hypertext see this same phenomenon as the beginning of a new electronic textuality full of possibility and promise. 

Both Bolter and Landau have come out with hypertext supplements to their books on disk. Both authors try to address the differences between their book and the hypertext version. Bolter makes a kind of shadow double to his published desk, much more quirky than his meditations in the book. He makes the linear structure of the printed text much harder to follow. Landow's changes are for the most part more didactic in content. He adds sections from the Johns Hopkins guide to Literary Criticism on the critics he names, such as Barthes, Baudrillard, Derrida, etc. He also includes reviews of his book, student contributions and parodies. Landau keeps overviews of his book handy, whereas Bolter slyly invites the reader to find the (non) center to his text. Both authors cross-reference sections which create reading loops when followed.