The Search for Reader-Centered Hypertext

It is now generally agreed that hypertext, as "nonsequential writing-- text that branches and allows choice to the reader, best read at an interactive screen,"^[c1] provides the reader greater control in the "creation" of the text, as well as more control in how it is read. This autonomy has, theorists argue, social consequences that reach far beyond the individual and the text. George Landow, whose work is among the most developed and recognized of these critics, explains that

a fully implemented hypertext system... obviously creates empowered readers, ones who have more power relative both to the texts they read and to the authors of these texts. The reader-author as student similarly has more power relative to the teacher and institution. This pattern of relative empowerment, which we must examine with more care and some skepticism, appears to support the notion that the logic of information technologies, which tend toward increasing dissemination of knowledge, implies increasing democratization and decentralization of power.^[c2]

Landow's acknowledgment (and subsequent neglect) of the more "relative" aspects of reader empowerment, alluded to in this passage, is common in literary approaches to describing hypertext. Most discussion of the design of hypertext applications in literary disciplines is confined to terms and issues current in critical theory, which inevitably revolve around the roles of reader, author, and printed text in creating meaning, and their "power relationships" in controlling this creation. Most literary hypertext theory does not carefully consider the role of the third agent in shaping the reader's understanding, the one introduced only in the electronic environment: the software designer. In view of some basic principles of human-computer interaction to be explored in this chapter, it is clear that electronic hypertext applications, like all software, necessarily constrain the reader in ways that fundamentally alter the nature of reading tasks. With a more systematic treatment of the constraining features in the design of hypertext software, one based on an examination of such features in the context of their use and not just their form, it is clear that such constraints play a role in the reader's understanding of the text.

Considering the role of these constraints requires a general conceptualization of hypertext. Hypertext is usually only associated with computer technology, but, as Michael Heim points out, the relationship between computers and hypertext is a pragmatic relationship rather than a logical one.^[cn3] Literary critics tend to focus on hypertext as an extension of current writing technologies, for them hypertext is a collaborative enterprise, a medium in which the roles of reader and writer merge into what Bolter calls the "wreader," Landow the "reader-author". This "wreader" is an empowered figure who physically rearranges or complements the text at will, and so achieves the status of co-author. Yet, hypertext should be understood in terms of "readerly" reception and comprehension as well. In fact, the experience of "hypertext" (the prefix "hyper-" meaning "over, above, or beyond"^[c4]) is already familiar to advanced readers of print who have engaged in comparative, analogical, or relational reading tasks. Examples of such tasks might include research, or the reading of a variorum edition of a literary work. In the former case, a reader begins with an original topic and constructs a series of readings that may share no inherent relationship outside of the author's exploration of that topic. In the latter, the reader examines a main text, may trace its development in manuscripts or earlier editions, and can explore the body of commentary that has grown around it. Thus, like the imagined reader of the electronic hypertext, such readers also read "non-sequentially". A hypertext, then, should be considered a representation of the reader's understanding, or a model for envisioning bodies of information. In whatever form it is represented, a hypertext is thus analogous to a mental scheme, it is a synthesis of ideas from diverse sources in the reader's memory. It is not necessarily a representation of the physical text, but the physical text and the reader's understanding of it. So we can speak of a "text" as an author's composition, and the "hypertext" of that composition as the text and the system of related texts and cross-references which the author and reader determine collaboratively, by association.

Various media, including books and computers, can "embody" hypertexts, but the nature of the physical materials always alter them in some ways. Because printed books must be separated into volumes and laid out on uniform pages to make them portable and readable, hypertext effects have to be "simulated" in print with symbols. Conventions of document design have evolved for representing logical units within hypertexts and relationships between them^[nc5]. In academic research-oriented publications, for instance, texts are brought together for comparison or synthesis with separated quotations, citations, footnotes, and endnotes. Chapter divisions, section headings, and paragraphs represent basic units of meaning. Though these systems often suggest meaningful structure, the symbols themselves are rather arbitrary; their purpose in the text is not obvious to anyone who has not been trained to use them. To read an endnote, for example, I have to recognize that one exists and I must take the initiative to flip through the pages until I find the appropriate reference. The second problem is that the reader has to do all the information management. A reference in an article may send me back to the library shelves, or I may have to move to another library to find all the related texts. When I have the texts assembled I then have to work through them, recognizing relationships important for my purposes and tracking these relationships on paper if I wish to remember them. Though I may impose a new physical order on the information by annotating or dog-earing pages, or by creating an index for reference, it is incumbent upon me to determine a means of managing this information with techniques that are never obvious affordances in the design of the materials themselves. Nevertheless, if I have a copy of a scholarly article and the other related texts I can explore the implicit "hypertext," limited only by my imagination and resourcefulness.

The power of the electronic hypertext is that it can represent such relationships concretely and meaningfully on the computer screen. Hypertext software can store and display more discreet units of text of variable length, and give the reader the ability to move between them automatically. It functions like a database in that it stores textual information in an author- or user-defined order and arrangement, and can retrieve and display the information based on the reader's decisions. This type of application has three basic components: a group of related textspaces or "nodes" that contain text, and a set of "links," which are pathways between nodes that allow the reader to rapidly "teleport" between them. Links, in Jeffrey Conklin's terms, are either "organizational" or "referential"^[nc6]: they can provide seamless coherence among members of an arbitrary group of nodes or draw the reader's attention to special relationships between them. The third element is a user-interface that represents the texts and links graphically and gives the user a set of controls for manipulating (or "navigating") the hypertext. Readers using such a network can choose their own "path" through the materials; and in many environments^[n7], the reader can add new nodes and links that are no different in appearance from the author's own. Critics like Jeffrey Conklin consider linking to be the single most important feature of hypertext, but the ability to view the text as nodes of information is also suggestive of meaning.

The "random access" afforded the reader in the ideal ("fully implemented") electronic hypertext, critics argue, enfranchises the reader with textual power stolen directly from the author. The author would normally use the "fixed" nature of print to their advantage in determining the reader's experience of the text. Landow and Delaney, for instance, explain that this medium would eliminate the illusion of the author's sole authority over the text:

Mixed with the generally democratic and even anarchic tendencies of hypertext is an opposing principle: that hypertext readers have more control over the most basic characteristics of this information medium. They are more free to determine the order within which they read individual passages than are readers of books: further, the reader's choices also define the boundaries of the text and even the identity of the author (if one can properly speak of such a unitary figure in this kind of dispersed medium.)^[c8]

This line of thinking is productive for social and literary theorists, but it rests on two tenuous assumptions. It assumes that these characteristics are inherent in the "fully implemented" hypertext design, and that this design is ideal for all reading tasks. It also assumes that readers are mostly helpless or passive in creating order in print, and that their actions are therefore predictable. Both of these claims demand further investigation.

Because of the attributes of printed texts and hypertext just described, theorists have assumed that the reader's control of the printed text is more limited and less productive for creating hypertexts than I have suggested. Print, they argue, affords only the creation of primitive "proto-hypertext"^[c9]. This idea arises from the view that print presentation delimits the reader's interaction and understanding of the hypertext. Landow and Delaney explain that throughout the history of print "as long as the text was married to the physical media, readers and writers took for granted three crucial attributes: that the text was linear, bounded, and fixed... Unlike the static form of the book, a hypertext can be composed, and read, non-sequentially"^[c10]. The physical text produces an uneven economy of social power tipped toward the author. Bolter explains,

The computer... makes visible the contest between author and reader that in previous technologies has gone on out of sight, 'behind' the page. The author has always had the upper hand, although each previous technology of writing allowed for a different degree of control... In general, authors in the age of print have exerted their authority in (subtle), but still effective ways. The electronic medium challenges all such attempts to circumscribe the reader's participation.^[c11]

According to this theory, then, readers of print may "rebel" while reading by selecting random pages or scanning running heads, but they are ultimately trapped to read and understand the text based on the structure that the author imposes. Gunnar Liest¯l has a less exclusive, but ultimately similar view: "A traditional text of print and book technology is 'prelinked' by the producer, and the reader is supposed to always behave within and be conditioned by the constraints and of that arrangement (and agreement) but is free to conceive of the text's signifieds in a variety of combinations"^[c12]. These formulations give only passing mention to the variety of reading tasks and associated techniques that "traditional readers" already use to create hypertexts with physical media, such as symbolic annotation, structured notes, research writing, and so on. They also fail to account for the fact that hypertext software also exerts subtle control over the reader's experience. Such software, it should be remembered, effectively puts the hypertext "under glass" by displaying it on screen and the software constrains manipulation with control menus and input devices. In an electronic hypertext, the editor/author may choose not to make all links available to the reader. Poorly designed or very complex link structures in some environments can also lead to the so-called "disorientation problem"^[cc13]. The "unboundedness," or selective boundedness, of the hypertext makes the reader much more, or at least just as, dependent on the design of the publication (and the software tools designed to cope with it) as the reader of printed books. Added to the assumption that constraint in the medium is undesirable, and that there are fewer constraints in hypertext environments, the suggestion that the reader of print becomes a "linear" thinker as a result of reading books is equally misleading. To characterize the proper limits of reading in print outside of this milieu of hypertext theory, and to understand the limits of electronic hypertext, it is instructive to examine the similarities between two popular researchers from the era just before the development of electronic hypertext: Mortimer Adler (who, as far as I know, has never received proper credit as a hypertext theorist) and Vannevar Bush.

While Adler wrote his How to Read a Book as a kind of "instructional" text, it also provides a portrait of advanced reading practices through the eyes of an experienced researcher and adamant bibliophile. Though Adler's medium of choice is print, the terms Adler uses are consistent with the ideals of hypertext theory, which Ted Nelson would not articulate fully until the 1960s. In his account of the syntopical reading process, Adler makes a distinction between remedial (viz. "linear") reading and more "active" forms. It is worth noting that Adler describes the "active" processes in terms that would capture the essence of hypertext itself:

When reading syntopically, the reader reads many books, not just one, and places them in relation to one another and to a subject around which they revolve. But mere comparison of texts is not enough... With the help of books read, the syntopical reader is able to construct an analysis of the subject that may not be in any of the books. It is obvious, therefore, that syntopical reading is the most active and effortful kind of reading.^[c14]

Admittedly, Adler's general conception of reading assumes a certain reverence for the author of works which possess literary greatness; hypertext theorists have suggested that this attitude toward texts arises naturally from the determined form and authoritative appearance of print^[ncc15]. Adler's syntopical reader eavesdrops on a discussion and cannot disrupt it by rearranging the text, as the user of the electronic hypertext can. According to Adler, syntopical readers nevertheless dissolve the boundaries of time, space, and culture between authors as the readers "bring the authors to their own terms": the reader can trace the dialog between authors, and ultimately participate in this dialog by creating new texts in research. Such readers may choose their own path through the materials and ultimately understand them in relation to each other^[nc16].

Adler also discusses ways of using the very design and conventions of printed materials to the reader's advantage in understanding, albeit much more slowly, to achieve exactly the same kind of mastery of the topic and its constellation of interrelated texts that Michael Heim expresses as the ultimate ideal of hypertext^[nc17], and Adler did not consider it particularly revolutionary. Yet, the steps Adler describes in this process involve precisely "frustrating the author's control" of the texts by reading "nonsequentially." His "active" reading process involves strategic and recursive stages of selective attention to the text. The reader begins by discovering the structure of the material with what Adler calls "x-raying the book" with a cursory reading, then prioritizes the content by "bringing the author to terms" in an "analytical" reading, and finally integrates the content with ideas assembled from other books when reading "syntopically." During the "analytical" stages Adler asks that readers "make the book their own" with a system of marginal annotation: "Marking a book," he writes, "is literally an expression of your differences or your agreement with the author"^[nc18]. Unlike the hypertext critics, Adler did not blame the design of the text for passive reading, only the reader's lack of training, practice, or occasional lack of interest. The power to "navigate" through the text was nothing that the reader should rely exclusively on the text to provide, the reader has simply to "master" the materials. At the same time, though Adler was able to invent many adaptive techniques to help readers cope with the stress of digesting information, he does not question the integrity of the printed text, or speculate whether, for some tasks, the book may be an inefficient means of storing information for easy access by the reader and the researcher.

Vannevar Bush took up exactly these questions: in his seminal proposal for "memex"^[nc19], Bush explores the possibility of adapting the design of the text to the needs of the reader. Bush bases his proposal on one fundamental insight that paved the way for all hypertext research that was to follow: the ways that books, and the libraries that house them, making information accessible is entirely unlike the way that readers use or process information. He noticed that the very bulk of materials stored in libraries exaggerates human inefficiency in managing them. In aggregate, this burden makes the research process prohibitively time-consuming. What makes the process even less efficient is that most libraries organize their collections alphabetically while, Bush argues, the human mind works "associatively"^[c20]. Bush's solution was technological: the researcher of the future that he envisioned would enjoy the benefits of what Douglas Engelbart would later describe as "intellect augmentation"^[nc21]. A set of mechanical tools would automate or enhance the effectiveness of repetitive research tasks: the researcher would wear a headdress containing a "cyclops camera" to photograph significant data he (for Bush, all researchers were assumed men) would find in books or discover in the world. He would speak his ideas into a "vocorder" recording device instead of taking notes. The researcher would then carry all of this data, and libraries of information, on tape and microfilm that he could organize and review with the "memex" machine. The user could maintain a "code book" of "trails" of associated segments of texts on cards that when placed into the memex machine would display the references on a screen. Each of these devices would extend the information handling capabilities of the researcher (Bush called his machine "memex" because it would extend human memory^[c22]), freeing time for tasks that require higher ordered thinking. In this way, Bush saw the potential for technology principally as a means for mechanization; the notion that hypertext is a "freer" medium would not materialize until Ted Nelson conceived the social theories associated with the Xanadu project.^[c23]

This comparison illustrates that the role of the reader is not as determined in print as we have been led to assume, and that envisioning hypertext is an idea that is independent of an electronic medium. While print presents information in a linear fashion, Adler shows that this is not necessarily the manner in which the reader envisions the information, or manages it. Putting aside the usual bibliophilic reservations about electronic hypertext, which usually appeal to the portability of books and their aesthetic beauty, readers of the paper volume have the advantage that they can manipulate the physical text in more subtle ways, albeit more laboriously, than in automated hypertext systems. Any experienced reader knows that the printed text supports an extremely wide range of reading activities and strategies. Every researcher develops subtle and entirely personal systems of annotation, note-taking, and text management (or may choose to ignore them) because of the text's physicality. Computer simulations of these very complex activities are, in many cases, unbearably primitive or compromising; and they must be. Like Bush's tools for expediting research, electronic toolsets are necessarily specialized^[c24].

Adler and Bush seem to agree that the thought processes required for research call for an interaction with texts that is ultimately "associative," and that the researcher must be able to make such connections quickly in order to cope with the bulk of existing research. The major difference between them, however, is that Adler solves the problem by evolving technique, and Bush solves it through technology. The disadvantages of both methods seem obvious. Adler's method requires fastidiousness for uncertain results. He recommends that his reader not get caught up in too many details in the early stages of reading, and yet not miss the critical details. Of course, how does one do that? Bush's method has the problem that his tools facilitate research with scientific literature or objective information, but they would seem superfluous or intrusive for reading tasks that do not involve associative thinking. Bush did not even suggest that his memex would replace the book in all of its uses. Presumably when Bush's "researcher" became "the reader," he turned off the augmentation tools.

The inevitability of specialization in mechanical hypertext applications becomes clear when considering the operational differences between print and mechanical hypertext as technologies for storage and retrieval. A book is, to borrow Donald A. Norman's^[c25] terms, only a "surface representation" of a narrative, reasoning, or understanding process. A book can only represent information on the surface of its pages, it cannot process information internally, and one cannot modify its structure without permanently altering the surface itself. Yet, the ability to manipulate symbols on this surface is an important advantage. Interaction with a book is governed by consistent physical laws that are affirmed in normal human experience. Machines offer, by contrast, "internal representations" of the same processes. Computers do not physically store words, they arrange electric impulses that represent binary numbers within circuits; and store groups of these numbers to represent words, sounds, or images. To make these representations comprehensible, computers have to translate them into surface representations by projecting them onto a monitor screen. To manipulate symbols within an internal representation, users are restricted by input devices that translate physical actions into electric impulses; for example: mice, keyboards, light pens, or touch screens. Moreover, there are a limited number of functions available within the user-interface: in hypertext applications, these functions might include the ability to follow a link or search for a group of words or to place the cursor and insert text. The transition from print to computer software, then, involves trading the affordance of subtler and more intuitive kinds of interactions for internal processing and enhanced capability for specialized operations. The user can temporarily or permanently alter both the contents and the arrangement of electronic representations of texts with the proverbial "touch of a button." The consequence of this flexibility is a necessary reliance on the mediating tools: both input devices and software tools available in the user-interface.

Internal representations can potentially cause abstraction in the users' activities between their intentions, their physical actions, and the reactions of the representing systems. Involved in all interactions with computers is what Donald Norman calls a "gulf of execution" and a "gulf of evaluation"^[c26]. Norman describes the "gulf of execution" as the combination of the cognitive difference between the user's goals or needs and the chain of actions necessary within the context of the representing system to achieve them. There is always a series of actions that the user must take to accomplish any specific task. What Norman calls the "gulf of evaluation" is the difference between the user's actions, and the user's recognition that the actions are producing results within the system. The "gulf" is widened when the user's actions produce unexpected results. The computer system bridges this gap by providing the user with "feedback": a noticeable visual or audio effect that appears to follow naturally from the users' actions and shows that data has been altered. The objective of design is to make the relationship between intentions, actions, the affordances of the input devices, and immediate feedback in the user interface, as obvious as possible. The result of these restrictions is what researchers call a "task domain," a range of tasks that the software environment is capable of supporting or representing.

The most important component of the user-interface that makes these cognitive gaps surmountable is its controlling visual and operational metaphor. Every graphical interface uses a metaphor to suggest a relationship between functions available in the interface and the user's worldly experience. An effective user-interface should imitate "physical" laws or follow a consistent set of operational rules. For the last two decades, software engineering has stressed such "intuitive" designs. The Macintosh's desktop metaphor (where, for instance, a user can use a pointer on the screen to "drag" an icon into an animated trash can to delete a file) is a classic example of a "natural mapping," an analogous functional relationship between software tools and the user's worldly experience (assuming that metal trash cans are part of most users' experience). The metaphor that StorySpace uses to match form and function, for example, is the cluster or flowchart, while the controlling metaphor for Netscape is the scrolling page. Thus the choice of metaphor for a user-interface must never be arbitrary. It does not just communicate a conceptual understanding of the data (it must be recognized), it represents a procedural understanding of the use of the information as well.

Hypertext software, then, does not just provide representations of relationships between texts and concepts, it represents the range of possible interactions, a "task domain," through which readers acquire or process ideas from texts. There is always a limited range of interactions that a software package can support, but a program must necessarily exclude capabilities for its function to remain comprehensible. The task domain of an electronic hypertext environment, then, is usually designed to conform to pedagogical practices or institutionally endorsed ways of understanding texts. Often the electronic hypertext can represent these ways of reading more suggestively than print. Because of the control that the designer has in delimiting the types of interactions available to the reader, the reader cannot avoid behaving in ways that follow these practices. While this flexible software environment can also make traditional ways of conducting literary research more efficient, it may also open texts to new questions that may be unimaginable to scholars and educators trained exclusively with books. Designers may be able to craft ways of reading with software applications that do not resemble either the ways we read books or even current ways we read electronic hypertexts; though, naturally, there are limits to human cognition.

It is clear from this discussion that the user of the hypertext application is never strictly "free" to interact with the text in ways that are entirely unexpected by either the author or the software designer. In fact, these constraints of the user environment alter, and in some ways determine, the nature of the task^[c27]. By using a StorySpace web, for instance, a student reader does not just follow links and add commentary, the student will also understand the concept of links by seeing them in action, and that part of understanding a text involves seeing relationships between it and other texts, or other parts of the same text.^[ncc28] In other words, electronic hypertext applications can provide readers with a system of understanding as well as a specific understanding of a hypertext.

The ways that software presents these ideas is not always explicit in the visual design, but they can be understood through interaction itself. Brenda Laurel uses the analogy of theatrical production to describe the interaction between the user and software. She explains that "the form of a play and that of an interface are similar in a fundamental way: Both are mimetic"^[c29]. A mimesis, she argues, is never an arbitrary representation, it has a direct resemblance to real experience, but it remains unmistakably artificial. A mimesis is also a closed system, ideally a user interface has its own consistent dramatic rules. When users act within the context of the interface, they are subject to the affordances and constraints of the "mimetic context" that are determined by the designer's choice of metaphors. Constraints are especially important because they "keep the user's activities within the bounds of the mimetic world." Laurel goes on to suggest that "in exchange for his complicity, the user experiences increased potential for effective agency"^[c30]. While it is clear that the user is forced into this world of play-acting, with a certain loss of freedom (though this type of "freedom" may be undesirable in many cases anyway) the experience of using the artificial environment presented in a graphical user interface gives the reader a way of understanding reading processes that printed texts (where the cognitive experience of the text is less pre-determined) cannot.

Because the networked electronic hypertext is comprised of simple text or graphic data, a variety of hardware platforms can represent the same hypertext, and each platform can access the hypertext through a variety of user-interfaces. As George Landow points out, each front end and its platform may have operational benefits to suit individual preferences much more powerfully than the printed book. This power may resolve the vagaries of technique and solve some of the problems of over-specialization. Landow writes, "A central fact about the digital word lies in its intrinsic separation of the text from the physical object by means of which it is read"^[c31]. The emergence of mark-up languages like SGML, and its HTML subset, have made possible "portable" hypertexts, capable of being transferred from one kind of document reader to another. The appearance of the text, and design of the interface that controls it, is different in each case, but the data and its structure are the same. Moreover, the interfaces themselves can be transported to different kinds of computers with their own operating systems and input devices^[c32]. Such a point may allay some bibliophilic suspicions, but it begs an important question that requires further investigation. Though electronic storage does afford this flexibility in representation and interaction, how does one discover the optimal representation of the text for the individual reader? Any proposal will require pragmatic consideration of the user, their tasks, and the user's environment.
 
 

Notes

¹ Theodor Nelson, Literary Machines (Swarthmore: Mindful P, 1981) N. pag. This is still the most quoted definition of hypertext. [Back]

² George Landow, Hypertext: The Convergence of Contemporary Critical Theory and Technology (Baltimore: Johns Hopkins UP, 1992) 175. [Back]

³ Michael Heim, The Metaphysics of Virtual Reality (New York: Oxford UP, 1993) 33. [Back]

⁴ "hyper-," The American Heritage Dictionary, 1993 ed. [Back]

⁵ David Jay Bolter provides a concise overview of the history of writing systems and physical media, as the roots of "the electronic book," in Writing Space: The Computer, Hypertext, and The History of Writing (Hillsdale, NJ: Lawrence Erlbaum Associates, 1991) 33-106. [Back]

⁶ These terms and the importance of links in defining hypertext, are explored much more completely in Jeffrey E. Conklin, "Hypertext: An Introduction and Survey," IEEE Computer 20 (1987): 18-19, 32-34. 17-40. [Back]

⁷ including Intermedia, the platform for Landow's research, and StorySpace, the environment which Jay Bolter helped design and was the kernel of his work in Writing Space. [Back]

⁸ George Landow and Paul Delaney, "Hypertext, Hypermedia, and Literary Studies: The State of the Art," eds. George Landow and Paul Delaney, Hypertext and Literary Studies (Cambridge, Mass: MIT P, 1991) 30. [Back]

⁹ Landow and Delaney 4. [Back]

¹⁰ Landow and Delaney 3. [Back]

¹¹ Bolter 154-155. [Back]

¹² Gunnar Liestøl, "Wittgenstein, Genette, and the Reader's Narrative in Hypertext," ed. George P. Landow, Hyper/Text/Theory (Baltimore: Johns Hopkins UP, 1995) 103. 87-120 [Back]

¹³ explored in Conklin 38-40. See Deborah Edwards and Lynda Hardman, "'Lost in Hyperspace': Cognitive Mapping and Navigation in a Hypertext Environment," ed. Ray McAleese, Hypertext: Theory into Practice (Norwood, NJ: Ablex, 1989) 105-125. [Back]

¹⁴ Mortimer Adler and Charles Van Doren, How to Read a Book (New York: Simon and Schuster, 1972) 20, my emphasis. [Back]

¹⁵ Both Landow and Bolter have written about the role of cultural assumptions that print presentation encourages in New Critical "fetishism" of the literary text. On New Criticism, the role of print in establishing notions of canon and hypertext as an institutional solution, see Landow, Convergence 149-161. See Bolter 147-168. [Back]

¹⁶ See Adler 333-334, where he argues against more traditional assumptions about the author and his "context." [Back]

¹⁷ Heim 38: "Hypertext emulates a divine access to things. Although God does not need to jump, the hypertext user leaps through the network of knowledge in something like an eternal present. The user feels intellectual distances melt away." [Back]

¹⁸ Adler 48-51. The agonistic tone of this suggestion is puzzling since Adler was writing about "great books" that were predominantly works of fiction (See "A Recommended Reading List," 347-362). [Back]

¹⁹ Vannevar Bush, "As We May Think," The Atlantic Monthly 176.1 (1945): 641-649, Rpt. eds. Dames M. Nyce and Paul Kahn, From Memex to Hypertext (Boston: Academic P, 1991) 86-107. Page references to Bush are from Nyce and Kahn. [Back]

²⁰ Bush 101. [Back]

²¹ See Douglas C. Engelbart, Augmenting Human Intellect: A Conceptual Framework (Summary Report AFOSR-3223. Menlo Park, CA: Stanford Research Institute, 1963). Bush's article had a profound influence on Engelbart, as his "Letter to Vannevar Bush and Program on Human Effectiveness" (Rpt. in Nyce and Kahn 235-244) shows. [Back]

²² Bush 102. [Back]

²³ Explained "non-sequentially" in Ted Nelson, Dream Machines (2nd Ed. Redmund, WA: Tempus Books, 1987) 141-152. [Back]

²⁴ While web readers like Netscape allow flexibility in the appearance of pages, it is interesting to note that their tools for manipulating the text have remained fundamentally the same since early versions. Their text handling capabilities are still primitive mostly because web browsers are designed for general users who are supposed to be most interested in commerce or downloading data. [Back]

²⁵ Donald A. Norman, "Cognitive Artifacts," ed. John M. Carroll, Designing Interaction: Psychology at the Human-Computer Interface (Cambridge: Cambridge UP, 1991), 25-26. [Back]

²⁶ Norman 23-24. The terms used in this discussion are mostly Norman's. For an overview of these concepts, see Donald A. Norman, The Design of Everyday Things (New York: Doubleday, 1988), 81-104. [Back]

²⁷ Norman, "Cognitive Artifacts" 21-22. [Back]

²⁸ There is a growing body of evidence in hypertext scholarship to support this claim. For two examples, see Michael Joyce, Of Two Minds: Hypertext Pedagogy and Poetics (Ann Arbor, MI: U of MI P, 1995) 49-59; and George P. Landow, "Hypertext in Literary Education, Criticism, and Scholarship," Computers and The Humanities 23 (1989): 173-198. [Back]

²⁹ Brenda K. Laurel, "Interface as Mimesis," ed. Donald A. Norman, User Centered System Design (Hillsdale, NJ: Lawrence Erlbaum, 1986) 70. 67-85. [Back]

³⁰ Laurel 80 [Back]

³¹ George P. Landow, "What's a Critic to Do?: Critical Theory in the Age of Hypertext," ed. George P. Landow, Hyper / Text / Theory. (Baltimore: Johns Hopkins UP, 1994) 4. [Back]

³² For some possible examples of other presentation media, see Landow, "What's a Critic to Do" 4-6. [Back]
 

 

---

Página creada y actualizada por grupo "mmm".
     Para cualquier cambio, sugerencia,etc. contactar con: fores@uv.es
     © a.r.e.a./Dr.Vicente Forés López
      Universitat de València Press
Creada: 15/09/2000 Última Actualización: 15/06/2001