2. Hypertext Theory
Familiarity with the concept of hypertext is necessary to understand the principle of nonlinearity and the structure of a website. The beginnings of hypertext date back to the year 1945 when Vannevar Bush, an electrical engineer, published his ideas on Memex (memory extander), a "mechanically linked"32 information storage system, where one would use microfilm to save and edit material.33 This system existed only theoretically and was never realised. The term hypertext first appeared twenty years later and was introduced in 1965 by Ted Nelson, "one of Bush's most prominent disciples".34 Between 1965 and 1991 a few hypertext applications were created especially for Apple computers. The year 1991 brought a major change in the use of hypertext with the invention of the World Wide Web, the first global hypertext which has by now become standard.35
There are two main ways of dealing with hypertext: the first is a scientific approach which points out the technical and practical aspects of this new medium. The second is a poststructuralist approach which defines the term from the point of view of a literary scholar by emphasising its theoretical functions and potential. Both ways are necessary to understand the structure of hypertext even though the theoretical ideas are of prime importance, since they also form the basis of the approach suggested by Jakob Nielsen, a distinguished engineer for strategic technology. He juxtaposes hypertext and traditional text and provides the following explanation:
All traditional text, whether in printed form or in computer files, is sequential, meaning that there is a single linear sequence defining the order in which the text is to be read. (...) Hypertext is nonsequential; there is no single order that determines the sequence in which the text is to be read.36
Simplified hypertext system with six nodes and nine links.38
Nielsen claims that "some nodes are related to many others and will therefore have many links, while other nodes serve only as destination for links but have no outgoing links of their own."39 In this case, the node may contain internal, or local links which lead to other parts within the document. Links which connect different nodes are called external, or global links.40
Thus, hypertext creates a network and requires active users who have to browse and navigate in this system in order to determine their own, individual way of reading the information. Since it is impossible to present an entire and usually very complex network on a computer screen, the user can only imagine how the system is organised. Therefore, the hypertext system has to supply means and devices which support navigation. The main problem is how to keep track: one may not be able to get back to the starting point,41 or to information that was encountered while browsing the network.42 These difficulties arise because most hypertext systems contain only one-directional, i.e. outgoing links which tell the user where to go next but not how s/he reached a certain location. The standard means of navigation are not efficient enough: guided tours help the user to move through a system but "they really bring us back (...) to the sequential linear form of information [and] cannot serve as the only navigation facility since the true purpose of hypertext is to provide an open exploratory information space".43 Backtrack, a device which takes one back to the previous node, is applied very inconsistently to hypertext systems. It is either not always available or does not take the user back to the very first anchor node. Most hypertext systems allow the user to set bookmarks at nodes to which s/he wants to return. The problem is, however, that some information might turn out to be worthwhile later but one may not be able to find the way back in case no bookmark was set. For this reason, a history list which allows access to all previously visited nodes is ideal to solve navigational difficulties. Overview diagrams are also quite efficient but get more complicated and confusing with the increasing complexity of hypertext systems.44 The best solution, according to K. Erik Drexler, would be "the use of bidirectional links"45 which provide a list of incoming and outgoing links so that it becomes evident from which direction one arrived at a location, i.e. a new document should always include a link back to the previous one. Nonverbal elements such as graphic signs and symbols can also help to make navigation easier. Hence the term hypertext, which implies that there is only plain text, should also be used for systems that support graphics. Nielsen states that graphics "can be the main form of information in the system (...), they can be used purely as illustrations (...), or they can be more actively involved (...) by also including anchors for the hypertext links."46 Graphics and especially colours do not necessarily have to be used for linking purposes, but are also helpful in order to remember in which document one has read something important.47
According to Campbell and Goodman, hypertext systems consist of three levels: database, hypertext abstract machine (HAM), and presentation level. Electronically saved data are the core of each document and have to be provided on the database level. Information is stored in computers that are connected to each other and constitute a network. The structure of this network resembles a database because it ensures quick and multiple access to various pieces of information. The hypertext abstract machine determines how the data shall be organised and arranged in links and nodes. This arrangement will then be displayed on the presentation level with the help of a user interface, i.e. the individual user's personal computer software and hardware.48 Not all computer software, however, needs to be converted into hypertext applications. Hypertext should only be used if the following rules, suggested by Ben Shneiderman, are fulfilled:
- A large body of information is organised into numerous fragments.
- The fragments relate to each other.
- The user needs only a small fraction at a time.49
The World Wide Web, the first global and currently most common hypertext application, follows these rules and is based on a three-level architecture: the whole internet serves as the Web's database. The HAM consists of HTML (HyperText Markup Language) and HTTP (HyperText Transfer Protocol). FTP (file transfer protocol) and gopher servers can also be part of the HAM. Graphic web browsers are eventually used to display documents on the computer screen.50 At present, Netscape Navigator and Communicator are the most popular user interfaces.
George P. Landow, professor of English and art history at Brown University, USA, suggests a poststructuralist approach to deal with hypertext. Thus, it is necessary to be familiar with the concepts of post-structuralism and deconstruction. In poststructuralist theory, the idea of textual order and structure is abandoned.51 The French structuralist critic Roland Barthes and his fellow-countryman, the philosopher Jacques Derrida, were two of the most innovative scholars in literary theory and criticism in the 1960s and early 1970s. According to Barthes, the author (or the concept of the author) is dead.52 He also claims that texts are either lisible (readerly) or scriptible (writerly) and can also be analysed and interpreted as such.53 'Readerly' means that a unilateral relationship between a text and a passive reader exists, in other words he or she accepts and can easily understand the meaning of a text without much intellectual effort, since a real world with real characters and events is presented. A 'writerly' text, on the other hand, establishes a bilateral or even multilateral relationship between text and reader, that is to say, the text requires an active and attentive reader who has to find out its meaning.54 Barthes's theory is important for the distinction of traditional text and hypertext. The former has an indirect communicative function: the author communicates with the readers but they cannot communicate with the author. The latter introduces a direct, or mutual process of communication which means that the readers are allowed to make alterations and contributions to the text. They are therefore involved in the production of a document and become authors themselves. The concept of deconstruction, which is a favourite idea of poststructuralism, was introduced by Jacques Derrida. He proclaims that a text has not just one but many different meanings, and should be seen as an endless stream of signifiers, with words only pointing to other words and without any final meaning.55 Hypertext has similar characteristics. Landow believes that "hypertext (...) may fulfill certain claims of (...) poststructuralist criticism [and] provides rich means of testing them",56 especially Barthes's ideas about the readerly and the writerly text.
Similar to Nielsen, Landow compares hypertext with traditional text and comes to the conclusion that it is necessary to give up "conceptual systems founded upon ideas of center, margin, hierarchy, and linearity and replace them with ones of multilinearity, nodes, links, and networks".57 In this respect, four meanings of the term network have to be taken into consideration. First, the printed text takes the form of blocks and nodes which are linked together in a network. Second, the collection of these blocks and nodes by a single author or by multiple ones forms a network. Third, the term network is related to the notion of an electronic system including computers, cables, and connections. The fourth meaning refers to the usage of the term in critical theory where all writing is called a network.58
He adds that hypertext is "to be read nonsequentially or in a nonlinear mode."59 Traditional text, however, can also have a nonlinear structure. Footnotes are the simplest examples of such a structure because they interrupt the linear order of texts. Nonlinear writing is the most complex instance because traditional concepts such as structure, beginning and end, or unity of a text are abandoned. Nonlinear narrative occurs in many literary works, for example in James Joyce's Ulysses and Finnegan's Wake, Laurence Sterne's Tristram Shandy, Graham Swift's Waterland,60 John Fowles's The French Lieutenant's Woman, or Thomas Pynchon's Gravity's Rainbow. Texts such as these are suitable for conversion into hypertext form. The easiest way is to preserve the linear structure and to make additions, e.g. commentaries, glossaries, or links to other related texts. Such additions change the textuality of the text which is now part of a network of documents and sources.61 Therefore, hypertext as well as traditional literary text are in intertextual relation to many other pieces of writing. Barthes, for example, believes that "'any text is an intertext; other texts are present in it, at varying levels, in more or less recognisable forms: the texts of the previous and surrounding culture'."62 He concludes that "this text is a galaxy of signifiers not a structure of signifieds; it has no beginning; it is reversible; we gain access to it by several entrances, none of which can be authoritatively declared to be the main one".63
Applied to the new medium, these ideas mean that a hypertext system has no centre. Decentralisation is in fact one of the main structural characteristics of hypertext. The system allows the users to create their own centre by reading a particular document - or part of that document - in a particular way, and so even marginal parts may become central. Consider the following excerpt from William Butler Yeats's poem The Second Coming which was written in 1919:
Turning and turning in the widening gyre
The falcon cannot hear the falconer;
Things fall apart; the centre cannot hold;
Mere anarchy is loosed upon the world,
The blood-dimmed tide is loosed, and everywhere
The ceremony of innocence is drowned;
The best lack all conviction, while the worst
Are full of passionate intensity.64
This passage can be interpreted in terms of hypertext: The negative connotation of "the centre cannot hold" becomes positive because a hypertext document has multiple centres. The meaning of "anarchy" is ambiguous: on the one hand, anarchy is a progressive aspect of hypertext since all users participate in an unstructured system which provides new possibilities, such as making alterations and free choice of navigation. On the other hand, chaos, loss of authority, uncontrollable expansion of the system ("the widening gyre", "Things fall apart"), and a lack of quality concerning the presented material ("The best lack all conviction, while the worst / Are full of passionate intensity") are also implied. In this respect, however, it is necessary to distinguish between read-only hypertext documents where readers can do nothing but choose their way of navigation, and systems which allow the readers to make contributions.65 Most of the currently available hypertext systems are read-only documents since their authors do not provide users with the possibility to edit pages or parts of the system directly. The facilities for the readers are restricted to sending emails to the authors who might, or might not, consider and enter submissions. For this reason, such documents are closer to printed text than to digital technology. Landow adds that hypertext systems have "two fundamentally different kinds of structures, the first closely reliant upon that of the linear book and the second realizing the dispersed, multiply centered network organization inherent in electronic linking."66
Links consist of verbal and nonverbal elements, i.e. words and phrases or graphic symbols and images. Hypertext systems include both. The cursor is, for example, converted into a hand when it is moved over a verbal link, or a graphic symbol constitutes the link to help the user to navigate in the system. Since hypertext documents are almost always judged by their visual appearance, graphics are not only used for navigational purposes but also "rhetorically to supplement system design".67 The use of visual elements in hypertext goes as far as answering Derrida's request "for a new pictographic writing as a way out of logocentrism".68 In short, the term logocentrism is paradigmatic of Western culture and suggests that the only true meaning is centred on words and on concepts (both means logos in Greek).69 Thus, hypertext systems including graphic signs and symbols become universal and multicultural means of communication,70 education, and research.
For scholars and students, hypertext offers the possibility of a reader-centred way of dealing with a text because they choose the material they want to use within the context of a large network. Quick and easy access to resources, electronic texts, and advanced material proves especially helpful for users who have only an inadequate library at their disposal. In such a case, the hypertext system becomes the library. For research purposes, Landow is in favour of "bidirectional links and efficient navigational devices"71 because one will otherwise be discouraged from using these systems. Any user should also be able to contribute material to the system.
Philippe C. Duchastel believes that "'hypermedia tools should be viewed not principally as teaching tools, but rather as learning tools'".72 Landow concludes that "hypertext demonstrates that its intrinsic capacity to join varying materials creates a learning environment in which materials supporting separate courses exist in closer relationship to one another than is possible with conventional educational technology."73 As a consequence, hypertext systems provide new resources and produce a new research environment. However, practical use will have to show whether the new medium will be more efficient than traditional devices.
32 Landow, Hypertext, p. 14.
33 Nielsen, cf. pp. 33ff.
34 Landow, Hypertext, p. 14.
35 Nielsen, cf. pp. 34, 65.
36 Ibid., p.1.
37 Ibid., p. 321.
38 Ibid., cf. p. 1.
39 Ibid., p. 2.
40 Peter Schnupp. Hypertext (München und Wien, 1992), cf. pp. 134 ff.
41 In the simplified system above, the user might not be able to return directly from node 'C' to node 'A', for example.
42 Kuhlen, cf. pp. 132 ff.
43 Nielsen, p. 249.
44 Ibid., cf. pp. 249 ff.
45 Ibid., p. 4.
46 Ibid., p. 6.
47 Ibid., cf. p. 157.
48 Ibid., cf. pp. 131 ff.
49 Ibid., p. 67.
50 Ibid., cf. p. 178.
51 Jacques Derrida, Writing and Difference (London, 1978), cf. pp. 157 f., 278-293.
52 Roland Barthes, Image, Music, Text (Hammersmith, 1977), cf. pp. 142-148.
53 Ibid., S/Z (New York, 1974), cf. p. 4.
54 Ibid., cf. pp. 4-6.
55 Derrida, Of Grammatology (Baltimore, 1976), cf. pp. 6-26.
56 Landow, Hypertext, p. 11.
57 Ibid., p. 2.
58 Ibid., cf. pp. 23 ff.
59 Ibid., p. 126.
60 Ibid., see the discussion of these novels on pp. 102 ff.
61 Ibid., cf. pp. 35.
62 H. P. Mai, "Bypassing Intertextuality." In: Heinrich F. Plett (ed.), Intertextuality (Berlin, 1991), p. 42.
63 Barthes, S/Z, p. 5.
64 W.B. Yeats, The Poems. Edited by Daniel Albright (London, 1990), p. 235.
65 Landow (ed.), Hyper/Text/Theory (Baltimore, 1994), cf. pp. 9.
66 Ibid., p. 23.
67 Landow, Hypertext, p. 46.
68 Ibid., p. 43.
69 John Peck and Martin Coyle (eds.), Literary Terms and Criticism (Houndmills, 1993), cf. p. 139.
70 Landow, Hypertext, cf. pp. 43 ff.
71 Ibid., p. 178.
72 Ibid., p. 121.
73 Ibid., p. 127.