Outside the environment of progressive, well-funded academia, hypertext made little impression on the "real world" until the early 1990's, when popular interest in the Internet boomed, and to many the World Wide Web has become synonymous with the Internet.⁵ The number of new users continues to rocket and claims that the Internet is a new mass medium may not be as far-fetched as they once appeared.

      All media operate within a specific code or set of codes. Codes can be defined as a "set of rules or interpretative device known by both transmitter and receiver, which assigns a certain meaning or content to a certain sign".⁶ Codes depend on specific knowledge, such as that gained passively from inhabiting a specific culture, or actively by studying a particular subject. Nobody wants to have to read a manual to understand a mass medium, indeed if that was necessary it probably would not be a mass medium. In Fiske's terms, the Web is in the process of transition from a narrowcast code, aimed at a limited audience and requiring active learning, to a broadcast code "shared by members of a mass audience...learned through experience".⁷ So how does this new medium make itself understood? Are the codes of Web navigation specific to the Web, or are they drawn from more familiar media? What knowledge is called upon during Web navigation?

      C. S. Peirce discusses the sign at the level of the object, identifying three categories: "...an Icon (where the sign relates to its object in some resemblance with it)...a Symbol (where the sign relates to its object by means of convention alone)...[and] an Index (where the sign relates to its object in terms of causation)".⁸ Iconic signs are motivated, as a natural bond exists between signifier and signified, whereas the symbol is the unmotivated, arbitrary signifier as discussed by Saussure, which is related to the signified only by convention.

      Some of the "icons" employed within the toolbar by the two most popular browsers (Netscape Navigator and Microsoft Internet Explorer ), are in fact symbols. The Back and Forward functions are signified by conventionalised symbols (arrows), the Back arrow pointing left due to the Western convention of reading left to right. Despite the global reach of the Web, it remains predominantly a Western/American product. These functions also have the effect of linearising the text which has already been visited, thus orientating the reader within the hypertextual network.

      Other icons within the toolbar are icons in the Peircian sense, as they are stylised representations of objects. However, the objects they represent are not the function that the icon actually signifies. Instead the object depicted is a metaphor.

      As Chandler states, "Metaphor expresses the unfamiliar...in terms of the familiar...".⁹ The unfamiliar, abstract concepts of Web navigation are concretised in icons depicting familiar concepts "...[evoking] meaning by transferring qualities from a referent to a new object through implied comparison..."¹⁰ The use of metaphor is widespread within computing, for example the "desktop" of the Windows operating system. Conventions of the traditional paper office have been extended to its replacement, the electronic office.

      Many metaphors encountered within Web navigation are of a spatial or geographical nature. The terms "navigate" and "explore" themselves employ a "travel" metaphor, as do "surf" and "information superhighway", though the user is static in front of a machine. Superficially an extension of the common spatial metaphor of communication, this terminology implies the motion of the reader, rather than the information. This motion metaphor spreads to the browser toolbar with Netscape's "Stop" icon, depicting a traffic light. A red light signifying "Stop" is common to many cultures, and employed in everyday situations, as is the abstracted, conventionalised red circle (as warning, halt or danger) in Internet Explorer. On the toolbar we also find the icon "Home". This choice of phrase, while expanding on the spatial/geographical metaphor described above, appears curiously loaded. The word "Home" connotes ownership, habitation, and belonging. Until otherwise instructed, most Web browsers use their company's site as the default home page, perhaps as a means of associating the company and the product with the qualities ascribed to "Home".

      Physical buttons are the primary way of interacting with electronic equipment in the "real world", and this has been extended metaphorically to the virtual space of the screen. Functional graphics, when "touched" with the cursor, become illusionistic raised buttons, appearing to depress when clicked. Although this is a good indication that the intended function has been selected, there are many conceivable ways that this could be achieved. Whilst drawing upon familiar concepts, button graphics also compensate for the lack of tactility of the electronic text. While touch is associated with acting upon electronic text, via a keyboard or mouse, these are physically distinct objects from the monitor. Even then, the image on the monitor is only an expression of the text, the text itself being stored in a form invisible and intangible without interpretation by the machine. This tactile metaphor may help the reader to engage with the text on a more "human" level, lessening the sense of alienation from the text, which is undoubtedly a problem in a culture accustomed to a physical, manipulable object, the printed text

      One indication of the presence of a hyperlink is the appearance of the cursor. The cursor usually takes the form of the generic symbolic pointing device, the arrow. However, when moved onto a hyperlink, the cursor becomes an iconic representation of a hand with one finger outstretched, implying pointing or touching. Again, the presence of a link could be indicated by other changes to the cursor. The iconic hand is more human, more active than the symbolic cursor, a tactile metaphor for the tactile click of the mouse.

      Links whose appearance changes when the cursor is passed over them are becoming common. This can be used to draw attention to a previously hidden link, but is often implemented within icons whose status as a link has already been made explicit by supplementary text. While being an attractive device, this animation gives the impression of the reader acting upon the text and, like the "active" cursor, expresses the (illusory?) concept of interactivity so central to the Web.

      It would be frustrating to have to run the cursor over every pixel of each Web page to locate hyperlinks, therefore, other cues are required within the text itself. By default, Web browsers display the simplest forms of hyperlink in a different colour from the main text (usually blue), and underlined. Simple hyperlinked images are bordered in the same blue. As Saussure describes, signs only mean by virtue of their difference from other signs. An opposition is set up between hyperlinked text and non-hyperlinked text, the linked section asserting its difference from the body of the text. Often, the text of the link is the phrase "Click here".¹² However, guides to Web design frown on such practice. Essentially, displaying the words "Click here" in blue and underlined is a tautology. This manner of distinguishing hyperlinks from ordinary text has become conventionalised to the extent that being blue and underlined signifies "Click here" in itself.

      The conventionalisation of the appearance of hyperlinks allows them to be embedded within the body of a text, indicating the presence of linked material without disturbing the narrative flow . This is probably the closest form to the idealised hypertext described earlier. Outside of academic texts, this form is rarer than one would think. In the late 90's commerce and leisure are higher on most Web users' agendas than the richness of the writely, decentred text.

      Identification of many forms of hyperlink relies on prior knowledge of titling. Titles act as signifiers, denoting the text to which they have been assigned. Titles generally appear above the text they refer to, in a larger script, different font, or underlined. Hyperlinks often consist of "floating" titles, sections of text visually distinct from the rest of the page, yet whose signified text is absent. This incompleteness and difference leads the reader familiar with the conventions of titling to extrapolate that the title itself is hyperlinked to the signified document.

      Despite the growth of multimedia content, the majority of Web page content takes the form of the written word, necessitating comparison with traditional print media. Long, dense "book-like" texts are rarely encountered, unless designed to be downloaded then printed out or read off-line.¹³ Most Web pages are constructed in a manner closer to the paradigm of illustrated magazines, combining short bursts of text, images and often advertising graphics.

      Hyperlinks used for navigating sub-pages of a Web site frequently take the form of a "Contents table", either as a standalone page or displayed down one side of each subpage . Contents pages consist of lists of titles, along with directions to locate the item signified by the title. This is also true of hyperlinked content lists. Tables of contents are employed in many other media, notably factual books and magazines. Such Web sites mimic the familiar act of reading a magazine, where short, discrete articles are read in order of interest rather than linear order. A contents list also has the effect of orientating the reader within the text, as if protecting against the "danger" of unlimited semiosis offered by embedded links. Contents pages define the boundaries of the text in a traditional fashion, and links to exterior texts are explicitly labelled as such, creating a conceptual punctuation of the text, akin to discarding one magazine and picking up another. Indeed, many large commercial Web sites offer few or no hyperlinks to documents exterior to the site, as if attempting to trap the reader within its circular, internal signification whilst maintaining the reader's illusion of self determination.

      As seen earlier, one way in which the Web makes itself comprehensible to new users is by drawing metaphorically upon familiar concepts. It seems ironic that the most pervasive metaphor is that of travel, due to the sedentary nature of Web browsing, yet the employment of this concept allows the reader to assign meaning to the unfamiliar actions required. It could also be hypothesised that this metaphor is a compensatory device, intellectually balancing the physical poverty of the pursuit, in the same manner in which tactile metaphors appear to compensate for the intangibility of the text.

      Discussion of hypertext from five or ten years ago now seems strangely idealistic. Although the possibilities identified by semioticians are still present, the implementation of hypertext within the Web closely resembles traditional print media. Indeed, the structure often imposed upon hyperlinks actively negates the radical, writerly, intertextual qualities previously envisaged. Undoubtedly, means of making the Web more accessible have been actively sought, drawing upon familiar paradigms. Without this familiarity the Web would still be an elitist, narrowcast code, indecipherable to the masses. The codes employed assume a literate audience, accustomed to varied forms of printed media, and to the basic operation of technological devices. This appears reasonable, considering the written linguistic content of the Web, and the fact that such codes have been naturalised within most "developed" cultures. It is easy to forget the relative affluence of a culture where the majority of inhabitants can access a computer, yet sobering to consider the percentage of the worlds population which excluded from this "global", "democratic" medium.

      The Web is so young and chaotic that generalisation and prediction is a risky business. Its growth has been accelerated not only by business and specialists, but by thousands of enthusiasts, meaning that any embryonic paradigms are no doubt broken frequently, and a representative sample of the millions of Web sites is unlikely to be gleaned by a lone individual. However, it appears that conventionalisation of some signifiers is beginning, as mentioned earlier with respect to hyperlinks. The non-mainstream browser Opera employs compact, abstract graphics, which are not augmented by text (though a description of the function appears at the touch of a cursor). Perhaps this is one indication of the beginnings of conventionalisation of these signs.

       Conventionalisation requires repetition, and a culture of "speakers" to recognise, stabilise and utilise the sign, and undoubtedly these conditions are in place. However, the tendency is still (perhaps rightly) to overstate the signification "...as if an underdeveloped reading public needed rigorous manipulation"¹⁴

References

      1.Quoted in Landow, George P., "Hypertext-The Convergence of Contemporary Critical Theory and Technology", 1992, p4
      2.Bolter, Jay David, "Electronic Signs", WWW document
      3.Each Interpretant (similar to Saussure's signified) becoming a further Representamen (signifier), thus producing another Interpretant ad infinitum.
      4.For further discussion of these issues see Landow, George P., "Hypertext-The Convergence of Contemporary Critical Theory and Technology" 1992, Chapter 1.
      5.The scope of the Internet is actually much wider, encompassing e-mail, newsgroups, Internet Relay Chat and many other network utilities.
      6.Dyer, Gillian, "Advertising as Communication" 1982, p131
      7.Chandler, Daniel, "Semiotics for Beginners" 1994, WWW document.
      8.Cobley, Paul and Jansz, Litza, "Semiotics for Beginners", 1997, p33
      9.Chandler, Daniel, "Semiotics for Beginners" 1994, WWW document
      10.Moriarty, Sandra, "An Interpretative Study of Visual Cues in Advertising", WWW document
      11.A section of the content of the page with the URL of the new page embedded within it, allowing a click on the hyperlink to automatically take the user to that page.
      12.This is often seen on simple Web pages, pages that have not been updated for some years and pages of large firms such as search engines and Internet service providers (which are designed as a first stop for new users).
      13.This is probably due to the physical discomfort of reading from a monitor for long periods, and the financial discomfort of the length of the phone call required reading a long text on-line.
      14.Eco, Umberto, "A Reading of Steve Canyon", 1965

Bibliography

      Barrett, Edward (ed.), " Sociomedia-Multimedia, Hypermedia and the Social Construction of Knowledge", MIT Press 1994
      Barthes, Roland, "Mythologies", Vintage, 1993 (first published 1954)
      Cobley, Paul, and Jansz, Litza, "Semiotics for Beginners", Icon, 1997
      Druckrey, Timothy (ed.), "Electronic Culture", Aperture, 1996
      Dyer, Gillian, "Advertising as Communication", chapter six of "Semiotics and Ideology", London, Methuen, 1982.
      Hawkes, Terence, "Structuralism and Semiotics", London, Methuen, 1977
      Eco, Umberto, "A Reading of Steve Canyon", 1965, from "Comic Iconoclasm", London, ICA, 1987, pp20-25.
      Fiske, J. and Hartley, J., "The Signs of Television", from "Reading Television", Methuen, London, 1980
      Hall, Stuart, "Encoding and Decoding in Television Discourse", in "The Cultural Studies Reader", Simon During (ed.), Routledge, London, 1993.
      Landow, George P., "Hypertext-The Converence of Contemporary Critical Theory and Technology", John Hopkins University Press, 1992
      Lane, Michael (ed.), "Structuralism-A Reader", Jonathan Cape, 1970
      Ryman, Geoff, "253-The Print Remix" Flamingo, 1998
      Silverman, Kaja, "The Subject of Semiotics", Oxford University Press, 1983
      Sim, Stuart (ed.), "The Icon Critical Dictionary of Postmodern Thought", Icon, 1998

World Wide Web Documents

      Bolter, Jay David, "Electronic Signs", http://www.rochester.edu/College/FS/Publications/Bolter Signs.html , Taken from Chapter Six of Writing Space: The Computer, Hypertext, and the History of Writing (Hillsdale, NJ: Lawrence Erlbaum Associates, 1991): 85-106.
      Chandler, Daniel, "Semiotics for Beginners", http://www.aber.ac.uk/~dgc/semiotic.html , 1994
      Colon, Carlos, "Semiotics in Hyperspace", http://php.indiana.edu/~ccolon/Semiotics/ccolon3.html, 1995
      Eco, Umberto, "From Internet to Gutenberg", http://www.hf.ntnu.no/Finnbo/tekster/Eco/ internet.htm ,transcript of lecture presented at the Italian Academy for Advanced Studies in America, 12 November 1996.
      "A Little Web History", http://www.newbie-u.com , author and date unknown.
      Gross, Larry, "Seeing Metaphor as Caricature", http://www.temple.edu/departments/anthro/ worth/sintro.html ,taken from Chapter Six of "Sol Worth and the Study of Visual Communication", date and publisher unknown.
      Moriarty, Sandra, "An Interpretative Study of Visual Cues in Advertising", http://spot.colorado. edu/~moriarts/viscueing.html , date unknown.
      Moriarty, Sandra, "Visemics: A Proposal for a Marriage Between Semiotics and Visual Communication", http://spot.colorado.edu/~moriarts/visemics.html, 1994.
      Sonesson, Gran, "The Multimediation of the Lifeworld", http://www.arthist.lu.se/kultsem/ sonesson/media_1.html, 1995