Prof. Dr. Mike Sandbothe
Media shape our image of reality. This applies for media in the broad, in the narrow, and in the narrowest sense. By media in the broad sense I understand space and time as forms of intuition. They function as the foundational media of our perception and cognition by making objects synthesizeable as identifiable entities. This insight lies at the root of the "Copernican revolution” with which Kant laid the foundation of modern philosophy. Post-Kantian philosophy, from Hegel, Nietzsche, Heidegger, Dewey, Cassirer and the late Wittgenstein, through to Derrida, Goodman, Gadamer and Rorty has demonstrated that the strength of this foundation lies not - as Kant thought - in its transcendental signature, but far more in its flexibility, openness and changeability. Our spatio-temporal "ways of worldmaking"1 are not a rigid, uniform and ahistorical apparatus. The media of human construction of reality are shaped far more by pictorial, linguistic and textual sign systems which are historically contingent and culturally divergent.
Images, language and writing are what's meant when I talk of media in the narrow sense. They have been at the centre of many philosophical discussions in the twentieth century which were mostly concerned with identifying one or more of these media as being a binding base structure for human understanding of reality altogether, or - at the very least - as the foundation of the world-picture characteristic of Western culture. The spectrum reaches from analytic philosophy's "linguistic turn"2 and the diverse misunderstandings triggered by Derrida's early concept of a philosophical "grammatology"3 in the realm of postmodern thinking, through to contemporary proclamations of a "pictorial turn"4.
Today it is becoming impossible to ignore the fact that neither media in the broad sense nor media in the narrow sense represent fixed, unchanging structures which offer a firm footing for philosophical theory. The way we deal with them is also dependent on institutional and technological developments which take place in the realm of media in the narrowest sense, that is, of technical media of transmission. This already holds for the influence which printed media, radio, and above all television have attained over our understanding of space and time as well as over our use of images, sounds and letters.5 The entangled conditions between media in the broad, narrow, and narrowest sense are becoming more significant still through the new importance which interactive data networks like the Internet are attaining for our perception and semiotic practice. With interactive data-networks the digital revolution is becoming the driving force of a transformation which is reorganizing our practices of symbolic actions. It is this transformation process that I will be looking into in the following deliberations by looking at the semiotic changes which our dealings with media in the narrow sense - that is, with images, language and writing - experience in the Internet.6
The Internet is not a radically new medium. It is far more a hybrid formed from media already familiar to us. The computers which are networked by high-speed telecommunications lines in the Internet link and transform applications, forms of usage, and contents familiar to us from television, telephone and face-to-face communication, from video, radio and the print media. All the same, the media-hybrid Internet does not consist - as the marketing slogan multimedia suggests - merely of a simple summation or a diffuse mixing of different media. The Internet is far more a highly complex and extremely sensitively organized transmedium in which a multitude of small innovations have condensed to create the impression on the whole of a 'new medium'. The transmedium Internet can be described semiotically as a digital mesh of images, language and writing - the sorts of sign which, until now, were distinctly separated from one another, but which in hypertextual conditions enter into precisely describable entanglements and whose specific features change in a scientifically reconstructable manner.
The World Wide Web's (WWW) graphical user interface comprises the functional centre of the Internet. It was developed in 1989 by Tim Berners-Lee and Robert Cailliau at the European laboratory for particle physics (CERN). The first PC versions of browser programms for navigating the World Wide Web, named Mosaic, were introduced in 1993 by the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana / Champaign. Netscape, currently the most widely used WWW browser, was developed in 1994. In browser programms the conventional text-based Internet services (e-mail, Netnews, IRC, MUDs and MOOs) are integrated in a graphically reworked form. Whereas these conventional services were oriented toward the model of linear textuality, in the World Wide Web a qualitative transition to nonlinear hypertextuality takes place. In order to make clear the changes in our sign usage which this transition brings, it is helpful to remind ourselves to begin with of the classical distinctions which have determined our dealings with signs until now.
In philosophy language and writing, as abstract and arbitrary sign systems, have traditionally been contrasted with images, as a concrete and natural medium of representation. In this images are afforded a peculiarly ambivalent status. On the one hand, in the tradition reaching from Plato to Hegel, they appear as "an imitation of looks"7, that is as intensified appearance. The media of language and writing, supposed to be more resistant to mere appearance, are opposed with the specious image. On the other hand the image functions in the mainstream of Western tradition, in which cognition is thought of as representation and truth as adequaetio, as a positive exemplar.8 Language has been interpreted ever since Aristotle as a tool for the arbritrary designation of mental images (ideas) representing reality which "are the same for all"9. Correspondingly, writing came to be degraded as a tertiary supplement. Being phonetic, it served, according to this tradition, only to materialize the phonological signs of spoken language and to make them storable. The ideal to which language and writing are therewith subjected in equal measure is the method of adequate, interpretation-neutral rendering derived from the model of images.10 Wherever language and writing are unable to fulfil this ideal they find themselves exposed to the suspicion of deception, a position occupied by the image in book X of Plato's Republic.
In recent discussion, prompted by Derrida, Goodman, Gadamer, Rorty and others, images are no longer apprehended as being demarcated from signs, but as sign systems in themselves, to be analysed according to the model of language and writing. Frequently, however, certain traditional presuppositions are adhered to. So, for example, it's usually assumed that the difference between linguistic, textual and pictorial signs is a difference founded in a binding manner in the semantic and / or syntactical structure of the respective system of signs.11 These assumptions contrast with the thesis, which goes back to the late Wittgenstein, that a sign is first defined through its usage as a sound, as a letter or as an image.12 Several authors however insist that even in conditions of a usage-theory of signs there is a unitary way of using something as image, as language or as writing.13 This view is based on the idea that certain features of the usage can be named which distinguish 'image games', 'language games' or 'writing games' as being image games, language games or writing games. These general features are supposed to permit the internally unitary definition of the various sign games and the clear division of the different sorts of sign from one another through a usage theory of signs. To this it must be objected that a consistent execution of a pragmatic usage theory of signs would indicate that we are dealing with complex bundles of image, language and writing games which will, even on the level of usage, exhibit no unitary feature common to all elements of the respective set. The metaphor of "family resemblances" was introduced by Wittgenstein to describe complex entangled relationships of this type.14
In addition to the internal entanglement of images, language and writing comes the external entanglement which determines the relation of the three sorts of sign to one another. Just as no general essential feature can be identified to define the image as an image, language as language, and writing as writing, no firm dividing lines can be fixed between the different types of sign. Pictures, sounds and letters are always intertwined or demarcated relative to, and dependent on the institutionalized technical media which set out the framework for their use. The modern media system, in which audiovisual and print media were clearly detached from one another, suggested strict demarcation between the sorts of sign. The World Wide Web's transmedia mesh of signs partly does away with this separation and redefines the relations.
Before I begin dealing with the transmedia mesh of image, language and writing which characterizes the World Wide Web, I would like to talk about text-oriented communications services.15 The Internet's text-based communications landscapes are older than the World Wide Web.16 Interesting changes in practical interactions with signs can already be ascertained by looking at the use of these simple services. In IRC, MUDs and MOOs writing functions as a medium of direct synchronous communication between two or more conversation partners who are physically separated and who, as a rule, have never seen each other. Up to a point the anonymity specific to the textual medium of the book is connected in "on-line Chat" with the synchronous interactivity and immediate presence of the conversational partner which characterizes spoken language in face to face communication. In "Computer Mediated Communication" features which previously served as a differential criteria for the distinction between language and writing are becoming entangled.17The transitions between language and writing become fluid. The traditional distinction of spoken language as a medium of presence is undermined by the 'appresent presence' of the participants in the written conversation of on-line Chat. It is this performative writing of a conversation in which language is interactively written instead of spoken, that I call the scriptualization of language.
I think it's important to point out expressly the proximity of interactive text-based communications landscapes to everyday face to face communication. For in current discussions this is not only often ignored, but more or less denied. Thus in her essay Interaction, Interactivity and the Personalization of the Mass Media18 Elena Esposito rejects the possibility that "telematic communication" enables "simultaneously personalizable and non-anonymous communication"19 from a system-theoretical perspective. In founding her thesis, one completely appropriate for interactive television and parts of the World Wide Web, the former Luhmann student highlights that in Chat fora only anonymous communication is found and that this is not personalizable because you would not be able to discern whether you are dealing with people or with so-called robots instead, i.e. with interactive programms.20
There is no doubt that, to begin with, a large part of Chat communication does not take place using real-identity names, but under the cover of a pseudonym. But it seems important to me to contrast clearly these forms of - as you could say - 'secondary' anonymization from the structural anonynimity which we are familiar with on the reception side of printed media or television. The Chat participant does not remain nameless, rather the condition for her participation is that she gives herself a name. To this extent Chat communication is, structurally speaking, personal communication. Even though the participant chooses a pseudonym as his name, he is nonetheless present as a 'persona', that is as a mask, as an acted identity. Of course there always exists, in addition, the possibility of replacing the acted identity with the real identity, i.e. to personalize the communication in an authentic sense. The danger, cited by Esposito, that in reality a machine might be behind the supposedly authentic person you think you're communicating with is negligible given the current state of AI development. Whoever's had any contact with a robot program, which lack precisely the capability for individual and context-sensitive communication, will know just how easily and quickly human-machine communication is to be recognized as such and to be discerned from human-human communication. This also applies to the 'Intelligent Agents' currently being developed, which, incidentally, are not primarily a matter of human-machine communication, but a form of machine-machine communication which can be programmed according to our individual interests.
A second example of the tendency towards media-theoretical overstatement of the differences between on-line communication and face to face communication to be observed in current debate is found in Sybille Krämer's essay From the Myth of 'Artificial Intelligence' to the Myth of 'Artificial Communication'.21 In this the Berlin-based philosopher expounds two theses, formulated as essential definitional traits of Internet communication. The first reads: "The electronic network, to the extent that it's used as a communications forum, has the character of a framework which ensures that a type of communication establishes itself on the Net which - in terms of a terminological distinction between 'play' which relieves from the day-to-day and 'seriousness' which reinforces the day-to-day - belongs to play"22. The second thesis, based on speech-act theory, serves to substantiate the first and states that "the communication in electronic networks" is based on "the redundancy of the illocutory and paracommunicative dimension of our symbolic dealings linked with personality and authorship"23.
Krämer's theses apply, if at all, only for certain forms of use of the Net which can be observed particularly in fictional communications landscapes such as MUDs and MOOs. In contrast, the everyday use of e-mail (e.g. for business communication), the use of mailing lists for the exchange of technical information or the application of the World Wide Web for teleteaching and for commercial uses such as teleshopping or telebanking show that there are a multitude of 'serious' forms of use of the Net which include paracommunicative dimensions in an essential manner. And even for MUDs and MOOs it must be underlined that in fictional contexts too play can very quickly become something serious, a personal conversation can very quickly arise from pseudonymous communication. From fictional roleplays virtual communities have often come about, and from these quite real and genuine friendships, and even marriages with church's blessings.
So the philosopher Eva Jelden, from Leipzig, is by no means wrong in emphasizing that the increasing reality of the virtual is decisive for the current development of the Internet. This, however, is characterized, according to Jelden, precisely by the inclusion of paracommunicative dimensions. Jelden writes: "With every click of the mouse I actually move something in reality, provide information about myself, transfer money, trade, and so on."24 Jelden's claim is admittedly to be limited to the extent that the influence of paracommunicative aspects is not, as Jelden implies, characteristic of 'every' communicative act in the Net, but rather for a certain, reality-related mode of Internet usage.25
The semiotic consequences which result in view of the World Wide Web altogether are more complex than the effects described in the domain of interactive communications services. By integrating within it the text-oriented Chat channels, it takes on board the textual variant of dialogue made possible by these services. Alongside the scriptualization of language which takes place in the communications services, there are two transformational tendencies in addition to this which are specifically characteristic for the transmediative and hypertextual World Wide Web: the 'pictorialization of writing' and the 'scriptualization of the image'. I will begin with the first of the two named tendencies. It is expressed in both the graphic handling of phonetic script and in the rehabilitation of non-phonetic script.
Both aspects of the pictorialization of script were anticipated by Jay David Bolter in his 1991 book Writing Space. The Computer, Hypertext, and the History of Writing. Attention is drawn to the first aspect - the pictorial use of phonetic script - by Bolter when he points out that the use of automatic structuring programms within the framework of word processing has the effect of making "text itself graphic by representing its structure graphically to the writer and the reader"26. The World Wide Web's networked hypertextual system radicalizes this tendency towards the pictorialization of script, one already fundamentally inset within "electronic writing". In hypertextual conditions writing and reading become pictorial operations. The writer creates a net-like structure on the screen, a rhizomatic image of his thoughts. This image is many-sided, associative and complex. It consists of a plurality of different paths and references which the reader forms into individually varying textual images, which result from the interplay between the text's open structure and the interests and perspectives of the reader.27 Hermeneutic operations and interpretory processes, which in reading a printed text take place in the reader's mind alone, become visible in hypertextual conditions as reading tracks which in navigated reading contribute to the constitution of the text on a software level. The entire hypertextual mesh of icons, digital images, audio and video sequences, as well as linear and nonlinear texts can, against this background, be metaphorically described as an image-like structure, i.e. as a 'textual image' or a 'text image'.
Bolter himself, however, in his essay The Internet in the History of Writing Technologies28, makes hardly any use in his semiotic analysis of the World Wide Web of the possible extensions, accentuated above, which follow from his 1991 book. Although he elicits with great clarity the way in which hypertext in the World Wide Web "is first produced in a process of rearrangement between the reader and the (absent) author(s) who built the corresponding links into the text"29. He does not, however, consider the pictorialization of script which takes place at a semiotic level in this interaction. Instead he emphasizes that for the relationship between image and writing: "All the same, the distinction between word and image does not break down completely in electronic writing. Or, better, the distinction collaborates - only to confirm itself once again anew"30.
Bolter is surely right when he emphasizes that in the World Wide Web too the difference between word and image doesn't break down completely. His conclusion, however, that "the distinction collaborates - only to confirm itself once again anew" seems problematic to me. In suggesting this, Bolter does not consider the possibility of the occurence of a semiotically relevant alteration in our handling of images, sounds and letters . If such a possibility is taken into account then the current shifts would not just be a matter of a simple iteration of a quasi-transcendental opposition, but of a medium-specific transformation of the semiotic base structure. The semiotic difference between images and writing neither breaks down completely, nor does it remain stiff and unaltered: rather it constitutes itself anew in the context of a medium-specifically altered usage, i.e. it formulates the semantic distinction it makes in a modified manner and accentuates aspects different from those emphasized until now.
Our talk of 'writing' in hypertextual conditions is indeed beginning to take on properties and aspects traditionally attributed to images. As such reading and writing in the World Wibe Web is not to be separated from visual design, theatrical production and aesthetic organization of graphically arranged textual signs. A well-composed hypertext is one put together from aphoristic thought images, each representing a scene meaningful in itself, even when absolved from the (variable) context, and at the same time offering significant transitions into other scenes with which interesting connections exist. The situation of the text in space, the tactile distinction of individual complexes of signs as clickable links, the variably formable structure of the text's background, or the opportunities provided by Java to set letters in motion and embed them in graphic scenes - these are all aspects of what, in summary, I call the pictorialization of phonetic writing.
The second aspect of the pictorialization of writing - the rehabilitation of nonphonetic writing - had also been identified by Bolter in his 1991 book as a fundamental feature of electronic writing space. Using the example of the Apple Macintosh Desktop he makes clear that icons function as "symbolic elements in a true picture writing"31. And he continues: "Electronic icons realize what magic signs in the past could only suggest, for electronic icons are functioning representations in computer writing"32. In summary Bolter finally underlines: "Electronic writing is a continuum in which many systems of representation can happily coexist"33. In his on-line article Degrees of Freedom,34 which is yet to appear in print, Bolter, with the World Wide Web in mind, modifies his coexistence thesis of 1991, one which resulted from the analysis of hypertext systems on stand-alone computers. Bolter writes, "If the World Wide Web system began as an exercise in hypertextual thinking, it is now a combination of the hypertextual and the virtual. But the hypertextual and the virtual do not always combine easily. Usually the graphics and photographs tend to muscle the words out of the way"35.
The new distinction introduced by Bolter between the 'hypertextual' and the 'virtual' indicates that in the Degrees of Freedom essay too a medium-specific transformation in the system of signs is not considered. According to Bolter the hypertextual is characterized by the preponderance of writing, the virtual by the dominance of pictures.36 Instead of acknowledging the complex entangled conditions between writing and images which are beginning to emerge in the World Wide Web, Bolter has replaced his old coexistence thesis with a new theory of rivalry, according to which the different sign systems, precisely by battling with one another for dominance in the networked computer's writing space, remain unchanged in their basic internal properties. On the one hand Bolter concedes that "the difference between the hypertextual and the virtual representation is not simply the difference between words and images"37. On the other hand, however, he maintains that even in cases in which "images (...) serve as electronic links, for example, in a World Wide Web page or as icons in a multimedia presentation (...) the sign remains iconic"38. What's problematic about this statement is the theoretical recourse to the conventional meaning of the word 'iconic'. Bolter attempts to describe a change in the use of signs without being prepared to relativize the evident nature of the old meanings of terms like 'image' or 'writing' which were established in nondigital media conditions. It will become particularly obvious that such a relativization is necessary in suitably discerning the changes which are taking place when we turn to the third transformation, which is becoming a fundamental experience in our use of signs: the scriptualization of the image.
In the World Wide Web too images often still function according to the traditional pattern as a kind of quasi-reference. They interrupt the flow of links and represent artificial end points of menus, i.e. impasses in hyperspace. But at the same time there are more skilled forms of image presentation on the Net which are more appropriate to the hypertext medium. This involves furnishing different areas of the image with "source anchors" which respectively refer to various "destination anchors". The image itself then functions as a hypertext. If I activate a link within an image I am referred to other images or texts. The image no longer appears as the referent and termination of a menu, but becomes a sign itself with references to other signs. In the same way as 'written' hypertexts, which are no longer organized internally in a linear but a rhizomatic and pictorial manner, the hypertextual image functions as a semiotic intersection in the unending referential framework of the "Docuverse"39.
If you consider not only the external scriptualization of the image but also the internal data structure of digital images then it becomes clear that the images composed of pixels have textual character as a result of their technological structure. With the corresponding editor programs, the elements which comprise a digital image can be exchanged, moved and altered just as written characters can. In this way images become flexibly editable scripts. In the digital mode the image loses its distinguished status as a representation of reality. It proves itself to be an aesthetic construction, a technological work of art whose semiotics result internally from the relations of pixels and externally through the hypertextual references to other documents.40
In the concluding part of his essay The Internet in the History of Writing Technologies Bolter looks at the significant developments which are taking place in graphic medium technology. In order to deal with the semiotic transformations which are dawning here, and yet be able to retain the terminological fixation on the established evident meanings of 'image' and 'writing', Bolter shifts the claim regarding the evident nature of the difference between image and writing for which he pleads from the theoretical to the object level. Thus he concedes in the concluding paragraph of his article that "more imaginative and intelligent forms of hypertextual communication exist in which words and images interact with each other as a matter of course"41. Hence, on the theoretical level the scope for a semantic reorganization of the image-writing difference is opened up. But the terminological consequences which result from this are nonetheless not drawn by Bolter. This omission is justified with recourse to deficits which, according to Bolter, can be observed on the object level: "Today too even the most refined observer of the World Wide Web can be tempted to forget the complex character of a Web page, so as to concentrate on the static or moving image as a direct representation of reality"42.
It is surely purposeful and important in theoretical semiotic and media-theoretical analyses to consider also the persistent force peculiar to the established meanings of terms such as 'image', 'language', and 'writing'. But to understand adequately the transition which is taking place continuously from old to new modes of use it is necessary, in addition, to retrace precisely the terminological shifts, through which in future the 'actual' meanings of the terms in question might be determined. This can be achieved if we pinpoint for examination the semiotic changes, which are perhaps only latently experienceable at the moment, on both the theoretical and objective level and in this way attempt to describe changes in meaning, so to speak, in an anticipatory manner. At the same time, from a future-oriented analysis such as this - one opposing the classical understanding of theory which bound academic study to the archaeological reconstruction of what's past - media-pragmatic and media-political fields of application can be inferred. The Internet's current development into a mass medium which is beginning to shape modern societies' communications behaviour and information practices far beyond the realm of academic elites43 marks out the largely open domain of a micropolitics of the sign - a semiopolitical domain in which it will be decided which dimensions of the new mass medium Internet will be made available to people and which won't.
Translated by Andrew Inkpin