2 Virtuality and mediation

Many questions can be raised about the impact of generalised abstraction on our lives. Human activities will increasingly be based on representations of reality rather than on reality itself. (...) Virtual life is not real life and the representation of reality is not reality. There is concern that the abstract nature of much ICT use leads to a similar abstraction of reality. (HLEG 1997: 24)

2.1 The problematic concept of 'virtuality'

The notion of 'virtuality' is extensively used in descriptions of CMC; for example, the term 'virtual community' is used much more frequently than 'electronic community'. Despite, or perhaps owing to, its popularity, there are a number of problems associated with it.

First of all, virtuality is a polysemic term, so it allows considerable flexibility in its interpretation. Some general meanings of virtual are: artificial, computer-mediated, on-line, simulated, imitated, non-real and imaginary. When used in conjunction with social relations, additional connotations are: 'non-physically linked', 'lacking face-to-face interaction' and 'bodiless' (Sudweeks & Simoff 1999: 52; Fornäs 1998; Donath 1996; May 1995). Despite this semantic fuzziness, there are not many articles on CMC where the meaning is more closely specified. A closer specification of what the term represents is hence desirable.

2.1.1 Reality, abstraction and transcendence

Whether explicitly stated or implicitly indicated, most articles on CMC take the antonym of 'virtual' to be 'real'. As the quote at the start of this chapter indicates, one source for this view is the notion that something that is computer-mediated is not reality, but rather a "representation of reality". Logically this is a little curious, since if some entity E is real and R is a representation of E, then R and E are obviously not the same, but R is nevertheless real. A map is not the land area it represents, but it is still a real thing.

On this background the claim that use of ICT leads to some kind of "generalised abstraction", an apparently novel and undesirable condition, seems to disregard the fact that such abstractions already permeate our culture. For what is traditional mass media, literature, art, and language (one may even include human sensory interpretation in general), if not abstractions or representations of reality? In this respect CMC seems to bring little new.

Regarding CMC and on-line social formations as non-real facilitates a commonly held view that they have the capacity of transcending the 'real world' and "existing in an independent reality" (Kendall 1999: 60). Although written "with characteristic grandiosity", John P. Barlow's Cyberspace Independence Declaration expresses ideas that many has come to associate with the Internet:

I declare the global social space we are building to be naturally independent of the tyrannies you [the governments of the industrial world] seek to impose on us. (...) Cyberspace does not lie within your borders. (...) You do not know our culture, our ethics, or the unwritten codes that already provide our society more order than could be obtained by any of your impositions. (Barlow 1996)

By physically or metaphysically separating "cyberspace" from the conventional world, there is a hope that one can build a new and better virtual world where negative features of the 'real' world are eliminated. Such a fantastic view is also inspired by the seminal Neuromancer novel by Gibson (1984), with its mythical concept of "cyberspace", described as a "consensual hallucination experienced daily by billions of legitimate operators, in every nation...".

2.1.2 Virtuality and imagination

As mentioned, few writers on CMC specify what is, or should be, meant by 'virtuality'. Fornäs (1998: Ch. 4) is a notable exception. He equates 'virtual' with 'imaginary' in the sense that each communicative act creates an imaginary world (in the interlocutors' minds) due to the unavoidable process of interpretation, like when reading a novel. At the same time, he claims, the virtual is material in the sense of being conveyed by material entities in the form of computer chips, cables, electrons, and so on.

But although Fornäs avoids the conflation of some important distinctions, which the dichotomy between virtual and real reflects, there remains one conflation, namely of the distinction between imagination and mental representation. This conflation arises from his claim that any kind of communication, and thereby interpretation, involves the creation of an imaginary world. But an imagined entity is a pure psycho-social construction with no direct roots in the material world (e.g. thinking about a unicorn), whereas a mental representation has a more or less close correspondence to some real entity in the world.

This 'close correspondence' does not rule out the possibility of interpreting the world differently, and is as such also partly a social construction. But the material properties of what is interpreted ultimately provide bounds for the scope of interpretation. This difference between interpretation and imagination is important in order to understand the different characteristics of different types of on-line social formations, a theme that will be elaborated in chapter 4.

The conflation of the distinction between imagination and mental representation leads to the view that computerised fantasy worlds are real due to their having a physical basis in the material technology. But if someone imagines a unicorn and describes this imaginary creature to someone else via ICT, this does not make the unicorn real, despite the fact that there are observable physical events taking place within the technology. For these events do not represent the unicorn, but the description of it. What is real, on the other hand, is the description. However, the realness of the description does not primarily arise due to the measurable technical phenomena which occur as the description is electronically conveyed; it acquires its realness from the (real) symbolic interaction which takes place between the human beings involved.

The technology naturally plays a crucial role in enabling this interaction, but referring to the transmission of electrons or zeros and ones as being the crucial difference in real vs. non-real communication leads to a techno-centric reductionist conception of communication where the question of whether the interlocutors actually transfer meaning between each other becomes irrelevant. Benschop (1998), for example, asks why it is that people get excited from merely transmitting zeros and ones between each other's computers. The rational workings of the technology that transforms the runs of binary digits into something humanly comprehensible then becomes mystified into being "the metaphysics of cyberspace" (ibid.).

To make a brief summary: There is an important distinction between imagination and mental representation. Fantasy worlds should remain in the imaginary realm (whether computer-mediated or not), whereas CMC in itself belongs to reality. But it still remains to clarify what is meant by 'virtuality'.

In order to tackle this it might be useful to consider a source where the concept is quite widely used, and used in a fairly well-defined way. One such source is textbooks dealing with technical aspects of computer networks. Of necessity, technical terms normally exhibit a certain degree of precision, since they represent features of actual (or potential) working machines. Hence, the technical meaning of 'virtuality' may shed considerable light on other contexts where the term is employed.

2.2 Towards a dénouement of 'virtuality'

The technical meaning of 'virtuality' is best explained by way of an example. Consider a so-called intranet. This is a computer network typically used for communication within a single firm. This form of communication often contains private and sensitive information, which must be protected by using appropriate data security technology (electronic firewalls, password protection etc.). Such a network can then be called a private network, both in the physical and the logical sense.

Now consider a multinational firm that wants to build an intranet. A physically private network now requires interconnections (cables and so on) to be laid from one country to another, an endeavour that is likely to be prohibitively expensive. A solution is then to use the existing public computer network infrastructure (e.g. the Internet), and then add required security measures, not to the physical network, but instead to the data that is transmitted. This is achieved by using cryptographic techniques.

This means that a physically public network is used in a way that is equivalent to having a private network; for the multinational firm it is thus logically a private network. This combination of public network infrastructure and data security techniques is called a virtual private network (VPN). 'Virtual' is thus used as an antonym to 'actual', rather than 'real' (Tanenbaum 1988: 10; Watson 1997: 129). Other terms used for 'virtual' in this sense are 'logical' or 'emulated'. Note also that the VPN is a real network, it is not an imaginary entity. (Unless it has not yet been built, and it only 'exists' as an idea in the minds of some engineers.)

2.2.2 Virtuality and mediation

Virtuality is strongly tied to the concept of mediation. As an example, consider a conversation between the Russian and the American presidents during a summit. The oral communication then normally goes through an interpreter; it is mediated by this interpreter. No actual oral communication goes directly from one president to the other; the actual communication goes from one president to the interpreter and from the interpreter to the other president. It would, however, be ridiculous to claim that the presidents do not communicate; they communicate in a logical (although not physical) sense. Using the terminology of the previous section, they communicate really, but virtually. If the interpreter is good, the communication takes place more or less as if the presidents spoke to each other directly. The interpreter becomes transparent, as it were.

This social phenomenon is schematically modelled in fig. 2.1: The virtual communication between the two humans, H1 and H2, is facilitated by the actual communication via the human interpreter, H3.

A similar mechanism is at work in CMC (fig. 2.2), where the mediating entity happens to be technology (T):

The similarity of the figures is not meant to indicate that the medium is irrelevant. The medium will in fact profoundly affect the nature of the communication, as will be explored later. Instead, the similarity indicates that virtuality is nothing that is exclusively confined to technology.

It can be seen that whenever there is a medium involved, there is virtuality in the sense that there arises a conceptual split between the physical and the logical communication. This interpretation of virtuality has the advantage that it provides a material justification for claiming that CMC is as real as other forms of mediated communication. And since, as will be shown below, even face-to-face communication is mediated (and thereby in a sense virtual), there is no longer any reason to associate virtual with non-real (as this will render all forms of communication non-real), and also no point in designating e-communities as 'virtual'. In this way the dichotomy between technology-mediated communication and 'conventional' communication becomes less pronounced.

2.3 A deconstruction of human and machine communication

A computer network model based on a conceptual functional hierarchy (of which fig. 1.1 represents a simple example) has proven to be a powerful tool for developing network technology, and such a model is used by system designers, standardisation bodies and textbook writers alike (see e.g. Tanenbaum 1988). The model involves breaking a system down into separate processing layers, where each layer performs a subset of network tasks.

Two computers that have implemented this functional hierarchy in compatible ways are able to communicate with each other; they speak the same language, so to speak. This 'language' is called a protocol. The layered model acquires its power from the fact that each layer seems to be communicating directly with the peer layer on the other computer; it does not have to care about what the other layers do. This modularization greatly simplifies both the analysis and synthesis of a communication system.

One (complex) example of a layered model is the OSI (Open Systems Interconnection) Reference Model developed by the International Standardisation Organization, ISO (Tanenbaum 1988). For our purposes a simpler model will suffice (in fact, the Internet itself is based on a simpler model). Consider fig. 2.3. If a person H1 wants to communicate with another person H2, say via e-mail, she uses a computer to access an e-mail application program, e.g. Eudora, via a user interface (normally a keyboard and monitor). Note that H1 does not actually communicate with H2; the actual (physical) communication is with the interface.

When a computer sends an e-mail to another, it cannot send the e-mail text directly. First of all, the source computer is unlikely to be directly connected to the destination computer. Therefore the message must pass through a number of intermediate computers. The address of the next computer must be found and the message sent there. This function is achieved by the network layer. Moreover, computers can only transmit binary digits (bits, i.e. zeros and ones), not letters. So the letters of the message must be translated into bits before it can be sent. This is performed by the physical layer. The bits (represented by voltage levels or light pulses) are then transmitted across the physical medium (wire or fibre-optic cable) to the first intermediate computer. This is the actual inter-computer communication. The process is repeated for the other intermediate computers until the final destination is reached. No actual communication has taken place directly between the original source and the final destination, only virtual communication. Between each layer there appears to have occurred direct communication, but this appearance is caused by the workings of the levels below.

Technical treatises of computer communication are limited to the ensemble of technological 'boxes' depicted in fig. 2.3. This thesis aims to extend this model to the social domain. This might be done in several ways, depending on which social aspect that one seeks to highlight. In this section the aim is to demonstrate the general applicability of the concepts of virtuality and mediation, and to do this, the model is augmented by the human elements shown in fig. 2.4 (the intermediate computer is omitted for simplicity). Here the material aspect is emphasised, so the human 'interface' represents senses (for input) and generators of physically transmittable symbols (for output), primarily the voice and hands (for writing and gesticulating), but also other parts of the body (facial expressions, movement and smell). Sensory inputs are converted to neural signals that can be processed by the brain, which ultimately forms some kind of meaning based on the inputs.

This model is not supposed to indicate that the human brain is comparable in operation to a computer. The brain is not constructed according to some layered communications model, so in the human domain the model is overly simplistic and schematic. The purpose of the model is rather to indicate the omnipresence of mediation in both the spheres of the human and the machine. "Mediation is everywhere", according to Fornäs (1998: Ch. 2). All human communication is subject to processing, by the brain and by technology, processing that influence the ultimate result of the communication, the psycho-socially constructed meaning. Even the most direct form of communication, face-to-face interaction, the communication can be said to be virtual, since there are no direct/immediate/unmediated ways of conveying meaning, as illustrated in fig. 2.5.

The physical transmission medium is in this case visible light, aural airwaves, olfactory molecules, and so on. These physical phenomena will be sensed and subjected to a socio-culturally dependent interpretative process.

This illustrates that the salient difference between CMC and non-technologically-mediated communication is not that one is virtual/mediated and the other real/immediate, for both forms have both a virtual and an actual aspect, and both are real and mediated. On this basis there is a case for abolishing the virtual/real and the mediate/immediate dichotomies. The focus may then be shifted to an analysis of the specific characteristics of the mediating entities, in order to clarify in what way these characteristics influence the nature of the interactions. Any a priori distinction between the technical and the social seems to be unfounded (what then for instance to make of the role of spectacles or hearing aids, essential artefacts for many when engaging in face-to-face interaction?).

Another point of the above elaboration is to demonstrate the unlikelihood that CMC exhibits the reality-transcending or otherwise revolutionary properties that many commentators ascribe to it. In fact, the only point in the model at which such capacities may arise is in the transformation from sensory input to meaning, which happens to be outside the scope of the technical. Any 'revolutionary' properties of CMC are thus to be regarded as social constructions. No part of the Internet technologies comprises any kind of "reality transcender". In the words of Argyle and Shields (1996: 61), "the communicative action is not 'in' the machinery. It is shared between persons and interpreted by each interactant".