acm - an acm publication


The non-autonomy of the virtual
philosophical reflections on contemporary virtuality

Ubiquity, Volume 2008 Issue May | BY Jeff Malpas 


Full citation in the ACM Digital Library

Much contemporary talk of virtual 'worlds' proceeds as if the virtual could somehow be considered as in competition with or as an alternative to the world of the 'nonvirtual' or the 'everyday'. This paper argues that such a contrast is fundamentally mistaken, and that the virtual is not autonomous with respect to the everyday, but is rather embedded within it, and an extension of it.

We often talk about virtual 'worlds', and about the 'reality' of those 'worlds', as if they could somehow be considered in competition with or alternatives to the world of the 'nonvirtual', of the 'everyday'. Yet widespread though this way of talking might be, it is also fundamentally misleading, if not mistaken. Indeed, a basic starting point for any serious discussion of the virtual must be recognition of the non-autonomy of the virtual - a recognition of the fact that the virtual does not constitute an autonomous, independent, or 'closed' system, but is instead always dependent, in a variety of ways, on the everyday world within which it is embedded.

It would be useful if this 'principle of the non-autonomy of the virtual' were thought controversial or contentious since controversy always attracts more attention, but whether or not the principle is taken to be so, it seems to me to express what is a fairly self-evident, and, in some respects, obvious and even mundane fact, although nonetheless an important one. In these few pages I want to explain something of what this principle of non-autonomy amounts to, what it means, and why it might indeed be a fundamental staring point for any discussion of the virtual.

So what is the nature of the non-autonomy that is at issue here? To begin with, virtual domains inevitably depend, in a straightforwardly causal and material sense, upon a large body of supporting infrastructure that exists in the everyday world and not in the virtual (notice that this is true even of imagined 'virtual' worlds such that of The Matrix). Even our interactions in the virtual as actors in a virtual 'space' depend first and foremost on our interaction within the everyday space in which is located the screen, keyboard, mouse or whatever other equipment enables our virtual activity (see Malpas, 2000). In this respect, one might view virtual space as a very specific modification of the space of the everyday - the non-autonomy of virtual space being a special instance of the non-autonomy of the virtual as such.

One way of putting the non-autonomy of the virtual is to say that the virtual is a product or artifact of the everyday - that it is produced through actions and events in the everyday world. Another way of putting it, to use some technical philosophical language, is to say that the virtual supervenes on the everyday, which means, in this case, that every virtual event matches up with some event or events in the everyday world, and that events in the domain of the virtual occur as a result of the events in the everyday that are their causes - thus every virtual event will be able, minimally, to be correlated with some event or events occurring in the computational system, including the hardware and software components, in which it is based, or more broadly, in the everyday world-environment that includes both the computational system and the user or users (as well as much more besides).

Of course, these ways of putting matters still retain the potential to mislead in one crucial respect: they suggest that the virtual is different from the everyday, whereas strictly speaking, the virtual is merely another part or aspect of the everyday world - and this is an important element in the very idea of the non-autonomy of the virtual. There is thus only the one world, and the virtual is a part of it. Saying this is not, of course, to reject the commonplace talk of different 'worlds' that occurs in a wide range of contexts (from 'the wide world of sports', and 'the world of the honey bee', to the 'World of Warcraft'), but only to insist that there is an important central sense of 'world' according to which there cannot be a plurality of such worlds - in this sense the world is precisely that which encompasses everything, and which cannot be wholly encompassed by anything within it.

The non-autonomy of the virtual is perhaps most obvious when one looks to the causal dependence of the virtual on the everyday - and not only its dependence on the physical structures that make it possible, but also on the various socio-economic and socio-cultural processes (including the process of design) out of which it arises. Yet the non-autonomy of the virtual is not simply a matter of causal or physical dependence. It is just as much a matter of its contentual dependence. The content that is embodied in the virtual - the narratives it presents, the significance of the individual events and elements that occur within it, the meanings that can be attached to the images and texts that it generates - is always dependent on the everyday world in which the virtual is embedded.

The most obvious way this is so is in terms of language - the languages that appear within virtual environments are derivative of the languages spoken outside of them. Moreover, even were one to design a virtual environment with its own 'in-built' language (as has been done - the Myst games present what is arguably a complete language, and certainly a comprehensive and novel base-five numbering system, as part of the D'ni civilization), still the language would derive from a design process external to the virtual domain itself, and there is no reason to suppose that such a language would necessarily remain restricted to its original virtual domain alone or that it could be insulated from linguistic change deriving from the everyday. Yet not only is the language of the virtual inextricably bound to the nonvirtual, so are the meanings of items and events that appear within the virtual invariably dependent on their embeddedness within the larger world-frame that is the everyday. The frames of meaning that give significance to the virtual are thus the frames of significance that users bring with them into the virtual and that encompass more than just the virtual alone - virtual experiences gain their content and significance through being embedded in those frames, and so through being connected to other experiences both virtual and everyday.

Both the causal and contentual dependence of the virtual on the nonvirtual is nowhere more evident than when we consider the fact that the agents who operate within virtual domains - the individuals who are able to exercise choice within those domains in realizing different possibilities - are always individuals whose existence is based in the everyday, and not in the virtual alone. Even those individuals who view their lives as primarily oriented around their virtual activities - individuals who may spend the majority of their waking hours in some on-line domain such as Second Life, and may even make a living from their activities there - still live everyday existences in the everyday world. Moreover, most of what matters to them in the virtual matters because of the way it connects with their everyday lives - whether that be in terms of certain personal gains such as recreational enjoyment, social interaction, or financial gain, or the broader social gains that come with scientific or creative advance.

One can, of course, imagine a situation in which an individual's conscious life was wholly restricted to events within a virtual environment (although one would almost certainly have to take elaborate steps to ensure the insulation of that internal world from any contamination by the everyday). In such a case, the experienced world of the individual could perhaps be viewed as constituting an autonomous and independent domain (so much so, one might argue, that from the individual's own perspective, there would be no reason to view that world as other than the everyday). This is, of course, just the situation that The Matrix envisages. Yet such a situation is not only far removed from that of any current virtual domain or its participants, but significantly, even were the conscious mental life of the individual to be restricted to the virtual and detached from the nonvirtual (and such a possibility is not without serious complications), that would not sever the ongoing causal existence of the individual in the same way. Just as the virtual is causally or physically dependent on the physical infrastructure that enables it, so too must the virtual activities and virtual lives of actors within the virtual supervene upon a set of everyday processes and structures that themselves underpin the actor's existence.

Even a wholly artificial intelligence whose conscious activities were focused on events in a virtual domain would still require both a virtual body that enabled it to act within that world (hence the importance of the avatar in contemporary domains such as Second Life) as well as some mode of physical realization that lay outside of the virtual - if not a human body, then some artificial surrogate, even if only a collection of wires and circuitry. For this reason it is deeply mistaken to suppose, as some do, that the virtual offers the possibility of any significant release from the limitations of embodiment. The Matrix itself provides peculiar confirmation of the primacy of the everyday in relation to the virtual, since although the movie places great emphasis on the apparent verisimilitude of the virtual domain, it is the everyday domain - the domain of a devastated Earth in which machines and humans fight for supremacy - that is clearly presented as the 'real' world, and, according to the movie, the only world that matters. (The one exception that complicates matters is the preference of one of the characters, Cypher, for the Matrix world over the world of the machines - suggesting not only a certain inconsistency in the thinking that underpins the movie, but also that what counts as 'real' is not always clear-cut).

To recapitulate: the non-autonomy of the virtual means that the virtual is both causally and contentually dependent on the nonvirtual, or as we may also put it, on the everyday. However, there is another form of dependence that should also be noted here. Just as it is common to talk of virtual domains as alternative 'worlds', so it is also common to view them as 'worlds' in which the constraints of the everyday no longer apply - in which there are unlimited possibilities and a new freedom. Yet here too the hyperbole goes rather further than the facts. For instance, the fact that I cannot fly unaided in the everyday world, whereas my avatar can be represented as flying in the virtual world of Second Life, might be thought to demonstrate the divergence in possibilities of the virtual from the everyday. That it is a divergence in possibilities depends, however, on treating the domain of Second Life as if it were indeed an alternative 'world' to the everyday (it also depends, as a corollary to this, on treating my avatar as identical with a human person). But this assumption is counter to the principle of the non-autonomy of the virtual according to which Second Life is itself embedded within the everyday. Moreover, inasmuch as Second Life is embedded in the world of the everyday, rather than running parallel to it, then nothing that occurs within Second Life can actually run counter to the physical or other constraints that operate in the everyday world. That an avatar in Second Life can be represented as flying is not counter to the laws of physics. Those laws govern all physical processes, including those that are relevant to Second Life just as they govern the physical processes of the everyday.

Yet it may be thought that this misses something important about virtual domains like Second Life, since even if there is a sense in which the laws of physics themselves do not change under the influence of the virtual, there is surely also a sense in which the virtual allows us to experience things as if the laws of physics were different. So the virtual may not enable an escape from the physical constraints on possibility that the everyday imposes, but it does enable an escape from the experiential constraints of the everyday. Even here, however, we need to be cautious.

First, the everyday itself allows for a great range in imaginative and experiential possibility without recourse to the technologies of the virtual - there are other ways, for instance, to simulate the experience of unaided flight (and not only in our dreams - skydiving, for instance, probably brings us pretty close). In this respect, virtual technology is perhaps best thought of as one among a number of means by which our experience can be elaborated and extended. Second, and more importantly, there are fundamental constraints on experience in the everyday that operate as conditions on the very possibility of experience, and those constraints necessarily carry over into the virtual.

The exact nature of these constraints is philosophically contentious (on one reading, philosophy itself can be seen, at least since the work of the 18th century German philosopher Immanuel Kant, as given over to the attempt to identify and explicate these constraints), but they almost certainly include, at the level of ordinary physical objects, constraints of consistency and coherence, of causal connectedness, and also constraints deriving from the structure of subjectivity, intersubjectivity and objectivity, and of spatiality and temporality (see Malpas, 1999, for an attempt to articulate some of these constraints within a 'topographical' framework). Perhaps we could consider the investigation of virtual worlds as one way of continuing this sort of exploration - although such an investigation would also need to be philosophically informed if it were not to lead us astray.

As a result, and in spite of the claims that are often made regarding the freedom that the virtual supposedly opens up, virtual domains are invariably structured in ways that closely parallel the everyday - the experiential variations typically do not constitute radical changes in the nature or possibility of experience, but variations upon the experiences with which we are already familiar. As a result, most virtual domains appear as somewhat impoverished and limited versions of the everyday. Indeed, the more the domain may appear to vary from the everyday - as in the case of fantasy domains such as World of Warcraft - the more it is typically constrained within a highly specific and limiting narrative, and so the more narrowly constrained is the variation. (In the case of World of Warcraft, the narrative is one already determined by highly specific narrative frameworks, including characters, creatures, landscapes, plots and themes, largely derived from pre-modern stories, myths and legends).

There is one area in which the autonomy of the virtual is sometimes asserted that I have so far not discussed - that which relates to ethics and law. This might be thought a particularly important area - certainly the issue of the legal status of virtual entities and events has been the focus for a great deal of discussion. Yet here it seems the principle of the non-autonomy of the virtual operates quite simply, self-evidently and directly: the non-autonomy of the virtual immediately rules out any idea that virtual domains could be regarded as independent, in any significant way, from existing legal or ethical frameworks.

Everything on which virtual domains supervene is already under some national jurisdiction, as is every individual who participates in those domains. Moreover, since what happens in those virtual domains derives much of its significance from the effects it has outside of the virtual - whether that be in terms of the emotional or attitudinal effects on individuals or the financial returns generated for individuals or organizations through virtual activities - so much of what happens within the virtual will, just in virtue of its everyday significance, also be open to everyday legal and ethical assessment (and existing legal frameworks, in particular, are already being deployed to deal with events and actions in virtual domains).

This is not to ignore the special problems that may arise as a result of the trans-national or trans-cultural character of the activities and relations that virtual domains can enable - but those problems are by no means unprecedented nor are they restricted to the virtual alone. Where those problems affect the commercial sphere, for instance, then it is as a feature of e-commerce in general, and not of virtual commerce in particular; where there is an apparent divergence in ethical standards, then this typically reflects a similar situation to that which also obtains in many everyday contexts. The virtual does not introduce any special legal or ethical problems that are peculiar to the virtual alone. Moreover, when it comes to ethics in particular, our virtual relations with others depend on the same common commitment to basic ethical notions such as overall honesty, trust and respect as do our relationships in the everyday. When such commitment fails, then so do our relationships - and this is just as true of relationships 'on-line' as those that are conducted 'face-to-face'.

Although the principle of the non-autonomy of the virtual embodies what ought to be taken as a fairly straightforward and uncontentious principle, the principle has consequences that nevertheless do connect with more obviously controversial issues. This suggests that the principle is neither trivial nor unimportant one, but does indeed constitute an important touchstone in thinking about the virtual.

On the one hand, the non-autonomy of the virtual implies that the virtual cannot be contrasted with the real in any simple fashion nor can it be treated as another 'reality' that stands in contrast to the everyday. One might take this to mean that virtual domains such as Second Life really ought to be understood as imaginative or 'fictional' domains ('constructs') that are different from the imaginative domains of movies and literature in their interactivity, but are similar in that they remain dependent on the everyday, and are elaborations or extension of it. On the other hand, the non-autonomy of the virtual means that the virtual is causally and contentually interconnected with the everyday. This not only means that virtual domains such as Second Life are causally and contentually dependent on the everyday, but that they are also causally and contentually effective on the everyday - that they have real, everyday consequences. In this way, recognizing the non-autonomy of the virtual also enables recognition of the reality of the virtual - a reality evident within the everyday.

The non-autonomy of the virtual allows us to grasp both the constructed or 'fictional' character of the virtual as well as the reality of the virtual. In this way, recognition of the principle enables a more complex and nuanced conception of the virtual, and its relation to the everyday, than is common in many discussions. While there can be no doubt of the sophistication and complexity of the technical thinking relating to the understanding of contemporary virtual domains, that understanding is often less well-grounded when it comes to its philosophical and conceptual underpinnings. Recognition of the non-autonomy of the virtual, and its implications, seems likely to constitute a crucial step in gaining greater philosophical clarity and insight into the nature of contemporary virtuality.

Author's Bio

Born in Sydney, but growing up in Auckland, New Zealand, Professor Jeff Malpas completed his undergraduate studies at the University of Auckland and his PhD at the Australian National University. His first academic position was at the University of New England from 1985 until 1989, and he was at Murdoch University from 1989 until 1997. Within philosophy, Professor Malpas is perhaps best known as one of a small number of philosophers who work across the analytic-continental divide, publishing one of the first books that drew attention to convergences in the thinking of the key twentieth century American philosopher Donald Davidson and the phenomenological and hermeneutic traditions, as exemplified in the work of Heidegger and Gadamer. More broadly, Professor Malpas' work has been characterized by considerable inter-disciplinary engagement, particularly in relation to issues of space and place. His 1999 book, Place and Experience: A Philosophical Topography was described by Professor Ed Casey of SUNY, as 'the most important book written on the topic of place' (Philosophy and Geography, 2001). Professor Malpas' most recent book, Heidegger's Topology: Being Place World is published in 2007 with MIT Press, further develops the analysis begun in Place and Experience. Professor Malpas also has considerable practical experience in ethics, working as a consultant to a number of private and public sector organizations, and currently Director of the Centre for Applied Philosophy and Ethics at the University of Tasmania, where he is also Coordinator of the University Theme Area, Community, Place and Change. Professor Malpas was a Humboldt Research Fellow at the University of Heidelberg from 1998-99 and at Munich University in 2004. He has also been a Visiting Scholar at universities in the United States, Britain and Scandinavia. He is currently Professor of Philosophy at the University of Tasmania.
Email: [email protected]

Full Profile of Dr. Jeff Malpas can be located at

Malpas, Jeff (1999). Place and Experience. Cambridge, Cambridge University Press.

Malpas, Jeff (2000). 'Acting at a Distance And Knowing from Afar: Agency and Knowledge on the World Wide Web'. In Ken Goldberg (ed.), The Robot in the Garden. Cambridge, Mass.: MIT Press, pp.108-125.

Slouka, Mark (1995). War of the Worlds: Cyberspace and the High-Tech Assault on Reality. Hew York, Basic Books.

Source: Ubiquity Volume 9, Issue 19 (May 13, 2008 - May 19, 2008)


Leave this field empty