Ghost Towns & Virtual Worlds

Tags:  | |

This paper was delivered at the sixth annual conference of the Society for Phenomenology & Media, held in May 2004 at Brigham Young University, Provo, Utah.


I teach multimedia at Arcada Polytechnic in greater Helsinki, in Finland. I am not a philosopher, although I have, from time to time, been accused of philosophising.

What I want to present here is a sketch of a problem that has arisen during the course of teaching multimedia to undergraduate students – a problem that has seemed to demand a philosophical approach. I will describe the problem and then indicate the ways we have been attempting to solve it. I am very aware that our approaches may, in many ways, be philosophically naïve, and I do not apologise for that.

There are two reasons for this. Firstly, the problem is important to me: not just because it is interesting, which (to me at least) it is, but also because the long term goals of a major multimedia project will be largely defined by the answers we arrive at. Secondly, I am hopeful that this paper may stimulate discussion will illuminate the issues further, and even prove practically helpful.

The Background

Our problem arose unexpectedly at the early stages of an attempt to build a virtual world at Arcada. I described and demonstrated an early version of this at last year’s conference in Helsinki, but let me briefly lay out the background of the project for those of you who have not already heard it.

Lars Lundsten, the head of the Media Department at Arcada, suggested to all the teaching staff that we should, in his words, “move from projects within courses to courses within projects”. His intention was that the video, sound, and multimedia courses should each have a single over-arching project that would be recognised outside the school, and would be used as the basis of both practical and theoretical classes.

The video department already had such a project – a television station that broadcast for one hour a week on a local cable channel. Although it barely existed for its potential viewers, for the students it meant that everything they did was done as though they had already graduated and were working professionally at a television station.

For reasons that I have described at length elsewhere, the multimedia staff and students decided to create a virtual world and then launch it on the web where it would function as a chat room, or an adventure game, or simply as a curiosity. We decided that, once the world was created, there were at least five to eight years’ worth of tasks that we could engage in to improve it, add elements to it, and attract users. These could be done while teaching students to make web pages, use Flash, manage databases, etc. It would, in effect, be our television station.

For a combination of perdagogical and technical reasons, that I discussed last year, we decided that the world would take the form of a fictitious island located at a spcific point at the south of the Meditteranean, half way between Crete and Malta and about one hundred miles north of Eygpt. This island is called Rosario, and the world would be a model of Marinetta, the walled city that serves as the island’s capital. We also decided that users of this world would enter the city as tourists.

The Problem

We chose some open-source modelling software, made some rough and ready decisions about what the architecture would look like, based on the island’s location, and set to work building the world.

It seemed sensible, at the start of the project, to give the students a research exercise investigating what kind of virtual worlds already existed on the web. We therefore gave a group of about 16 students a list of URLs and asked them to report on their feelings about the worlds.

We expected a series of technical comparisons and suggestions: comments like “can we have animated avatars? Can users choose the clothes for their avatars?” Indeed, I expected to spend the first year or so dealing with technical problems of one sort or another. Instead I spent most of it reading and rereading philosophy books, and grappling with the ontological concepts that underpin ideas of virtuality.

The students, almost without exception, came back and reported that there was something wrong with the worlds. They were not at all specific; they simply agreed that the worlds “didn’t feel right”. We sent them to explore more.
We asked them to be precise about what they meant. Eventually we got two enlightening answers. One: there is something hollow at the heart of the worlds. Two: it feels like you are wandering around stage sets.

The consensus seemed to be that the worlds felt like cardboard boxes painted in one way or another. As one student said, “When I went to the Venice world there was no Venice-ness about it”.

That, stated as briefly as possible, is the problem we have spent a year wrestling with. We have been trying to establish whether it is possible to put ‘worldliness’ into a virtual world, and this, of course, has meant trying to define precisely what we meant by worldliness, by virtual and, indeed, by world. It has also meant trying to establish why ‘worldliness’ appears to be lacking from all the other worlds we have looked at.


We began by trying to define our terms, and we started by looking back over the history of computer gaming to try to see how others had used the terms.

Richard Bartle, the man who is credited with inventing the first MUD – the text-based multi-user dungeons that were the precursor of virtual worlds, with Roy Trubshaw at Essex University in 1980, has defined virtual in his book Designing Virtual Worlds as “that which is not, having the form or effect of that which is”. The philosopher CS Peirce said almost the same thing somewhat more technically, in How to make our ideas clear: “a virtual X (where X is a common noun) is something, not an X, which has the efficiency (virtus) of an X”.

Both of these definitions are functional: broadly speaking they paraphrase the cliché that suggests that if it looks like a duck and quacks like a duck, then it is a duck. If you are wandering through a virtual world and come across something that reminds you of a piano, and is capable of being played and producing musical sounds then, Bartle and Pierce would seem to say, it makes sense to refer to it as a virtual piano.

It follows from this that a virtual world should have the efficiency, the virtus, of a world.

Heidegger points out in Chapter 4 of Being and Time that defining world is far from simple. He offers four definitions, but for our purposes the first, which is also the simplest, will suffice. Here Heidegger suggests that world “is used as an ontical concept and signifies the totality of those things which can be present-to-hand within the world” (Heidegger, 1962). I should note here that Heidegger himself places this use within single quote marks, and uses the word world as defined in his third definition; in which it has “a pre-ontological existential signification”.

Having mentioned Heidegger, I will now begin my definition of worldliness by making clear that this is not a synonym for what Heidegger refers to as worldhood. If I understand Heidegger correctly, and I will be the first to admit that I may not, then only the world itself can have worldhood, and the chances of imbuing any artificial, or virtual, world with worldhood are less than none.

Worldhood is a function of the world. What I mean by worldliness, on the other hand, seems to be a function of perception: the sum total of real and intentional objects whose apparent self-presentation would be sufficient to allow a participatory setting aside and translocation of belief.

Charles Harvey explained in his paper yesterday the suspension of disbelief that occurs when we relax into watching a film – when, perhaps, we make a phenomonological reduction to see things without our taken-for-granted beliefs and attitudes. He suggested that this is a foundationally passive experience, because we let the events in the film show themselves from themselves as they themselves are.

In entering a virtual world, however, we are one of the objects present-to-hand in the world for the other users, while they themselves are present-to-hand for us. The world exists for us, as we exist in it for others. We do not merely have to make a reduction. We also have to actively habituate ourselves to a somewhere else that is where we agree to believe we are.

That is, if such an activity should prove to be possible.

Games and worlds

My first thought was that, since none of the worlds we had seen had exhibited worldliness then perhaps, for whatever reasons, it was simply not possible. However, it was pointed out to me by several of the less philosophically inclined students that this was so self-evidently untrue as to be not worth discussing. Students offered the observation, from personal experience, that they felt much more absorbed into the game of Tomb Raider than they did the movie based on the game. People move bodily when playing driving games like Colin McRae Rally. They can be observed twitching and jerking to dodge the bullets in shoot-em-ups.

This was offered to me as prima facie evidence, and it reminded me of Jonathan Weidenbaum’s paper Race Cars, Hallways and Monsters, in the 2000 edition of Glimpse. In this he discusses the ecological psychology of James Gibson and points out that “computer games cleverly utiliise the sysem of perceptual invariants in order to simulate the same optical patterns as would be experienced in the real world”.

However games are not worlds, because games are task-based or goal-oriented in a way that the world is not. Games come pre-laden with criteria for success and failure. Only people of deep and teleological religious beliefs would view the world as task-based in this way. In a racing game I concentrate on the single task of getting and staying ahead of the other cars. The surrounding world flashes past me as scarcely noticed background. In Tomb Raider I am pursuing an overall objective that involves completing a number of tasks. I may be able to vary the order of these, and I may not need to complete them all to win the game, but I cannot make Lara Croft decide to give up the chase, retire, and open a little beach cafe.

In a world as opposed to a game, virtual or real, that is precisely the sort of career decision I might expect to be able to make.

Categorising virtuality

We next attempted to categorise virtuality in order to be able to ask what other qualities we might expect a virtual world to display.

Virtual objects are the building blocks of virtuality: recognisable models of things, that can be viewed from any angle. They differ from representations of things in photographs, movies, and on television in that you can walk around the back, or turn the object around. You can also move nearer, or further away, and have the object appear to change its appearance in ways that an image on a screen could not.

Virtual cells are small spaces. Their key feature is that they are limited by something such as a wall or a fence, and the user can easily overlook the whole area. Such spaces could, for example, be rooms in buildings, or fenced gardens.
Virtual landscapes are larger spaces that are big enough that the user cannot oversee the entire space at once without having any intrusive limitations. If there are limitations they will appear as ‘natural’: borders such as a coastline or waterfront. Landscapes will contain cells and objects.

Virtual worlds, in this context, is a term similar to “the roman world” or “the world of high finance” (Bartle, 2003). A world is of a different order from a landscape although it will contain one or more virtual landscapes.
There is not a single consensual definition of the concept ‘virtual world’ (Jakobsson, 2003, p. 2), although many of the definitions have very similar attributes. Broadly, definitions of ‘virtual worlds’ propose four criteria that have to be fulfilled for the term to be applicable:

  • The environment has to support the feeling of presence, which means that the user must be able to maintain the illusion that he is ‘in’ the environment.
  • The environment has to be persistent, so that, when the user leaves the environment, its existence continues and it evolves during the time that the user is away.
  • The environment has to support interaction between users, and between users and the environment in the form, for example, of communication with other users.
  • Finally, there has to be a representation of the user, so that other users can tell when she is present. (Jakobsson & Skog 2001, p. 2, See also Manninen & Pirkkola 1998, p. 1)

Klastrup’s definition, in his paper A Poetics of Virtual Worlds is one of the most descriptive:

A virtual world is a persistent online representation, which contains the possibility of synchronous interaction between users and between user and world within the framework of a space designed as a navigable universe.

“Virtual worlds” are worlds you can move in, through persistent representation(s) of the user, in contrast to the represented worlds of traditional fictions, which are worlds presented as inhabited by real people, but not actually inhabitable.” (Klastrup 2003, p. 101)

However I want to argue that a vital aspect of specificity is missing from all of these definitions.
When suggesting that the word ‘world’ in the phrase ‘virual world’ should be used in the same sense as in a phrase like “the world of ancient Rome” Bartle (2003) implies that a virtal world must support the feeling of being a specific kind of world.

A game ‘world’ only offers a player the possibility to play a previously defined game, while an inhabitant of the “real world” may feel that they can choose from an apparently almost limitless range of options. Indeed the worldliness of “the Roman world” can be said to lie in the multitude of different persons and objects involved in making it a specific world, a world that could not easily be confused with, for example, nineteenth century Provo.

Time, Space & Culture

People and objects live in time as well as space. This may seem a banal observation but I believe that it has vital consequences for any attempt to design of virtual worlds. People age and the results are at best ambivalent. There is a lot of inconvenience attached to the process. Landscapes, that is the collection of objects that together comprise a specific place, age in the same way, generating inconvenience as they do.

The Dutch architect NJ Habraken has explored this at great length and in great depth in his book The structure of the ordinary: form and control in the built environment. Here he looks at the way several cities have changed their form over several hundred years. He points out the ways in which the siting of objects like churches in the past have forced later development to fit itself around them.

In most European cities and towns streets are not laid out in the grid-like fashion you find in cities like Provo and New York, but wind around churches, market halls, palaces, courts and old landmarks. The result of this is a specificity which could be said to be one of the essential factors in, for example, the ‘Brussels-ness’ of Brussels.

The beliefs, habits, customs and laws that determine how and where objects are built in a specific place, and how and by whom they are decorated, might be described as the culture of that place, if we also include in our definition the cumulative effects of the accumulation of decisions mad over time. This kind of learned culture has been described by Miraglia as “like a lens which filters all the information we perceive through our senses; sensory information passes through this lens of culture and is filtered, or interpreted, into a recognizable pattern that has meaning.”

The presence of a learned culture, being contended and extended, is what gives a place its specificity, and makes it what it is and not what it is not. By way of a example, we might discuss the controversy of General Mannerheim’s statue in the not-too distant past in Helsinki.

When the plans for Kiasma, the modern art museum in Helsinki, were proposed, it was realised that the statue of General Mannerheim would have to be moved some short distance. This provoked a large and public controversy, and for a time it looked as though the plans for the museum would have to be changed. My point is this: to understand this controversy you would have to understand something of finnish culture, and specifically something about the totemic importance of Mannerheim in that culture. Without that understanding, that is, from a position outside that culture, the bitterness of the argument, and the reasons for it occurring at all, are without meaning.

Ghost Towns & Virtual Worlds

The virtual worlds we had looked at were devoid of any evidence of the inconveniences of history and the impositions of outmoded beliefs. The streets did not wind around the edges of an old plague burial ground, and nor was there a park to commemorate the founder of the timber mill. Indeed, they all looked and felt as though they had been designed from scratch for the convenience of unknown potential users. They were ghost towns not because there was no evidence of people but because there was no evidence of a past.

The result of this realisation was that we set aside the building of Marinetta, leaving the original test-model in place, and concentrated on creating a history and a culture for the island. From this came the first student thesis derived from the project, as Niki Weckstrom coordinated and documented the attempt to understand the nature of worldliness.

He also supervised the process through which we created 2000 years of history, that was later published in three e-books (Kelly & Macoti, 2003). Rosario’s history was built around gaps and vagueness in the official history books. We borrowed a language for the island – the almost moribund artificial language Ido. We created a religion for the island, that is a variant of historic gnosticism, and draws from the documents of the original Council of Nicea found in 1945 in Nag Hammadi (Robinson, 1981).

We have now begun to tackle questions of function. If we accept the definitions of virtual that I have given above, then a virtual cafe must function, for the user, differently from a virtual bar. This raises at least two questions: why does the user need any kind of virtual liquid, and what can we possibly mean by virtual drunkenness?

That is apparently where our philosophical quest is leading us next.


Owen Kelly would like to thank Niki Weckstrom and Camie Lindeberg, his colleague in the Marinetta Ombro project, for their assistance and encouragement in the ongoing development of these ideas.

Works Cited

Bartle, R. “Designing Virtual Worlds” New Riders Publishing. 2003. ISBN 0-7357-1351-0.
Habraken, N.J. “The structure of the ordinary: form and control in the built environment”. United States of America, Graphic Composition, Inc. 1998. ISBN 0-262-08260-8
Harvey Charles. paper in Glimpse. 2004.
Heidegger, Martin. “Being and Time”. Translation by John Macquarrie and Edward Robinson. Blackwell. 1962.
Jakobsson, M. “A virtual realist primer to virtual world design”. In Ehn, Pelle & Löwgren, Jonas (eds.). “Searching voices – towards a canon for interaction design. Studies in Arts and Communication #01”. Malmö: Malmö University. 2003.
Jakobsson, M. & Skog, D. “What makes a house? Approaching architecture in virtual worlds”. Proceedings of IRIS 24, Norway. 2001.
Kelly, O. & Macoti, T. “Hike Ni Esas, Tomo 1: Anciena epoki”. La Redaktaji Kolegio, Rosario. 2003.
Klastrup, L. “A Poetics of Virtual Worlds” Melbourne DAC2003 digital arts and culture conference. 2003.
Manninen, T. & Pirkola, J. “Comparative Classification of Multi-User Virtual Worlds”. 1998
Orliaguet, J-M. 2001. “Design, Virtual Reality and Peircean Phenomenology”. 9th International Conference on Human-Computer Interaction. 2001.
Peirce, C. S. “How to make our ideas clear” in N. Houser, & C. Kloesel (Eds.), “The essential Peirce: selected philosophical writings” Bloomington, IN: Indiana University Press. 1993.
Robinson, James M. “The Nag Hammadi Library in English”. Harper & Row. 1981
Weidenbaum, Jonathan. “Race Cars, Hallways and Monsters”. Glimpse. 2000.

This essay was first published on July 11, 2004