Networks, Margins and Centres

Models of Online Cultural Geographies as Dispersions and Hierarchies.

by Simon Biggs, 2003

Originally commissioned for and published in "Bandbreite - Medienzwischen Kunst und Politik"
edited by Rudolf Frieling, published by Transmediale, Berlin 2004

for Micky Kwella: explorer and supporter of cultural practice in communities of diverse character and happenstance.

There is a well known and fascinating website, established and maintained by Martin Dodge of the University of London, known as An Atlas of Cyberspaces ( This site seeks to document the kaleidoscopic character of the internet and, beyond that, the network of all networks as it might be mapped and modelled. The site illuminates and documents the wide range of solutions, the diversity of approaches and the avalanche of disciplines involved.

However, to employ any one, or more, of the models or maps available on the site to gain an insight into the cultural topography of the internet, and how this might map onto actual demographics and the manner in which various cultures represent themselves or come to interact with one another through the internet, could prove problematic with only a partial interpretation ever possible. Either there is no method by which such a "geography" can be explicitly mapped or, as yet, nobody has been troubled to undertake such a project.

How might the various maps that do exist, as documented on the Atlas of Cyberspace's website, be interpreted and how might they point towards a mapping strategy which could function to reveal a cultural topography of the net? In what ways can such a topography allow us to consider the social forces operating upon the net at this time, and within our broader culture(s)?

The Atlas of Cyberspaces website, as mentioned, documents numerous methods and strategies for modelling and mapping the internet. It would be impossible to discuss all, or even a reasonable sample, of these in this context. Rather, it might be practical and useful to look at just a couple examples.

Many formally conventional maps of the internet can to be found on the Atlas of Cyberspaces website under the "geographic", "census" and "info" sub-sections of the site. Here we find maps that document the actual infrastructure of cables, satellites and computers that make up the back-bone of the internet. We can also study maps showing us the relative distribution of web site addresses around the world, or of internet trace routes and Internet Domain Name registrations, which tellingly allow us to see the density of internet usage around the globe. Not surprisingly we find exceptionally high densities in North America and Europe, rather less dense network usage in East Asia, Southern Africa and Eastern South America and then huge tracts of apparently empty territory across Africa, Amazonia, Oceania and Asia.

Such maps function to reinforce our expectations and preconceptions; that the internet, as a product of the post-Industrial West, having been developed principally within the military, industrial and academic communities of the USA and its close economic allies, is little used and difficult to access outside those locales where late-Capitalist culture is established as the dominant cultural mode.

It is only in the 1990's, with the consumer driven commercialisation of the internet and the rapid uptake of access into the home and the advent of web-related mobile communications technologies, that we see the North Atlantic geocentric map shifting tectonically towards a more global distribution. The history section of the Atlas of Cyberspaces website reveals that some of the areas where network usage seems so light have, in recent years, also seen some of the fastest growth in internet access. That this growth has begun from extremely low levels, and only very recently, functions to explain why this does not show up more generally on other maps and does suggest a view on how communications, and the "information economy" model, are being globalised. Nevertheless, it is difficult to dispel the over-riding sense that established interests still dominate the net and that there appears to be little to threaten these hegemonies.

Quite distinct from the "straight" demographic maps that dominate certain sections of the Atlas of Cyberspaces, and representing a very particular approach to modelling the network, is London artist collective I/O/D's "Webstalker" application. This is as much an alternative web browser as a mapping system. It functions not by displaying the contents of web pages but rather the links embedded within them, and the links that are then embedded in those linked pages, ad infinitum. "Webstalker" does this by graphically mapping a web address as a node and all links from it to other addresses as radiating connections, at the end of which are further nodes, again with radiating links. It is conceivable that this application could eventually map the whole of the internet as it traces every link available at each iteration of the process.

"Webstalker" functions to reveal not the content of the web, nor the use that its parts are put to, but rather an evolving structure where content associated with specific sites has been determined and then related, by embedding links, to the content of other sites. "Webstalker" is a real-time mapping tool designed to reveal the topology of links that underlie the structure of the web, revealing the values and interests of the owners/authors of the websites concerned. As such, it reveals the network not only in its technical nature but also in its social structure.

How different might the web appear depending on where you initiate the "Webstalker" application? Does the web always appear to be more or less the same or do different aspects of the web, as reflected in the cultural topographies that determine who might link to whom, actually generate profoundly different syntactic, and thus visual, structures? Do different parts of the web have distinct vernaculars of linking? Would a section of the net dominated by business be characterised by more formally defined hierarchical structures and thinner densities of links, reflecting concern as to the perceived value of who you might be linked to, contrasting with the more dense and less "vertically" structured linking habits practised in music swapping and similar Peer to Peer internet communities? What of the pornographic web, where sites compulsively link to one another in an orgy of connectivity whilst seeking to do so "under the radar" of authority?

An internet mapping application which functions in almost the opposite fashion to "Webstalker", but equally functions to confound our expectations, is "1:1", developed by Californian artist Lisa Jebvratt. Here is an implementation where no concepts of difference are allowed to disturb the purity of representation as the entire network is mapped as a clearly ordered hierarchy of Internet Protocol Numbers. Using "1:1" it is possible, with only four mouse clicks, to find any and every site on the net, simply by hierarchically navigating the four levels of the IP numbering protocol (eg: 123.321.12.3). In this respect it must be the most efficient and ergonomic web navigating tool ever invented. "1:1" functions to expose the simple and austere beauty of the internet's primary protocol.

Whilst "1:1" functions to represent the strictly hierarchical numerical structure of the internet it also establishes that whilst you can find any site by its numerical address this in no way assists in establishing what the content of that site might be, what its relationship with other sites is nor whether the site actually exists. "1:1" is a formalist mapping exercise gone mad, to the point where little meaning regarding the subject can be determined from the model other than the driest and emptiest of significatory syntax. "1:1" exposes a rupture between form and content in the fundamental nature of the internet which mirrors a similar rupture in a society of information overload, where our means of diffusing and storing data has led us to know the location and specifications for everything and the value of nothing.

Whilst neither "Webstalker" nor "1:1" are intended to specifically map the demographics of the net they do function to draw our attention to a topography where there can be no centre and thus no margins...models where the network functions alternatively in a non-hierarchical and multi-nodal manner or according to the strictest, but utterly meaningless, hierarchical organisation. Do these mappings suggest that the net is dispersed and non-hierarchical or that it is structured and hierarchic, similar to and reflective of the socio-economic forces that dominate global culture and the formation of the internet itself? It is possible that the maps found on the Atlas of Cyberspaces site only function to model the preconceptions and values of those who created them, not dissimilar to those medieval maps that functioned less as an aid to navigation but rather as a means to reflect and evangelise the value and glory of the subject of the map (eg: maps with Jerusalem at the centre of a world where the further a place from the centre the closer to Hell it appeared)?

Going beyond the issue of mapping, to address network protocols themselves, it might be possible to find an alternative to what, till now, seems like a self-fulfilling prophecy. In the case of Napster, or similar network implementations such as Gnutella, we have what is known as a Peer to Peer networking paradigm. Unlike most of the internet, which is based on an infrastructure of corporate or institutionally maintained servers through which users gain access to the net, a Peer to Peer network relies instead on the direct linking of users machines to other users machines, the network emerging out of the totality of this linking. At least this is the idea in principle. In reality, the protocols that make it all work are still routed through the same internet network structures that other protocols use, so Peer to Peer must conform to the same protocols as conventional network models.

Whilst it is questionable as to what degree Peer to Peer networks differ in their status and value from the dominant modes of the network it is clear that such developments show us that it is possible to both create cultural maps, expressions and even new network structures through the combination of new mapping techniques (in this case, how the internet is mapped so that a series of network connections can be established) and radical software development. It is also clear that this can then flow into the formation of new social networks and even models of exchange, as exemplified with the free music swapping scene and, more profoundly, the open source, tactical media and Copyleft movements.

What is also clear is that when the interests of those who are responsible for establishing the actual infrastructure, which Peer to Peer and similar networks secondarily rely upon, are threatened they will move to remove that threat. They might seek to do this via the courts, through mergers and acquisitions or by establishing rival networks with added value in the hope that what is a trade outside their control will come once again within their orbit (eg: the Apple Music Store).

The point here is that the structures of power that preceded the formation of the net, and which undoubtedly contributed to the net's actual development, will seek to re-establish their control. The hierarchical socio-cultural tropes that are already so well established are thus perpetuated. In a sense the structure of the net, as an evolving model and means for the exchange of knowledge, reflects the structures of the social forces that contributed to its genesis. In this respect any map of the internet is also going to function, to a greater or lesser degree, as a map of the society from which it arises. Without initiatives which function to problematise and re-map the system the inertia exerted by social conventions upon any system will remain unchallenged and will thus prevail. Like those medieval Christians, we will continue to map the world only as we have known it.

copyright Simon Biggs 2003