INTERVIEW WITH I/O/D


17TH JUNE 1999

Isla Saint-Claire

Isla Saint-Claire:  I/O/D is a three person group, (Matthew Fuller, Colin Green and Simon Pope) based in London. They are the producers of I/O/D 4: The Web Stalker, a piece of software that you can use to surf the internet's global web of worldwide information and also of I/O/D 1, I/O/D 2 and I/O/D 3, which did not have cute names and stuff. What were you doing in the first three issues?

Matthew Fuller:  They ranged from the first issue which was a full, take-your-machine-over interactive zine, to number two which used a sound-based interface for navigation and so on, and the third issue which operated on the level of the operating system - as a parallel finder. Every issue so far has been able to fit on a disc - And we're

All of these issues were produced in collaboration with a range of invited artists, writers, people working in sound. We took their material and worked it in various ways in order to get to deal with the core problematics we are trying to deal with: the culture of software and of interaction and its implicit political and phenomenological dimensions.

ISC:  'Phenomenological', now there's a good word to begin my next question with! What is I/O/D 4: The Web Stalker all about?

Colin Green:  Well, thanks for that question Isla. Let's start with what it actually does rather than what it 'means'.

The user opens the program. They marquee a box in the blank screen, then apply a function to it. The first function necessary is a 'crawler'. This calls up a web site and proceeds to follow every link on that site - and every subsequent link. A 'map' window can be opened which traces the progress of the crawler. The map is a basic real-time dynamic visualisation of the underlying link-node structure of the web.

Nodes from the map can be dragged into an 'extract' window which pulls any text from them. You can also view the 'HTML stream' as the crawler works its way through the web or 'dismantle' sites to analyse their contents.

MF:  It is a way of using the web that resonates with the actual qualities of the network, rather than attempting to repeat the conventions of graphic design for paper - as is being done with style sheets. We believe that the World Wide Web Consortium, Netscape and Microsoft have closed down the possibilities of what the web might be. This is seen even in much artists' use of the web, where the conventions enforced by the software manufacturers are simply unconsciously repeated.

ISC:  You call yourselves artists and this is a big art book here that this interview is in right now, but...

Simon Pope:  The various media that art operates through: galleries, magazines and so on can cope with something that is relatively static such as an installation, CD ROM or a web site. They do however seem to have problems with incorporating something that is as essentially tied up with process, with actual use and a contemplative involvement over time - perhaps even with simultaneous multiple use, such as software. Where's the artwork? Is it in the software, the socially produced situation, the revisualisation of the network, the network itself?

We see the institutions and discourse-networks of art being important to I/O/D because they provide a particular range of social and aesthetic devices for the consideration and circulation of information about projects and particularly as a medium in itself for the distribution of software.

At the same time, we are equally as keen to operate through different media systems such as getting our work distributed through the CD ROMs of free and demo software on the covers of computer magazines.

ISC:  What would you say constitute the main areas of work in I/O/D? In the sense of being your underlying concerns? Can you remember?

CG:  Well, I guess there are two key areas. Although these are relatively messy and operate only on one level of focus.

ISC:  Of course - you are all so multi-layered.

SP:  Yes, as Colin was saying, our two key areas of work are, on one level, to open up the machines - to produce software that involves developing transversal relationships between different, normally masked, elements of computing process and users - see-through software. To ask what types of subjectivation, both of 'groups' and 'individuals', but also what types of machines (their embodiment of different forms of movement, relations and intelligence) what types of networks of information elements are being produced and to produce a software context that allows for the development of other types of formation and circulation. - speculative software

ISC:  That sounds very exciting and intelligent. What is your relationship to more standard methods of software production - traditional Computer Science say?

SP:  In CS terms we are operating at the points where social protocols mesh with technical protocols. For computer science however these areas are quite distinct except for their point of touch at the interface. We sense however that there is a greater deal of technical bleed into the social and a great deal of social, political, economic and aesthetic involvement in the production of the subjectivation machines of software and of the networks than CS is prepared to acknowledge.

MF:  For standard Human Computer Interface designŠ the principle axis in simple single-user software is between psychology-computing. The equivalent axis in CSCW (computer supported co-operative work), groupware, is sociology-computing. Our work is in a sense to return the psychological to the social and to disturb the production of subjectivity within the neat and tidy models it is reported as conforming to by psychology.

We want an end to the anthropomorphisation of humans via the machines!

Whilst CS cripples itself by the compulsion to become a science - that is to focus solely on the objectively observable and repeatable and is by and large driven by commercial and technical opportunism we do not seek to replace CS with cultural studies approaches to networks which manage by and large to be not only credulous and inane, but shallow enough to guarantee that every academic with a research output rating to boost can retrofit the same old theoretical approaches onto a new, under-researched context simply by virtue of their publisher's hope that they might possibly flog more than five hundred copies for once.

ISC:  You What?

SP:  I/O/D has the same relation to the software industry as the production of hot-rods has to the car industry. I/O/D deliberately sets out to do something that it knows it has no clue about perhaps offers a little more interest in terms of a working process.

MF:  All software is the result of interpretative and productive operations on the entire realm of possible programming, operations that stipulate what is desirable, what is necessary, what is profitable, what does not disturb the pre-existing settlement, or that which being technically fascinating or opportune perhaps encourages escape from it.

ISC:  Now you must tell me some secrets! What is I/O/D planning for the future?

CG:  We are currently building a simple component-based real-time communications tool - I/O/D 5

MF:  a concrete problem for investigation is seemingly very simple: how do you deal with text transcripts? In a situation where there are at least two users typing at each other live down the telephone wires, how do you establish what constitutes a unit of meaning suitable for entry into the shared context?

CG:  Some of these problems are very deep within the technology. ASCII (American Standard Code for Information Interchange) a particular protocol is for instance one of the underlying determinates of much software that brings with it a whole range of assumptions that to greater or lesser extents allocate what is possible with it. For instance, anyone who has knocked up any web pages using HTML will have noticed the restrictive amount of Œspecial characters¹ and the odd codes necessary to display anything but beyond the restricted character range of ASCII.

Typically of such a situation, the problem isn't reducible to one protocol however, but the interplay of many. Browsers could have been designed to interpret simpler natural language descriptions of symbols in the same way that they do for a limited range of colours for instance. At the same time, when it comes to displaying characters such as Kanji, programmers and users are compelled to work-round the architecture of the computer's hardware in order to get anywhere. Problems such as this are immediately political and economic as well as technical.

ISC:  Mmmm... This relates of course back to the Web Stalker?

SP:  Yes Isla. These problems are ramified all the more when the software is used. As well as the complicated position of the software within an individual user¹s machine: the hardware; the operating system. They immediately become implicated in a wider - complex rather than simply complicated - set of interrelated fields: the cost of telephone calls; the speed of modems; how diffuse or focused the range of the typing is (in both semantic and ;data-processing terms) what 'technical' information about the exchange is considered to be important or not (i.e. at the IP network level - data about routing or the sequential arrival of packets could be given or arranged differently).

At a hardware level, (if telephone connections are being used) perhaps whether information about the connection is suppressed or amplified as sound after the connection has been established. ((('internet radio' consideration of internet as Audio media))).

How the software is distributed (whether a critical mass of users is being developed which allows it to tip over into some kind of relation of continuity and effect), how the software is changeable or not, how the language it is written in sits within different practices and economies.

Within the operating system, whether the software is usable alongside other programs or whether it monopolises the system.

In terms of users, at a range of variables including the intelligibility of the help files, to whether their mates are using the software, to the responsivity of the program to their habitual ways of sensing and doing in the world at levels ranging from the phenomenological to the legal or habitual.

MF:  Given the complexity of the situation, the universalising culture of software and the implicit and explicit ways in which it constructs communication should be opened up. Given this complexity however, to try and produce another counter-universalisation would be a mistake. Speculative software adopts a strictly local procedure.

CG:  It finds out about itself by doing.

ISC:  Doing - yes! What we want to find out about though is when I/O/D 5 is going to be out. I know you've lied about it in the past. Tell us the truth.

CG:  I/O/D 5 will be out soon enough. Promise.


Isla Saint-Claire is a cultural theorist.