The Linker software comes out of a specific need. A key part of what Mongrel does are workshops, dialogues to produce fast artefacts of digital culture with other mongrels. A crucial thing about these workshops is that people want to produce something that looks good, and means something, but don't want to have to invest months in teaching themselves up to know something like Photoshop and Director. We don't particularly want to knock these programs, but they're cultured up to be useful really only to experts.

The way the workshops run is based around the structure of the Linker. There's one or more computers running the program and ways of getting material into it. A video camera and some cheap digital cameras to produce the still or moving images. Sound comes in from off of the video or straight into the computer's mic. A simple sound editor program can be used to edit and add effects. The text is banged out directly on the machine. Other stuff can come off the net, CDs, wherever. It's quick.

What Linker does is essentially what Director (pretty much the standard multimedia authoring tool and the program whose internal language, Lingo, Linker is written in) does already. But Linker does this in a more restricted set of ways. Director is built to process any form of data type and provide a way of working them together, usually by relatively complicated programming. Many multimedia programs make giant baroque concatenations. Linker by comparison is a very slight offering. Basically an opened up versioning of an editing tool that has been used to make a number of CD ROM based artworks (such as 'Rehearsal of Memory'), it is deliberately constrained.

The constraint is what makes it quick. It is also what pulls things together formally. Linker is the multimedia equivalent of a throw-away camera. Other culture technologies it relates to: a page on a photo album; print club; advent calendar; photo frames for several pictures of family members. Your Granny might have a dinky set of display shelves full of glass animals, direct from the Franklin Mint and souvenirs. You can do the same thing on a computer, but make it modernlike and multimedia. On the other hand, you can make something tight, vicious and full of a whole load of other kinds of memories or new sensations. Kids setting out collections of stones or little soldiers on a table are at the same time attentive to their aesthetic detail, how realistic, how variable the faces, as to how great their ability to inflict real or ideas of damage.

The software feeds into the natural delight that people take in analytical sorting or in extrapolating from imagined or actual patterns. The eye-hand strays and picks at this and that, producing connotations of meaning by the simple fact of linkage, of little tokens, tenderness, tat. Whilst Linker aligns itself with the delights of opening, of skimming over, and of fiddling it also presupposes a sense of montage which, like fiddling, goes nowhere in particular.*It is a machine for producing a form of montage which is more like a game of dominoes than a dialectic. That whilst it does not assign the particular to nowhere, in providing a device for its extrapolation, is not also dependent on cell a clashing with cell b to progress to cell c.

This is not to say that Linker could not be used as a space for the operation of such procedures. Perhaps we even remain in hope that things could be that algorithmically simple. The software even explicitly provides a domain in which such devices, such rule-fantasies can be played out without harming the public. Constraints, of formal rules, of available materials, of rhyme and metre, of rythms are familiar tools for meanings to trick themselves into being made. Adoption of a constraint paradoxically allows creation to become a process. There is a sense in which software, something which is often said to be 'open', can produce similar effects.

Software constructs a sensorium: a set of ways of sensing, knowing and doing in the world. A few examples: command line interfaces such as UNIX transposing the 'memory' of the operating system onto the user, what elements of HTML documents are deemed to be important for crawlers; how search engines incorporate semantic judgements; the spatialisation of memory relied upon and reinforced by the WIMP (Windows, Icon, Mouse, Pointer) genera of interfaces; the inclusion of unique machine identifiers within the Pentium III chip, and the wider questions of value-production in information economies commonly, but in some ways less than usefully, grouped under the heading of 'privacy' that go with it; and so on.

It is utterly lame to suggest, as David Gellernter, a Professor of Computer Science, does in his extended homily 'The Aesthetics of Computing', that software as a science is not formulated by currents other than 'itself'. There is a twin movement. Involving scientific, that is to say rational, methods in wider fields that may be political, social, conceptual, aesthetic. At the same time, teasing out those ways in which the internal configurations of practiced rationalities (such as software production) already operate within and produce these domains of influence. They present a possibility for actually enriching rationality and making it, as a particular kind of knowledge machine, more productive.

Software that is not socially constructed, not only gets no users (for it does not hook into or effect any of the involvements they might have for it), but even as an 'orphan', something that has no preconditions, something that is solely of itself, it is formed, as an impossible object, at least in part as a negative imprint of what is already in existence. The point is not whether software is socially constructed or not, or put the other way, whether rationality has its uses. This is a minor argument. Rather, what kind of currents, what kind of machine, numerical, social and other dynamics it feeds in and out of, and what others can be brought into being? This is not just a question of, 'putting software at the service of the community' or some other farce of repurposing but of developing modes of study, innovation, production and use that acknowledge that, "To be 'technical' simply is to be a response to a history of conflicts."

fly-note:

Each of the nine image cells that form the map are split into a further 16 'hot areas'. Each of these has a further eight possibilities of actions: sound; map; scale; text; video; jump; image; chat. The maths can be expressed as follows: 9 x 16 x 8 = 1152 possibilities for each map. A graphical representation of Linker's basic algorithm is pasted on the front of the piece as the software starts up. This map of linkage forms a direct symmetry between the interface and the algorithms working below. The lines then shift, according to how the data in that particular Linker sorts itself. This allows the user to gain a graphical representation of the links between the elements they have placed in their Linker. The interface is constructed to represent the code and the limited possibilities of its use, nothing more. This goes against the grain of much proprietary software which attempts to acheive the most narrow kind of practicality, "There's a job to do. Let's just get it done. Don't think about what it means."* at the same time as subsuming every possible function or way of treating data within it.

: We are transfixed by the outcome of our interaction with applications. We forget the program in order to get on with the task. If we can reach clarity about what software does, how it offers us a limited range of objects as a menu of 'creativity' or of process, we can begin to see what is missing. There are three by three cells in the first layer of the interface. From the Three Little Pigs to the Holy Trinity, three has an interesting position, always beyond duality - here on in things get complex it promises. The third in a series always suggests the onset of a series, elaborating a something between the preceding numbers causing things to move on - a factorial, a function, a game. Constraining the number of image cells in the Linker allows it to be filled fast at a basic level. It also forces users to make choices, to discriminate about the use of a particular graphic in relationship to the others within the fixed number of cells available.

Formalism becomes a machine for affect. Adding sounds, images, words and video together in a pattern for the first time is really quite a powerful experience for many people. Importantly also, viewing a Linker with elements made by or about them, their peoples, creates a very intimate relation with the process of using and viewing. This might have something to do with being able to create a cluster of media with strong presentational authority in terms of coherence of design and function. Alternately, for instance, when used as a kind of miniature family album with sound recordings, photos and so on it might also have something to do with the sense of openess inherent in the formal system of the database.

Formal constraints repress what is underneath them at the same time as allowing their articulation in certain ways. A little thing such as this, the software, whilst it includes constraints, does not allow access to the Law or a state of numbered grace. Nor does a 'full' Linker map form a final will and testament, a chance to speak that will only be given the once. Instead, a spread palm-load of sleights of mind that people can play upon themselves, upon memory.

Lev Manovich, in his useful essay on the 'Database as a Symbolic Form' , suggests that what is often found in actual usage of databases is that what has been assembled is, "A collection, not a story". In Linker, forcing a limited number, but no more, of image cells to be filled before the thing can be used encourages a certain amount of syntagmatic relations between data elements in the constellation of many which the database is composed of. As Manovich suggests, this can be like putting together a sentence in a natural language. It also suggests what he calls the conflict between database and narrative, between more or less open arrays of elements, paths and strata and the timelapsed results of particular routes through them congealed as a story.

Another difference between narrative and database is between signs, the base constituent element of narrative, and the digitised elements, cast members, sprites, objects, whatever that are actuated in a database. Whilst this is a difference of degree and not of exclusivity, simple material factors such as the amount of processor cycles needed to call up an element also have their effect in terms of composition. Linker makes use of a material factor like this in a determining way in that each cell changes size according to the dimensions of the image file. Thus, it allows the overall visual pattern of the first layer of the database to emerge as a result of the properties of its constituent elements. This is a small thing, but in the unusual context of artists producing a system rather than its content, one that presents processed documents rather than perhaps open up the process of their construction, the interface is essentially all there is. It has to be thought, and sensed, through.

Mathew Fuller info@mongrel.org.uk