»A smart artist makes the machine do the work«. This is the motto of the »Net Art Generator«1 by Cornelia Sollfrank.1 This program samples and recombines Internet sites at the click of a mouse and purports to elevate all users to the status of art producers thanks to the obediently working machine. The net-art generator, which is effortless to operate and yet always produces unpredictable images, gives an ironic slant to the programmed individualism of the information era – a little machine for each of us, and, hey presto, we're interactive. But it also shows that the artist has not disappeared because of the machine or the cooperation with various technicians: instead, as the author of a conceptional program, she is »smart«.
Sollfrank's slogan encapsulates the discourses about machines taking over the creative process and the politics of software that have again become topical in the media-art scene over the past few years. The key words here are algorithmic art or generative art, software art or code art. The main venues for these art debates were and are, for example, the Transmediale festival, which has awarded a software prize since 2001, the travelling exhibition »Generator« (2002), the Read-Me Festival with the Net archive runme.org, which has been taking place since 2003, the Ars Electronica festival with the theme »Code« and the second part of the online exhibition »CODEeDOC«, which began in the Whitney Museum, and, at present, the long-term exhibition in the Karlruhe ZKM , which opened at the end of October and has the title »The Algorithmic Revolution – The History of Interactive Art»2
These key words delineate the current scene for the more than one-hundred-year-old history of the break with the (romantic) image of the artist prevalent in the modern age, and its transformation into that of the artist's being the creator of concepts based on automatistic aesthetic strategies. But even though artists' working methods have changed completely or been machinised, the central role of the author as the guarantor for artistic (as opposed to technical) methods of production has not gone out of currency even in the alternative media-art scene; indeed, it can even be said that the name of the author or the collective of authors has gained in importance as a guarantor for the idea parallel to their gradual departure as the creators of works.
So what is so controversial and urgent about the discourses about the algorithmic and generative – that is, the regulated and automatistic - that have re-arisen in the field of media art over the past few years, if the author has at the same time not really been dispensed with as the fundamental pillar of the system of art? And what is the point of claims that, in principle, software art and code art already came onto the scene with those artistic trends in which rules and concepts, structural parameters and performance directions are the most important aesthetic determinants instead of creative spirit?

The algorithmic as performance instruction
Among other things, the issue is highly relevant because electronic media art, which has up to now unjustly been relegated to a niche, has begun claiming authorship and taking over the historical continuity and urgency of the questions it raises. One means of doing this is to reclaim the algorithm as an artistic method: »By algorithm, one understands a decision-making procedure, an instruction consisting of a finite number of rules designed to solve a specific problem.. […] Algorithms have been used intuitively for centuries as systems of rules, instructions, game regulations, as plans and scores in architecture and music. Even the art books of the Renaissance were nothing but instructions for producing paintings, sculptures and buildings«. This is the definition of algorithm given in the text accompanying the exhibition »The Algorithmic Revolution – The History of Interactive Art«, which, with this definition and reclamation of the algorithmic as an integral part of a conceptional, historically comprehensive, anti-subjectivist concept of art, basically reflects the current discourse on this topic.3
Software is understood here as a written performance instruction that controls the computer processor by means of code. From this it can be concluded that every piece of software or code is per se based upon algorithms and that they thus always refer back to the algorithmic processes that have (historically) preceded them. In this sense, the algorithm is at the same time a product and a forerunner of computer programs and therefore not only eminently suitable for becoming a metaphor for processes and concepts, but also the forerunner of computer programs and therefore not only eminently suitable as a metaphor for processes and concepts, but also for becoming the very »mythos« of the digital system, and for »somehow« shifting forward its traces historically. The critical potential of the discourse on the algorithm would consist in exposing and occupying this mythic aspect, making the algorithmic a category that was always present, but one where the consequences of its reductionistic one-dimensionality are only comprehensible today. This makes it a category that needs to be addressed all the more urgently, and one that stands for something very different from merely calculatory or technological processes: it stands for the ideological subjugation of the world to the law of the numeric and the automatic.
However, most discussions focusing on the continuity of the algorithmic concentrate chiefly on the art-historical embedment of current trends in computer art and software art. A current example of this is the ZKM exhibition »The Algorithmic Revolution«. This reference allows it for the first time to place the ZKM's own collection, with its focus on interactive media art of the early nineties, within a framework that encompasses various media and fields and includes works of the fifties, sixties and seventies from the areas of art, architecture, music and literature, as well as contemporary software art. The exhibition demonstrates clearly that an engagement with the »essential nature« and effects of the computer does not necessarily have to take computer art as its point of departure, but that every medium can address these issues.
In keeping with this approach, the exhibition starts with the presentation of a Zuse computer coupled with kinetic and optical art; in other words, with artistic styles whose tendency to formalism would have obvious parallels to current visualisations of fractals, cellular automata, L systems and evolutionary algorithms. However, this reference is not made. The »works« that are based on written and drawn instructions seem virtually to be algorithmic showpieces, such as John Cage's scores for musical events or the performance directions by George Brecht, which are disconcertingly concise: »NO SMOKING EVENT. Arrange to observe a NO SMOKING sign./smoking/no smoking«. These works are made special not by what is calculated, as in kinetics or in the early computer art of people like Frieder Nake or George Nees, in which simple input-output processes take place, but on the contrary by their structural openness and the interpretability of these instructions – the fact that every instruction can turn out differently depending on its execution. This means, however, that it is the elements of shift in the instruction that count, and not complete compliance with it. In short, the algorithmic in fluxus is a »program« for participation, for occupying the gaps, and not for programming an input-output system. Chance and unpredictability, emergence and contingency4 play a central role in generative art today, but they are, in the final analysis, always the output of a calculation – something which hackers also use when they utilise random generators for their purposes.

The algorithmic as an abstraction
The examples of contemporary software art in the ZKM exhibition provide a spectrum ranging from rather formal, innovative approaches to those with politically engaged content, and can give the exhibition a point of convergence. But in the end there are too few for them to be able to represent contemporary discussions in a comprehensive manner. For example, the most important initiators of the discourse on »generative art« are missing: Geoff Cox, Adrian Ward and Alex McLean, all from England, who run the eu-gene mailing list on this topic and who organised the exhibition »Generator« two years ago – an exhibition that actually anticipated the ZKM exhibition in its conceptional approach, but focused mainly on software art.5 Also missing are some works, such as the »Auto-Illustrator« by Ade Ward, an adaptation of Adobe Photoshop, which received an award at the Transmediale.01 and allows one to experience the automatistic pitfalls, holes and rigid licences of this commercial software.6
One example of innovative interface software shown at the ZKM is the reactive, audio-based Net work »Yellowtail« (2000) by Golan Levin,7 for which he developed the software AVES (Audiovisual Environment Suite). This software reacts to movements of the hand with sound and images and produces a display that permanently coils and moves, and sometimes glows auratically while producing sounds. In the field of audio and sound, and above all in the area of clip culture, the algorithmic is a popular method of producing synaesthesias and minimalist aesthetics that remind one of hidden machine rhythms. The exhibition »Abstraction Now« and the film programme »Maths in Motion« organized by sixpackfilm in the Vienna Künstlerhaus provided a really good and comprehensive overall view here.8
Another example of similar abstract trends is the work »pedigree« by Annja Krautgasser and Rainer Mandl, which was displayed this year at the Viper festival in Basle, among other venues. It was made for the second part of the »CODEeDOC« exhibition at Ars Electronica 2003, whose main theme was not a theme at all, but an instruction. A work was to be created where at least three dots and three lines resulted; in addition, the source code had to be disclosed. Krautgasser and Mandl revived a sort of Oedipus myth in the digital era. Instead of the narration of his life, an abstract family history is generated on the basis of dots and lines that grow, become denser and are permanently in motion. The process continues until you click on »reset« and the story starts again from the beginning. The myth has become a generative process of geometric elements. This can be understood as a proposal for reading the basic building blocks as new myths, as an engagement with the algorithmic reduction of the world to numeric entities, as a game with the current fascination with evolutionary algorithms that at the same time makes clear its limitations.

The political aspect of algorithms
Interestingly enough, there is a growing tendency in the context of abstract software art to read minimalist works and those based on formal methods as politics of form and a critical reflection on software.9 In its praise of self-referentiality, the argumentation follows a similar line to that already used in the net-art debate at the end of the nineties: as soon as software is used as a tool in a manner for which it was not intended, so that it at best generates and reveals its own regulatory mechanisms, it is interpreted as artistic, critical and political. The more abstract it is, the more remote the graphic interface appears to be from commercial, object-oriented software and the more proximate to the source code, which is frequently stylised into the very »essence« of software and graphic interfaces. The sympathy artists feel with the political objectives of the Free Software Foundation has in the past few years led to a sort of code fetishism finding its way into software art. It is already seen as avant-garde when, instead of a flood of images, the much more insubstantial program code is shown as an image. The exhibition »CODEeDOC«, which tries to focus on the code and its execution, is probably the most extreme example of this.
It would be necessary to examine separately the questions of how far the tendency towards formalism, which was thought to have been dispensed with, has wormed its way into politically engaged software-art contexts as well, and whether the »wilfulness« of the code has not been overestimated.10 However, as long as the connections between the »control society« (Gilles Deleuze) and its basis – the algorithms – are not constantly concretised and contextualised, the apparently direct visualisations and automatisations of code must always be read as renderings of themselves as well and not as instructions for a critique of the possibility of the powerful reduction of life to mere code. So where is the advantage today – that is, in the golden age of the »informatics of domination« (Donna Haraway) and almost a century after the theories and practices of »art in the era of its technological reproducibility« (Walter Benjamin) – in still constantly propagating strategies of automatism, if they do not only want to transform the history and ideologies of the undead author?
In order to give at least a partial answer to this question, I should like to return once more to the »smart artist«. During the preparations for her exhibition »legal perspective« at plug.in in Basle11, Cornelia Sollfrank was told that, owing to possible copyright suits, the exhibition of the flower pictures by Andy Warhol as sampled and transformed by the net-art generator had been judged to be too dangerous and was not permitted. Instead, she showed interviews that she had held with lawyers about the legal implications of the net-art generator. It became clear that the definition of art not only increasingly threatens to become a matter for jurists, but that there are actually no clear regulations; and that precisely this lack of guidelines does not lead to moderation, but to a much more total mobilisation of the legal machine. In this case, the algorithmic as an artistic method makes visible the algorithmic in our society and shows, contrary to my initial scepticism, that automatist artistic practices are more than important, as one never knows how real they will become.

 

Translated by Timothy Jones

 

---