Software art and poetry: Graham Harwood and William Blake's 'London'

Lecture written for the Electrohype conference 2002, in Malmo Sweden. In this text I present a new work by Graham Harwood, a piece of software that is based on a poem written by William Blake. The software can be used to calculate the total amount of last breaths of people dying in a inhumane situation over a specific period of time, in order to then use the outcome of this calculation to determine the length of time a horn or siren should be played throughout a city. The last breaths play the horn.

harwood s

I have to admit that when I first heard about 'software art' I was a bit skeptical about it. Why invent a new term for something that is in fact already part of older notions of art, like net art, media art or digital art? But then again, net art is closely related to media art and so is digital art. Often different terms do point at the same or similar art practices. We simply need to be able to address quite specific aspects of new media art practices clearly and we need to be able to focus on details of those same practices. We are entangled in a mesh of terminology, a labyrinth of words, through which we have to communicate our experiences to others. Like software art is now becoming a way to reveal and emphasize the deeper meaning and workings of digital art, net art and media art, so does the language we use to discuss it serve as a way to reveal and emphasize the connections between all present day media art practices and art at large. Software art and a focus on code helps us to see and to show an aspect of art which, after a strong presence in the 1960s and 1970s, has become slightly obscured again in the last two decades: the close relationship between art concept and the art work itself, or in other words: the connection between our use of language and our art practices. The domination of the visual in the arts is once again opposed, yet often in a less conscious way then in the conceptual arts of the sixties and seventies.

It is an age old battle, the philosophical battle between the material and the immaterial, the struggle with the question what is more important: that which we see or that which is (mostly) invisible. This is not just a philosophical issue though, it is most of all a matter of knowledge and therefore also of science. The importance of hidden or obscure matters depends on our knowledge about it. We gain knowledge through experience and through education. Most of the knowledge we have of things 'invisible', be it about for example atoms, the universe or evolution, is taught to us through education and not through personal experience. We trust our teachers and scientists, and we accept what they say. We base judgments on what we know through mental experience of a history of science, not just on what we actually see for ourselves. One could maybe say knowledge is a way of seeing without eyes, a way of seeing things internally, a way of being able to envision things and situations. Knowledge is therefore partly dependent on our imagination. It is related to the imaginary, because a lot of knowledge, like the imaginary, depends on our capacity to experience something with our minds alone. In daily life our mental and our physical experiences become one. When we experience an art work, we depend for judgment of it on both our mental and physical experiences of earlier encounters with art. The theoretical, historical and conceptual context of art is part of our experience of art. With digital art, net art, media art or software art we still do not have enough of both the mental and the physical experience to make a sound judgment of these art practices. We lack an internal experience with these art practices most of all though. We have not met enough with the minds of the makers yet.

With the computer we have (if we follow the imagination of Marshall McLuhan for a moment) created an extension of our brain that helps us with many of our brain's functions, but not with all. The computer can help us remember, it can help us calculate and it can help us connect and exchange information. When following the lead of the computer's primary functions some artists can easily get stuck in a quite formal approach of art. In fact, a lot of so called digital art explores and admires the beauty of the computer's calculations and processes, like for example the popular fractal art. In his text for the catalogue of the Read_Me festival in Moscow the German writer and software art critic Florian Cramer writes: "..such aesthetic conservatism is widespread in engineering and hard-science cultures; fractal graphics are just one example of Neo-Pythagorean digital kitsch they promote." To reach beyond the level of predictability, decoration and kitsch when using a computer within an art process asks not just for knowledge of the way a computer works, but it asks first of all for creativity and imagination. Florian Cramer then continues: "As a contemporary art, the aesthetics of software art includes ugliness and monstrosity just as much as beauty, not to mention plain dysfunctionality, pretension and political incorrectness." In other words, it asks for human rather then technological complexity.

Earlier in this lecture I made a connection between the use of language and the use of code, or rather the relationship between creating meaning with language and creating actions and objects with programming. What interests me very much is that we discuss this topic inside a complex structure of code and meaning, namely the internet and the developing structures of media art institutions and criticism connected to the internet. In some ways the discourse around software art has replaced or taken over the critical debate around art in networks. The focus on code rather then a focus on the network is once again forcing people to pay attention to deeper complexities of technology and culture, instead of gazing away at the hypnotizing shine of on screen fashions in visual design. But like with net art, software art suffers a bit from the expectations that seem to accompany any development in media art almost by default: that of an inherent technological innovation being part of media art practice. People tend to think technological skill and innovation are at the basis of electronic art always (and the value of a media art work is in their eyes also connected to the novelty it represents), but personal media, and especially video camera's, cheap sound recording tools, personal computers and the internet, have 'de-professionalized' the usage of electronic media to the extend that most contemporary art practices, be they expressionist, conceptual or formal, are quite easily combined with the use of electronic media. In a correspondence I had with Florian Cramer before this conference he says the same about the essence of new media, computer code: "Since networked computers have become as a mass commodity, code is no longer a clean-room construct, but flows in abundance and of course can be artistically used also in ways that would have scandalized everyone from Pythagoras to Donald Knuth, i.e. in incorrect, ugly, eclectic and whatever fashion." Like with painting, sculpture or performance the value of a media art work does not necessarily lie in its innovative character at all, but it lies in its artistic value, just like with any other art work.

Artists can make complex works with electronic media nowadays without too much trouble, yet media art works are still approached from an angle of design (a combination of aesthetics and skill) mostly, which in turn makes potential new media artists, art professionals and art audiences shy away from them unnecessarily. For the Jodi exhibition at the Plug In gallery in Basel I was in contact with Joan Heemskerk and Dirk Paesmans. To help me write a text about their way of working for the catalogue they told me some fascinating stories about their work and the way they see themselves. What was most interesting for me was that Jodi are, like some other net artists (Heath Bunting, Alexei Shulgin), absolutely not interested in presenting their work as a result of technical skill, on the contrary! Inspired by the programmers' term 'cargo cult code' they turned the negative interpretation of this term by programmers (namely: unskillful and messy coding) upside down to use it to their advantage. Taking the original meaning of the term 'cargo cult' they see themselves as artists -mimicking- the old notion of media art, evoking media art 'support' and acceptance for their work without actually following its rules. The work of Jodi is like the work of many present day media artists, much more intuitive and expressionist then has been acknowledged until now. Like Florian Cramer already said: "As a contemporary art, the aesthetics of software art includes ugliness and monstrosity just as much as beauty, not to mention plain dysfunctionality, pretension and political incorrectness." The emancipatory processes of art and their liberating effects on identity and personality in electronic media are also part of digital art, network art, media art and/or software art. Two years ago Dirk Paesmans stunned the rhizome mailing list with an abundance of seemingly machinic mails that mostly showed one crooked line scribbled over entire mails. At a presentation in Amsterdam Dirk Paesmans revealed these hundreds of mails were not computer generated, but manually created by him, one by one, in a sleepless night. Text, code and therefore also email has been very important in Jodi's work from the beginning. End of 1995 they were best known for mysterious, strange mails filled with a landscape of signs and letters. Whereas many critics have marveled over Jodi's skillful play with web visuals and source code, they liked to compare their work to concrete poetry.

In his own way, the artist Graham Harwood does the same. Harwood likes to look at art and culture from a broad, historical perspective. For Harwood everything is part of a bigger story, a context in which politics and culture can not be separated. A museum for instance is not just a building in which a collection of art is preserved and presented. As a member of Mongrel (consisting of Graham Harwood, Matsuko Yokokoji, Richard Pierre-Davis, Mervin Jarman and occasionally Matthew Fuller) Graham Harwood looked not just at the Tate Modern art collection, but also at the history of the building of the Tate Modern in London in a way the Tate Modern itself had never done before, at least not publicly. Suddenly the building is not just a brick structure housing one of the most important contemporary art collections, but it is a historical site built with a specific purpose: it was a prison. This first and original purpose has inscribed itself into the building, and has been of considerable impact on british culture at large. For Graham Harwood and Mongrel the Tate Modern is a monument for a culture inscribed by elitism, racism and other undemocratic forces. To them this same culture has inscribed itself in many cultural products, also, or maybe -especially- in software.

The history of software and computers is one result of centuries of mathematical sciences. In order to progress human kind has always searched for escape from chaos and confusion in the apparent, but often false simplicity of linearity and calculation. Hidden tendencies and desires behind specific types of calculation and logic only rarely get discussed. The pursuit of rationality and logic separated from morals or emotion can create monstrous events and situations. By using the personality and work of the 18th century poet William Blake Graham Harwood attempts to relive and emphasize centuries of conflicts between rationality and emotion. In a period of his life when he was away from his home country (the year 2002 he has been artist in residence at De Waag, in Amsterdam), Graham Harwood was inspired by Blake's poem 'London', written in the second half of the 18th century and dealing with the death of poor children in an expanding London at that time. He started to transform the Blake poem into a perl poem (written in the programming language perl), at first without having a real idea about what the code he thus developed would or should be able to perform. In other words: Harwood started to sketch or write intuitively.

Harwood is floating between homesickness and home hatred. He is deeply affected by social issues, such as the stunning fact that even today 33% of all children in the UK live below the poverty level and on average 1 child a week dies of neglect (his words). Already in 1996 he produced, together with his students at Artec in London, a CD-ROM called Rehearsal of Memory in which he dealt with the personal history of his father, who grew up in an orphanage after his parents did not take proper care of him. Images of Harwood's father and mother were also used in the Mongrelized Tate site. In the Blake perl poem the words of Blake are transformed into a code to "manipulate London". Whereas William Blake used his skills and imagination to create a text which would affect the emotional and cultural consciousness of an 18th century audience to the best of his abilities, Graham Harwood as Blake in the 21st century uses the text to perform on two levels at once: that of literature (or the purely imaginary) and that of physical, be it audible space.

It is the literary side of software art that deserves most of our attention. In the CodeDoc exhibition, launched September 2002 at the Whitney Museum New York, curator Christiane Paul's aim was "to create transparency for an artistic practice that seems to be fairly unique: in software art and many net art projects, the artists write a purely verbal 'description' of a work that ultimately reveals itself to the audience as visuals or a 'communication process' in the broadest sense." For CodeDoc "a dozen artists coded a specific assignment in a [programming] language of their choice and were asked to exchange the code with each other for comments". The exhibition, which is available only on line, is not just a aesthetic experience, but it is also educational. Christiane Paul about this aspect of the exhibition: "I'm quite often confronted with an audience's notion that 'the computer does it all' or with the dismissal of projects because they don't present themselves as (colorful) visual forms. People sometimes approach this art with the language of painting, which can only lead to profound misunderstandings. I believe that in order to understand an art form you have to understand the very basics of its practice." In the Read_Me text Florian Cramer puts it more bluntly: "The history of the digital and computer-aided arts could be told as a history of ignorance against programming and programmers." The on line introduction to the CodeDoc exhibition elaborates on what it means to look at people writing in programming languages and creating art with computers: "In traditional art forms, the 'signature' and 'voice' of an artist manifests itself in aesthetics of visuals and execution. Every medium may have its specific language but in digital art, this language has a quite literal rather than figurative manifestation. In software art, the visual results of the artwork are derived from the language of code. Languages are defined by grammar and complex rules and at the same time leave space for individual forms of creative expression. Our identity and the roles we play are expressed in our use of language. One might assume that the aesthetics of artists who write their own source code manifest themselves both in the code itself and its visual results". Yet, they cannot be separated. What distinguishes software art from other digital, net or media art is the conscious involvement of the code into the art work by the artist.

In the case of "Perl routines to Manipulate London" we can see the specific way Graham Harwood perceives and uses code. In an interview conducted for the newsletter Cream in spring 2002 Harwood reveals his view of technology (and an art criticism that has to deal with technology) in an indirect way: "One problem we have at the moment is that we have these beautiful systems for collecting tax, these marvelous giant algorithms that are able to extract money out of our pockets. They have a great poetic. How they work, how they configure and bring all these different people together and how they structure them in these big things. It is a huge and marvelous thing. It is a beautiful monster. And you want to make it work right, you want to be able to get into every person's pocket. You want to be able to take all this money out and spend it on arms and all these other things. Programmers sit there completely fascinated by this giant monster, this universe of taking tax. The thing is, you need to be able to describe that. They only do this because it is beautiful. People don't do things that don't have this kind of beauty. It is the same beauty as the pilot dropping the bomb on people. It is very beautiful from above. He looks down and it is aesthetically pleasing, it is elegant, it is powerful, it is beautiful. That is why he can drop the bomb. It is the same thing with the tax system or the social security system or all these other giant systems. They are all beautiful. We need to be able to critique that. The net art critic needs to be able to deal with that poetry. When you're dealing with that poetry, then the place of people making specific statements about these cultures has a resonance and a place. Then we are dealing with structures that occupy our lives. Then we have much more context I think." From this long quote it seems quite clear that Graham Harwood is admiring and maybe even taming the beast, he is full of fascination and horror over the (almost) perfect systems we design to run our world. Also Harwood's specific use of code in the presented poem here is to "emphasize the dryness, the coldness of code and rationalism in an unexpected place". It is a way to create consciousness within a routine process.

In his text 'Software and Concept Notations, Software in the Arts' Florian Cramer describes the perl programming language as "a programming language of which the instruction syntax has been consciously developed as close to the english everyday speech as possible by its developer the computer linguist Larry Wall, which [means it] contains a great number of instruction words and leaves programmers great freedom in syntax and notation". Combining the freedom of the perl programming language with the rigidity of a mathematical approach to the actual theme or focus of the poem Graham Harwood is creating a comparable chilling experience for both the readers of the poem itself and the audience who will witness the actual 'physical' execution (just mark the morbidity of this particular computer science term in this context) of the code. The poem starts by describing its context, and the reason for its existence or development. It then continues with a method for calculating the lung volume of children according to class, age and height. From this lung volume the program then goes on to calculate the volume of air replaced by the last scream of the child. The amount of air replaced by all last screams of dying children in London from 1792 until now together will finally be emitted through a horn or siren, to be heard throughout London.

While Graham Harwood is working on it and talking about it, the Blake project develops further. Calculating the lung volume from dying children to transform their collective cry into one long audible wail, an almost endless scream, he started to think about an actual application of the code, about actually realizing the project in an existing city. London is not the only place where people have died from the results of pure greed. Being in the Netherlands Harwood started to think of Dutch history, in which plenty moments of shame can be found. From the children in London Harwood has moved to the holocaust (in which 90% of all jews in the Netherlands died because of the efficiency of the Dutch bureaucracy) and (more recent) the massacre at Srebrenica, former Yugoslavia where Dutch UN soldiers found themselves unable to raise a finger when thousands of men were separated from their families and killed in a town the Dutch were supposed to protect. The 'perl routine to manipulate London' is turning into a morbid and frightening method to raise consciousness.

But it all started with Blake's London. Graham Harwood does not just identify with Blake because of his connection to England. He is also fascinated by Blake's apparent artistic resistance to the rigidity he perceived in the new philosophies of the late 18th century which culminated in the enlightenment. William Blake lived at a time when the work of dutch philosopher Spinoza was of great influence. Spinoza's philosophy (in which he reasons how God can not be separate from the world, but how He must in fact be in everything all the time) created a huge freedom of thought on the one hand, but it also created a cultural nihilism on the other. If everything is God, then there is no force from beyond us that controls everything. More importantly: there are no sacred, authoritative rules or laws that bind and guide us anymore. Reason and calculation started to replace a world full of mysticism and pure imagination. Graham Harwood sees this period in western history as one of the most influential for western culture and the rationalism that is inherent to it. Like he says William Blake did more then two centuries ago, Graham Harwood aims to bring back the power of imagination into a rational environment. The power of an internal vision that is created by not just rationalist calculation alone. Harwood as Blake is wrestling with the conflict in western culture between imaginary and rational visual cultures.

Reading Harwood's 'London' is like getting trapped in a perverse universe of misplaced calculation. The words represent horror and the way they are ordered to form a piece of code, a simple routine to calculate one isolated aspect out of the entire human history told by the choice of words, has a nauseating effect on the reader. One does not even have to understand the perl programming language to feel a chill running down one's spine. At least I did. Yet this perl poem is not a technological novelty or innovation. This work is not a revolution of any order in the technological realms of media art. In fact, the first perl poem ever was written by the inventor of the perl programming language himself, Larry Wall, in 1990, 12 years ago. According to Florian Cramer perl poetry became very popular amongst programmers shortly after this. A lot of these perl poems were written in purposely dysfunctional code. To make the text turn against itself inside the code was and still is a popular and effective art form. So was and is making words part of program instructions, thus often giving words double meanings (read: sub merge my senses; ASCII art, Rekursion, Lyrik in Programmiersprachen, Florian Cramer 2002) or emphasizing meaning inside the poem. Harwood does nothing of the kind. He lets the rigidity of code reveal itself by combining it with a very sensitive subject: the death of children, jews or innocent fathers and sons in former Yugoslavia. His is a creative innovation, a successful artistic intervention in the realm of programming.