i was indicated this great piece of... receipt data art, called "Data Burger". Who is googling "data burger" anyway? (I am not telling...-------------------------------> Justin Katko!) This is beautiful:
More data stuff here: http://www.stat.ku.dk/~rahbek/OxfordHilary2006/Data2006 , which are the teaching archives of the statistics division of the Institute for Mathematical Sciences in Copenhagen
It reminded me of a few poems collected in the Conceptual Writing Anthology by Ubuweb, and in particular a dyptic instruction piece by Dan Graham entitled "Schema/Poem". It displays the same differential structures that are used by poetry/text generators. But while the texts generated by these programs are boring if not reprocessed by the human hand (RACTER for instance), the instructions themselves are cool to read: as a list with blanks to fill (cf. constraint literature like OULIPO, or more recently Olivier Cadiot's Art Poetic'), or as a tree structure, or even, of course, as code. Here is one piece of the dyptic:
I recently read a great essay by British poet and theorist Eric Mottram: "The Triumph of the Mobile: The Structure of Information, the Language of Computers and Contemporary Poetry". He proposes to read language experimentations within system theory and information contexts as "adequate retrieval systems", an idea that appeals to me: "exchange information systems or games of exploration".
In his argument, Eric Mottram quotes Frederic Nake's "There Should BE No Computer Art" (part of the critical publication that followed legendary exhibition Cybernetic Serenpidity): "The interest in computers and art should be the investigation of aesthetics information as part of the investigation of communication. This investigation should be directed by the needs of the people".
The conceptual writings don't deal directly with data, but I think it can be interesting to read them within the frame of a referential critical system (not only self-referential), because they can process, potentially, data. Information theory seems useful to read these.
The British conceptual group Art & Language provides the idea of indexation, to read their work, but it is so difficult to find examples of their works on the web that I can't really say much for now. I am reading a book called Essays on Art and Language by Charles Harrisson and hope to know more soon.
Still on Ubuweb, two artists from Art & Language are included in the conceptual writing anthology: Terry Atkinson and Michael Baldwin, with wheir piece, "22 Sentences: The French Army". This is the hilarious beginning of the text :
KEY:
FA -French Army
CMM -Collection of Men and Machines
GR -Group of Regiments
Assertions. Explicata.
The context of identity statements in which collection of men and machines appears as a covering concept is a relativistic one. Identity is not simply built into that concept. The "sense" of identity is contrasted with the constitutive one.
The FA is regarded as the same CMM as the GR and the GR is the same CMM as (e.g.) "a new order" FA (e.g., morphologically a member of another class of objects): by transitivity the FA is the same CMM as the "new Shape/Order one." It's all in the support of the constitutive sense that the FA is the same CMM as the GR. The interference is that the FA and the CMM have the same life history (both the FA and the CMM are decimated) and in which case CMM fails as a covering concept. If the CMM isn't decimated (no identity) the predicate fails. The "constitutive" concept stays. And its durability doesn't come from a distorted construction of "Collection."
[...] (full piece on:http://www.ubu.com/concept/atkinson_22.html)
2 networked writers would appreciate the "data burger" piece for sure. Alan Sondheim and Noemata (Bjorn Magnihldoen). They themselves have a pratice of what I would call data art, but in a much looser sense than the conceptual artists. It is fascinating to get lost in the archives of these two writers, archives that have been compiled from found, generated, distorted data in all formats, and stored in the most minimalistic way (Gopher interface like).
But I feel uncomfortable dealing with the theoretical issues of these works. I don't like this idea of textual noise as a form of drift/chaos, in my sense building up a very useless idea of a contemporary sublime.
I recalled of an essay I read a while ago, by Italian Mario Costa, and who actually argues in favour of what he calls a "technological sublime", issued by the experience of digital networked writing. Costa interprets this loss (or drift) as an experience of drowning in an "artifical sublime", an environment that he describes as a "landscape" characterized by a behaviour of indigestion and boulimia. But this sublime is concluded from some a priori that are not necessarily true. Costa presupposes that we approach electronic textuality (digital networked writing) with an interpretive and comprehensive behaviour. He doesn't take into account at all of the fact that new ways of reading have been generated by and simultaneously with the new textualities. Actually, they were sure existing way before, because the "informational mode" (Mark Poster) is less an era, a period, than a context which is reproducible and alterable through time. Think about the amount of information provided by newspapers. How do you handle it? how do you process text? Our reading of the internet textualities has a lot to do with that.
But Mario Costa talks about the "technological sublime" as a new "threat of alienation" that the self is subjected too (after having passed successfully the test of the scattering of the self imposed by the modern sublime-!). Cyborg theories. I guess, following Costa and the Sondheim data art genre, you could say that the 'noise' produced by artists is our saviour... from the "most alarming danger for what is human to be expropriated by the neo-technological" (Mario Costa, Internet et globalisation esthétique, L'Harmattan, "Ouverture philosophique", 2003, p.23). It functions as a 'counter-attack". Not chaos against order (this is an old modernist idea I guess), but chaos again entropy. Hyper-subjectivity for hyper-media is entropic, and chaos produced by artists is chaos generated to turn this alienation into "sublime" (yeah!). Hmmm... It is frightening to see that the only example Costa provides of a successful attempt at the "technological sublime" are the NASA experiments.
Eric Mottram, in the essay previously quoted, has another perception of what noise means in an information/system theory context. It is maybe more pragmatic and low profile, and doesn't end up programming space rockets. And doesn't even include the tiniest amount of sublime, which feels good.
The human system, like the computer, "can make relationships and refer them to its habitual programming unless it is obstructed. When the human system is obstructed, and when the computer system is abused, they have nervous breakdowns. Neither system find it easy or automatic to see or hear chaos, or make it, or recognize it. Two random elements juxtaposed tend to make a cell of energy fusion. Whatever the computer produces, the human system will want to transform into meaning if it is alert and tuned. Random accumulations tend to move away from chaos and become their own order".
Eric Mottram ends up making two propositions, arguing for his idea of art in a cybernetic context as an extension of cultural experience:
1. The Function of poetry is not to order the object but to open us up to kinds of experience which printed words or word signs or heard words can generate.
2. The nature of chaos and entropy in a poem is only conceivable if we acknowledge that chaos is impossible for a human being to create, since all he does is form and all energy has form.
Keywords: alan sondheim, conceptual art, dan graham, data, data art, data burger, database, electronic text, electronic writing, eric mottram, mario costa, network, noemata
In a room, a text that is written by nobody runs on a projection screen. The keys of the keyboard move swiftly as if by a ghost's hand and a mechanical voice reads out the generated text, sentence by sentence. 
In this interaction of machine words and human text, «Poetry Machine» creates a new écriture automatique, where language is no longer the exclusive domain of human thought but also that of the internal logic of computers.