Search This Blog

Friday, November 8, 2013

Examples of Datamade Remixes? Or digital art by another name? by Joe Nalven

The recent article posted by Patricia Frischer about  West, Malina, Lewis et al) suggests a digital path to Duchamp's readymades. Consider the analogy:

In both process and outcomes, the “datamades” resulting from DataRemix are envisioned to function analogously to Duchamp’s readymades. Their ultimate objective is to destabilize the framing narratives of data creation and representation in order to generate the possibility for new forms to arise in hopes of allowing us to see and know beyond what our instruments, algorithms, representational schemas and prevailing culture enable us to see and know. Yet, reappropriation and recombination also bring with them the framing narratives of artistic traditions from the early 20th Century that continue to evolve in our digital culture.  (excerpt from DataRemix: Designing The Datamade - Through ArtScience Collaboration by Ruth West, Roger Malina, John Lewis, Member, IEEE, Scot Gresham-Lancaster, Alejandro Borsani, Brian Merlo, and Lifan Wang)

Unfortunately, there are no examplars in the article (or that I could find on the web using their  terminology):  What kind of art this might be?  In the spirit of the digital zeitgeist, I will propose several examples (below) that might fill this void.

However, before doing so, an important and final disclaimer in the article is worth noting - can these art objects be created from a 'neutral' exercise in finding datamades or must there be a deus ex machine (or an artist ex machina or artist ex computer)? Simply put, if there is no artist making a decision in whatever field of experience, on whatever planet in the universe, then there is no art. If a tree falls in the forest, etc. etc.

Let us remind ourselves that Marcel Duchamp's Fountain, and other readymades, did not magically get transported into a gallery. There must have been: (1) a recognition by Duchamp that this was the object he wanted to put in the exhibit; (2) Duchamp's purchasing the object; (3) Duchamp's reorienting it for display; (4) Duchamp's signing the object with a pseudonym; (5) Duchamp's submitting it for exhibition; (6) a rejection by the art committee and hiding of it during the show; and (7) photographing of the object by Alfred Stieglitz. (There are other versions of what happened, but that is simply another path of decision-making of taking the 'fountain (urinal)' and transforming it into a 'readymade.'

Obviously, to get an analogous object - in whatever form(s) - there needs to be a similar set of interventions by the artist and others.
But, why even bother about an analogy to finding readymades in Big Data, genomics, astrophysics and the like -- and then calling them datamades? There still must be an artist making a decision to call 'X' (whatever 'X' is) something other than what it seems to be in the real world and labeling it 'art.'

The abstract provides an answer of sorts. A crisis is imagined and the datamade is offered up as a solution. Does it work?

We propose a role for ArtScience research and creative work in contributing to the necessary shifts to go beyond the current crisis of representation. We specifically describe DataRemix, a recombination and reappropreation (sic) practice intended to trigger novel subjective experiences and associations. The narratives framing data creation and representation circumscribe what we can see and know, and how we see and know. How do we see and know beyond what our instruments, algorithms, representational schemas and training guide us to see and know? How do we look for what we don’t know we’re looking for when we can only examine at most a tiny fraction of the available data? Our argument is grounded in and will be illustrated by experience with several ArtScience collaborations spanning genomics, astrophysics, new media, and holographic sound design.

I might note in passing that Protagoras' famous quote suggests that we are stuck with ourselves and the tools of measurement we use in representing what we know and perceive. Clearly, our tools and methods change, but these newfangled tools are still a human view of reality. Big data, genomics, etc. are still 'human tools and human measures.' So, how are we to get outside of the human condition - a NHI type of art? (NHI = no humans involved)

I appreciate the concluding words which admit of an impossibility to the proposed tasks.

Yet, reappropriation and recombination also bring with them the framing narratives of artistic traditions from the early 20th Century that continue to evolve in our digital culture. These carry an aura of arbitrariness that runs counter to the functioning of science which requires reproducibility and validity. This very contradiction is at the heart of our working definition of DataRemix. In proposing DataRemix we hope to contribute to the dialog about arbitrariness already ongoing in the visualization community. Maintaining the dichotomy of artistic approaches as devoid of meaning, decorative or subjective and non-artistic approaches as meaningful, valid and objective eschews the practical reality that, as Monroe observes, visualization is inherently aesthetic and created for an intended audience, and iterates towards the audience as part of the analytic process. Additionally, familiarity with a representational schema enables us to forget that at one point elements of its design were also based on arbitrary yet repeatable mappings that lead to their utility and meaning. Stylistic and aesthetic concerns are increasingly a subject of study in the VIS and HCI communities. As Viegas and Wattenberg reflect, the power of artistic data visualization arises from artists “committing various sins of visual analytics” and directly engaging and guiding an audience towards a point of view. 

[Ah yes, their are artistic interventions and somehow geared to an audience!]

They remind us that even with dispassionate analysis as its goal, creating a visualization that is truly neutral is “generally impossible” and propose further exploration of the value of artistic explorations. In this light, we propose to explore DataRemix as a mechanism for artistic approaches to engage empirical approaches in creating new ways of seeing and knowing. (References are in the original article.)

Well, what is really possible and fruitful?  There are any number of artists that play with randomness. Such randomness can be applied to the incorporation of any field of data.  I would think that if one allows impurity (namely, human and artistic interventions) then the model works. Without those 'impurities' (that's me and you and all the other humans reading this narrative), we get Platonic zip. Idealized nothings.  But then, that's my point of view. 

Here goes with some impure artistic inventions that incorporate randomness into their methods. 

These examples would be DataRemixes that yield datamades.  (I suppose if there is an objection to my use of these neologisms, I can call my examples 'DataRemixes2' and 'datamades2.' A rose by any other name is still a rose.) 

Paul Reiners Cellular Automata

I don't pretend to understand the mechanics of cellular automata, but Reiners has been incorporating this method into music and visualization. While the approach is intentional, the results provide randomness. 

Paul Reiners, Cellular Automata in Van Gogh's Sunflowers
Reiners:   A CA consists of:

    A matrix, or grid, of cells, each of which can be in one of a finite number of states
    A rule that defines how the cells' states are updated over time

The matrix of cells can have any number of dimensions. Given a cell's state and the state of its neighbors at time t, the rule determines the cell's state at time t + 1. (This will become clearer after you look at some concrete examples.)

So, how would this look.  If you go online you can watch (enlarge the screen) the cells within Van Gogh's Sunflowers move as if they were little insects.

Don Relyea's Random Art Generator

What I was able to do with Relyea's random art generator is to control its source data - either limiting it to a predesignated folder on my computer or allowing it to mix with randomly selected objects on the web.

Paradox created with Relyea's Random Art Generator limited to file folder on my computer. (Upper)  Konecni Random mixes an image provided by Vladimir Konečni with an object selected at random from the web. (Lower) Layout of items is randomly assigned by Random Art Generator and then reworked by the artist.

These two examples, it might be objected, begin with an artist's intent with randomness subsumed to the original purpose. The examples are not just found.  But then, neither was Duchamp's readymades just found.  Their is a conceit in each approach that gives the appearance of some found object as if it were randomly encountered. Hardly. While the readymade is not studio art or a plein air painting or an intended photograph, it is carefully selected with an artistic purpose.

The proposed DataRemix must also rely on some conceit to give it the air of being found - as the analog to a readymade, the newly minted datamade.

We are trapped inside the human condition, the human measure of things. The artist can create a novelty as if it were found out there somewhere - a philosophy of 'as if.'

I am arguing for a lower threshhold, a less ambitious approach to finding art in the vast hurricane of data on the net and in our minds.

And all of these are wonderfully included in the making of contemporary art in digital media (might even call it, digital art).

In that regard, I am pleased to see further advances with all things digital.


As Joe Nalven obliquely points out, essentially all of the concepts discussed by the authors of the article are old hat – as concepts. To realize this, one needs only to examine developments in electronic music since 1950s in which a variety of approaches to randomness has been utilized. In some of the allegedly aleatoric work of John Cage, there is actually a great deal of composer's intervention. Iannis Xenakis sometimes intervened very minimally indeed, using the random-field stochastic processes.

May I briefly go off the subject and introduce the Chinese man whose portrait I took? (Joe combined it with some allegedly randomly chosen object from the web.) I met him on one of the many thousands of stone stairs leading to the peak of Tai Shan, a holy mountain (5,000+ feet) in Shandong province. Off to the side of the giant staircase, the charming man ran a miniature fertility counseling center!

Vladimir Konečni

Saturday, November 2, 2013

New Ways of Seeing and Knowing by Patricia Frischer

DataRemix: Designing The Datamade Through ArtScience

Ruth West just presented this data remix paper at IEEE VIS Arts Program (VISAP), Atlanta, Georgia, October 2013 the full paper is available on line

Visual artists are comfortable with collage and ready mades and over the years have gained an audience for these works which combine disparate objects to generate new meanings. In the music word this process is called remix or mashup. Using a computer to aid in this process , we use tools like copy and paste. In fact, every one who uses copy and paste is in fact, remixing. 

Ruth West (one of the judges for the DNA of Creativity project, and her colleagues (Roger Malina, John Lewis, Member, IEEE, Scot Gresham-Lancaster, Alejandro Borsani, Brian Merlo, and Lifan Wang ) are attempting to use scientific data in this way. They are calling this "datamix" generated art "datamades".

The goal is to encourage the scientific community that usually relies on feasibility through reproducibility to throw caution to the wind and see what happens if arbitration is embraced. The hope is that the results will have one big advantage that the arts can claim and that is relevance to an audience. 

Please read the entire article for extensive explanation of their whole process.  We would love to hear your comments and observations on this subject 
Locations of visitors to this page