Diagramming a New Theory of Picture

The goals, uses, materials, and processes of science and art are not necessarily exclusive, but often mirror each other. In both fields, a premium is placed on freedom of enquiry and instrumentalizing both physical and metaphysical data, all in the service of hypothetically reciprocal (but as often competing) social and theoretical ends. The differences in presentation and interpretation often lie as much in the chosen application of visualization technology and the expected terms of service.

*Caption at foot of article.

Put differently, consider the more recent question of “image” versus “file.” Most scientific imagery is produced for professional journals in the form of reductive quantitative diagrams—tables, graphs, Muller plots, etc., that ostensibly reflect an “objectively” procured set of source data (the inherent scientific content). The “file” of data is the nominal work product. In reality, many of these graphs are so graphically essentialized as to be almost useless scientifically, while the contextual information such as spatial co-ordinates, time and duration is easily manipulated to help the simple data visualization conform to the research premise. But the premise remains that the source information is directly accessible and subject to review.

In a few data rich fields, like astronomy, network theory, medicine, and biology, more complex data visualization methods can produce more highly developed visual forms.  But here the contextual information is often stripped away. The essential keys to their scientific sources (such as spatial coordinates, time and duration) are removed and the primary file is doubly compressed, from a data table into a fixed image, now more resistant to future informational decompression and more accessible to aesthetic compression.

Personally, I appreciate the visual economy of the image as much as the complexity of the file. But in many cases of the first type, instrumentalizing one kind of knowledge, the primary database or file (assumed to be the “science”), with a simple visualization technology table, graph, or diagram produces an image that is not “art.” It retains its ability to be decompressed or be “useful.” In many cases of the second type, the combination of a given visualization technology and the secondary database, or re-processed file (still science?) more thoroughly instrumentalizes the second kind of knowledge, and produces another, tertiary kind—an image (art?).

In art, the mirror premise is that the subjectively procured source information, is, like the source data in a science experiment, somehow still directly accessible, either as “raw” content or through material form. In reality, Frederik Stjernfelt’s analysis of the sketch describes how the original terms of any given hypostatic abstraction are progressively obscured as they are aestheticized through a similar three-stage process. This should be no surprise, as all information accessible to what Antonio Damasio calls the “core-self” must undergo a similar three-stage compression—to present the information in a form that is accessible to the viewer.

There is another, ongoing version of this discussion in the art world between the terms “photograph,” “image,” and “file”—all of which also produce “pictures.” We lack a meaningful grammar to analyze distortions in the translation, or gauge the inherent mirror symmetries that might govern the aesthetic compression  (how information decompresses into and out of pictures).  We don’t know how to evaluate the exchange between science and art. “Sci-art” seems to fall into this space.

Can multiple forms of enquiry, research and knowledge production—whether, intuitive, deductive, scientific or artistic—be engaged and represented simultaneously in a way that allows them to be coherently transposed, decompressed and usefully compared inside a common theory of picture? A promising form to carefully consider might be the diagram, or “informational drawing,” which re-emerged in the 17th century as the essential tool of scientific research, and whose ability to concretize process can be clearly distinguished from Foucault’s concept of the “table” or closed disciplinary array, by its ability to both cross boundaries and to produce pictures of thought structures—or theories of picture.

Can this be done? I’ve developed a simple visual grammar of how these theories might interact diagrammatically that visually unifies diverse approaches in the histories of science, aesthetic theory, network theory, and ontological philosophy. The dynamics of a complex system can be described as a walk, or game, drawn on a high-dimensional free energy surface. More knowledge of the properties of the system can be obtained if one knows the distribution and properties of all the local minima of the surface. Just as there are minima for physical reality (the four constants), there are ontological minima for the human framing of knowledge, and spatio-temporal minima for the useful proliferation of information. However, most of the time, any system is confined in its deep local minimum, as the high transaction costs of moving between ontological and temporal energy minima mean that any new form trying to escape tends to revert to the local informational minima—hence the difficulty of defining a combinatory position between science and art!

  • Some of the shared forms of enquiry in a common theory of picture are shown here using a simple visual grammar, projected over a simplified contour map standing in for a hypothetical higher dimensional information/energy translation surface. The four ontological minima are shown at the corners, the four spatio-temporal minima are shown along the sides. Just as in a real landscape, there are transaction costs involved in moving from one area to another by climbing over 'saddles' in the energy surface. Terms of informational enquiry, such as imagination, anticipation, prediction or representation, have varying transaction costs and ergodic recurrences, depending on the topology of the surface and the valences of the local minima. Most of the time, due to the high transaction costs, enquiries will tend to revert back to the nearest ontological minima.

Although this diagram shows the minima symmetrically organized, we do not yet have an informational gauge symmetry. However, even if the phase space were actually symmetrical, the ergodic hypothesis allows for all possible states to be reached over time, and that they may recur.


Matthew Ritchie

Matthew Ritchie is an artist and writer based in New York.