Reviewer biography

Current Reviews

Review Articles

Book Reviews Archive




Aesthetic Computing

by Paul Fishwick, Editor
The MIT Press, Cambridge, 2006
488 pp. Trade, $42.50
ISBN: 978-0-262-06250-3.

Reviewed by Michael Kelly
Department of Philosophy
University of North Carolina at Charlotte

Let me start with two brief disclaimers to make it clear what aesthetic computing is not since it is a new field and there is naturally some unclarity about its identity. Aesthetic computing is not the application of computer artifacts — models, programs, data, codes, interfaces, and the like — to art or aesthetics. There is such a field, and it’s called computer art or computer aesthetics. Also, aesthetic computing is not directly concerned with the development of new art mediums such as interactive art, software art, internet art, and the like, though these mediums may enter the discussion because they embody some of the results of aesthetic computing.

Rather, aesthetic computing is about the application of the arts and aesthetics to computing. According to Paul Fishwick, aesthetic computing takes the computing discipline itself as its raw material and explores how aesthetics might productively shape computing (including programming languages, AI, HCI, graphics, visualization) (pp. 7-8). Or in the words of Roger Malina, the aim of aesthetic computing is "to transfer ideas and techniques from the arts to computer science and engineering" (p. 44).

In elaborating on the impact and scope of this transfer, Malina highlights a dichotomy within aesthetic computing, or indeed within computing as a whole: Is the computer to be understood as a transparent "information appliance" or as a "medium for reshaping perception and cognition" (p. 44). If the computer is an appliance, aesthetic computing is a matter of design aimed at making the computer as transparent as possible so that we can achieve the desired results, such as effective communication or legible visualization. But if the computer is capable of shaping perception and cognition, aesthetic computing is a way to understand how perception and cognition can be shaped by and, in turn, shape, technology.

Following the structure of this dichotomy, Malina outlines two kinds of claim, weak or strong, that can be made on behalf of aesthetic computing, depending on whether we’re talking about the design of the finished products of computer technology or the codes underlying computer software. "The weak claim is that by stimulating the flow of ideas and methods from the arts to computing, computer scientists and engineers will achieve their objectives more easily, quickly, or elegantly" (p. 47; italics added). For example, artists can demonstrate how computing devices are more likely to be adopted by the public if they are found aesthetically appealing; these insights might, in turn, inspire innovation in future research projects (with the Apple iMac or iPod often sited as exemplary success stories). By contrast, the strong claim about aesthetic computing is "that by introducing ideas and methods from art and design into computing, new practices and approaches will emerge responding to new objectives that would not naturally have evolved within the computing sciences and engineering" (p. 48). Here, the claim is that aesthetic insights gained from artistic practice do not merely allow computer scientists to achieve ends formed without taking aesthetic considerations into account but that these insights actually shape the objectives of computing enough "to redirect the future development of computing, provoking new developments and inventions that would otherwise have been impossible. A different computer science and engineering may emerge" (p. 50). This is a strong claim, indeed, which Fishwick corroborates by claiming that one of the "core goals" of aesthetic computing is "to modify computer science through the catalysis of aesthetics" (p. 11).

To answer which, if either, claim about aesthetic computing can be supported, we first need to clarify what aesthetics is. Fishwick offers some clarification by dividing aesthetics into three concerns: modality, or "ways in which we interface and interact with objects"; culture, meaning genres, movements, and such in the history of the arts; and quality, referring to symmetry, complexity, parsimony, beauty, etc. (pp. 12-13). Although this division is helpful, the inclusion of "quality" (or, better, "property") requires some clarification because it determines how we approach modality and culture. So let me add yet another disclaimer. Aesthetics is not merely about symmetry, harmony, elegance, optimality, and other similar properties of the artifacts of computing, whether they are used in computing or created by it. It’s not that these properties aren’t relevant in aesthetic computing; it’s just that aesthetics is a philosophical discipline and these properties are not, by themselves, philosophical issues. In fact, aesthetics is not about the specific properties of any particular objects, whether works of art, natural objects, or artifacts of computing [1]. If I can use the term ‘Beauty" with a big ‘B’ to stand for the set of all such properties, including the particular property of beauty with a small ‘b’, Beauty is not a property of any object. This does not mean, intentionally or unwittingly, that aesthetics is subjective or that, as we often hear, Beauty is in the eye of the beholder. Aesthetics isn’t subjective any more than it’s objective, since beauty is not in the subject any more than it’s in the object [2].

Then what is aesthetics, or, where is Beauty? In the language of eighteenth-century aesthetics, Beauty is a relational property, that is, a property resulting from relations between human subjects and certain objects in art or nature. Or, in the language of contemporary computing, Beauty is an interactive property between human subjects and the artifacts of computing. What this means is that when aestheticians take up the question of Beauty, they concern themselves with the nature and structure of the cognitive and affective relationships between human subjects and certain objects in the world, to which we can now add computers. The objects here are at the same time occasions for interactions not only between humans and objects but among humans. To take a simple example that does not necessarily involve computers, when several people take pleasure in a painting, opera, or pop song, the artwork is an occasion for these individuals to discover something they have in common. The philosophical issue this discovery provokes is what, at a deeper level, makes it possible for people to have a work of art in common. This deeper level involves human emotions, passions, and the like, as well as their effects on human perception and cognition. Insofar as aesthetics is the interdisciplinary study of the complex commonality that underlies our shared experiences of art, it is necessarily connected to other disciplines that are also concerned with human emotions, perception, and their interactions [3].

In contrast to this account of aesthetics, many contributors to this volume seem to attribute Beauty to artworks and thus to computers. For example, Laurent Mignonneau and Christa Sommerer emphasize complexity, diversity, and emergence as the properties in HCI, with a special focus on "users’ interaction input" (pp.169-183); Jonas Löwgren identifies a set of nineteen qualities tied to HCI (pp. 383-403); Stephan Diehl and Carsten Görg understand beauty in terms of the sum of elementary properties (pp. 229-37); and, finally, Michael Leyton develops aesthetic rules: maximization of transfer and maximization of recoverability (pp. 289-313). But this focus on Beauty as a property is what I’m claiming is problematic. Beauty is a property of relations or interactions among humans (which may very well be what the above authors have in mind) rather than of the works that occasion such relations or interactions. Aesthetics is the understanding of what makes such relations or interactions possible, not just what makes them more effective, more pleasurable, and the like, though by understanding what makes them possible, we’ll presumably be in a better position to address these other concerns. Aesthetic computing is the same type of understanding connected directly to computers [4]. In a word, if aestheticians now work with computer scientists, as I now expect they will, it will be a natural extension of what they’ve been doing all along.

Now, to return to the weak and strong claims for aesthetic computing, it’s helpful, following Fishwick, to narrow computer science to three areas and to identify what aesthetic computing might involve in each case. First, on the level of computer programming, there are questions about whether and, if so, how to represent programs and data structures with "customized, culturally specific notations." Second, there are issues about how to incorporate "artistic methods in typically computing-intensive activities." And third, in connection with HCI, there are issues about how to improve "the emotional and cultural level of interaction with the computer" (p. 6).

Fishwick provides a good example of the first case, for he argues that aesthetics will alter not only the design of computer software at the point that users begin to interface with it, but also the very programming that makes software possible (pp. 9, 13-20). The rationale for this strong claim is that programming will change as computer scientists alter their objectives as a result of attaining a better understanding of the aesthetics of HCI. Put simply, programming will have to change to create the desired interface — an obvious point, but one that is now coming with an aesthetic imperative attached. Norm Tractinsky’s and Dror Zmiri’s research on "skinnability" (alternate interfaces to commonly used applications) is a good example here because they focus on interaction, while taking consumers’ interest in skinnability as evidence of their interest in the aesthetics of computing (pp. 405-22).

Concerning visualization, there are two types which fall under the general heading of data visualization: scientific visualization, which is the creation of visual representations of scientific data from physics, biology, or any of the natural or social sciences; and information visualization, which involves visual models of information from all sorts of sources: business, government, the sciences, or elsewhere. Both types involve aesthetics since visualization is, in Donna Cox’s words, "the creative translation of data into visual representation" (p. 94). She provides a systematic and clear analysis of the aesthetics of visualization by explaining the basic metaphorical structure of the translation of data into visual models (pp. 89- 114).

Now, some people also speak of knowledge visualization, which, if I understand it, is a meta-level of visualization that articulates the epistemological implications of the two types. For in knowledge visualization the claim is that you’re not just visualizing or illustrating what is already known; rather, in the words of Monika Fleischmann and Wolfgang Strauss, "artistic works in the area of aesthetic computing must lead to a synthesis of sensory perception and cognitive insight, yielding new ways of thinking and models of experience" (p. 131). How this perceptual/cognitive interface works is a basic subject matter of aesthetics. For example, Aaron Quigley uses the expression "relational information," which is very similar to the idea of beauty as a "relational property" or "interactive property" (pp. 316-33). So there’s a natural role for aesthetic computing in visualization.

Finally, in the third area of computer science, HCI, we have the following picture, to quote from Frieder Nake and Susanne Grabowski: "Interface aesthetics is different from the aesthetics of packaging," the design approach to aesthetic computing, "in that the interface to software belongs to the software. Software never appears without its interface. The human-computer interface is, first of all, the face of its software" (p. 67). In this light, the weak and strong claims about aesthetic computing would be better characterized, as they are by Jay Bolter and Diane Gromala, as the inside and outside of computers, meaning the code and the interface (pp. 369-82). So we don’t have to choose between the weak or strong claims any more than we have to choose between the code and the interface. Rather, the interaction between the code and the interface is the basis of HCI and, in turn, the basis of aesthetic computing.

At the end of his introductory essay, Fishwick asks whether aesthetic computing is something new or whether it just "rehashed old material." He and his expert contributors argue that technology has developed to the point today where it is not only possible to pay attention to aesthetics, but there is now a sense of urgency coming from computing. In Fishwick’s words: "We have had to wait for the technology to become available to leverage the arts" (p. 13). If this is accurate, what we have here is a new field called aesthetic computing. And what we have in this collection is an excellent contribution to aesthetic computing, an extremely valuable text for aestheticians and computer scientists alike.

Works Consulted

[1]Fishwick claims that computer interface "should be as much about quality as it is about quantitative performance" (p. 21). My turn away from "quality" seems at odds with this claim. But I think we are proposing something very similar because he seems interested in quality only as it relates to the affective as well as cognitive dimensions of HCI rather than to the properties of artifacts (e.g., a computer or a graphic user interface) that would occasion such interaction.

[2] Frieder Nake and Susanne Grabowski (pp. 53-70) add semiotics to the aesthetics and computing mix, apparently on the belief that aesthetics is subjective (p. 55) and needs to be offset by the more objective semiotics. As I understand aesthetics, however, semiotics does not add anything that couldn’t be included within aesthetics. Umberto Eco’s combination of aesthetics and semiotics is an example of what I have in mind here.

[3] Jane Prophet and Mark d’Inverno (pp. 185-96) prefer to use the term "transdisciplinarity" in place of "interdisciplinarity" or "multidisciplinarity," because they think the first term emphasizes that something new emerges from the interactions among these disciplines.

[4] Elsewhere [e.g., in my Encyclopedia of Aesthetics (New York: Oxford University Press, 1998) or Iconoclasm in Aesthetics (New York: Cambridge University Press, 2003)], I characterize aesthetics as critical reflection on art, culture, and nature. In this light, aesthetic computing is critical reflection on — or philosophical analysis of — the aesthetic theories, principles, beliefs, ideas, and the like underlying computing once it is governed not only by technological concerns but by artistic practices.



Updated 1st January 2007

Contact LDR: ldr@leonardo.org

Contact Leonardo: isast@sfsu.edu

copyright © 2007 ISAST