Computational Media and Information Design
In Visualizing Data, designer Ben Fry (2008) provides a process for doing computational information design that is structured through seven stages— acquire, parse, filter, mine, represent, refine, and interact. This process is robust enough to guide a practitioner through many computational information design problems. It provides a flexible operational framework for working with data that is informed by Fry's extensive experience in using computation as a medium for information design. But for both the practicing designer and the design scholar, such processes need to be complemented with an understanding of the qualities of computation that underlie such an operational framework and that make computation a distinctive medium for design.
For this discussion of computation and information design, I draw on digital media studies, where much of the richest discussion of computation as a medium has taken place. In part, this may be due to the legacy of media studies and its concern with the characteristics of a given medium, whether radio, film, video, or digital. Extending this concern to the work of identifying the qualities of computation, there is value in understanding what distinguishes a medium, while avoiding essentializing the medium.
Drawing from digital media studies, the three principal qualities of computation that characterize computational information design are pro- cedurality, transcoding, and the network as a medium of storage, access, and exchange. In computational information design, these qualities combine to render data into information in new ways and to produce new forms of expression. The network as a medium of storage, access, and exchange enables the culling and referencing of an enormous diversity of data; transcoding enables those data to be converted into and across a variety of formats; and procedurality enables the authoring of code to perform these conversions and structure representations algorithmically.
The term procedurality refers to the operational characteristic of computation: computation works by executing a set of rules for symbol manipulation. These rules most commonly come in the form of software or, more colloquially, code, which formalizes relationships between symbols and determines how the rules are to be executed, thereby defining the capacities and form of a given computational expression. For many digital media scholars, procedurality is the foundation of computational media. For Janet Murray (1997, 71), procedurality is one of the four essential properties of the digital environment, and she characterizes it as "the defining ability [of the computer] to execute a series of rules." For Ian Bogost (2007), pro- cedurality is central to computation and the basis of a new form of rhetoric performed with computational media, which he calls procedural rhetoric. Both Murray and Bogost identify procedurality as the prime factor in the expressive capabilities of computational media. As Bogost (2007, 5) states, "Computation is representation, and procedurality in the computational sense is a means to produce that expression."
Although Lev Manovich does not use the term procedurality, his notion of the "new media object" references procedurality. Manovich (2001, 47) describes the new media object as being "subject to algorithmic manipulation" via programming and then goes on to claim programmability as "the most fundamental quality of new media that has no historical precedent." In effect, Manovich, echoes the centrality of procedurality to computational media. Manovich (2001, 47) further develops the notion of the new media object through a set of general principles: numerical representation, modularity, variability, automation, and transcoding. Of these, transcoding—the capacity of a new media object to be converted from one format to another—is particularly salient to practices and products in computational information design. As an example, consider any number of interactive map products or location-based services that integrate spatial data, photographs, and user-generated content.6 These products and services are made possible by the shared structure of the code and the subsequent capacity to weave together digital content, dynamically integrating strings of text with the vector graphics of maps along with animated bitmaps of graphs and images. Such transcoding depends on and expresses Manovich's other four principles. Because of numerical representation and modularity, transcoding can occur, and through transcoding the variability of computational objects is expressed, increasingly in an automated or semiautomated manner.
Finally, the practice of information design today is, in large part, fashioned by the growth of the network as a medium of storage, access, and exchange. The Internet as both a repository of digital data and a medium for the transmission of data has significantly influenced the practice of computational information design by providing access to a massive amount and diversity of data. Photo-sharing services, text chats, and social network sites are all used as sources of data. Reports on stock market activity, environmental conditions, and the weather are also available in real time or near real time. Many databases produced by government agencies and nongovernmental organizations are readily available, covering the broadest of perspectives, from the Central Intelligence Agency's World Factbook7 database of countries and territories to the Greenpeace Blacklist8 database, which registers fishing vessels and companies engaged in illegal, unregulated, and unreported fishing. In addition, commercial databases are readily obtainable for purchase, with cost, not content, being the constraining factor in what kinds of data can be obtained as source material. As Alex Galloway (2007, 566) notes, the use of the Internet as "one giant database, an input stream that may be spidered, scanned, and parsed" constitutes a distinctive methodology that is notable for its awareness of a principal quality of computational media—"the fundamental mutability of data." The creative use of the network as a medium of storage, access, and exchange is thus characteristic of computational information design and suggests yet another facet of what it means to do design with computation.
Taken together, procedurality, transcoding, and the network as a medium of storage, access, and exchange are primary qualities of computational media. To do computational information design requires an understanding of and agility with these qualities. Regardless of whether one uses Ben Fry's seven stages or another model of information design, these qualities of procedurality, transcoding, and the network as a medium of storage, access, and exchange have affected information design by providing new means of acquiring, organizing, transforming, and presenting data. For the most part, the uses and purposes of information design remain the same—providing structure and form to data to impart greater comprehension in communications across formats as different as news media and scientific publications. Some examples of information design, however, use these qualities of computation for political ends, demonstrating the possibilities of an agonistic information design.