By Sarah Doyle
At the second Code Acts in Education seminar, only a handful of participants declared themselves able to code. The gathered researchers hailed from fields such as education, sociology, science and technology, and geography. Some of our roundtable discussions focused on whether and how we can find ways of knowing code without (everyone) being able to code, and which approaches and methodologies open up ways of examining code. Data visualization is an interesting example of code at work, and it is one that provides insights into some of the interdisciplinary dimensions of coding.
Part of the rationale for exploring this interdisciplinary angle lies in the recognition of coding as socio-digital practice(s). Instead of thinking about coding as a detached, neutral sequence of technological instructions, coding is understood as inherently imbued with sociality. Using the example of data visualization, we can see that coding entails working on the data to produce something new. The visual artefact emerges through a process of highlighting, leaving out and smoothing over. Visualizing data might be understood as a kind of translating, but like all translations, the work requires decisions about form and content. Who or what does the deciding about what to code in and what to code out?
In data visualization, socio-digital practice(s) are especially apparent: finding the coded algorithm in the software is not the whole story. This is a point made by Jeremy Knox, soon-to-be-completed doctoral researcher, in his Code Acts talk on multimodality and digital literacy. Jeremy used the example of word clouds to remind us that the visual artefact produced is not simply a mechanical movement of text from one format to another. On the contrary, the algorithms work on the text according to pre-selected criteria such as identifying which words are used most frequently. Producing algorithms that encode particular criteria is at least as much a social practice as it is a digital one.
A post-event conversation with Maureen Michael – another soon-to-be-completed doctoral researcher – helped me begin to unravel some of these (en)coded practices of visualizing data. Coming from an art and design background, Maureen’s eye is drawn to the composition of visuals, such as the judicious and deliberate use of line, shape and form. For those who know how, the purposeful selection of colour engages the audience in particular ways. Producing text in different fonts is a loaded process: traditional fonts invoke history and seriousness while newer ones portray simplicity. The positioning of graphics and the ways that captions are used (or not) can lead an audience in a particular direction or privilege a certain message. Web pages and documents disclose traces of these decisions.
These artistic practices encode specific assumptions and values. What happens when they are put to work in digital spaces? In what ways do they tangle with algorithms? Gillian Rose is investigating these and other issues in her work on digital technologies and architectural design. Her interest is in the ways that these technologies are specifying both the form and social use of buildings, as well as the work practices of the architects involved. In educational contexts, how do these socio-digital entanglements change data through visualizing processes? How does this visualization work to influence debates about which knowledge(s) are most worthwhile, which learning is most desirable or which institutions are most highly evaluated?
In her own research, Maureen is using visual methods to investigate art practices and often finds her data being created differently in the various software programmes that she uses. For example, her hundreds of observational photographs are viewed and manipulated in the digital camera, computer directory, Windows Live Photo Gallery, Adobe Photoshop, Word (and SmartArt Graphics), PowerPoint and Photobox. Each one affords different ways of viewing, grouping and manipulating the photographs. Each one has different aesthetics and tools to be used or sometimes subverted. Maureen’s viewing of the data is not just her professional looking. Instead, the looking is actively constructed in each one of the different digital spaces. For Maureen, the worlds of art and design are unavoidably implicated in data visualization.
Code acts: it participates. Code has effects. Coding encapsulates educational and political (and in data visualization, artistic) assumptions, preferences and priorities. In turn, these are entangled with the affordances of digital technologies and data servers as well as the availability and participation of digital literacies. While we might think we deploy code in ways of our choosing, code weaves throughout the fabric of our lives, changing and changed by the collaborations encountered. Code deploys us, too.
Unraveling the specific nature of how code acts in particular settings means attending to the work that code does as well as what code is. It means, for example, identifying and examining the amplifications, deletions and obfuscations effected by code. This is interdisciplinary work. We can find source codes and crack algorithms, and we can bring these into dialogue with critiques of power relations and social inequities. We can ask about the mathematical and computational details of code, and we can also ask about the embedded artistic, sociological and ethical traditions.