- In “Digital Media Archaeology: Interpreting Computational Processes,” Noah Wardrip-Fruin argues that the processes behind digital works, rather than merely the outputs, are paramount to our understanding. Outputs are the surface level manifestations of complex data and processes, as seen in Christopher Strachey’s 1952 love letter generator. When we analyze these processes, especially in their historical and cultural contexts, we are better able to see the greater system at work. Regarding the love letter generator, we can see the project as a carefully constructed parody of the normative expressions of desire found in letter-writing rather than simply as a humorous arrangement of words.
- One of Wardrip-Fruin’s first main points is that “the resulting letters are not really the interesting part of the project” (306). If we only conduct a surface-level analysis of the love letter generator, then all we will see are the letters themselves. While they are humorous, there is not a lot of interpretation to be gleamed from them alone. Wardrip-Fruin writes, “The views of the generator that include its data and processes, as well as its output, are views that consider the work as a system” (307). An analysis of the generator as a system from a literary perspective leads to a greater interpretation.
Wardrip-Fruin proceeds to examine how data and process are used in the generator. The data consists of lists of words, sentence structures, and various rules pertaining to form. In this case, “the vocabulary is largely based on Roget’s Thesaurus,” meaning that there does not appear to be any pattern (309). He asks, “For what sort of processes would one choose to copy the data from a thesaurus, rather than carefully select each element?” (309). He continues, “In process-intensive work, data is interesting primarily when considered for how it will be employed in processes” (309). After comparing the love letter generator to two other works, it becomes apparent that the processes used are not as complex as they could be and that the resulting structure suffers. If the processes were more complex, then the love letters would be better.
In light of the data and processes and knowing that this project was carefully built, Wardrip-Fruin asks, “How should we consider the love letter generator’s deliberate simplicity, its statelessness and randomness, and the fact that its vocabulary is a transcription from a thesaurus?” (314-5). He writes, “To put it another way, the love letter generator—in the way it operated—was a blunt parody of normative expressions of desire. It played the role of the lover as an inept spouter of barely meaningful, barely literate sentences, composed with repetitive randomness while one finger still rested in the thesaurus” (315). Strachey’s conscious decisions in the construction of the generator thus serve as a critique of “conservative elements in English culture” (318). Especially in light of his and Turing’s homosexuality, the love letter generator can be seen as a parody of the oppressive heterosexual norms prevalent at the time. By seemingly not doing well what it is supposed to do, the generator does exactly what it means to do.
- In this situation, I understand the importance of seeing the project as a system; the author calls this analysis a “literary” perspective. However, in many classes professors have discouraged students from thinking too much about the historical, social, and cultural contexts or the writing process, proclaiming that in literature all we need to analyze is the text itself and nothing more. How am I supposed to rectify the argument made in this essay with my experiences in the classroom?
What are some more reasons that critics have largely failed to analyze digital artifacts from a literary perspective? Wardrip-Fruin states that the reason is because the digital does not feel human. How are we to explain this disconnect?
- Strachey seemed to be meticulous in documenting his development of the love letter generator, which certainly made the examination of the data and processes much easier. What about digital artifacts of which we have no knowledge of the development and do not have access to the processes behind the outputs? Is knowledge of these processes necessary in order to interpret the artifact as a system, or are we left to analyze the outputs only?