Plain text social science is great. The usual justification for plain text work is that it encourages interoperability, sustainability, and replicability in scholarship. It’s true: it does. And if you’re even a tiny bit committed to the values associated with the ‘Invisible College’ of academia, then this justification is a pretty compelling reason to choose markdown and pandoc over Word wherever possible. We owe it to each other to show and share our work in open and sustainable ways; plain text formats/workflows are better at achieving this than their alternatives. But I also think there’s a more intimate benefit to working in plain text: I spend a shocking amount of time in the terminal because it changes my thinking for the better. I mean this in three senses.
One. The terminal doesn’t make me more productive—I’m suspicious of people who claim that it does—but embracing the mindset associated with the modern unix terminal is emancipatory. The classic ‘unix power tools’ that people tend to be familiar with (grep, awk, sed, cat, vi, etc.) all enact a simple philosophy in which:
- Almost every thing is plain text;
- Almost very small tool is agglutinative (and most power comes from piping the output of one tool into the input of another);
- Basic tools can be used to build more complex/powerful tools, which you then you can (and should) share generously with others.
Spending any serious amount of time with these tools, even as an amateur like me, involves repeatedly exposing yourself to an informal design philosophy and aesthetic that prizes openness, generosity, and functional simplicity.
Two. There are modes of thought that seem, to me, nearly impossible without a REPL and the strange genre of the computational essay. The terminal is an inherently iterative, fluid, and build-centric space. When I’m using it—when, to borrow a phrase from Evelyn Tribble and John Sutton, my “mental activities spread or smear across the boundaries of skull and skin” to include my laptop, its terminal, its tools—I have iterative, fluid, build-centric, ideas. In this, the experience of having what I sometimes call ‘REPL thoughts’ aligns closely with what people in fine arts and design sometimes call practice-led research. We know that there are thoughts that are hard (or impossible) to have outside of a laboratory filled with certain instruments and apparatuses. We know, too, that there are certain thoughts that are native to the material acts of crafting and creation. So, too, with plain text: sitting at my laptop, with the terminal open, I have programmatic, plain text thoughts.
Three. The iterative and cyclical experience of writing/modelling in the terminal has, for me, pretty conclusively destroyed my sense of having a continuous or strictly bounded ‘self’ that is stable across time and space. In a recent discussion of humanities modelling and the ontological turn in ethnographic theory, Willard McCarty gestured towards a methodological project of “recursively involving the machine’s point of view with our own as both develop in interaction with each other” (38). The ontological turn is exciting and productive, and I think Willard is onto something great, here, but I quibble with the “machine’s point of view”. My sense is that, when you cut through the frustratingly obscurantist language of much ‘ontological’ ethnographic work, one of the central messages is that bounded ‘points of view’ are suspect. The thing that it is to eat the apple of contemporary posthumanist and ‘new materialist’ philosophy, and (in my experience) to have ‘REPL thoughts’, is to attend to the fact that the material universe is not human-centric and cannot be carved into ‘passive matter’ and ‘vibrant life’. Cognition blurs across bodies, laptops, and tools. When I spend time in the terminal, I confront the degree to which the apparently-strict boundaries of subjectivity are, in fact, persistent illusions. And I confront the degree to which the material reality of the universe (and cognition, which is and must be, also, stubbornly material) does not seem to care much for any division between ‘mind’ and ‘thing’.