To be entirely honest, I don’t have much of a position about whether students of the humanities should learn how to code. I think they, like any other student or individual, should learn to code if that knowledge would help them achieve their own scholarly, professional or personal goals or if coding is of particular interest to them. I don’t think it is–or should be–required for anyone interested in engaging with central questions around new and emerging technologies to learn coding first (if at all), as there are many possible paths and access points to those conversations. Essentially, I agree with Donahue’s point, but just as he says about Kirschenbaum’s essay, I feel “the manner in which he says it may be misleading if it is taken at its face value.”

Donahue rightfully asserts that Kirschenbaum’s essay bizarrely minimizes linguistic disciplinary knowledge of the concepts of models and world building–both are essential aspects of modern syntactic and semantic theory, and Kirschenbaum’s uncertainty whether linguists would even find the concepts “familiar” betrays a limited perspective on intersections between the humanities and the computer sciences. But I feel like Donahue himself oversells it a bit when he asserts that Natural Language Processing is one of the “concerns to which computer sciences and humanities have equal access.” Because in reality, Natural Language Processing is functionally more a sub-discipline of computer science than it is an interdisciplinary pursuit attended to by linguists and computer scientists alike. There is plenty of gate keeping in that domain that functionally barrs many students from entering that field, and others like it, even if they are willing to learn how to code.

Take, for example, the only class on Natural Language Processing offered at my alma mater, Michigan State University, where I studied linguistics. The class is an 800 graduate-level course in the Computer Science department. The only prerequisites required are programming, probability and statistic skills and the only content covered in the course is a crash review of several linguistic domains and plenty of applied practice. So if you’re college-Michelle, an undergraduate student of humanities without existing knowledge of coding, you certainly don’t have “equal access” to the domain of Natural Language Processing, and neither does anyone else who has interest but not experience. I suppose maybe Donahue didn’t mean for that assertion to be considered literally, but as someone impacted by this conversation, I find myself more concerned with the reality faced by humanities students in this situation than the scholarly conceptualization of the topic.

One of the things that strike me most about this conversation is the way in which both authors consider the question of whether or not humanities students should learn to code devoid of any consideration of whether or not that option is accessible to students. I find it telling that the onus of responsibility for learning to code here is placed on the students, but there is no equal responsibility for professors or programs to teach that skill in accessible or engaging way. Therefore the authors must be assuming that students can easily get that knowledge elsewhere, but for students with historically marginalized identities, neither STEM classrooms nor online platforms tend to be welcoming learning environments. The implicit bias in STEM, described here in an article from Carleton’s own SERC, and the internet’s both overt hostility and casual marginalizations (see: an HTMLDog tutorial unnecessarily including a term often used as a slur against intersex people in their code example) are very real concerns for me when encouraging students to learn to code without also offering support or at least a warning of what that pursuit might entail. But I guess blogs, like Kirschenbaum says about worlds, “embody their authors’ biases, blind spots, ideologies, prejudices, and opinions.”

As I’ve mentioned in other posts, my own experience with coding is pretty limited, taking place mostly in the early 2000s on a handful of now largely defunct early and proto-social media sites. This coding was less of ‘coding’ in the traditional sense, and more how we talk about it in this class: learning how to ‘hack’ something, how to take it apart and change it and put it back together in a different way to suit my own needs and ideas. For example, one thing I remember searching for a code for was a comment box, so that there was a more accessible way for friends to compliment my MySpace layout.

The world where I began learning to code was the 1.0 Web world. As Kate Wagner describes it in this piece for The Baffler, “Web 1.0 required users to manually code, design, and manage their own spaces on the net and interact with others on the web in ways such as clicking links, copying an email address, or manually linking to other webpages.” This project-based model of coding that centered my own interests and preferences was a compelling way to learn, but that option rapidly faded as we transitioned to Web 2.0 and the ability to manipulate the inner workings of websites took a backseat to things like branding and user experience. The loss of this personal connection to coding made it difficult to sustain interest in learning, especially as a teen girl who was not interested in other common access points like videogames and was still too young to take classes or locate other resources to teach myself. And by the time those things became options to me, I had already been deterred by bias from what I saw as a uniquely STEM pursuit anyways.

I remember the moment that I decided I wasn’t going to become a linguist–it was the same moment that I decided I wasn’t going to learn how to code. I was sitting in the sociolinguistics lab at MSU, watching my supervisor and PI give a tutorial on a computer program we would need for our upcoming project on the Northern Cities Shift in the Lansing area. She explained that when she was in school she had to do much of this analysis work manually, as this program, which she kept referring to as a particular linguist’s program, hadn’t been developed yet. Curious about the program’s association with that linguist, I asked my supervisor why it was named after him–did he use it a lot, or maybe was it popularized through his research? Nope, it was named after him because it was his–he made it, created it, coded it. “Lots of linguists learn how to code” she casually explained, unknowingly shattering my conceptualization of my possible future as a linguist. While I was wholly engaged with what I studied and deeply interested in questions that coding could help answer, I didn’t see myself as the kind of person that would succeed in the environment of an MSU computer science program–my sister had actually just transferred out of her own computer science program, and I was unwilling to subject myself to similar experiences. So if successful linguists knew how to code, I was not interested in being one anymore.

Looking back on that decision now, I can’t say I still think it’s the right one, but I also can’t say I would have been able to find a better solution to that problem in 2013. Even though I was encouraged to learn how to code, that recommendation, in absence of a supportive environment in which to learn, was effectively an impostor syndrome beacon, making me wonder whether I was even cut out for the humanities. Nowadays there are more initiatives and programs working to fight the factors that deter people with marginalized identities from engaging with STEM disciplines in academia and on the internet at large, but back in 2010 when those blog posts were being written, and I was switching from MySpace to Facebook, there were fewer people thinking and writing on that topic.

“But what about linguistics and gender???” might seem like an unfair critique to have of what are admittedly two nearly decade old blog posts by men studying neither topic, but I do not ask that question in order to disparage, but to complicate. Those asking questions about what students should or should not be doing should also be asking questions about what opportunities students have to learn, what supportive environments they have access to, and other factors that need to be incorporated into their concept or model of the world for the conversation around this question to truly reflect the realities students face when deciding whether or not they should learn to code. Not just about the scholarly concerns.

But that’s just one linguist who doesn’t know how to code’s take on it.