Hey Coders! Do You Speech Computer Goodly?
"In Paris they simply stared when I spoke to them in French; I never did succeed in making those idiots understand their language." - Mark Twain
A research paper1 I read recently established that natural language aptitude was a better predictor than numeracy for who would eventually be a good software coder. This reinforces something I’ve known from experience for a long, long time.
An uncountable number of times I’ve had conversations with incredulous software developers/engineers who are adamant that a computer science degree is the sine qua non for coding software. That I had two non-computer science degrees and could code circles around most of them wasn’t persuasive. So I’d explain that chemistry is actually more difficult than writing instructions for a computer. In chemistry, understanding every single step in the chemical process was vital. If something went awry and you’re lucky, you might end up with a beaker of smelly sludge. If it didn’t go well, the “program” could quite literally blow up in your face. This teaches a level of discipline absent in any computer science degree.
Over the course of my IT career, I’d also encountered a fair number of excellent coders with non-CS degrees - geology, literature, and economics are a few that I remember. One characteristic that set them apart from the archetypal CS degreed geek was their ability to communicate with team mates as well as employees in other areas of the business.
Having gained this insight had a significant impact on how I approached interviewing potential candidates for a coding position. Building on ideas I got from Joel Spolsky, I would ask questions like “How would you design a spice rack for a blind person, a trash can for an outdoor mall, or a hiking backpack for carrying a 1 year old child?” Of course, there is no right answer to these questions. What I wanted to see was how they responded. Did they complain or refuse? Did the question stump them? Did they dive in and give it a good effort? I would also give them real-life scenarios and ask them how they would resolve the presenting problem. For example, “The senior VP for marketing steps into your cube and gives you a set of directions for how you should change the code you’re working on. How do you respond?”
Questions and challenges like this revealed how well a candidate could think, how they approached solving problems, and how they responded to challenges. If they were good at these things, I could teach them whatever coding I needed them to do.
The work by Prat, Madhyastha, Mottarella, et al. (2020) reinforces my position by validating that a coders natural language abilities position them to be successful more so than someone skilled in the ways of math.
“The research described herein is motivated by a conceptual paradigm shift, namely, that learning to use modern programming languages resembles learning a natural language, such as French or Chinese, in adulthood. Specifically, we argue that research on the neurocognitive bases of programming aptitude has largely missed the fact that computer programming languages are designed to resemble the communication structure of the programmer (human languages).”
The paper goes on to lay out the researchers methodology and results based on teaching Python to non-coders. The results are interesting to say the least. Those with natural language capabilities outperformed those with mathematical capabilities. I believe the findings from this and subsequent research will have a significant impact on the future of software development in the age of AI prompts and “vibe coding.”
If you have any questions, need anything clarified, or have something else on your mind, please send a DM or email me directly. Please consider becoming a paid subscriber. Doing so supports my writing efforts and grants me the most precious of all resources - dedicated time for writing. And for that, I am deeply grateful. If you prefer, buy me a cup of coffee if there is an article here and there that you enjoyed or found valuable.
Visit The Remnant’s Way to read more of my writing on non-tech topics.
Image credit: Grok 4
Footnotes
Prat, C.S., Madhyastha, T.M., Mottarella, M.J. et al. Relating Natural Language Aptitude to Individual Differences in Learning Programming Languages. Sci Rep 10, 3817 (2020). https://doi.org/10.1038/s41598-020-60661-8