Language Choice in CS vs. Linguistics: Is One Overplayed and the Other Underplayed?
A recent Hacker News discussion, prompted by the question "Ask HN: Does CS underplay language choice while linguistics overplays it?", delved into the perceived attitudes of these two fields towards the importance of specific languages. The original poster suggested that computer scientists often dismiss language choice due to the Turing completeness of most programming languages, while linguists, conversely, emphasize the unique impact of different natural languages. However, the ensuing commentary largely challenged these initial assertions from multiple angles.
The Computer Science Angle: Beyond Turing Completeness
The idea that computer scientists view language choice as unimportant was directly contested. One commenter, ottaborra, pointed out that while most programming languages might be Turing complete (meaning they can theoretically compute anything computable), this doesn't negate the practical importance of choosing the right language. They argued:
Efficiency and the real world will eat them (comp scientists who think like this) up for breakfast
This highlights that factors such as performance, development speed, available libraries, ecosystem support, maintainability, and suitability for a specific problem domain are critical in real-world software development. Theoretical completeness offers little guidance when faced with these practical demands.
The Linguistics Angle: The Significance of Language Differences
The assertion that linguistics "overplays" language choice was also questioned. User slwvx expressed skepticism, asking for a reference to support this idea. ottaborra introduced the concept of linguistic relativity (often associated with the Sapir-Whorf hypothesis), suggesting:
Linguistic relativity seems to be a thing. Because of the differences I would think it could be inferred that an ordering amongst languages exists. In the least a partial ordering exists, so maybe some "specific" languages that are better exists
Linguistic relativity posits that the structure of a language can influence the way its speakers perceive and conceptualize the world. If true, this would imply that the choice or nature of a natural language is indeed highly significant, as different languages might offer different conceptual toolkits or predispose speakers to certain modes of thought. This doesn't necessarily mean a simple hierarchy of "better" languages, but it does underscore that differences are meaningful.
Overall Consensus: Language Choice Matters in Both Fields
User pavel_lishin succinctly stated, "I don't think either of those generalizations are accurate," a sentiment that resonated through the other comments. The discussion, though brief, leaned towards the conclusion that language choice is significant in both computer science and linguistics, albeit for different reasons and in different contexts.
It's also worth noting a meta-commentary thread initiated by Jtsummers, who questioned the premise of the original post based on the submitter's history. This aside, the substantive part of the discussion effectively deconstructed the OP's initial dichotomy.
In essence, the dialogue suggested that a nuanced understanding is required: in CS, practical utility and efficiency drive language choice, while in linguistics, the potential impact of language structure on cognition and expression makes specific languages a subject of deep importance.