Organic shift (?)

Remember the time when a linguist would be asked by a non-linguist what they did for a living? Remember that the reluctant response, “I’m a linguist,” was (extremely) often followed by: “Ah, so how many languages do you speak?”? Well, dear all, the moment when this common misconception is finally dispelled is here, only to be succeeded by another misconception: “Ah, so then you must work with LLMs?” Of course, we do! Everyone does! It is like asking a researcher in any field if they use Google. But, in any case, there are different ways to read this. First of all, we are talking about a historical shift, where computational methods are no longer limited to computational linguists and NLP engineers. That being said, solely the application of LLM suggestions do not make one a domain expert. Yet, LLMs still allow at least a small degree of domain expansion. Second, it highlights the inseparable relationship between language and modern technology. Who has the most intimate relationship with language if not a linguist? Technology today is about language: it feeds on language and produces language. Therefore, the role of the linguist becomes that of someone who understands, uses, and helps control these systems, as well as someone who can anticipate and mitigate their consequences. And this does not mean that the misconception is correct. LLMs are usually the focus of computational linguists and NLP engineers (in terms of language), but thinking about the effects LLMs have on human language (because they do have them) is a concern that spans all branches of linguistics, from neuro- and psycholinguistics to pragmatics and syntax. So the verdict is that there is some truth in this misconception. If someone asks you how you can’t possibly be a linguist without speaking many languages, you can be annoyed. If they ask you whether you work with LLMs, indirectly, you probably do so.