How to draw the line between philology and historical linguistics when it comes to etymology?

Question

When a user comes to Ling.SE with an etymology question about a word or words that can be traced back to a language with a major written tradition like Sanskrit, Greek or Latin (say), it seems that some users tend to close the question as off-topic as they are more philological than they are linguistic, while others tend not to (as they edit the question without voting to close).

Intuitively, I agree that such questions do not seem to be particularly linguistic, as they are more likely to be studied in departments like classics or area studies, rather than linguistics departments. However, I often have doubts as to whether I should vote to close these questions. This is because if someone were to ask for the etymology of a word from an 'exotic' language family or branch such as Pama-Nyungan, Algonquian, Gur or Polynesian, etc., my intuition would be to accept such a question (even though I've never seen such questions here myself). This is because linguists in linguistics departments do study etymology in these languages.

Another possible reason for rejecting etymology questions is that they aren't useful for other visitors to the site. But while syntax trees and language identification clearly fall into this category, etymology seems less clear-cut. It could well be that more than one person wants to know, say, the common origin of the different words for 'night' in Indo-European languages.

And finally, I found this in the FAQs, listed among the acceptable question types:

• a single word in relation to multiple languages or a single word/single language

Does this imply that etymology questions should ordinarily be acceptable?

Answer 1

This is a consequence of the interpretation of the "single language" rule, which exists as a widespread but non-uniform interpretation of something that was supposed to be mentioned as off topic. As a grammarian, I personally have no interest in some of the minutia of historical change especially in semantics, but that doesn't mean these aren't linguistic questions. I also have no interest in the NLP questions that get asked here, which I think would be better suited to an NLP site because they are basically narrow programming questions, but they are allowed here. Some people have a different view. Ultimately, a user who can vote to close can vote to close purely out of spite, which is why it takes 5 votes (or one moderator) to close a question. It would not be inappropriate, IMO, to solve the "single language" problem, but there isn't much interest in that.

Answer 2

My perspective is that linguistics is about the study of systems in language. The history of individual words is a fluke of, well, history. Words are borrowed across languages constantly with little rhyme or reason.

If you can connect a word to a system of borrowings or a system of meaning changes, then that is the foundation for a good question. But if you're just after the history of a word, well that's what a good dictionary is for. If you can't demonstrate in your question that there is a reasonable assumption of systematicity, then I think that makes your question off-topic.

So questions about inflectional or derivational morphology will almost always be on-topic. It is the questions about lexical/content words which need to make their case. Show that the word is part of a group or family of words which have changed together.