Saturday, September 2, 2017

Diachronic conlangs

Yeah, I've been really lazy about updating. Still doing plenty of conlanging, as can be seen on reddit, just not updating here. I've been working on a big "papualang" project. But I'm getting on to do a minor, inconsequential rant about conlanging.

One thing that many conlangers like to do is diachronics, so deriving languages from other languages. That's cool and all, especially when done a posteriori (and well) or for a conlang family. Even I'm part of a diachronics project. My problem isn't with diachronics.

No it's the whole thing where you create a proto-language for only one daughter language and even more the idea that this makes the daughter lang inherently better. Why does this bug me? Because it doesn't make the language actually more realistic, especially since many of the sound changes and grammatical changes found in the daughter langs might have been a stretch to occur naturally. Also it leads to this weird idea that proto-languages were more regular than daughter langs, which isn't actually true. The other thing that bugs me about this is that the proto-lang itself is still a conlang. It's not like you made the conlang less constructed. If you have no plan on making a language family, why do the protolang. You're getting all the features you wanted anyway, but now you are taking extra steps. Just make the language without those steps.

It's a minor gripe. I'm not going to discourage people from making proto-langs, not at all. It's mostly harmless. But I do wish more people understood that proto-langs are reconstructions and aren't what people actually spoken. Are they close? In many cases, yes, probably, at least somewhat. Of course there's dialects in real protolangs, something not often reflected in constructions, academic or otherwise (at least as far as I have seen). That's okay, it's difficult to do, but people gotta remember this.

No comments:

Post a Comment