#another fun index diachronica find

[velaraffricate]

ok

[dedalvs]

Hi, Mr. Peterson! So, I'm currently creating a conlang and I have a question about evolving lexicon. For example, in my conlang the proto word for 'house' is /*falħʊ/ & 'houses' is /*falħʊm/. Could I evolve /*falħʊ/ to [falhu] and /*falħʊm/ to [falhom]?

Yes, easily. You could also not do that, or do something different.

If I can, though, this is not a root question, but a branch question (hoping that metaphor makes sense). For example, I could have just responded “yes” and been done with it. And that might have been all the help you needed, if there’s literally nothing else you’re going to evolve that you have questions about. But if there is, I mean, theoretically, you could follow this up with an almost identical ask asking about /*talħʊ/ “mouse” and /*talħʊm/ “mice” and then /*talsʊ/ “louse” and /*talsʊm/ “lice”, etc. Let me try to head this off.

If you have questions about the kinds of evolutionary paths sounds can take, there’s a great resource here. It’s a four page .pdf by William Annis that charts, visually, common changes between vowels and consonants in which environments. He’s synthesized a lot of material and consolidated it into four simple charts. It’s awesome. It is not a representation of every sound change ever, of course, but it gives you a good sense of the major evolutionary trends. If you want more than that, there’s the Index Diachronica, which is a sincere attempt to catalogue every single sound change. It’s not complete (and never will be, because there’s so much we’ll never know about the languages of the past), but it’s as thorough as it could be. That’s definitely the place to go to see what can happen to what sound in what environment.

Having said that, it’s not as if you simply need to think of a sound change and go to any authority, be it a list, a chart, or a person like me, and see if it approves. First, if you peruse the Index Diachronica, you’ll find many sound changes where the complete opposite thing happens in the same context. One may be more likely than another, but it’ll happen. That doesn’t mean anything can happen anywhere (necessarily): it just means there’s always a story. If you’re creating a naturalistic conlang, you’re crafting that story, and it needs to hold together the way any story does. It’s not a matter of grabbing random sound changes and throwing them at some data (though that too can be fun).

For example, with the sound change you listed, sure, that can happen, but is it commensurate with other sound changes in your language? What’s happening to the ends of words in general? Or vowels in general? Or before nasals in general, or before codas in general? What’s happening to your language over the course of time—and your speakers? What’s causing it to change, and how?

Also, even changes that look totally bizarre are explicable with either more steps, via reference to other sound changes, or via an influx of non-native speakers for whom the change makes sense, because of their primary language. It’s not really the job of a conlanger to ask, “Does this work?” It’s the job of the conlanger to say, “Here’s what happened, and here’s why.”