Interesting linguistics post, all about the r/s thing.

[conuly]

Now with 25% more bathroom graffiti!

Please thank squittycat for this.

What's really cool is that something vaguely similar happened in Latin (and I was just discussing this with my mother, too!) See, in Latin, the third declinsion kindasorta has the ending -s for the singular nominative case. So rex, regis is really regs, regis (with the stem being reg) except that voiced stops assimilate to unvoiced stops before s (so urbs is pronounced urps). This was true for the word honos, honosis, which meant honor. However, there was a language-wide change where s between vowels became r, so for a long time we had the irregular honos, honoris. And then the noun regularized into honor, honoris, giving us the word honor that we have today. It's still kinda irregular, but it's the third declension, nobody cares.

Comments

[conuly]

I'm replying to both of these together. The h/ng thing was a joke, because one only occurs initially and the other NEVER occurs initially, so you can say that they're allophones of the same phoneme. If you're crazy.

There's no way to ever find out for sure how our current sounds were pronounced in dead languages (unless the speakers of those languages kindly left us linguistic notes - that's how we reconstructed the pronunciation of classical latin) but we can make some educated guesses based on their descendents. By educated I mean "slightly less wild than rolling a die".

I'm interested in linguistics for a few different reasons. 1. I got interested in phonology when I had to have speech lessons as a kid. 2. I got interested in dialects when most of the books I read were set in Britian. 3. I got interested in conlanging when I found out about esperanto. Mix 'em and I have a profound interest in linguistics.