Zompist Bboard Again

How *did* it happen that Proto-Eastern *Endanor became Axunašin Inbamu? That d -> b change seems unlikely, and (maybe I'm missing something) but I don't see anything on the Proto-Eastern page that would suggest how this occurred.

Not all sound changes are regular. This is an assimilation to the following m.

zompist wrote: ↑Fri Sep 10, 2021 5:24 am Not all sound changes are regular. This is an assimilation to the following m.

I think a more accurate way of framing it is that all changes are regular within a certain speech community. However, certain changes can happen within separate speech communities, then one becomes dominant but what amounts to loans from the secondary community get integrated into the majority's speech. So there may have been a small enclave that had a regular d > b / _Vm but only a handful of those changed words become mainstream.

In Icelandic for example, there was a regular change ö > e / [+palatal]_ (derounding) that seems to have occurred in the southeast sometime in the mid-nineteenth century. This change never became mainstream and was quashed by the school system within a few decades, yet a few words remain in ö/e couplets, sometimes with slightly different meaning or connotation, where ö is replaced with e: smjör/smér (butter), kjöt/ket (meat), gjöra/gera (do). The latter one is actually an example where the derounded version became more dominant, while smjör and kjöt remain more common for the other two. But there are written records of people which show that they had completely lost the phoneme /œ/ after palatals.

vegfarandi wrote: ↑Fri Sep 10, 2021 12:37 pm

zompist wrote: ↑Fri Sep 10, 2021 5:24 am Not all sound changes are regular. This is an assimilation to the following m.

I think a more accurate way of framing it is that all changes are regular within a certain speech community.

I may be out of date, but everyone I've read-- Lass, Trask, Hock, Labov-- concludes that sound change is not perfectly regular. Labov, I believe, presents data on sound changes being gradual, hitting some words early. Often the process completes, sometimes it just doesn't. And a few changes, notably dissimilation, are very likely to be irregular.

I agree, of course, that a particular word may be borrowed from a dialect where a particular sound change worked differently than the standard.

zompist wrote: ↑Fri Sep 10, 2021 11:53 pm

vegfarandi wrote: ↑Fri Sep 10, 2021 12:37 pm

zompist wrote: ↑Fri Sep 10, 2021 5:24 am Not all sound changes are regular. This is an assimilation to the following m.

I think a more accurate way of framing it is that all changes are regular within a certain speech community.

I may be out of date, but everyone I've read-- Lass, Trask, Hock, Labov-- concludes that sound change is not perfectly regular. Labov, I believe, presents data on sound changes being gradual, hitting some words early. Often the process completes, sometimes it just doesn't.

"Sound change is phonologically distinct, but lexically diffuse." I remember the quote but not the citation - sure it's Labov though.

And as Zomp is saying, diffusion processes intersect and interrupt. I think Labov also tries to examine sociolinguistic factors that explain how and why certain lexemes have resisted certain sound changes, and how certain other sound changes have died early on their diffusion journey.

My memory of this is a bit hazy, and possibly outdated now, but I used to have a handy bank of examples to whip out in exams of (a) lexemes which argue for rare sound changes which 'died' before diffusing far either through a lexicon or through a speech community (Veg's examples), and (b) lexemes which are still widely resistant to sound changes which are clearly still active. (Edit: Staff/gaff(e) may be a serviceable example for RP/southern British English: you'll always hear /stɑːf/ but never /ɡɑːf/.)

I don't know about whether certain changes "are likely to be irregular" - irregularity seems like a fuzzy concept here. What is the type of 'regular' this is opposed to? Does it mean regular across all speakers of a language, or in a particular speech community - but not necessarily all the way through a lexicon? Or throughout the entire lexicon of an individual, or throughout the entire lexicon of a speech community, or the entire lexicon of all speakers of a language...? I feel like the diffusion model is clearer. Perhaps some types of sound changes are less 'sticky' or something - less likely to diffuse easily.

The other thing is that some classes of words just have different (or no defineable) rules - place names, for instance. These really are "irregular", I guess.

The assimilation of /f/s in four and five is another example.

God names are definitely a special category. My mum pronounces "Yahweh" [ˈjɒx.we͜ɪ] or something. This is about the only [x] I can think of in her speech - she pronounces 'Loch' and 'Bach' [lɒk] [bɑːk].

The assimilation of d > b to match a following m, in the context of a god name, sounds exactly like one of these true "irregularities".

zompist wrote: ↑Fri Sep 10, 2021 11:53 pmI may be out of date, but everyone I've read-- Lass, Trask, Hock, Labov-- concludes that sound change is not perfectly regular. Labov, I believe, presents data on sound changes being gradual, hitting some words early. Often the process completes, sometimes it just doesn't. And a few changes, notably dissimilation, are very likely to be irregular.

Interesting! And makes sense.

sasasha wrote: ↑Sat Sep 11, 2021 2:18 amThe assimilation of /f/s in four and five is another example.

This is a really good example actually. But you could think of that as a sound change that isn't so much phonologically conditioned as pragmatically conditioned.