zompist bboard

Let me begin like that:
German had an /f/ allophone [v] for a moment which then changed back to [f], but the spelling remained in some words like Vater. The Dutch language voiced it's word initial fricatives, but nowadays they're both back voiceless for most of the speakers in the Netherlands, the Belgians seem to keep the distinction e.g. between [s] and [z] (jal told me the distinction in the north is more fortis vs. lenis). There's also the mostly fortis vs. lenis distinction between Germanic stops. Is that only a coincidence or is does it come even from Proto-Germanic?

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:Let me begin like that:
German had an /f/ allophone [v] for a moment which then changed back to [f], but the spelling remained in some words like Vater. The Dutch language voiced it's word initial fricatives, but nowadays they're both back voiceless for most of the speakers in the Netherlands, the Belgians seem to keep the distinction e.g. between [s] and [z] (jal told me the distinction in the north is more fortis vs. lenis). There's also the mostly fortis vs. lenis distinction between Germanic stops. Is that only a coincidence or is does it come even from Proto-Germanic?

...is what only a coincidence?

[regarding voicing of stops: most English dialects voice intervocalic fricatives, and some (in the West Country) voiced word-initial fricatives as well ('Zommerzet Zider', for normal 'sommerset cider'; these dialects are where English gets words like 'vixen']

I think he's asking whether the fact that all this devoicing happened is a coincidence. I don't see why it would be, though. Aren't all of these cross-linguistically common sound changes?

Vijay wrote:I think he's asking whether the fact that all this devoicing happened is a coincidence. I don't see why it would be, though. Aren't all of these cross-linguistically common sound changes?

I'm not familar with unconditional word-initial voicing of fricatives happening in non-Germanic languages, although I suppose just the presence of voiced fricative phonemes like /z/ /v/ is somewhat more common in Europe than elsewhere.

(It wouldn't surprise me that much to learn that some Italian dialects had something similar, although that kind of sound change would be less surprising in Italian since so many Italian words are vowel-final, so voicing a word-initial /s/ to [z] could be seen in many contexts as just intervocalic voicing. That's just hypothetical--I don't actually know of any Italian dialects that voice word-initial /s/.)

I mean the devoicing, though.

Vijay wrote:I think he's asking whether the fact that all this devoicing happened is a coincidence. I don't see why it would be, though. Aren't all of these cross-linguistically common sound changes?

Yes, you got it, I didn't expect my message to be that messy. I just spotted there is one specific group of languages in Europe prone to obstruent devoicing and this is the Germanic branch of Indo-European family. I'm just wondering if anyone thought of assuming Ger voiced stops were just lenis in the first place.

That wouldn't seem to make all that much sense given Grimm's Law.

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:

Vijay wrote:I think he's asking whether the fact that all this devoicing happened is a coincidence. I don't see why it would be, though. Aren't all of these cross-linguistically common sound changes?

Yes, you got it, I didn't expect my message to be that messy. I just spotted there is one specific group of languages in Europe prone to obstruent devoicing and this is the Germanic branch of Indo-European family. I'm just wondering if anyone thought of assuming Ger voiced stops were just lenis in the first place.

I don't see how anyone could tell, retrospectively, whether something was truly voiced or merely voiceless lenis, hundreds or thousands of years ago. It can be controversial even with living languages. Although if you're suggesting a Proto-Germanic with voiceless stops, voiceless fortis fricatives and voiceless lenis fricatives (the "voiced stops"), that would seem typologically uncommon.


One approach might be from the opposite direction. Kortlandt thinks that the 'voiceless stops' in PGmc were in some way still glottalised. This might explain why weird stuff happens to them (preaspiration in Icelandic, affrication in High German, glottal reinforcement and glottal replacement in English and elsewhere). And if there were something 'odd' (i.e. distinctive, other than voicelessness) about the voiceless stops, that might help encourage voiced stops to devoice (because voicing wouldn't be the only distinction, so there's less pressure to maintain it).


Then again, I think you harm your own case, by noting multiple instances. We've got:
- devoicing of stops in Icelandic
- devoicing of stops in some forms of German
- devoicing of fricatives in some forms of German
- devoicing of final stops all over the place
- devoicing of fricatives in Dutch, including ones not originally voiced

That doesn't look it's just the result of one set having been voiceless all along. It looks like multiple devoicings.

Ok, thank you all. I haven't linked all of this to the glottalic theory, which I believe is true, but now it seems reasonable to me that Germanic languages preserved some weird PIE feature instead of having created their own out of nothing similar.

Salmoneus wrote:- devoicing of fricatives in Dutch, including ones not originally voiced

Just want to make everything clear for me, what do you mean by that?

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:Ok, thank you all. I haven't linked all of this to the glottalic theory, which I believe is true, but now it seems reasonable to me that Germanic languages preserved some weird PIE feature instead of having created their own out of nothing similar.

Salmoneus wrote:- devoicing of fricatives in Dutch, including ones not originally voiced

Just want to make everything clear for me, what do you mean by that?

As you say yourself:
- voiceless fricatives were voiced
- voiced fricatives, including the ones that originally were voiced, were devoiced.

This shows that it's not a unique ancient property of the originally 'voiced' fricatives that lead to them devoicing, because the ones that were voiced later were also devoiced. I mean, it doesn't prove anything (maybe the fricatives 'voiced' in a way that gave them whatever quality the originally voiced ones had) but it makes it more improbable.

Well, the only voiced fricative left is /ɣ/, which is gone in the Netherlands, too, but all three /z/, /v/ and /ɣ/ remain distinct from their counterparts /s/, /f/ and /x/ by means of voicing in Belgium. They all also assimilate in voicing with voiceless obstruents from prefixes, e.g. gaan [ɣaːn] afgaan [ˈɑfxaːn], zeggen [ˈzɛɣə] ontzeggen [ɔntˈsɛɣə]. Because of this I just think /v/ and /z/ should be treated analogically to /ɣ/.

AFAIK the only precedents for ejectives spontaneously affricating are in Khoisan, so I'm not sure what's gained by assuming PGmc voiceless stops were glottalized rather than aspirated.

Nortaneous wrote:AFAIK the only precedents for ejectives spontaneously affricating are in Khoisan, so I'm not sure what's gained by assuming PGmc voiceless stops were glottalized rather than aspirated.

Maybe Semitic, assuming tsade was once upon a time /sʼ/ rather than /ʦʼ/. There's no direct evidence for that, but most Semitic languages have /s z/ and /sʼ/ is not an impossible sound, especially given that ḍad was once /ɬʼ/.

I think Tigrinya and some Amharic-speakers have [sʼ].

Yeah but, those are fricatives. The Germanic ejectives and I believe the khoisan ones were all stops(if they existed).

Germanic gamma was apparently retained as a fricative wehere b&d became stops, so it was the sole voiced fricative in some states of Dutch. F>v And s>z may have been therefore compensatory shifts to make all the frics behave alike.

I think Amharic might also have had *t' > [t͡ʃʼ] in some environments, but I'm not sure yet (and I don't know how that worked since it also has /t'/).

EDIT: I guess it was before front vowels? All I've found so far is this, really.