zompist bboard

Nortaneous wrote:03:03 < nort> mundari nasal release only occurs in monosyllables
03:03 < nort> /ub/ [u?bm] but /udub/ [udu?b]
03:24 <@vlad> southeastern tepehuan also has b > ?m word-finally but not just in monosyllables
03:24 <@vlad> and I thought that was weird

Just noticed this for some reason. Kharia, another Munda language, collapses all coda stops (plain, voiced, aspirated, and breathy) to voiced, preglottalized, nasally-released sounds, except the velars which just become ?. Though apparently this is only in careful speech, more often they're unreleased... but still voiced

Vijay wrote:There's this set of sound changes in Dravidian languages that I've been wondering about for a while now involving nasals. I'm not even sure what the exact progression of the change was, but I get the impression that it went something like this:

*nr > ndr > nd (> nd̪?) > n̪d̪ > n̪n̪ > nn

In particular, the part that always looked kind of weird to me was nd > nd̪. I've never really been sure why the [d] would just suddenly become dental

As far as I know, this is usually reconstructed with the alveolar stop *t in Proto-Dravidian, which in Tamil developed into /r/, but may have elsewhere in Dravidian rather merged with the more common *t̪; which probably happened due to *t being kind of a rare phoneme.

/nn/ in dialectal Tamil would have to be, at least genealogically, an independent development from Malayalam (otherwise that variety would of course be a Malayalam dialect and not a Tamil dialect). Nasal+voiced stop > geminate nasal is a fairly simple type of sound change after all.

Well...I don't know how much of this you already know; for all I know, you may know it all already, but in Malayalam - like, in the orthography - we have റ, which is [ra]. Then when you double it (basically), you get റ്റ, which is [tta]. (Or maybe just [ta]. I'm not sure it's always geminate). Also, we have ന്റെ (i.e. ൻ [n] + റെ [re]), which occurs on a lot of possessives (I'll go into more detail on that if you want me to lol ).

In Centhamil (so "literary Tamil," which is misleading at least in a modern context since these days, literature is written in regional dialects, too), they have ற [ra], ற்ற which AFAIK is pronounced [tra], and ன்ற [ndra]. Also AFAIK Centhamil [r] corresponds to Malayalam [r], Centhamil [tr] corresponds to Malayalam (or whatever), and Centhamil [ndr] corresponds to Malayalam [n̪n̪].

Kannada seems to have [n̪d̪] corresponding to Malayalam [n̪n̪] and Chenthamil [ndr] (and [nn] in dialectal Tamil - is that in all dialects of Tamil? I'm pretty sure Madurai Tamil has it).

And yes, I also assume /nn/ in dialectal Tamil is independent from Malayalam. But I'd still think it evolved from an earlier [n̪n̪], which is no longer attested in Tamil, or am I completely wrong about that?

Vijay wrote:And yes, I also assume /nn/ in dialectal Tamil is independent from Malayalam. But I'd still think it evolved from an earlier [n̪n̪], which is no longer attested in Tamil, or am I completely wrong about that?

Taken in isolation, the order could have been just as well either *nd > *n̪d̪ > *n̪n̪ > nn, or straight *nd > *nn. Do the Tamil dialects you have in mind have /nn/ also in place of Centhamil /n̪d̪/ (நத in the native script, I believe) though? If they do, that's surely a point in favor of the first scenario; if they don't, the second.

An alveolarization *n̪n̪ > nn would sound like Indo-Aryan influence actually. I've seen several IA languages (e.g. Bengali, Gujarati, Hindi, Nepali) described as having only /n/ and not /n̪/. Though now that I check, Marathi closest by to the Dravidian-speaking area has switched over to /n̪/, so maybe that's a red herring.

Tropylium wrote:Taken in isolation, the order could have been just as well either *nd > *n̪d̪ > *n̪n̪ > nn, or straight *nd > *nn. Do the Tamil dialects you have in mind have /nn/ also in place of Centhamil /n̪d̪/ (நத in the native script, I believe) though? If they do, that's surely a point in favor of the first scenario; if they don't, the second.

OK, I see your point, thanks! (It's ந்த btw, so like you had it, except with a dot over the first letter to show that there isn't an /a/ following the consonant ).

An alveolarization *n̪n̪ > nn would sound like Indo-Aryan influence actually.

Maybe, but couldn't it just as well be an independent innovation?

/u(:) o(:) a(:)/ > /ʉ(:) ɵ(:) ɛ(:)/ in all open syllables in Khaling (source).

Machvano Vlax Romani is a variety of (Vlax) Romani formerly spoken in Mačva, Serbia but now spoken almost exclusively in the Americas (and especially in Central and South America) IIRC. In Machvano Vlax, apparently, this sound change took place (and the pronunciation of these sounds is a shibboleth of Machvano Vlax; I'm not aware of any other variety of Romani in which this particular sound change took place):

/t͡ʃʰ d͡ʒ/ > [ʈr ɖr]

I wonder how often something like that happens.

Straits Salish has a split in the labial consonants, which occurs without any conditioning as far as anyone can tell:

p pʼ m > p~tʃ pʼ~tʃʼ m~ŋ

Vijay wrote:Machvano Vlax Romani is a variety of (Vlax) Romani formerly spoken in Mačva, Serbia but now spoken almost exclusively in the Americas (and especially in Central and South America) IIRC. In Machvano Vlax, apparently, this sound change took place (and the pronunciation of these sounds is a shibboleth of Machvano Vlax; I'm not aware of any other variety of Romani in which this particular sound change took place):

/t͡ʃʰ d͡ʒ/ > [ʈr ɖr]

I wonder how often something like that happens.

That looks like it might a backwards sound change, like when acrolectal varieties of Réunion Creole use unetymological high rounded vowels for etymological high unrounded vowels e.g. [føv] for fèv (bean).

Because of influence from (Parisian?) French? But then I wonder what language would have influenced this development in this particular variety of Romani.

Because of influence from French, yes. I think the French dialects were still important enough in the early 19th century to continue their earlier influence though, but to lesser extent.

I can't find evidence for a palatalisation of /tr/ in the Balkans yet, but I have found palatalisation of clusters ending in /l/ in Gheg Albanian and Romanian.

There are some sound changes in Indo-Aryan languages that strike me as pretty weird, too, like spontaneous nasalization (e.g. Old Indo-Aryan ákṣi > Pali/Prakrit akkhi/acchi > Hindi ã̄kh). How often do words in various languages come to have nasalization on their vowels that apparently came out of nowhere?

Without more examples I can't be sure of whether this comment is valuable but spontaneous nasalization of low vowels is a kind of common process.

Really? That's neat! Can you give me an example? I'm just curious.

I guess it doesn't only happen to low vowels in Indo-Aryan since it can also happen to (for example, "mung" (as in "mung beans") is a loanword from Hindi mū̃g, but that's descended from Prakrit mugga), but that's valuable enough of an answer as far as I'm concerned anyway.

I think a similar process occurred in French, rien coming from res, where polysyllabic words are lacking the nasal.

jmcd wrote:I think a similar process occurred in French, rien coming from res, where polysyllabic words are lacking the nasal.

AFAIK, it goes back to the old accusative rem (like in most French nouns, which are based on the old accusative). The number of syllables plays a role - final /n/ (from Latin /m/) was normally dropped in Romance except for some monosyllabic words (cf. French mon, ton, son from Latin meum, tuum, suum which must have become monosyllabic at some point on the way to French, or Spanish quien from Latin quem).

Isn't the low vowel obligatorily nasalized in Iau?

One of the defining features of Southern New England Algonquian is the nasalization of proto-Algonquian *ā, so Massachusett squnck 'skunk' (probably [skɐ̃k] or thereabouts) from PA *šekākwa.

In Sani, palatalized labial plosives became alveolar lateral affricates:

*bya2 > dɮa(ma) 'bee'
*byam1 > tɬɪ 'fly'

(from the Handbook of PTB)

Some Levantine Arabic varieties: short *i and *u are neutralised to schwa in stressed syllables but remain separate in unstressed syllables (for examples see Wikipedia EDIT: and also this recording of the Iraqi national anthem).

According to Fortescue's 2011 Comparative Chukotko-Kamchatkan Dictionary, PCK *lː > Chukchi tl.

I don't think that change is particularly odd. Didn't that also happen in Icelandic and Faroese?

Vijay wrote:I don't think that change is particularly odd. Didn't that also happen in Icelandic and Faroese?

I classified it as "odd" because the sequence /tl/ is in general dispreferred among languages.

It's /tɬ/, not /tl/, unless /tl/ was an intermediary for another sound. Chukchi doesn't have any laterals but /ɬ/. In fact the Chukchi grammar in the Grammar Pile gives the rule instead as ɬ > t / _ɬ.

Vijay wrote:There are some sound changes in Indo-Aryan languages that strike me as pretty weird, too, like spontaneous nasalization (e.g. Old Indo-Aryan ákṣi > Pali/Prakrit akkhi/acchi > Hindi ã̄kh). How often do words in various languages come to have nasalization on their vowels that apparently came out of nowhere?

I read a paper years ago on spontaneous nasalization. From the little I can remember, the argument was that it's related to aspiration, and IIRC both aspiration and nasalization blur the lower formants of vowels. I should probably look that paper up...