ჱ Georgian წ

[Io]

Anyone with good resources on its phonology or sufficient practical experience? I've just returned from a trip to the country and having my own observations after listening to Georgian more than I ever wanted to I'd like to have them confirmed or rebuked.

— [v] doesn't seem to exist at all, it's either [w], [β] or merely labialization. So, I wonder where does the convention [v] for ვ come from?! In «The Consonant Phonotactics of Georgian» Butskhrikidze says /v/ can be [v] but not a single time did I hear that sound from any speaker.

— what exactly is ყ really? There is something, uhm, fricative (?) about it, it's certainly not a plain [q'].

— is ც aspirated at all? Now, I have to admit I'm pretty much deaf to aspiration and most of the time I have to pay attention to hear it, but in the case with ც I can't really detect it. Incidentally, the fount of all wisdom says «Opinions also differ on the aspiration of /t͡sʰ, t͡ʃʰ/», just what exactly is meant here, whether they're aspirated at all or something else?!

— ღ surely isn't [ɣ] from what I heard, the above source says about it «[...]Aronson (1990) classifies them as post-velar[...]» and I tend to agree. Here is a sound sample for ღომი (a type of dish) http://1drv.ms/1TU32o0, tell me what you think.

— does Geo have creaky voice? I sure did notice something with the vowels after ejectives, at first I thought it sounds like pharyngealization but I'm not so sure if, say, პ can actually be realised as [pˁʼ] in Georgian, so creaky voice maybe?!

[ná'oolkiłí]

Rad that you just were in Georgia! I'm actually going to Tbilisi in a few weeks myself. Do you speak the language?

• ვ /v/ is a tricky one, both phonetically and phonologically. I think you do occasionally come across [v], especially word-initially before a vowel or voiced consonant. [f] is another allophone, found before and sometimes after voiceless consonants. But I agree that more often than not it's [w ~ β]. There seems to be some degree of dialectal/inter-speaker variation about this, though; I understand that some people have [v] more often, and some practically never have it.

Phonologically/phonotactically, ვ has properties of both sonorants and obstruents, and that's probably related to why it's so hard to pin down.

Also: in old Georgian texts you'll see both ვ and უ̂ (usually transliterated w), suggesting at one point there were distinct /v/ and /w/ phonemes. I'm not an expert in the diachrony of the language, so don't quote me.

• Yeah, in my experience never hear ყ as a genuine [q']. I'd transcribe it as [χ' ~ qχ'], and intervocalically it can be reduced to something like [ʔʶ], especially in faster speech.

• Phonologically, ც /tsʰ/ and ჩ /tʃʰ/ are definitely aspirated; you can tell by how they pattern in consonant clusters. Phonetically, the fact that they have a fricative release obscures the delayed voice onset, so the [ʰ] part of it is not nearly as salient as it would be for the other aspirated stops. I think this is a common property of aspirated affricates in general, not just a Georgian-specific thing.

• I agree; ღ is closer to [ʁ] than [ɣ]. Likewise for ხ, which I'd transcribe [χ] rather than [x].

• I've never noticed pharyngealization or creaky voice myself, but certainly the latter is crosslinguistically pretty common on vowels around ejectives. I wouldn't be surprised if creak pops up to enhance the contrast between ejectives and the other laryngeal categories, especially since in Georgian ejection on stops is typically not produced very forcefully (ejectives can often just sound like unaspirated stops).

[vokzhen]

Definitely seen /q'/ transcribed as [χ'].

It's very common for "velar fricatives" to be a term of convenience, and I've seen Georgian transcribed as post-velar or varying between velar and uvular. Not too surprising given the presence of /q'/ without other uvulars, that uvular-glottal seems be acoustically farther apart than velar-glottal, and that /x/ comes from both *x and *q.

Aspirated affricates are often difficult to tell perceptually because the friction of the release obscures the lengthened VOT of the vowel, and in some cases "aspirated affricates" have a longer-than-normal fricative release rather than actual aspiration. My /tS/ in English, for example, is aspirated initially, but I wouldn't be able to tell except comparing it to medially after stress where it's clearly not.

I've definitely heard of creakiness on following vowels being one of the markers of ejectives in Georgian.

[CatDoom]

With regard to creaky voice, some authors (particularly those studying the languages of the Americas) have noted that ejectives come in more than one flavor. "Strong" ejectives are like the ones you hear in the Wikipedia IPA pronunciation sound bites, and are articulated with a distinctive "popping" sound, followed by a clear vowel. "Weak" ejectives, on the other hand, may not involve complete glottal closure, and are closer to plain voiceless consonants but impart creaky voice on the following vowel. I would guess that the Georgian ejectives are closer to the second type.

Americanist Phonetic Notation actually used to use separate symbols for the two types of ejective, but no natural language is known to contrast the two, so the convention was abandoned.

[vokzhen]

Americanist Phonetic Notation actually used to use separate symbols for the two types of ejective, but no natural language is known to contrast the two, so the convention was abandoned.

Actually, while not exactly the same, it sounds extremely similar to what was said for Taa's "voiced ejective" in another thread: /t' d'/ contrast with the latter having a voiced onset before becoming voiceless, with a light ejective release and heavy laryngealization on the vowel.
EDIT: I got it slightly wrong, I'll just quote:

Wait, Taa has voiced ejectives? I thought those were impossible.

The voiced aspirate and ejectives are said to be mixed-voice, at least phonetically -- they start out as a normal voiced consonant, then devoice in the middle before the release.

Thus: [dtʰ, dtsʼ].

It's a bit more complicated than that. Voiced ejective stops in !Xoon have a voiced closure, but it is different from a normal voiced closure-- the voicing is weaker and less regular (the article I have just calls it "glottal pulses of decreasing amplitude"). The release is also different from that of a regular ejective: there is no audible posterior release of the glottal closure, and the vowel onset is strongly laryngealized, i.e. the glottal closure slowly releases instead of there being a burst release. There's also a less prominent pause between the release of the consonant and the onset of the vowel.

[Io]

ná'oolkiłí, sadly no, I don't speak it. As for [v], its absence was the first thing I noticed, as I was staying in ვარკეთილი; the name ვაჟა is also with a [β], so not even word-initially before a vowel I could hear [v]. Indeed, in «The Consonant Phonotactics of Georgian» page 95, it says According to some studies (e.g. Gamkrelidze & Machavariani 1965), there was a distinction between syllabic /u/ and non-syllabic /u̯/ in Proto-Kartvelian. The merger of the non-syllabic /u̯/ with the sonorant /v/ happened in a later period of the development of Georgian. The Modern Georgian sonorant /v/ corresponds to Old Georgian non-syllabic /u̯/.

>[qχʼ]

Yeah, I think that's about the best description of it. Now, [ʔʶ] would be really interesting to hear in Geo, do you have any examples or hints in what environments it can occur?

I think the stops are fairly well 'ejectified', especially [kʼ], on the other hand I found the ejectiveness of [tʃʼ] quite week.

>Likewise for ხ, which I'd transcribe [χ] rather than [x].

I think that goes without saying, I've no idea why it's generally transcribed as [x].

P.S. Beware of the traffic in Tbilisi or anywhere else you go in Georgia, if it's the first time you're going.