zompist bboard

I was trying to think if there are any unrelated languages that nonetheless have identical phonologies. Excluding that, what unrelated languages have the most similar phonologies? Off the top of my head I can think of Latin and Japanese (if you allow for different pronunciations of <r> and <u>), but that still differs in that Latin has two liquids, and allows far more clusters and final consonants. Is there another pairing that's closer?

The phonology of Ainu is way more similar to that of Japanese than is Latin or any Indo-European language. The main differences are that Ainu allows obstruents in final position (how many depends on the language) and lacks a voiced/voiceless contrast in its consonants and a length distinction in its vowels.

If you're disqualifying Ainu because, either due to prolonged contact or genetic relatedness or both, the resemblances can't be considered "coincidental", then looks at Austronesian languages. Plenty of languages in that family with small phoneme inventories and a pronounced preference for open syllables.

Europe clusters fairly well with the Austronesian languages of Southeast Asia in terms of phonotactics, though I wouldnt be surprised if you could go through the entire list of 1200+ languages and not find a single exact match to any European language's phonology, primarily because of the vowel systems.

Spanish to Greek is a fairly close match, though the languages sound quite different to the ear. Since the two languages' phonologies diverge away from each other as you go back in time, this could be called a coincidence, not due to language contact or genetic relationship, though it's far from an exact match. Other than that I think the best hope is to go for simple systems. There might be a language out in the Pacific Ocean that has re-evolved a voicing contrast and can therefore be matched with Japanese. Perhaps if prenasalized stops are considered to be phonemically identical with plain voiced stops it'll be easier. (Old Japanese had prenasalized stops.)

In a broad sense, the Northwest Caucasian Languages wouldn't be too out of place in the Pacific Northwest. While most PNW languages lack voicing distinctions, Coast Tsimshian (for example) has it.

Publipis wrote:Spanish to Greek is a fairly close match, though the languages sound quite different to the ear.

I often mix them up if I only hear a short snippet. Greek to me, sounds like European Spanish but with fewer recognisable words and with a lot of [ç], [ps] and with [is] at the end of words.

I am not completely convinced that German didn't come to Europe from somewhere in the Pacific.

Then again, Austronesian and Papuan languages are the most similar to European ones in general, with the important difference that Papuan languages often prenasalize their voiced plosives and Austronesian doesn't always have a voice distinction.

It's easier to find a set of rare and salient characteristics that all show up in two different places, like bizarre consonant clusters and implosives in Tsou and Khmer or an unusual rhotic and a strange vowel system full of diphthongs in English and Mandarin.

IIRC there is a language spoken in the Chinese or Burmese highlands that has a vowel inventory almost identical to that of American English. I can't remember the name, though - I saw it on the forum here a while ago.

Are you counting phonotactics, or only phoneme inventories?

Porphyrogenitos wrote:IIRC there is a language spoken in the Chinese or Burmese highlands that has a vowel inventory almost identical to that of American English. I can't remember the name, though - I saw it on the forum here a while ago.

I thought there was one that was linked some time ago that was West African, rhotic is an approximant, /e o/ are diphthongized, and a sample of speech sounds basically like nonsense English words. It's possible I'm confusing that for what you're thinking of, though.

Neon Fox wrote:Are you counting phonotactics, or only phoneme inventories?

This is an important question. If you're just looking at a table of phonemes, it probably won't be too difficult. Include phonotactics, allophony, and distribution and it's going to be much harder.

Those two languages are Angami (or maybe Pochuri, but if you thought it was African it's probably Angami, which has /pf bv/ everywhere) and Sorbung.

Sorbung's vowel inventory: /a aː e eː o oː i iː ʉ ʉː u uː/ [ɐ aː ɛ ei ɔ ou ɪ iː ʉ ʉː ʊ uː].

Proto-Dravidian have bunch of PoA, no fricatives, and voicing distinction... just like stereotypical australian languages.

Proto-Bantu is also reconstructed without fricatives too, but it has voicing distinction.

I was surprised to see that German and Hebrew have fairly similar phoneme inventories—apart from some minor things…

Gaspard wrote:I was surprised to see that German and Hebrew have fairly similar phoneme inventories—apart from some minor things…

As far as Jewish language goes Yiddish is the one most similar with German.

Gaspard wrote:I was surprised to see that German and Hebrew have fairly similar phoneme inventories—apart from some minor things…

This is not at all surprising if you consider the history of modern Hebrew. The language was artificially revived and initially spoken exclusively as an L2, mostly by German Ashkenazi jews.

As for the OP's question: If by identical you mean "phoneme charts with the exact same symbols", it's pretty unlikely because of differing linguistic traditions and a certain tendency to overspecify segments in a phonological analysis. If you're looking for languages that have the same set of segmental contrasts and are willing to overlook phonetic differences, then you might have more luck -- I'm not sure to what extent you're going to find pairs of true coincidences as opposed to languages from a given linguistic area though.

Yaali Annar wrote:Proto-Dravidian have bunch of PoA, no fricatives, and voicing distinction... just like stereotypical australian languages.

Yeah and its really tempting to connect this with the new evidence that sailors from India settled Australia around 2500 BC, right around the time Proto-Pama-Nyungan was spreading itself around the whole continent, but it just doesnt quite fit right. For one thing, there's obviously no shared language family, and 2500BC really isnt a long time ago. PIE is probably older than that. So where's proto-Pama-Dravidian?

The common answer is that it wasnt Dravidians who sailed over, but a now-lost language family with a very similar phonology, which also influenced Dravidian. But who are these people that had the military might to conquer Australia but couldnt stand up to the Dravidians? I can still think of some more ideas, but each one of them has many weaknesses. It may be that it is all just a huge coincidence that the only language families in the entire world with that exact type of phonological setup are unrelated even though there is evidence that in recorded history speakers of one family migrated to live with speakers of the other.

Yaali Annar wrote:Proto-Dravidian have bunch of PoA, no fricatives, and voicing distinction... just like stereotypical australian languages.

Which of course proves the existence of Lemuria

But more seriously, this has been noticed by real linguists, too. There is indeed genetic evidence of an immigration into Australia from India about 4000 years ago, which coincides with the introduction of the dingo and a number of cultural innovations in Australia, and these people may just as well have brought Proto-Pama-Nyungan to Australia. And if PN originated in India, it may have been in a kind of Sprachbund situation with Proto-Dravidian, or even share a common ancestor.

But while Dravidian and Australian languages are similar in their consonant inventories, the vowels are different. PD had */i e a o u/ plus phonemic length, while the stereotypical Australian vowel inventory is just /a i u/ and nothing else.

Actually length contrast is quite widespread in Australia: http://i.imgur.com/biJoGF4.png

Not that it matters too much one way or the other. Length is easy enough to loose and reinvent and both /a e i o u/ and /a i u/ are too common systems to show relatedness. A relation between Dravidian and the Australian languages still seems far fetched but it makes you wonder what the linguistic landscape looked like in the south Asia before the families now dominating it took over.

Publipis wrote: But who are these people that had the military might to conquer Australia but couldnt stand up to the Dravidians?

Most of the Australian continent is terrain that doesn't lend itself to empire building. Newcomers could have spread with no trouble just by having a couple of new technological innovations with them.

Aren't the facts backward from what we'd expect if PN were Indian in origin?

If PNs came from India, where did they land? Presumably not on the northwest coast, since this is the only place that's reliably NOT PN. While they could have landed there, gone through the current non-PN territory and then expanded, why wouldn't they have expanded when they first arrived?

Presumably then they landed on the west coast and moved east. But the west is not where there is most diversity in PN. OK, that could be due to the terrain, or later expansions, but it's not a good sign.

And why did the PN speakers expand so widely? Presumably they had some substantial cultural advantages. But why did Indians have cultural advantages specifically useful in the desert - but that weren't advantageous in the northwest?

Whereas if we look at it the other way, we have one group on the northern edge of the desert, among many, that developed some advantage and then expanded across the desert to the south, leaving beside some potentially-related sister languages in the north. This would look like a very normal sort of expansion. It would still be a bit odd that multiple groups of PN-speakers had then crossed the mountains to the eastern coast independently, leaving no native groups... but that's a problem no matter what, isn't it?

If there is some concrete connection between Pama Nyungan and Dravidian, it's unlikely that there weren't any other languages within this same areal or genetic group. Such group could have been spread quite far in the south Asia and there's no pressing linguistic reason why Pama Nyungan would have to be derived directly from the Indian subcontinent and not just from somewhere in the Malay archipelago.

The internal diversity of Macro Pama Nyungan would actually be in line with the hypothesis that this family represents a different migration into Australia from the rest of the north Australian languages as it's also highest in the north. Though, it's also possible that the rest of Australia has just experienced strong linguistic levelling masking away the ancient spread patterns.

But why do we specifically try to compare Pama Nyungan with Dravidian? If the motivator is the similarity between the phonologies, I don't see why the north Australian families would be any worse candidates.

gach wrote:If there is some concrete connection between Pama Nyungan and Dravidian, it's unlikely that there weren't any other languages within this same areal or genetic group. Such group could have been spread quite far in the south Asia and there's no pressing linguistic reason why Pama Nyungan would have to be derived directly from the Indian subcontinent and not just from somewhere in the Malay archipelago.

Sure. There would have been other languages of that kind; alas, we don't find any. AFAIK, there is nothing like a Dravidian-style or PN-style consonant inventory anywhere between India and Australia.

gach wrote:The internal diversity of Macro Pama Nyungan would actually be in line with the hypothesis that this family represents a different migration into Australia from the rest of the north Australian languages as it's also highest in the north. Though, it's also possible that the rest of Australia has just experienced strong linguistic levelling masking away the ancient spread patterns.

I don't know about the actual time depth of (Macro-)Pama-Nyungan. Is it actually compatible with the immigration scenario discussed here?

gach wrote:But why do we specifically try to compare Pama Nyungan with Dravidian? If the motivator is the similarity between the phonologies, I don't see why the north Australian families would be any worse candidates.

Point taken. These consonant inventories are also found in many non-PN languages, this is true. Yet, the whole set of Australian languages is too diverse to descend from a common ancestor that wass spoken about 2500 BC, or we could reconstruct Proto-Australian, which we can't.

At any rate, there is no reconstructible "Proto-Dravidian-Pama-Nyungan", so if Pama-Nyungan is related to Dravidian, that relationship would have to be very deep, well beyond 2500 BC. Apparently, Elamite seems to be a better candidate for the nearest known kin of Dravidian, and that doesn't show a "Dravidian-Australian"-type consonant inventory at all!

WeepingElf wrote:Apparently, Elamite seems to be a better candidate for the nearest known kin of Dravidian, and that doesn't show a "Dravidian-Australian"-type consonant inventory at all!

Even more so, the phonological consistency across Australia shows that it's been actively maintained through language contact. If that's the case, wouldn't it be more likely to assume that it represents an equilibrium point that's a native Australian innovation.

gach wrote:

WeepingElf wrote:Apparently, Elamite seems to be a better candidate for the nearest known kin of Dravidian, and that doesn't show a "Dravidian-Australian"-type consonant inventory at all!

Even more so, the phonological consistency across Australia shows that it's been actively maintained through language contact. If that's the case, wouldn't it be more likely to assume that it represents an equilibrium point that's a native Australian innovation.

Yes. The similarity between Dravidian and Australian consonant inventories probably is nothing but an amusing coincidence. Especially given the fact that the Australian languages show no signs of affinity to Dravidian in any other regard.