I thought this model applied to some Balkan countries, as in a book of mine it was used as an example of characteristics of the Balkan Sprachbund.
I'm sorry to spoil your expectation, but I believe that this model applies to all Slavic languages - at least all I was able to check. The only difference was a level of distortion from the original structure - which in case of Bulgarian is very close, while in the North they were more severely changed, albeit not beyond recognition. Let's take twelve as an example:
Indeed, the Balkan languages (some linguists seem to include also Greek, Slavic Macedonian, as well as a bunch of minority languages or dialects) share a whole lot of structural similarities. But apparently the structure of the numerals is shared on a much wider area.Actually, I didn't have any expectation, because I know nothing on this subject.
Reading about the Balkan Sprachbung I saw that three Balkan languages (Bulgarian, Romanian, Albanian), that don't belong to the same family, form some numbers (e.g. 11, 12) in the same way. This motivated my desire to open this thread. Thank you for you thorough reply!
(from German Tausend of course)
Probably so, but who said that certain features of the Balkan Sprachbund cannot be induced by Slavic languages?
It is worth adding that the Bulgarians in everyday speech use the reduced form of these numerals, so instead of единадесет, дванадесет ... двадесет ... they say: единайсет, дванайсет ... двайсет ... respectively.
Phonetically unmotivated shortenings of very frequent words and, in particular, numerals are entirely common. Cf. Rus тысяча ['tɨsʲəʨə] > ['tɨɕ:ə] ("тыща"); пятьдесят [pʲɪdʲ:ɪ'sʲat] > [pʲɪj'sʲat], шестьдесят [ʂɨ̞zʲdʲɪ'sʲat] > [ʂɨ̞j'sʲat] etc. Old Russian "один на десяте", "дъвѣ на десяте" etc. also wouldn't have produced Russian "одиннадцать", "двенадцать" etc. with regular shifts and morphological analogies only - that also required an irregular shift from е to ь, which then regularly led to -дьсять > -цять (the fall of the yers and resulting assimilation) > -цать (the hardening of ц).
[As usual, I pronounce it in a less reduced way than you, namely [pʲɪɪ'sʲat] and [ʂɨ̞ɨ̞'sʲat], with three distinct syllables.]
[I periodically encounter these войны and гуманойды in the net, but yet have to hear such a pronunciation in person to ask the speaker about it. So far I strongly suspect these -oйд- are not phonetic changes but rather plain anglicisms, and воин->войн- are results of interference with война. Otherwise why don't we find **читайт or **думайт?]
Er... how? Okay, it might work for Android OS, but I positively cannot imagine a situation when a person becomes acquainted with the word астероид from the English pronunciation of "asteroid" and not from spoken or written Russian. On the contrary, I'd normally expect the Russian -оид to influence the English pronunciation of -oid among Russians (compare how they still occasionally pronounse English "c" [s] as [ʦ]).
That "interference" actually results in a merger. Looks far too complex an explanation.
Because of the absolutely transparent morphology here.
[I need actual speakers to investigate: as I have written, I have never heard this in person. There is also выйграть etc. in the net, which looks to me a regional pronunciation: again, never heard by me in the real life.Er... how? Okay, it might work for Android OS, but I positively cannot imagine a situation when a person becomes acquainted with the word астероид from the English pronunciation of "asteroid" and not from spoken or written Russian. On the contrary, I'd normally expect the Russian -оид to influence the English pronunciation of -oid among Russians (compare how they still occasionally pronounse English "c" [s] as [ʦ]).
That "interference" actually results in a merger. Looks far too complex an explanation.
Because of the absolutely transparent morphology here.
You just wouldn't recognize it as such, I suppose. The purely acoustic difference between [V.ɪC] and [VjC] is negligible, so the brain simply reconstructs the "correct" structure, whatever it may be.
I don't think it's fully correct. Stressed /ViC/ and /VjC/ do tend to merge, but the result of this merger is closer to the former than the latter. [aë] can pass as a relaxed version of [ai̯], but not the other way around.
As I said, the purely acoustic difference is negligible anyway. The numer of syllables, however, does count phonologically.
What do you mean by this?
That there is a pronounced phonological difference which, however, may have little or, potentially, even zero effect on the actual acoustic realization.
