WeepingElf wrote:
Yep. If the two families had a common ancestor in India about 4000 or 5000 years ago, the resemblances would be readily apparent and the relationship firmly established by now. It isn't.
Not necessarily--if they had a common ancestor about 4-5kya, then the resemblances would be discoverable, but not necessarily readily apparent. Anyway, there are some "readily apparent similarities" already. The consonant inventories for example. But there are languages which have been known to academic linguistics for some time, and their (lack of) relation to other languages was quite obvious until it was shown quite rigorously that there wasn't actually any such lack. Some examples exist in Austronesian, for instance, that diverged from the rest of the family much later than 4kya and yet are barely recognizable because of various influential factors.
EDIT: also, where is the person who wrote this article getting their numbers? 200 is not "larger than Indo-European" or "Almost as large as Sino-Tibetan" by any counting of those language families that I've ever seen. Lolo-Burmese alone is like 100.
thetha wrote:Not necessarily--if they had a common ancestor about 4-5kya, then the resemblances would be discoverable, but not necessarily readily apparent. Anyway, there are some "readily apparent similarities" already. The consonant inventories for example.
Well, at that kind of time depth, I wouldn't put much store by common consonant inventories. Phonological inventories and typological features are much more likely to be areal features than to be inherited (the situation is of course more complicated when genetically related languages hang around in geographically contingent areas, like Semitic). If someone could show common inherited vocabulary and morphemes, that would convince me.
thetha wrote:EDIT: also, where is the person who wrote this article getting their numbers? 200 is not "larger than Indo-European" or "Almost as large as Sino-Tibetan" by any counting of those language families that I've ever seen. Lolo-Burmese alone is like 100.
Older classifications, probably. Ruhlen has 144 IE languages, 258 Sino-Tibetan. (Living languages only.) These numbers were in line with Voegelin & Voegelin and the Ethnologue.
Actually upon doing some closer examination of my personal sources on Sino-Tibetan stuff (mainly Matisoff's work) 258 (or something in the higher 200s) does seems to be a reasonable count of them.
hwhatting wrote:
thetha wrote:Not necessarily--if they had a common ancestor about 4-5kya, then the resemblances would be discoverable, but not necessarily readily apparent. Anyway, there are some "readily apparent similarities" already. The consonant inventories for example.
Well, at that kind of time depth, I wouldn't put much store by common consonant inventories. Phonological inventories and typological features are much more likely to be areal features than to be inherited (the situation is of course more complicated when genetically related languages hang around in geographically contingent areas, like Semitic). If someone could show common inherited vocabulary and morphemes, that would convince me.
For sure, but that's my point. We do have readily apparent similarities, they've just done nothing for us.
Soap wrote:Well, American English has plenty of words with initial /kj/, such as "cute", "cube", "cuneiform", etc. But none with initial /tj/, because yod=dropping only happened after coronals. Granted, that /kj/ in itself only appears before /u/, but Im not sure that matters. I remember hearing a news reporter say "kə'tu.šə" for Katyusha on TV once and being surprised.
Right, AmEng only tolerates Cj before /u or ɚ/. (In dialects where /ur/ is preserved, Cj is only tolerated before /u ur/.)