Hi all, I just returned from a fantastic, somewhat adventurous trip to Ladakh, the northern region of India. As of course one can't expect that I go to a place with a funny language without me diving into it, I have dived into Ladakhi.
Ladakhi is technically a 'dialect' of Modern Tibetan but is considerably different, and generally more archaic when it comes to the pronunciation of words.
The word for language is a nice example.
Which is pronounced as:
/kε/ in Tibetan but as /skat/ in Ladakhi.
As you can see, Ladakhi is more archaic.
In Tibetan the prefixes /s-/, /r-/ and /l-/ are lost while Ladakhi retains them. Ladakhi in some cases even retains the prefix /m-/.
On top of that, in Tibetan final /-d/, /-s/ are lost which causes preceding /a/, /o/ and /u/ to shift to /ε/, /ø/ and /y/ respectively. The nasal final /-n/ causes a similar shift, but also nasalises the preceding vowel.
Quite a few other developments take place, but I would like to focus on the vowel change and loss of /-s/ in Modern Tibetan.
As pointed out, in Ladakhi this development generally doesn't occur, but strangely, this shift does occur in the case system.
The Ergative marker -(y)is is realised as /-(y)i/ and the ablative marker -nas is realised as /ne/.
Also the ergative form of 'I' ŋa-s is realised as /ŋe/.
There are plenty of lexical items that retain the s in similar contexts. For example gñis `two' is realised as /ñis/.and las 'work' is realised as /las/.
I have found one example of a lexical item of a final s that is not pronounced in the common name bkra-śis which is pronounced as /ʈa-śi/. But as this is also a very common name in Tibet, it seems likely to me that this name is simply loaned from Modern Tibetan.
Other morphology that involves a final s behave normaly. za /za/ 'to eat' in the perfect becomes za-s /zas/ not /ze/.
So now we have a situation where a sound law seems to have been applied solely to case suffixes. Sound laws are generally considered to be 'blind' to grammatical categories, the cases in Ladakhi do not seem to care.
The suffix -s in za-s could still be due to analogy from verbs that end in consonants like skang perf. skang-s but such an analogy explanation fails with words like las or gñis.
I have considered that due to some odd development the Ladakhi Genitive has taken over the function of the Ergative, as in modern Ladakhi, despite being written differently, they are completely homophonous in every context. While this wors pretty well, it still does not explain why the Ablative case is pronounced /ne/ and not /nas/.
So in fact, it almost looks as if Ladakhi retained its final /s/ but due to some bizarre twist of history it was deemed necessary to loan the Ergative and Ablative case forms from Modern Tibetan.
Bizarre. I'm very curious if someone else can come up with a likelier scenario.