Print

Print


How frequent are languages in which 
- geminate Cs (that arise in morphology, say) are preserved as distinct from
single Cs, but
- there is no sort of compensatory lengthening otherwise?  e.g. when C1 is
lost in VC1C2, it becomes neither V:C2 nor VC2: (nor anything else such)?

I'm thinking about how to implement this stuff, and moraic behaviour, in
Gleb.  Perhaps the cleanest way would just be to say that consonants in some
positions have a mora and others don't, and the two sorts of behaviours
above happen because the mora is being preserved.  But if those two
behaviours frequently fail to co-occur, I need to tweak this. 

Alex