[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >
Mike S., On 25/08/2012 23:41:
What about a compromise? What if the form of roots were constrained as follows: root := (S) (F) (P) (P) (F) (S) P:= p | b | t | d | k | g F:= f | v | s | z | c | j | x S:= m | n | l | r ...and predicates were combined from one or more roots: predicate-stem:= ( � root )( � root )*
but what happens to stems or C sequences that don't start with S or don't end with S or contain FF or PP? Again, for reasons of economy, we want all possible C sequences to be valid stems. Remember, I'm not anti-compounding; I just don't think there needs to be unambiguous marking of stem-internal boundaries.
A further optional compromise is this: The schwa-placement rule is to be only weakly enforced. That it, it is still licit to insert or delete [�] between consonants at will. In those cases, the correct predicate is grokked from context.
I think I misunderstood. I was thinking that stem-internal boundaries would be signalled by stem-internal SS, not by schwas.
Interestingly, I understand that there are languages that contrast /k?/ and /k'/ (by which I mean plosive + glottal stop cluster versus ejective plosive, in case I have the symbols wrong).
I didn't know that. At any rate, in Livagian, /qk/ gives [?k] or (unaspirated) [k] (/k/ gives [g], /kh/ gives [kh], /kg/ gives [kG, qR], /qh/ gives [kx, qX]; and /kq/ gives [k'] as mentioned earlier; and then /kqh/ gives [k'x'] and so forth). --And.