Triggeriness ...
From: | Andreas Johansson <andjo@...> |
Date: | Tuesday, December 9, 2003, 17:18 |
This discussion of trigger languages is making me confused - I thought I had a
decent idea what they were about, but apparently not. Anyway, it seems to me
whether a language use a trigger system or not should be orthogonal to whether
it's accusative, ergative, active, clairvoyant, MRL or tripartite, so please
shoot the following down:
Assume we want to translate the English sentences "I bathed _in the pool_"
and "I killed a shark _in the pool_" into a trigger lang. According to my
(apparently erroneous) understanding, these would become something like
the_pool-TRIG bathed-LOC 1st.sg-S (i)
and
the_pool-TRIG killed-LOC 1st.sg-A a_shark-P (ii)
and therefore it would a perfectly well-defined question which, if any, of the
markers S, A and P are identified. Say that the markings S and A are the same,
and we'd have a nominative trigger language; say A and P are the same, and
we'd have a MRL trigger language; and so on.
It would still apply if we retopicalize:
1st.sg-TRIG bathed-S' the_pool-LOC' (iii)
1st.sg-TRIG killed-A' the_pool-LOC' a_shark-P (iv)
a_shark-TRIG killed-P' the_pool-LOC' 1st.sg-S (v)
since we simply ask which, if any, of S', A' and P' are identified.
Also, since this is apparently NOT how a trigger language works, what would
one call a language that DOES work like this, and are there any?
Replies