A noun classifier marks on the verb some information about the form of one of its arguments. For example, Atili has a handful of "stationary" classifiers that produce semantic distinctions:
Emanyo bu-vahkin-us azven<k>o.
Emanyo ball-cook-P/3;PFV cow<3>
Emanyo cooked the meatball.
vs
Emanyo mal-vahkin-us azven<k>o.
Emanyo board-cook-P/3;PFV cow<3>
Emanyo cooked the brisket.
Atili also has "motile" classifiers that indicate manner of motion:
Yeredol many-az-a ranhal<k>o la.
Yeredol run-go-P;IPFV beach<3> to
Yeredol runs to the beach.
vs
Yeredol ib-az-a ranhal<k>o la.
Yeredol climb-go-P;IPFV beach<3> to.
Yeredol climbs (down) to the beach.
Is it attested for a language to use a similar system for encoding whether motion is towards or away from the speaker, as in the following hypothetical examples?
Yeredol ko-az-a ranhal<k>o la.
Yeredol go-go-P;IPFV beach<3> to
Yeredol goes (away) to the beach.
vs
Yeredol kom-az-a ranhal<k>o la.
Yeredol come-go-P;IPFV beach<3> to
Yeredol comes to the beach.
ko
andkom
occupy the same slot in the verb template asbu (ball)
andib (climb)
? Can you combine them giving something like? Emanyo bukom[throw]a azvenko (Emanyo threw the meatball towards me.)
Also, dobu (ball)
andib (climb)
occupy the same slot in the verb template? – Gregory Nisbet Oct 4 '20 at 22:45