Kymna has developed a fundamental perfective/imperfective aspect distinction much like the Slavic languages. The lexical verb forms are all perfective as they were in Proto-Tallic and also are in Talmit. The ways of derivation are very limited:
The most common way to derive imperfective forms was to attach the prefix ṇ-. It seems to have conveyed a sense of extension in space for nouns and its reflexes an-, na-, in-, ni- are widerly used in this way in Talmit. Applying it to verbs in order to express an extension in time seems to have been a Kymna innovation.
Syllabic ṇ- has become a- in Kymna and thus coalesced with the intensive prefix a- (former positive signum) which may have contributed to its development. It is never stressed, persumably because ṇ- could not be stressed either. In Moluma, initial ṇ- additionally led to a nasalization of a following stop: *ṇp-, *ṇt-/*ṇd-, *ṇk-/*ṇg- > am-, an-, aŋ-. Hence:
Due to medial lenition of -p- and -g-, verbs with initial p-, g- have seemingly irregular forms. In Moluma, *p-, *g- > am-, aŋ- and further: ŋ > n before e, i, ë, y; ŋ > m before o, u; dropped before r, l. Also bear in mind that a becomes o by u-mutation except in Moluma where analogical levelling led to uniform a-.
Verbs from ”light roots” (which ended in a vowel or r, l, s, n, Q) can fully reduplicate their roots for the imperfective aspect. This is not found in Talmit, where partial reduplication forms an iterative aspect for all verbs.
The underlying derivational patterns can be conveniently organized into several paradigms.
The first and simplest one is the atelic paradigm for verbs which have no specified endpoint (τέλος). The simplex is perfective, the imperfective is obtained by prefixing a- (causing sound changes):
| ||||||||||||
The stative paradigm, used for stative verbs, is a three-step ladder. The simplex is perfective, the imperfective is obtained by prefixing a- or by reduplication. The resulting imperfective can be made perfective again by prefixing s- (causing sound changes):
| ||||||||||||||||||||||||||||||
The telic paradigm branches a perfective simplex into two imperfective variants: one in which the τέλος has not been achieved yet, and ’the action as such’, without any restrictions. The latter is formed by reduplication or by a compound with the verbalized preverb (see below and 4.3) porūm ’to be above’3:
| ||||||||||||||||||||||||||||||
Thus sahi-porūm is literally ’to be upon eating’. In English, ’burn up’ and ’eat up’ use an image schema where something accumulates and rises in height up over time (e.g. pouring water into a cup), but in Kymna the underlying image schema is the carrying or throwing of an object – the action lasts as long as this object is up in the air. Also, when performing a task, there is a metaphorical rise in effort and concentration.
Finally, there is the preverb (see 4.3) paradigm which is a bit different from the others. One has to point out that preverb simplexes are always imperfective (while other verbs are always perfective) and usually denote position in a location. Again, the change of state into that position is expressed by prefixing s-. A secondary imperfective denoting that this goal has not been reached yet is expressed by a compound with another preverb tënum ’to be headed for’:
| |||||||||||||||
Of all the verb paradigms, the telic verbs are the most complicated. The reduplicated T.r form can cross over to the S.r → S.r.s pattern with a new T.r.s skachodzum ’start to burn’4. The T.c form can cross over to the P pattern and form T.c.s sahi-sorūm ’start to eat’ and T.c.c sahi-porutënum ’prepare to eat’. Using pyrūm ’to be below’ instead, one obtains sahi-hyrūm ’finish eating’ (pf.) and sahi-pyrutënum ’to be finishing eating’ (ipf.)5.
Unlike Talmit, which has no aspect marking, the inconclusive form distinguishes aspect (but not tense) in Kymna:
In the conclusive form, present perfective has a gnomic sense in declarative sentences and a future sense in conditional and consecutive clauses: