If _ninqitá-_ 'whiten' has a causative suffix, which seems to be
the case, why do we have _tulta-_ 'send for, fetch, summon'
(< *_tultâ-_ 'make come' < TUL- (cf. VT46:20) and not **_tultá-_?
[I would hesitate to call what we see in _ninqitá-_ a "suffix". Given
the frequent and apparently free variation we see in prehistoric Eldarin
forms between long and short vowels in word- and stem-final position
(denoted by Tolkien as a macron with a breve over it), I would guess that
the actual case is one of _selection_ of variant forms in _association_
with distinguished meaning: thus, in this scenario, the long form was
selected and associated with the causitive and/or inchoative sense (the
gloss 'whiten' is itself ambiguous: it could mean "to cause something to
become white" or the process of a thing itself becoming whiter), and the
short for the intransitive/stative sense; and the length difference then
retained _because_ it maintained the distinction in meaning. To what
extent this selection formed a pattern in other cases is unclear. CFH]