The perfect prefixes as erstwhile verbs
The next piece of supporting evidence involves the diachronic origin of the three perfective prefixes nu =, su=, and ghu=. Like other TAM markers, perfect, perfective or past markers usually grammaticalize from erstwhile verbs, as is evident in the current set of younger TAM suffixes (slot-1). Is there any evidence for that in Tolowa? Given that the vowel in the three perfective prefixes is the neutral, most bleached vowel in Tolowa (/u=/, schwa), the task of relating the three prefixes to extant verb stems is difficult. But at least one of them, ghu=, may have come from-gha ‘come’, ‘arrive’ (see earlier above).
There is, however, one strong piece of evidence to suggest that all three may have come from verbs. This has to do with their special form for first — but particularly third — person subject. As noted above, we have good reason for reconstructing a perfective nasal suffix for an earlier stage of Tolowa. Now, two of the perfective prefixes — ghu= and nu= — display a nasalization of their vowel in the 3rd person singular. And all three display an aberrant 1st person subject pronoun-i- (rather than the more regular-sh-). Thus compare:
(1)
(2)
(3)
The most reasonable explanation of a nasal suffix on the perfective morphemes is that it represents the last vestige of the old perfective verb suffix. But this must mean that the current perfective prefixes ghu=- and nu =- —and most likely also su= — must have been originally verbs. It is their perfective form that became the perfective marker, and they simply pulled the perfective nasal suffix with them.
