(fwd) from VIctor Golla re. Agent/Patient

[David_Tuggy at SIL.ORG]

Victor asked for examples of impersonals becoming specific person subject
markers. Here's one:

In some towns of the Orizaba Nawatl (Nahuatl de la Sierra de Zongolica)
area, the subject prefix se- (or see-) has pretty clearly developed from an
impersonal subject meaning to mean 'we'.  This phenomenon was first, to my
knowledge, reported by Jeff Burnham, who described it in a grammatical
sketch of Rafael Delgado Nahuatl which he distributed in the late 70's at
the Friends of Uto-Aztecan meeting).

se- is transparently related to the numeral se or see (the length is very
slippery) 'one'. In other towns in the dialect area 'impersonal subject' is
its primary meaning. The verbal morphology associated with it is that
appropriate to a singular subject, even when the meaning is clearly 'we'.
The towns that use it to mean 'we' still conserve, at least in that they
understand in their neighbors' speech, the standard Nahuatl ti- 'we' prefix
(which has plural morphology, as you would expect. ti- with singular
morphology means 'you sg.') Even in the towns which do not use it to
clearly mean 'we', there are of course many contexts in which either
meaning would be appropriate, especially in procedural discourses: "that's
the way we do it/it is done".

In most Nahuatl se / see does not function as a subject prefix, but can be
an independent subject preceding the zero 3rd person subject marker. In
other words, the acoustic difference between se 0-kineki 'one wants it' and
sekineki 'impersonal wants it' is somewhere between slight and
non-existent.  In certain tenses an o- other prefixes precede the subject
prefixes, so it is the innovation of words like y-o-sekinek 'impersonal
already wanted it' (as opposed to se yosekinek) that seals se-'s status as
a subject prefix.

--David Tuggy