accusative and ergative languages

[Jens Elmegaard Rasmussen]

On Tue, 22 Jun 1999, Patrick C. Ryan quoted Larry Trask
<larryt at cogs.susx.ac.uk> to have written on
Sunday, June 20, 1999 7:31 AM:

[...]
>> [S]peaking generally, the `passive'
>> interpretation of ergativity, once so popular, was discredited years
>> ago.
[...]

Dear discussants and List,

pardon the intrusion, but this has always been a great question in my
head: Has the passive interpretation of the ergative really been
disproved? Does it not still provide a smooth and unproblematic answer?
And are not passive circumlocutions known to underlie some of the cases of
ergative structure arising in the light of historical records, as e.g. in
Indic?
   Was the discrediting not rather aimed at making the passive
understanding synchronically valid? Of course that is not sensible, for it
is only when a passive periphrasis has lost its marked status as a passive
that it can be perceived as the "normal" verb which is then construed with
an ergative syntax. If the old normal verb disappears, and a new way of
expressing the passive - used when that is meant - if created, one would
presumably have to speak of an ergative structure. It is my impression
that the fierce opposition against the passsive analysis of the ergative
has often (mostly? - always?) been part of a semi-political crusade
against the desire to explain everything in terms of linguistic history.
And, of course, if one does not care about the origin of the ergative
structure, there is no point in deriving it from something different from
what it is in pure synchrony (actually there is no point in deriving it at
all). But if one _does_ want to find out how it came into being, is the
passive solution not still the best guess around - and is it not known to
be true in a number of cases?
   A truly innocent question for information: Are there other avenues that
are _known_ to have led to the creation of an ergative than the one
starting from an old passive?

Jens