Re: Human language (long)

Michael McBroom (
Thu, 23 Jan 1997 00:36:50 -0800

John A. Halloran wrote:
> In article <> (Gregoire) writes:
> >->
> >Apparently, the passive voice--not centered in the self as chief
> >controller of events--was common in many tribal languages (like Dakota)
> >[see Werner M€ller,
> >"The 'Passivity' of Language amd the Experience of Nature: A Study in the
> >Structure of the Primitive Mind" (1968). It's old and but one example.]
> Thanks for the reference - by passive voice do you mean the same as the
> ergative construction? In an ergative language, instead of saying 'John rolls
> the ball down the hill' where John is marked as the subject as in an
> accusative language, one says in effect 'John causes the ball to roll down the
> hill', where the ball is marked as the subject and John is just a causative
> agent.

The above example does not illustrate ergativity. Generally, in an
ergative language, the case-form marking of the subject of an
intransitive verb is identical to that of the object of a transitive
verb. It is somewhat difficult to give examples in English since our
language essentially lacks case markings (the exceptions are a few
overtly marked pronouns, such as I/me, they/them, he/him, she/her), but
a crude analogy might go something like this:

Intransitive: Me arrived on time.
Transitive: She preceded me.

Please note that neither sentence uses passive construction.

There is a certain elegance to the grammatical construction of ergative
languages that non-ergative languages lack, and it is this: by using
what would normally be accusative case morphology in a non-ergative
language for both the nominative-intransitive and accusative case in an
ergative language, the grammar has been simplified without increasing
ambiguity. The speaker or listener will know immediately that a subject
noun in the nominative case will be used only with a transitive verb and
a direct object, whereas a subject noun in the accusative case must be
associated with an intransitive verb. There is no equivalent to this in
English -- we must make do with a syntax that is inherently less clear.

It should also be mentioned that languages as unrelated as Basque and
Djirbal are ergative, and share these same basic similarities in
grammatical construction. This seems to me to argue that, while we
don't have a rigid template for grammar, there is a finite number of
pathways that can be followed in order to develop a grammar, and
ergativity is one of them.


Michael McBroom
CSUF Linguistics