Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/17/84; site bcsaic.UUCP
Path: utzoo!linus!decvax!tektronix!uw-beaver!ssc-vax!bcsaic!michaelm
From: michaelm@bcsaic.UUCP (michael b maxwell)
Newsgroups: net.nlang
Subject: Re:  Query on ergative languages
Message-ID: <195@bcsaic.UUCP>
Date: Thu, 18-Jul-85 17:38:45 EDT
Article-I.D.: bcsaic.195
Posted: Thu Jul 18 17:38:45 1985
Date-Received: Sun, 21-Jul-85 04:17:49 EDT
References: <6312@ucla-cs.ARPA>
Organization: Boeing Computer Services AI Center, Seattle
Lines: 92
Keywords: Ergativity syntax Mayan

Re Rich Wales' request for info on ergative languages (I'm posting this
thinking there may be some general interest, although I apologize for
the length...)
First let me say that I wouldn't describe ergative languages in terms of
patient/agent terms as opposed to subject/object terms; the two ideas
are quite different (patient, agent... are semantic terms, subject,
object... are syntactic terms).  There are languages which (it has been
claimed) use notions like patient etc. rather than subject, such as the
languages of the Philipines. These are not ergative.
The basic idea of an ergative language is that it (in some way) treats
the subjects of intransitive verbs like the objects of transitive verbs.
Let's try an example--in English, the subject precedes the verb in both
transitive and intransitive verbs, whereas the direct object (which only
transitive verbs have) follows the verb:
	John (subj.) sees Bill (obj.).  =transitive
	John (subj.) runs.   =intransitive
Now let's suppose that English were ergative. Then we might have the
following sentences:
	John (subj.) sees Bill (obj.).
	Runs John (subj.).
Actually, there are certain constructions of English that resemble ergatives. 
If there is a prepositional phrase, that PP can often precede an intransitive 
verb, and the subject follows the verb:
	Into the room ran John.
	In the garden stands a statute.
An article in a recent issue of the journal Linguistic Inquiry claimed that 
certain other constructions of English were ergative, although they don't look
like what I'd call ergative sentences.  At any rate, English is usually 
considered nonergative.
Now--what are some other ways languages can treat subjects of
intransitive verbs like objects of transitive verbs?  In many languages,
which NP (=Noun Phrase) of the sentence is the subj. and which is the
object is marked by case markers (typically suffixes) on the nouns,
rather than by word order (the word order typically being rather free in
such languages).  Latin is such a language, and we have a vestige of
case marking in English in the pronominal system ("he" is the subject,
"him" is typically the object, "his" is the possessive).  An ergative
case marking language would use the same case marker for the subject of
an intransitive verb and the object of a transitive verb.  I'm not sure,
but I think Basque displays this kind of ergativity.
Finally, a language might use person markers on verbs to indicate which
NP is the subject and which the object.  (Most Indoeuropean languages do
this; again, we have a vestige of this in the third person marker -(e)s
on most verbs, and more completely in the present tense paradigm of the
verb "be.")  In some of these languages, this is the main way of telling
which NP is which, since either they have fairly free word order, or the
word order isn't enough (as in an SOV--i.e. Subject-Object-Verb--language
where neither the subject nor the object was obligatory).  Tzeltal, a
Mayan language from the state of Chiapas in Mexico, is an example of an 
ergative language using person marking on verbs.  The subjects of
intransitive verbs and the objects of transitive verbs are marked as
suffixes on verbs, while the subjects of transitive verbs are marked as
prefixes on verbs.  Some examples (with spelling regularized; the word
yash/ya is an incompletive marker, somewhat like one of our "helping"
verbs; it has a different form for transitive and intransitive verbs,
but this has nothing to do with Tzeltal being ergative):
	Yash     talon    ta   sna      Antonio.
	incompl. you-come from house-of Antonio
	"I come from Antonio's house."

	Yash     talat    ta   sna      Antonio.
	incompl. you-come from house-of Antonio
	"You come from Antonio's house."

	Ya       'awilon.
	incompl. you-see-me
	"You see me."

	Ya       kilat.
	incompl. I-see-you.
	"I see you."
I believe most Mayan languages are ergative.  Niuean, a Polynesian
language, is also ergative; a fairly extensive syntax is given in a
paper by William Seiter which appeared in Linguistic Inquiry about 1979
(sorry, I don't have the exact reference, but I can get it if you want).
There are many other ergative languages, although they are in the
minority.  
A question remains: are ergative languages *really* that different, or
is it just a matter of "superficial" differences like case marking?  I
suspect the latter--in fact there are processes in language which appear
to happen prototypically to subjects (Example: in infinitivals, the
missing argument of the verb is the subject: "John tries to run" =>
John is the runner; you can't say something like "John tries for Mary to
see" meaning that John is the one seen without using the passive.
Tzeltal is no different from English in this respect--it is always the
subject which is "missing".)  An interesting article in this regard
is Stephen Anderson 1976 "On the Notion of Subject in Ergative
Languages," in Subject and Topic, edited by Charles Li, pg. 1-23.  I
believe he discusses Caucasian languages, although it's been awhile
since I read it...
-- 
Mike Maxwell