accusative and ergative languages

[Larry Trask]

On Fri, 18 Jun 1999, Patrick C. Ryan wrote:

> I am almost sure that you will want to refine my understanding of
> the word "simplicity" but for whatever it may be worth, I would
> characterize an isolating language as simpler than a flectional one.

If this were true, then a good reference grammar of Chinese or
Vietnamese ought to be shorter than a good reference grammar of Russian
or Latin.  And this does not appear to be the case.

> In keeping with the schema you presented recently, I think an
> isolating type language, which Klimov would connect with his neutral
> and active "types", must have preceded agglutinating, flectional,
> and analytic "types".

Why?  No evidence.  And what distinction are you drawing between
`isolating' and `analytic' languages?  The terms commonly mean the same
thing.

> That is, of course, not to say that an analytic language might not
> be able to revert (is that better than "regress" or "devolve"?) to
> an isolating type.

Why not just say `change'?  Nobody will object to that verb, though I
still don't understand how an analytic language is different from an
isolating one.

> I assert that I conider it impossible for an inflecting type
> language to have been what wer should expect to see at the very
> earliest stage.

Why?  We have no data on the earliest human language(s).  But we do have
data on languages which have come into existence very recently.  Many
creoles, such as Tok Pisin, have developed grammatical inflections
within a generation or two of coming into existence.  Apparently the
same is true of Nicaraguan Sign Language, which has only existed as a
mother tongue for about one generation -- though I don't have good
information on NSL.

[on my objection to Pat's claim that an object/non-object distinction is
tantamount to ergativity]

> First, let me say that I am well aware that, your being a specialist
> in Basque, gives you a deep perspective on ergativity that few
> members of the list will share through work in the own particular
> specialities but still I must tentatively disagree.

I can't claim any special expertise in ergativity just because I work on
Basque.  What is true of Basque is not necessarily true of any other
language exhibiting ergativity.  Basque, Georgian, Hindi, Dyirbal and
Nass-Gitksan all exhibit ergativity, but they differ substantially in
the circumstances in which ergativity appears and in the manner in which
ergativity is expressed.  Apart from the trivial observation that all
exhibit some measure of ergativity, there is probably nothing which is
true of all five beyond what is true of languages generally.

> And, of course, it seems to me that this is another case of
> differening definitions. I prefer a functional one, you, I believe,
> prefer a formal one.

I don't see that this matters.  A binary division between objects and
non-objects is accusativity by definition, regardless of how it is
expressed.  Ergativity is a binary division between transitive subjects
and all else, also by definition.

Anyway, as Ralf-Stefan Georg has already pointed out at length, there is
probably no language which is 100% ergative.  Ergativity may be present
in a language in various circumstances and to varying degrees, ranging
from none at all to quite a bit.

English, for example, is marginally ergative in its word-formation:
certain word-forming suffixes, such as the <-ee> of `standee' and
`employee', work in an ergative manner.  Basque, unusually, is totally
ergative in its inflectional morphology (apart from two minor wrinkles),
but totally accusative in its syntax and also in its word-formation.
For example, transitive and intransitive subjects exhibit identical
control properties, while direct objects do not, and the agent suffixes
can be added indifferently to intransitive verbs like `go' and `sleep'
and to transitive verbs like `make' and `watch', producing in every case
a noun denoting the performer of the action.

I don't think the concept of an `ergative language' has any real value
in linguistic analysis.

> For me, a passive of the form "the man is being slandered", displays
> an underlying object of a transitive verbal action, whatever the
> formal marking of "the man" may be.

I don't suppose anyone would disagree with this.

> I will speak only of Sumerian if you do not mind. In that language,
> so far as I know considered by most as an ergative language, a
> two-element sentence of the form Noun + Verb(transitive) will, in
> nearly all cases, have to be interpreted as a passive. And frankly,
> I am not sure that this analysis is not more appropriate even to
> Verbs which would normally be considered intransitive or stative ---
> but let us not get off onto a side-topic.

I know no Sumerian.  But, speaking generally, the `passive'
interpretation of ergativity, once so popular, was discredited years
ago.  In very many ergative languages, it is trivial to demonstrate that
transitive sentences are active, not passive.  I myself have done this
for Basque.

> I assert that the relationship is closer between an object and
> transitive verb in an ergative language because the relationship
> (frequently OV) and is immediate and direct. The ergative agent,
> IMHO, can best be regarded as a *missible* adverbial adjunct to the
> verbal phrase, consisting of the Obj + V, having no more importance
> to the verb than an adverbial phrase denoting the target or manner
> of an intransitive verb of motion.

This may or may not be so in some ergative languages, but it is
certainly not true for ergative languages generally.

> Pat continued:

>>> I would also assert that, at least once, an "ergative stage" must
>>> precede any "accusative stage" or a mixed system.

> Larry objected:

>> Unsubstantiated assertion.  You might, with equal justification, assert
>> that accusativity must precede ergativity in all cases.  If anything, it
>> is this last statement which is better supported by the evidence.

> Pat responds:

> I find it no more unlikely than to assert that a synthetic language
> is not going to develop "directly" from an isolating one.

Surely you mean `fusional', not `synthetic': agglutinative languages are
synthetic, but can readily arise directly from isolating ones.

Fusional languages do not ordinarily arise directly from isolating ones
because there appears to be no possible pathway other than one leading
through agglutination.

> An adverbial phrase specificying the agent seems to me to be an
> integral step that must be taken before a nominative is developed.

If that were true, then we might expect to see children acquiring
English pass through an ergative stage before they grasp accusativity.
But we don't.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk