LINGUIST List 3.332

Thu 09 Apr 1992

Disc: Ergativity, Gender

Editor for this issue: <>

Directory

1. , 3.316 Ergativity, Agentivity and Classification
2. , 3.314 Gender
3. John Cowan, Re: 3.314 "Neuterm"

In response to the query about ergatives and locatives, I am
moved to report a discovery I made a few years ago but have
yet to publish concerning the etymology of the word 'ergative'.
While it is commonly reported to come from Greek 'ergon' ('work'),
in fact its earliest attested usage is for a kind of locative
case which apparently means something like French 'chez', German
'bei', and ... Latin 'erga' (i.e., approximately 'in close proximity
to', 'within the sphere of'). While there is no specific comment
as to its etymology, it seems clear, therefore, that it was
derived from 'erga' (either by Sidney Ray or by someone he got
it from).

Despite what you might think, this case did not have an agentive
(or ergative, in our sense of the word) function in the language
of the Torres Straits Islanders, in a description of which I found
this first usage. A few years later, though, the celebrated pater
Schmidt used the term in the way we now do. His usage was then
picked up by the monogeneticist Trombetti and from him by the
Caucasianist Dirr (the latter is the person to whom the term is
usually attributed).

Interestingly, Schmidt appears to have used the term just that
once (later reverting to the older 'casus auctoris' or 'casus
agentis' I forget which at the moment). And Ray also gave it up
in his subsequent description of the same language.

Thus, apparently, Schmidt must have provided the term with
a folk etymology (!) deriving it from the Greek and used it
in a sense Ray never intended. The two of them must then
have both decided to give it up, because of the ambiguity.
But in the meantime it had been picked up by others, and the
rest is history.

A language which has not yet been mentioned but which has
a feminine that is unmarked in the case of non-animate nouns
is Khasi, a Mon-Khmer language which is the official tongue
of the Indian state of Meghalaya. Of course, one could also
say that this is the neuter gender and that females are grammatically
neuter in this language! As I see it, there is the same problem
I have commented on in at least one other context of applying
grammatical terms to different languages.

It also might be of interest that in Polish a group of mixed
sex (i.e., at least one man and at least one woman) is referred
to in the neuter.

Finally, it is not really accurate to describe Russian d'ad'a
'uncle' as having a "feminine" declension. While it is true
that most of the nouns that have this particular declensional
pattern are feminine, there plenty of masculines like d'ad'a
(which, moreover, is historically probably a neuter!).

Message 3: Re: 3.314 "Neuterm"

Henry Churchyard <LIFY436orange.cc.utexas.edu> quotes:

> ... from E.E. Evans-Pritchard, _Social_
> _Anthropology_and_Other_Essays_ (1962)

> p.311 (quoting a discussion of whether a certain vague culture-hero is
> conceived of as a supernatural entity):

[...]

> using the masculine _ko_ and not the supernatural neuterm
> _u_.' ''

I suppose that in fact "neuterm" is haplology by Churchyard for "neuter term",
but what a great portmanteau word for pronouns like "heesh" and "sie"
discussed on this List earlier!

"What neuterm do >you< prefer?"

--
cowansnark.thyrsus.com		...!uunet!cbmvax!snark!cowan
		e'osai ko sarji la lojban