Re: How many proto-languages?
From: John Atkinson (johnacko_at_bigpond.com)
Date: 09/15/04
- Next message: Mxsmanic: "Re: History of French"
- Previous message: Mxsmanic: "Re: History of French"
- In reply to: r: "How many proto-languages?"
- Next in thread: Jacques Guy: "Re: How many proto-languages?"
- Reply: Jacques Guy: "Re: How many proto-languages?"
- Messages sorted by: [ date ] [ thread ]
Date: Wed, 15 Sep 2004 03:34:40 GMT
"r" <nospam@thank.yu> wrote ...
> I've never been convinced by the arguments for a single proto-language.
> I'm not well enough informed to comment on Nostratic, but it's
> certainly an interesting hypothesis.
Nostratic, if it existed, was a *long* time after proto-world. (10^4 years
BP as against 10^5).
> What I have noticed from the small number of modern languages I have
> looked at even cursorily, is thay they seem to fall into three groups:
>
> 1) Ones without gender, e.g. Finnish
>
> 2) Ones with approximately three genders, of which one is always
> masculine/superior, and another feminine/inferior. E.g., Most IE ones
> I've come across, Arabic, Tamil
Animate/inanimate (or human/nonhuman) is about as common as
masculine/feminine in two-class languages.
> 3) Ones with lots of classes, or with some classification of one or more
> of the parts of the speech according to whether they are: human, liquid,
> long, alive, and sometimes many more.
[snip a bit]
> It seems to me that there is a fundamental contrast between languages
> which divide things by a gender (which may be grammatical or natural or
> as in Irish both) and those which construct words by reference to a
> class. If, for example, Bantu languages had 24 classes of which two
> were clearly related to male/superior and female/inferior,
For what it's worth, there *are* M/F systems where F is the unmarked or
dominant gender. (But, yes, they *are* a distinct minority.)
> I could
> accept that gender languages have simplified a large number of classes
> to a small number of genders. But when a geographically-contiguous
> group uses a few genders, and another geographically-contiguous group
> uses a lot of classes, and there's no overlap between genders and
> classes,
Some Amazonian languges do have *both* genders (M/F/etc) and "classes", both
of which are involved in agreement patterns, but in different ways.
> I begin to feel that looking for a common origin of all
> languages is a vain quest.
>
> The only way I can make any sense of this to myself is to assume that
> class-using languages spread across the world, and later a gender-using
> group pushed them out of the geographically-contiguous areas which are
> now occupied by gender-using groups. This argument, however, has as
> many holes in it as a mosquito net,
The main one being the relevant time-scales (see below).
> and I don't advance it seriously.
> Holes of ignorance (I know nothing of class and/or gender in Khoi/San,
> Caucasian or Fulani-group languages,
"Khoi-San" isn't a genetic group. The Khoe languages (central Khoi-San,
including Nama) have gender (masculine/feminine/common). Some non-Khoe
languages have noun classes (four in Jul|'hoan). The isolates Sandawe and
Hazda both have gender (masculine/feminine).
Fulani, like many other Niger-Congo languages (including the Bantu group)
has a pretty full noun-class system.
Noun classes are widespread in the North-East and North-Central Caucasian
families -- two to eight classes. In North-West Caucasian, just one
language has classes (human/nonhuman). The unrelated South Caucasian
(Kartevelian) languages do not have noun classes or gender.
World-wide, about a quarter of languages have noun class/gender.
> to take only three important
> groups), holes of probability (the gender-using peoples don't have a
> long history, and are more probably immigrants into Africa rather than
> emigrant from there, so either they appeared out of nowhere or suddenly
> decided to use gender rather than class) and holes of history (the
> languages I know are modern languages, not proto-IE, proto-Iroquian,
> proto-Bantu).
As a physicist, you know about timescales. Human language has been around
10^5 years, more or less (but not much less). Noun class systems seem to
last 10^3-10^4 years. In maybe 5*10^3 years, IE languages have gone from
animate/inanimate to M/F/N to M/F (in the Romance languages), or
common/neuter (many Germanic languages) or no-gender (a few Germanic and
many Indo-Iranian languages). AFAIK, the most conservative family in this
respect is Afroasiatic, which has maintained a M/F system for perhaps 8*10^3
years. My point is, that languages cycle in and out of having noun classes
an order of magnitude faster than the time since proto-world, so it's *very*
unlikely the present distribution of NC/genders can tell us anything about
what was going on away back when.
A couple of references for you:
Greville Corbett, "Gender", Cambridge, 1991 [Describes NCs of all kinds and
how they work].
Johanna Nicholls "Linguistic Diversity in Space and Time", Chicago UP, 1992
[World-wide distribution of things like this, and how they might have got
that way].
John.
- Next message: Mxsmanic: "Re: History of French"
- Previous message: Mxsmanic: "Re: History of French"
- In reply to: r: "How many proto-languages?"
- Next in thread: Jacques Guy: "Re: How many proto-languages?"
- Reply: Jacques Guy: "Re: How many proto-languages?"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
