Re[2]: [tied] Re: IE right & 10

From: Brian M. Scott
Message: 34346
Date: 2004-09-28

At 3:39:34 PM on Tuesday, September 28, 2004,
enlil@... wrote:

> Harald:
>> Don't be silly. Most hunter-gatherers in e.g Papua and
>> Amazonia still trade as much today, and with little
>> correlation to their numeral system. One does not even
>> need numerals to trade.

> Me being silly? What exactly are you talking about when
> you say Amazonia?

Presumably the usual meaning: the Amazon Basin, home of some
300 languages belonging (apart from several isolates) to
some 20 families.

> The Piraha perhaps? That looks to me to be nothing short
> of a linguistic oddity. They don't even draw, supposedly.
> Dr Gordon (no relation) apparently concludes that the
> Piraha have no sign of mental retardation. How nice of him
> to conclude that :O

> Now why would one use that as a fair example of the
> majority of hunter-gather bands or tribes, most of which
> no longer even exist?!

There are quite a few hunter-gatherer tribes in Amazonia
(and for that matter Papua); I see nothing in Harald's
comment that would even suggest that he was thinking
primarily (or at all) of the Piraha. You're ignoring his
stated point: trade by hunter-gatherers is not correlated
with the nature of their numeral system.

[...]

> You may have had a different ethnology professor but the
> one that I had when I took a filler course while in
> CompSci had established in class clearly that
> hunter-gatherers are an overwhelming minority in the world
> today compared with other kinds of societies.

Of course. So what?

> In terms of world population, what is the percentage of
> people living as hunter-gatherers today? One percent?
> Clearly this is very different than in the paleolithic.

It is not at all clear, however, that it is different in any
relevant way. To show relevance, you'd have to show that
the numeral systems of hunter-gatherers are somehow
correlated with the percentage of humans living as
hunter-gatherers.

> In statistical terms, the sampling size may be indeed too
> low to yield an appropriate picture of paleolithic
> linguistics.

Elementary statistics will tell you that the sample size
would be quite adequate if the sample were random.
Arguments from sample size will get you nowhere. You should
instead be trying to argue that the sample isn't
representative. This might be a bit difficult, though: it's
not obvious that there should be a connection between the
survival of a hunter-gatherer culture down to early-modern
times or later and the nature of its numeral system.

> So to reason that hunter-gatherers wouldn't normally have
> number systems based on a handful of hunter-gatherer
> societies that still exist appears to me to be completely
> assumptive.

That's a pretty large handful; several hundred, if I'm not
mistaken.

>> Besides, the tendency is exactly same in vocabularies
>> gathered in the 17th to 19th centuries.

> Do you think the percentage of societies that were
> hunter-gatherers was terribly much higher only a few
> centuries ago? What do you base this on?

Why on earth do you think that this percentage has any
relevance at all?

> What's silly here is that we seriously reconstruct ten
> numbers and even a hundred for Proto-Indo-European and
> yet, being that the speakers of the language were probably
> mostly pastoral, they weren't exactly far removed from
> hunter-gatherer lifestyles themselves.

On the contrary, there's an enormous difference between
hunter-gatherer and pastoral lifestyles.

>> Look at 19th century Australian aboriginal languages.
>> Almost nowhere is anything but a 2-4 numeral system
>> attested (compared to 300 or so cases to the contrary).

> And perhaps, I could cede, that in a majority of cases
> hunter-gatherer societies have few numbers at their
> disposal, but we cannot say logically that it's not
> possible.

As I recall, that wasn't the question. Rather, the question
was whether a 2-6 system was at all plausible.

Brian