Re: glossogenesis (was: Indo-European question)

An all-in-one reply to several messages, not chronologically.

On Sun, 24 Jun 2001 01:10:25 -0400, John Cowan <cowan@...>
wrote:

>Andreas Johansson scripsit:
>
>> I think that the term "Proto-World" implies that we're talking about a
>> language stage BEFORE any branching etc had taken place.
>
>Not necessarily.  It is merely the hypothetical common ancestor of all
>*extant* languages.
Yes, I like John's definition better. Kinda more practical: the system
which comparative linguists can attempt at reconstructing some day.

I also prefer to think it was already "real human"; but the latter isn't
mandatory for my reasoning: it's enough to assume that a previous stage
had *some* grammar - for the evolution to start playing with...

On Fri, 22 Jun 2001 23:32:33 -0700, Tommie L Powell <tommiepowell@...>
wrote:

>Thanks for being quite clear about the time frame that you have
>in mind for "Proto-World" language -- namely, that you mean
>language as spoken before neolithic times began. (Neolithic
>times began approximately 40,000 years ago.)
Wait... I thought it happened later, some 10 millennia ago. Where do
you take this figure from? It looks more like the dating for earliest
H. sapiens sapiens finds (if "non-specialized Neanderthalians" are
not included in H. s. s.). Or am I missing any recent discoveries?

>There's plenty evidence that the neolithic revolution -- with its
>myriad technological innovations and population explosion --
>wasn't due to any biological change in our species, and that
>our brains have almost certainly been capable of  all linguistic
>tricks ever since our subspecies -- homo sapiens sapiens --
>began (about 150,000 years ago).
Which looks like the alternative age of our (sub)species - also
acceptable for me, but I'm not sure if generally accepted.

>So I basically agree with
>your reasoning.
Precipitating it, I'd say my point was close to the opposite of
yours ;)

Namely, that langs started to evolve (and split) *before* they
became "real human" langs (and their speakers - "real humans")...

> But I'd augment it, as follows:
>
>A big change in our lifestyle began about 40,000 years ago,
>and directly caused the neolithic revolution.  Until then, our
>largest social unit was the band: A band consisted of from
>50 to 150 people, who lived together and followed a yearly
>circuit of migration to gather food sources.  They had to,
>because no locality could feed so many people year-round,
>and because different foods became plentiful at different
>localities at different times of the year.
I'm afraid I don't quite ubderstand this. Certain subarctic
peoples live (or lived recently) in groups of about that size
or even smaller. Why call them 'bands'?

Seasonal migrations are also typical of some of such peoples
(e. g. all the deer-breeding ones). This doesn't prevent their
dialects from diverging. In fact, low population densities
seem to force their dialects to split even faster than in
more populated areas (I mean e. g. the dialectal diversity of
Khanty, Mansi, Saami, Evenki... - with remarkably few speakers
of many dissimilar dialects)

I don't think scarcity of food was more critical with Stone Agers
than with modern(/recent) peoples of the North.

(long snip, to which my doubts about the datings apply as well...)

> So we couldn't have spoken distinct
>languages before tribes formed: We could only have
>spoken a language that varied slightly and gradually within
>any span of a couple hundred miles (though it may well
>have varied too much across any thousand-mile span to
>be readily recognizable as the "same" language).
With these estimates - just count how many mutually incomprehensible
langs could fit in, say, southern Europe alone ;)

>Tribalization changed that: Since each member of a tribe
>spent his whole life speaking only with other members of
>his tribe, each tribe could develop a language of its own.
>And such a language could be wonderfully complex,
>because its speakers didn't have to talk with people whose
>linguistic tricks were slightly different (but different enough
>to create confusion if one tried to express a thought in a
>fairly complex manner).
So do you say that langs can't change before they've split?

(and if they can, I don't see why they can't become complex)

On Sat, 23 Jun 2001 15:40:03 -0700, Tommie L Powell <tommiepowell@...>
wrote:

>But language could not have originated with our subspecies:
Aha, now I see one point where we do agree... Indeed, if we don't
restrict the concept of 'language' to its "real human" state...

<...>
>Tribalization just insured that you'd spend your life talking
>only with people who knew and used all the same linguistic
>tricks as you, and hence that a tribe could develop a much
>more complex language than had previously been practical.
Are you saying that "difficult" features couldn't spread? Why?

Just consider the list of today's most widespread langs. Among top
10 (and candidates whose inclusion in top 10 depends on how you
define 'language') we see Russian and Arabic; also German, Spanish
and Portuguese (do you think verbal paradigms of the latter two are
especially simple?); only English and Chinese are relatively "easy";
French and Hindi/Urdu seem to be somewhere in between, and I'm not sure
where to put Japanese, Bengali and Indonesian...

On Sat, 23 Jun 2001 14:22:11 -0400, Andreas Johansson <and_yo@...>
wrote:
<...>
> I think you're probably right, but
>it does still not really address the question whether inflection etc has
>been around as long as content morphemes*.
I believe content morphemes are older. I think signals used by animals
can be roughly identified with our content morphemes (if such signals
can be compared with language at all), but by no means with our
grammatical morphemes. Which OTOH has no relation to the question of
Proto-World: my reasoning was exactly about Pre-Proto-World having had
many millennia of its own evolution behind it (after grammar as such first
emerged).

>I still find it rather natural to assume that things like number, tense and
>mood haven't been arround quite as long as nouns and verbs. After-all, you
>won't need, say, tense until you've introduced verbs.
AFAIK the noun vs. verb thing is within the reach of chimps using a sign
lang. Am I wrong? At any rate, I tried to argue that the last addition(s)
that made our language "real human" must've affected less fundamental
layers.

On Sun, 24 Jun 2001 01:56:11 -0700, Tommie L Powell <tommiepowell@...>
wrote:

> any tribal language embodies a large
>base of sophisticated techniques in its grammar.  Such a vocabulary,
>and such a grammar, surely takes many generations to build up.
How many generations, to be more specific?

Recently I read a couple publications about English-based creoles of (The)
Surinam(e). I wouldn't say their "techniques" appear too transparent
(you know, morphologically conditioned tonal sandhi, 'embedded nasality'
and suchlike...).

And these creoles are less than 5 centuries old. Not very impressive when
you're manipulating dozens of millennia... ;)

Thanks for interesting discussion!


Basilius