Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site rtp47.UUCP
Path: utzoo!linus!philabs!cmcl2!seismo!harvard!think!mit-eddie!genrad!decvax!mcnc!rti-sel!rtp47!throopw
From: throopw@rtp47.UUCP (Wayne Throop)
Newsgroups: net.origins
Subject: Language
Message-ID: <54@rtp47.UUCP>
Date: Fri, 31-May-85 17:12:03 EDT
Article-I.D.: rtp47.54
Posted: Fri May 31 17:12:03 1985
Date-Received: Thu, 6-Jun-85 12:27:38 EDT
References: <1168@uwmacc.UUCP>
Organization: Data General, RTP, NC
Lines: 169

I'd like to thank Paul for reposting some articles about language that I
missed.  I've always been interested (in an amature sort of way) in
linguistics.  I do have some comments on these postings now that I've
seen them.  These comments mainly go into the relationship between
facts, assumptions, and explanations, and get most interesting (I think)
at the end of the article.  Bored readers might want to skip ahead to
Paul's "One might postulate" note and my response to that.  *Very* bored
readers have hit "n" already, so I won't worry about them. :-)

In <1168@uwmacc.UUCP> Paul DuBois
> [regarding the notion that the earliest languages are the most
>          complex]
> Baugh evidently had some doubts about this statement himself, as it was
> dropped from the second edition.  I shouldn't wonder.  The logical
> conclusion would be that the most complex languages are the most
> primitive, and the very simplest the most advanced.  This might be a
> difficult proposition to defend in view of its manifest absurdity.

Why is it manifestly absurd?  It seems quite reasonable to me, for
reasons I've gone into before.
(See the recent posting pair from me entitled "Re: Misc Language")

(I realize this is a reposting, and Paul's position may have changed in
the meantime, but if it has not, I'm curious as to why this notion is
considered manifestly absurd.)

> [excelent exposition of the details of the normal direction of
>     linguistic change, with clear references, and so on]
> An inference to be made from this, given the observed trend towards
> simplification in civilized languages, and the high degree of
> complexity of the languages of primitive cultures, is that the
> languages of the latter cultures are more closely related to their
> original languages than are the civilized ones.  This is entirely
> consistent with the hypothesis of an origin of language that was
> complex and completely functional from the start.  It is also
> inconsistent with a developmental theory of language origin positing
> increasing complexity arising from non-linguistic sources.

Again, *why* is it inconsistent?  It doesn't seem strange to me at all,
for two reasons.  1) the phase of increasing complexity (if any ) is
necessarily lost to us completely,  2) there are many examples of newly
"invented" languages which are initially complex and get simpler (see
previous postings).

Note: clearly this is not inconsistent with the view that languages
arose "instantly" at some time in the past, nor is it inconsistant with
the notion that somebody *taught* language to humans and that language
did not spontaneously arise.  However, I can't see how this is
inconsistant with the "spontaneous" explaination of language origin
either.

> It appears that languages become progressively simpler and that the
> oldest languages are the most complex.  This is all wrong for a theory
> which must have language coming from non-language, although it is
> perfectly consistent with a model positing language beginning
> full-blown from nothing.  The data also suggest that degeneration (or
> simplification) processes be must be a component of such a model.

Again, *why* is it all wrong.  I just don't see it.

> I daresay that it would not be impossible to come up with an
> evolutionary interpretation that would incorporate these facts.
> However, I doubt that the facts themselves would derive the
> interpretation.  An interpretation must be derived into which the facts
> will fit.

What, exactly, is the difference between "facts themselves ... derive
the interpretation", and "an interpretation [is] derived into which the
facts fit"?  The first case has the "facts" "derive" an
"interpretation".  I take this to mean that some actor (clearly not the
facts, I hope), creates an interpretation which does not conflict with
the facts.  This also seems to be exactly what is going on in the second
case.  What is the difference besides the possibly misleading syntax?

>            Such information as we have is difficult to explain on
> evolutionary grounds, since language must develop from non-language,
> i.e., become qualitatively different and quantitatively more complex.

A clarification: *Human* language does not have to arise from
non-language (though clearly, at some point language must (under
evolutionary assumptions) have arisen from non-language).  Human
language has only to have arisen from (depending on one's viewpoint)
proto-human language, or human proto-language.  Neither of these steps
(nor the required previous steps in less sophisticated species) is so
large as to boggle the evolutionary hypothesis.

I think that even Noam Chomsky would allow that difference between human
speech and primate sign language is not all *that* great (though from
his writings he clearly thinks there is a distinct and important
difference). (I may, of course, be mistaken on this point. :-)

> One might postulate an increase of complexity to a level higher than
> currently, followed by a period of decline, but the facts do not
> suggest this.  It is an assumption that is required by the evolutionary
> interpretation.

I'm not ceartain what is being claimed here.  "The facts" I take to mean
that older languages are more complex, and languages become simpler over
time.  I take "it" to mean that an increase in complexity (of language)
occured at some time in the past.  Now then, why is "it" an unwarranted
assumption, and in what way do "the facts" not support it?  The only
real assumption (that is not supported by the facts) is that language
was once simple.  However, the creationist position makes that same
assumption (or the related assumption that language didn't exist at all
at one time).

It seems to me that *assumptions* are (or should be) supported by facts.
Any *explaination of* these facts are not (and do not need to be)
*supported by* the facts.


One more note about language.  I think that Paul is quite correct in
stating that there is no connection between "language evolution" and
"biological evolution" except analogy.  Therefore, to say that either
one "supports" the other one is not correct.  However, I think it is
also not correct to draw the opposite conclusion: that this lack of
support for evolution is support for creation.  (Obvious disclaimer: I
do not accuse anyone of making this point explicitly.  I am trying to
steer readers away from taking this point implicitly from what Paul had
to say.)


Philosophical side issues:

It is useful to keep track of exactly what are the facts, what are the
assumptions, and what are the explanations involved in a discussion.

  - facts are things that are known by observation
  - assumptions are things that (ought to be) supported by facts, and
    are necessary (and taken as true) for explanations.
  - explanations account for facts, given that assumptions are true.

Two parallel examples (yes, the creationist one may be a parody, but I
don't know how else to illustrate my point, so feel free to correct it):

  - facts: languages now are simpler than they once were
  - evolutionist assumption: languages were at one time simple, and
        selection pressures can cause languages to get more complicated
    creationist assumption: languages were at one time simple (or
        non-existant), and more complicated languages must have been
        taught (or "magically" introduced)
  - evolutionist explaination: early humans developed
        complicated structures in response to communication needs,
        and simplified them over time
    creationist explaination: God taught language to (or "magically"
        introduced language to) early humans, and then language use
        "degenerated"


In general, the relative merits of two hypotheses are often judged by
explainatory power (how many facts can a hypothesis account for, and how
economically).  It is on the grounds of explainatory power that I prefer
evolutionary explainations over creationist ones, because (as I see it)
creationist explainations *explain fewer facts*, and they do it a the
expense of *a greater cost in newly introduced assumptions*.

(The last there might seem a little strange.  Creationists seem to make
one assumption: God.  Evolutionists make many assumptions (this arose,
there is a mechanism at this obscure point, etc).  However, I rate the
*cost* of the God assumption high, since it is invoked at all the same
points as the evolutionary assumptions, and some more besides, and is
even invoked at times to deny the validity of many things that seem to
be valid facts.  Note that the God assumption is *not* bad in and of
itself (as far as I can see), but the way it is used in creationist
explainations of observations is unconvincing.)

> Paul DuBois     {allegra,ihnp4,seismo}!uwvax!uwmacc!dubois
-- 
Wayne Throop at Data General, RTP, NC
<the-known-world>!mcnc!rti-sel!rtp47!throopw