At 4:26 am +0000 18/10/98, Tommie Powell wrote:
>Nik Taylor wrote:
>> Tommie Powell wrote:
>> > And I don't think
>> > we should ignore them, when the only natural languages that resemble
>> > programming languages are languages of Stone Age people.
>> I'm having a hard time believing this.  Could you point me to the
>> research that leads to your characterizing these languages as resembling

>Sure, Nik.  Such languages are very easy to spot: Just look for an example
>of how a
>language actually says something -- anything -- and if that expression is
>as a one-word sentence, you're almost certainly looking at what I call a
>Stone Age

With respect - rubbish.

What you're looking at is a language with POLYSYNTHETIC structure.  I know
of NO evidence whatever that such languages are "Stone Age".

Indeed, it has often been observed, rightly IMHO, that although the
_written_ French pretends that the language is an inflexional one that's
hardly changed since the 13th century, the modern _spoken_ language is
essentially polysynthetic.  But I don't observe that French society has
been moving towards the Stone Age since the 13th century!
>I found an example on the Internet today, at
>The example is a Cheyenne "word" 18 syllables long, and means "I truly do not
>pronounce Cheyenne well."  The reason it's a single "word" is that you
>cannot break
>its "morphenes" apart and rearrange them to express that thought in any
>other way.

There is no need to put 'word' or 'morpheme' in double quotes.  18 syllable
long words are not exactly uncommon and the morphemes _are_ morphemes; it's
just that most morphemes are bound in polysynthetic languages.

>na=I, ohke=regularly, saa=not (first half), oneseome=truly, peheve=good/well,
>tsehest=Cheyenne, o'ane=pronounce, he=not (last half).
>The first half and last half of "not" ("saa" and "he") are almost
>certainly in the
>string of syllables that dictate the type of expression,

First half & second half of 'not' separated by a string of morphemes is
hardly "Stone Age" or unknown over here in the old world, e.g.
French:  ne..........pas
Welsh:   ni(d)/ <initial mutation>........ddim
Breton:  ne.......ket

>o-type morphene (pronounce) in the third blank.  Then the speaker has the
>option of
>inserting the other morphenes where he/she did.

The pro-complements in a verbal string in modern French are all bound
morphemes and come in a very fixed order; the speaker has the option of
inserting the bound morphemes in the required places.   That neither makes
modern French a "Stone Age" language nor 'computer-like'.

Indeed, many linguists now prefer to treat the constructions found in
"polysynthetic languages" as a complex of agglutinative & fusional

That speakers of mutually incomprehensible polynthetic languages should
adopt an more isolating structure as they derive a trade pidgin is hardly
surprising.  Indeed, it'd be very surprised if it were not so.  That
doesn't mean that the pidgin is "more sloppy" - just has a different

Of course, it must be remembered that terms such as 'isolating', 'fusional'
and 'agglutinative' show typological _tendencies_ rather than describe
individual languages.  Most natlangs seems to show varying aspects of these

Personally, I still have little doubt whatever that if I were to be
transported back 4000 years or more I'd still find a variety of isolating,
fusional & agglutinatives complexes in a wide variety of 'mixes'.