Hallo!

On Mon, 30 Mar 2009 08:13:43 +0100, R A Brown wrote:

> Jörg Rhiemeier wrote:
> > Hallo!
> > 
> > On Sun, 29 Mar 2009 18:18:29 +0100, R A Brown wrote:
> [snip]
> > 
> >>       Certainly there is no  
> >> doubt one can define an 'oligosynthetic' _formal_ language.
> > 
> > Yes.  Formal languages have little to do with human languages,
> > if anything at all.  For one point, the definition of a formal
> > language does not say anything at all about *meanings*.  
> 
> Yet, *meaning* is surely fundamental to what is usually understood to be 
> an 'oligosynthetic' language, isn't it?

It is.  The whole notion of "morpheme" depends on the concept of
"meaning": morphemes are the smalles meaningful elements of a
language.  If there is no way of breaking it up without losing
its meaning entirely, it is a morpheme.  If you have just a set
of meaningless strings generated by a Turing machine or whatever,
you don't have morphemes, and cannot meaningfully apply the term
"oligosynthetic" to it.

> I've always been a little puzzled why, in fact, the term 
> 'oligosynthetic' is in fact used for these putative languages. Surely in 
> the strict meaning of its two compounds (i.e. oligo- + synthetic) it 
> ought to be (more or less) synonymous with 'isolating' or 'analytic'? 
> Wouldn't 'oligosemantic' or 'oligomorphemic' be more descriptive?

Blame it on the dreadful Benjamin Lee Whorf.  He invented
the term.  Indeed, "oligomorphemic" would be better.

> [...]
> > Right.  Human languages, in order to have full expressive power,
> > need at least one *open* class of lexemes, i.e. one to which new
> > lexemes can be added whenever needed.  And that is precisely
> > what an oligosynthetic language is not: "oligosynthetic" means
> > that the set of lexemes is *closed*, i.e. no new lexemes can
> > be added to it.  
> 
> It seems to me significant that neither Crystal nor Trask give 
> 'oligosynthetic' as a lemma in their linguistics dictionaries.

Yes, because linguists deal only with real languages (only
few take conlangs into account), and such languages are never
oligosynthetic, "oligosynthetic" is strictly spoken not a
linguistic term, and thus not listed in dictionaries of
linguistics.

> [...]
> I see oligosynthesis working only with a community that is isolated from 
> the rest of humanity and retains a conservative world-view that 
> understands everything in terms of a closed set of semantic primes. That 
> is  possibility in an alternate history or a science fiction scenario.

That community would also have to be inhumanly hidebound
- unable to invent or imagine anything beyond their very
restricted world view.  I could imagine a race of non-sapient
beings such as _Homo erectus_ speak an oligosynthetic language,
but I find it hard to imagine a culture of _Homo sapiens_ get
along with such a language.

> But, as I have observed before, such a set of _semantic_ primes cannot 
> IMO possibly be culturally neutral. This has certainly *not* been the 
> case with any of the oligosynthetic languages invented in the past.

Yes.

... brought to you by the Weeping Elf