Re: Sapir-WhorFreakiness

John Cowan said:
> Mark P. Line scripsit:
>
>> A genetic defect is one possibility, I suppose. Perhaps something very
>> odd
>> happened during the evolution of this language. (I dunno, pidginization
>> followed by stunted creolization, with lexifiers and substrates all lost
>> in the meantime. Or something. *shrug*)
>
> Well, that itself would be informative: it would bash whatever remains
> of the Bickerton bioprogram into a bloody pulp.
Absolutely. And it couldn't happen to a nicer bioprogram.


>> Alternatively, maybe there is an additional cultural constraint that
>> cannot, due to its nature, be discovered by outsiders: "Don't talk
>> straight with outsiders."
>
> That was my first thought.  But how could they be so consistent,
> including even the children?  I'd rather believe in a pervasive
> genetic defect than a pervasive conspiracy.
There is no conspiracy underlying the refusal of Americans (including
American children) to eat horse meat.


>> 2. Never mention color.
>
> This is the one of your six rules I just can't swallow.  How would they
> know how important basic color terms are to us?
Perhaps by having talked to neighboring communities who had been inundated
by Western researchers whose sole interest was in names for colors. (That
presupposes that there are different grades of outsider, though.)

I know the baby-talk-with-outsiders explanation is a stretch, and it's
only my cynicism that makes me want it to be true: I have a long-standing
beef with linguistic fieldworkers who drop themselves into a village for
two years and then consider themselves as having done very nearly
exhaustive research on the language. (I have a similar beef with foreign
language teachers who lead kids to believe that they've learned "all the
grammar rules" of a language like French or Spanish in two years of high
school.)

To be honest, my money is on the contact-weirdness scenario. It would be
fun to set up an agent-based model to see what constraints and extrinsic
events might get it to evolve anything like Piraha, starting from two or
more languages in contact that are more clearly within the usual
evolutionary space. Of course, it would be quite a project to build a
reasonably coherent model even for a clear case of pidginization. My plan
would be to use complete grammars of conlangs rather than incomplete
grammars of natlangs in any such model (at least in the first instance),
so if there are any conlangers interested in collaboration on this, we
might be able to engage in some synergetic puttering.


-- Mark

Reply