Re: Error rate, Circumlocution, and Cappucino

Hi all,

On Mon, 26 Sep 2005, Paul Bennett wrote:
> Am I alone in hating this kind of linguistic journalism completely lacking
> any kind of linguistic know-how?
>
> http://news.bbc.co.uk/1/hi/magazine/4248494.stm
No, I love it!  At least they're talking about language, getting people to
perhaps think outside the linguistic square ...

Of course, their Malay example is back-to -front!  :-)  Wouldn't expect
too much accuracy from the BBC, now would you?

To be fair, let me quote:
'These fabulous examples have been collected by author Adam Jacot de Boinod into The
Meaning Of Tingo - a collection of words and phrases from around the world.

"What I'm really trying to do is celebrate the joy of foreign words (in a totally
unjudgmental way) and say that while English is a great language, one shouldn't
be surprised there are many others having, as they do, words with no English
equivalent," he says.'

I think this aim is totally unexceptionable, and can't see
that getting more people interested in how language works
could possibly harm either the science of linguistics nor
our ability to conlang.  I also don't understand why you
seem to believe that a journalist need to be an expert in
order to write a review, for a general audience, of a book
intended for a general audience.

"... the joy of foreign words ... " ... and their utility, too!
How, without these foreign words, could the news editor
have ever composed the banner headline:
   "SEPOY RUNS AMOK; BERSERKER IN BARRACKS"?

Well, OK, "RUNS" is arguably English ... :-)


...[snipt]

> What can you say about the acceptable error rate within your conlang(s)?
> Does it easily tolerate sloppy grammar, or unusual accents, or poor
> articulation?
Well, I've only just begun ... but!

Here's my thought - every natlang that I know of permits many different
ways to say much the same thing.  Some aspects (!) of grammar may be
quite forgiving, yet others are almost inviolable.  Agreement of case, gender,
number and tense are essential in most IE languages, but almost entirely
optional (where expressible) in most Austronesian languages.  These have a
strong tendency to express meaning thru word order.  Therefore, it's my
guess that most natlangs have evolved and preserved distinctions which
were at one time (and may still be) important to their speakers.

To the extent that a conlang is naturalistic - that is, models a natlang -
it would want to tolerate some kinds of error much more readily than
others.  Building redundancy into the conlang would seem a sensible
approach - after all, we humans tolerate low-level inconsistencies much
better than high-level ones, so the occasional slip in say, number agreement
or tense should not totally destroy the meaning of an utterance.  For
example, if a speaker has been referring to a plural referent, such as
"les maisons", giving attributes such as "grandes" and "oranges", a human
listener could readily forgive calling them also "orgeilleuse" in the singular
instead of "orgeilleuses" in the plural.  Perhaps the listener would be more
nonplussed by them also changing gender from feminine to masculine, if
the speaker said instead "orgeilleux".

Any natlang that has absorbed many speakers of different origins will,
I think, has perforce adapted itself to the sound distinctions they were
accustomed to make in their original languages.  Accordingly, it may have
adopted some sounds that were once foreign to it, or it may have
gradually assimilated one sound to another, as the pair lost its distinctive
force.  For example, in English we now pronounce "pair" and "pear"
identically, yet the spelling tells us it was not always so.  My ideal conlang
would have few rather than many distinct phonemes, to more readily
accommodate speakers of different origins and abilities.  There's no point
in having a distinction between a trilled and a flapped /r/ if half the
population is unable to produce the trill.  Each phoneme, and again, each of
its allophones, would have a broad range of acceptable articulations.  An
example of this in a natlang is Malay, which has essentially three vowels
/a/, {/i/, /e/}, and {/u/,/o/}, despite its orthographic five: a, e ,i ,o ,u.
You could call for "Porridge!" in Malay by yelling "Bubor!", "Bubur!",
"Bobor!" or "Bobur!" - where the vowels u and o have approximately the
same values as in Italian or Spanish.  Or you could use a vowel intermediate
between the two.  The difference between all these forms of the {/u/,/o/}
vowel is not phonemic; the phoneme has wide range of acceptable
articulations.  This to me is desirable in a conlang that will be a lingua
franca.  However, I would expect a more isolated society and language to
distinguish between articulations in ways that others find difficult.
Most Europeans find four or more of the Arabic consonants very difficult
to learn to pronounce; yet almost everyone everywhere has trouble learning
to pronounce almost every consonant in the languages of the Kalahari
Bushmen.

In short - I think a trade language should be much more accommodating
than the language of an enclave, or indeed of a priestly sect.  The language
should fit the clime ...


In his reply to you, I note that Doug Dee makes the point that Esperanto,
among others, makes important distinctions with minimal means, often just
changing one vowel.  This was one of the reasons I gave up on Esperanto
long ago, as being unsuited to we frail, error-prone folk.  I don't think I'd
want to make a trade language use any less than two phonemic changes per
morphemic distinction.  For example, I'd contrast -am with -in, but not with
-im or -an.  But such precision and subtlety might be useful in a priestly or
courtly language, readily differentiating the initiates from the hoi polloi.

Which leads me to:
...[snipt ... on Circumlocution]

Interestingly, the extinct initiation language of Mornington Island, Damin,
used many fewer morphemes than the parent language.  Most of the parent
language's words were strictly taboo.  This required a greater level of
abstract thinking, and more circumlocution, than the parent language.  It
also required, I think, great self-control to speak it correctly, since the
grammar was essentially that of the parent, but the grammar's rules were
applied to a highly restricted vocabulary.  The teaching of self-control, so
necessary in cooperative hunting, was a prime goal of most Australian
initiation practices.

Perhaps, when I finally build a conlang, it too will have a special-purpose
ritual dialect (or two) with little tolerance of - but great scope for -
error.  The heightened danger of breaking a taboo should serve to induce
fear - or is that a proper awe of the supernatural? - in the neophyte.  It
would certainly mark the supernatural as entirely distinct from the
profane world in which the neophyte had grown up.  Having distinct
languages for "women's business" and "men's business", as many Austra-
lian languages do, would serve to further define the social structure.

Regards,
Yahya


--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.11.8/113 - Release Date: 27/9/05