On Sat, Jul 30, 2011 at 3:05 PM, R A Brown <[log in to unmask]> wrote:
> On 30/07/2011 17:12, Jim Henry wrote:
>> On Sat, Jul 30, 2011 at 5:34 AM, David
>> Peterson<[log in to unmask]>  wrote:
>>> Rather reminds me of the Esperanto question. That is,

>> It's not a natlang per se, but it's more natlangy than
>> any other conlang;

> Really? Many of its features seem very artificial to me;
> whereas languages like Quenya, Sindarin, Tepa and some

Esperanto is one of only a handful of conlangs where we can ask "how
does it work?" and not be implicitly asking "how did its creator
intend it to work?" -- we're asking a question that has an empirical
answer that can be arrived at pretty much the way linguists study
natlangs, by statistically analyzing corpora or asking a fluent
speaker whether a given utterance is well-formed.  And given the size
and age of its speaker community, it has this natlangy property in a
greater degree than Toki Pona, Lojban, etc., probably somewhat more
even than Interlingua or Ido, though they all have it to some degree
as well.

Of course most fluent speakers of Esperanto (or other conlangs) are
more or less consciously influenced by models of good usage, some of
which may even overlap with the set of documents included in the
_Fundamento_; but even with many natlangs you sometimes have
confounding factors of that type.  Is your informant's judgment of an
utterance's well-formedness purely based on their language intuition,
or also on what they've learned in school?

This property, of being defined by usage rather than standards
documents, is a matter of degree, like the other properties we talked
about recently.  Esperanto has it in a greater degree than Lojban,
French spoken by educated people has it in a greater degree than
Esperanto, natlangs with no academy or written standard have it in a
greater degree than formal French.

Many artlangs are more naturalistic than Esperanto in every other
respect, but less so in this one, because there's so little usage that
in most cases you have to look at the grammar specification rather
than the corpus for answers to linguistic questions, and if you ask
the most fluent speaker a question which neither the corpus nor the
grammar documents can answer, they're as likely as not to say "I
haven't figured that out yet".

Jim Henry