Hallo conlangers!

On 01.06.2016 22:56, David Peterson wrote:

[in reply to two messages from And Rosta]

> After reading a number of your messages, I think I can sum up our differences as follows:
>
> (1) The sufficient description of a natlang (and also a conlang) you describe is either (a) not theoretically possible, or (b) well within the capabilities of a human being.

Certainly, it is an either-or, as both obviously cannot be true at the 
same time.

> In the case of (1b), though…
>
> (2) I don’t believe you have studied either an “authored” conlang or “pidgin” conlang, as you call them, in enough detail to make any claims about them or their descriptions.

Nor do I. I don't know how many languages, nat or con, And has studied, 
but I seem to remember himself saying that he knows only one at any 
reasonable depth: English.

> Consequently, a claim that something like Esperanto is less “deep” than a natural language leaves me cold. I don’t think you know Esperanto (not merely as a language, but as a linguistic community/phenomenon) well enough to make any claims about it, other than hypothetical ones. And I think that goes double for “authored” conlangs. Thus, I remain unconvinced.

Esperanto certainly has a shorter history than natlangs: while natlangs 
evolved over thousands of years, Esperanto is less than two centuries 
old. But natlangs do not just aggregate features over their history, 
they also lose much (which is why reconstructing ancestral languages 
such as Proto-Indo-European is so difficult). There is certainly some 
kind of limit to the complexity of a human language: beyond that limit, 
human children no longer can cope with the language and will 
inadvertably simplify it. Most conlangs are probably less complex than 
the average natlang (indeed, in IAL circles, it is the main argument for 
an artificial language and against using a widespread natlang such as 
English, that the artificial language is simpler and thus easier to 
learn [which may or may not be the case - "simple" and "easy" are 
different notions]), but when a conlang is used, it *will* aggregate 
some complexity that wasn't in the initial design.

> I also think the term “model” simply cannot be used when it comes to a created language. This is my English speaker’s intuition about the way the words “model” and “language” are used. I think a model language could be a thing, but it would be a description without any actual tokens, e.g.:
>
> Model Language A
>
> Has 7 noun cases: nominative, accusative, dative, allative, ablative, locative, vocative.
> Has 2 tenses: past, non-past.
> Has 2 aspects: perfect, imperfect.
> Has a vocabulary of 90,000 words.
>
> Etc. That’s a partial model of a potential language. You could even specify how the cases work, how the tenses work, etc., but without the material (actual phonology, actual words), it’s purely theoretical. I’ll grant there’s a possible definition of model that could apply to conlangs (as a technical word, I’m sure “model” has lots of very specific uses in disciplines I have no knowledge of), but as an English speaker, it doesn’t pass the smell test.

Yes, that could be called a "model" of a language, and certainly is not 
a language - it is a *plan* of a language. Many of us begin with such a 
plan, and then build their language by filling it in with material.

> As a small side note, I also agree with Christophe (though this is purely opinion, in my case) that language is likely not something where it’s possible to write down every single rule of it. I suspect that’s a misunderstanding of the nature of the object of study. Though perhaps one day it will be proved that there actually is no free will, and every single action every human and creature takes can be predicted exactly. Kind of a depressing future, but if that comes to pass, then surely language will be a part of it.

The debate about free will is something better not touched here.

On 02.06.2016 11:08, And Rosta wrote:

> On 1 June 2016 at 20:00, Jörg Rhiemeier <[log in to unmask]> wrote:
>
>> Hallo conlangers!
>>
>> [...]
>> Fair. Many conlangs are not sufficiently developed to be actually used -
>> but unless the design is flawed (or contains ideas that hinder its
>> usability on purpose, such as Fith), such an insufficiently developed
>> conlang has the potential of being expanded such that it becomes a usable
>> language. Insofar, a conlang cannot be aptly compared to a scale model of a
>> building, which will never be usable as a building.
>>
> A scale model of a cathedral might be usable as a garden shed...

OK, but that's besides the point. It still doesn't function as a cathedral.

> I suppose
> one could analogize that to conlangs by saying that even a most
> insufficiently developed conlang can be used for rudimentary communication.

Is there anything that is not usable for *anything*? The 17th-century 
"philosophical" languages are often (and IMHO justly) held to be 
unusable as languages, but they may perhaps be useful to librarians as 
classification schemes for books.

> But a more pertinent analogy would be modelling a cathedral by just
> building its facade, so that seen from outside (from a certain perspective)
> it looks like a cathedral, but to use it as a cathedral you have to go and
> build the rest of the cathedral.

This is perhaps a better analogy of a conlang than the scale model. An 
incomplete conlang gives an *image* of a language, but there is too much 
missing to actually use it. Yet, if the conlang is well-designed, the 
missing parts may be added later, just as the missing parts of the 
cathedral can be added later to its hollow facade.

>> Now that I think of it, I no longer take a middling position between
>> Padraic (who considers use - either actual use or at least potential
>> usability - to be a major criterion in conlanging) and And (who sees
>> conlangs as scale models for which usability is not an issue), but side
>> more or less with Padraic (though I'd rather say that potential usability
>> matters more than actual use).
>>
> This completely misrepresents both Padraic and me: it may describe your own
> views, but it wholly fails to accurately describe Padraic's (I think) and
> mine. Padraic feels that, according to his aesthetic, the highest form of
> conlang artistry models not only langue but also parole. Completely
> orthogonally, I think that the existence of (nonfictional) parole is not
> criterial for Full languagehood. Neither Padraic (afaik) nor I had opined
> about usability.

I apologize for misrepresenting both of you (if I actually did). But I 
don't see how Padraic "did not opine about usability" when he pointed 
out the importance of _parole_ - writing texts in a conlang certainly is 
a use of the conlang, and requires at least a certain degree of 
usability. But it is true that actual use is not criterial. What I feel 
necessary for a conlang to be well-designed, though, is that it *can* be 
used, or at least developed further to a usable state.

> And without wanting to add to the "Espe-rants" in this thread that indeed
>> belong not here but on AUXLANG,
> That is very unfair to Christa, who wrote ardently about what Esperanto
> means to her; and we should call a stop to this habit of vocally wishing
> the merest mention (or is it just the merest impassioned mention?) of any
> auxlang to be banished to Auxlang.

The way I perceived this debate, it bordered on auxlang policy. Your 
mileage may vary; I apologize to anybody who may feel that I was unfair 
to them.

--
... brought to you by the Weeping Elf
http://www.joerg-rhiemeier.de/Conlang/index.html
"Bêsel asa Éam, a Éam atha cvanthal a cvanth atha Éamal." - SiM 1:1