Hallo conlangers!

On 12.02.2016 20:44, R A Brown wrote:

> Ah well, since this thread, which I inadvertently (and with
> *no* intention of nitpicking) began, seems to be going on ...
>
> On 12/02/2016 17:12, Jeffrey Brown wrote:
>> Perhaps, it makes sense to speak of the sources of the
>> conlang, rather than using "a priori" or "a posteriori".
>>  For example, Olivier specifically says that his
>> language, Sambahsa, is derived from Indo-European
>> sources. That says a lot more than "a posteriori".
>
> Yes it does; and saying that Afrihili is derived principally
> from Swahili(of east Africa) and Akan (of Ghana) says a good
> more than saying it's "a posteriori."

Sure. "A posteriori" says so little - from *which* languages? and which 
way? Also, the diachronic method used in languages such as Brithenig is 
very different from the selection method used in languages such as 
Esperanto. Indeed, as a diachronic conlang, Brithenig has much more in 
common with "a priori" Quenya than with "a posteriori" Esperanto.

>> As for engelangs, there is always some overriding
>> principle that the conlanger has applied in the design
>> of the language; knowing that is more helpful than the
>> term "a priori".
>
> AMEN!
>
> Simply calling George Dalgarno's 'Ars Signorum', Mark Line's
> 'Classical Yiklamu' and Tolkien's Quenya "a priori" tells us
> nothing about the vocabulary except that it's not (directly)
> derived from natlangs.  Indeed, in the case of Quenya, see
> below, IMO it is misleading rather than helpful.
>
> Ars Signorum's vocabulary is derived through a process of
> thesaurus-like taxonomic classification - see:
> http://www.frathwiki.com/Ars_signorum#Word_formation

Yes, this is the method Couturat and Leau had in mind when writing of "a 
priori" languages. It is vastly different from the two methods you write 
of below, which also differ hugely from each other.

> The vocabulary of Classical Yiklamu is computer generated;
> the phonotactics of the language was fed into the computer
> so that it could (pseudo)randomly generate separate
> permissible combinations, each being associated with one of
> the 99K+ entries in WordNet.

In some ways, this is the *opposite* of the taxonomic method. While in 
Ars signorum, semantically related words have similar shapes even where 
natlangs use entirely different words (such as _dog_ and _fox_), in 
Classical Yiklamu even most closely related words (such as 'bake' and 
'baker') have utterly different shapes (don't ask me about the actual 
words, I haven't seen the Classical Yiklamu vocabulary yet).

> Quenya's vocabulary was the creation of Tolkien who wrote:
> "I am personally most interested perhaps in word-form
> itself, and in word-form in relation to meaning (so-called
> phonetic fitness) than in any other department.  Of great
> interest to me is the attempt to disentangle - if possible -
> among the elements in this predilection and in their
> association (1) the personal form from (2) the traditional.
>  The two are doubtless much interwoven ... "
>
> Where is the traditional phonetic fitness to be found?  Why,
> in natlangs!  That's is why I added _directly_ in
> parentheses above.  Clearly Tolkien's method of vocab
> generation is _quite different_ from the machine generation
> of Classical Yiklamu or the taxonomic approach of Dalgarno.

Of course it is! Quenya is a *naturalistic* language, in the sense this 
term is used among artlangers (which is *not the same* as the one in 
which is used among auxlangers!), while the other two are engelangs, 
though of very different kinds. Calling all three "a priori" says so 
little. Moreover, Quenya is a diachronic conlang, thereby in a class 
with Brithenig or Modern Gothic, even though the latter two are "a 
posteriori".

>> On Fri, Feb 12, 2016 at 12:56 AM, Pete Bleackley wrote:
>>
>>> One thing I've noticed is that the terms "a priori" and
>>> "a posteriori" have both largely fallen out of favour
>>> in the community at large.
>
> ... and IMO a good thing too.

Yep. They are part of an old classification scheme for auxlang that is 
obsolete even in that field, and never applied usefully to artlangs or 
engelangs. It is like classifying locomotives by some construction 
detail of their steam engines, which fails at diesel or electric 
locomotives ;)

--
... 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