Hallo conlangers!

On 12.06.2015 11:24, R A Brown wrote:

> With respect, I think we should remind ourselves that
> _artlang_ derives from "art-language", a term used by J.R.R.
> Tolkien in his essay "A Secret Vice" where he says an
> artlang is not concerned with "base considerations of the
> 'practical', the easiest for the 'modern mind', or for the
> million - only a question of taste, a satisfaction of
> personal pleasure, a private sense of fitness."
>
> In other words, an artlang is a work of art. It is,
> therefore, IMO quite barmy to say that an artlang must be
> derived in this way or that.  All we can do is give pointers.

Sure!  Not all artlangs are diachronic and naturalistic!  It is just 
that the *majority* of artlangs are *fictional* language, and the 
*majority* of fictional languages are naturalistic, and that the 
diachronic method is considered the *best* (but not the *only*) method 
for arriving at a good naturalistic artlang.

> On 11/06/2015 20:06, Mark J. Reed wrote:
>> DP=David Peterson RAB=R.A. Brown
>>
>> DP>This is the ultimate failing of the diachronic
>> approach
>>
>> RAB> Do you then consider that Tolkien's languages
>> ultimately failed?
>>
>> I don't want to speak for David, but "a failing" is not
>> the same as "failure". In this case, I take it to mean
>> only that the diachronic approach is limited, in that
>> even if you could perfectly simulate the evolution of a
>> language from an earlier protolinguistic phase, you still
>> aren't getting the full picture of glossogenesis, because
>> we still don't know exactly how the initial spark of
>> language forms.
>
> But with respect, this is treating the derivation of an
> artlang as something mechanical and taking it to absurd
> lengths.

Yep!  Just mechanically running word forms through a sound change 
application program won't give you a good diachronic artlang!

> I said in my previous email that the protolanguage
> need only be a sketch.  I did not say you had to have a
> fully formed protolanguage and then derive your conlang
> entirely from this by a simulated diachrony. 

Indeed, you did not.  In fact, with Hesperic, I do not work with a fully 
formed protolanguage at all.  I just have some basic ideas of what 
Proto-Hesperic was like - a phoneme inventory, some inflectional 
paradigms, a handful of roots - and when I find that I need to know more 
about Proto-Hesperic, I add it "on the fly", projecting back onto 
Proto-Hesperic what I have found out about the Hesperic language I 
currently chiefly work on, Old Albic.

> But if your
> language aims at _naturalism_ then it will have had a
> history and one should be aware of this. Once again, look at
> Tolkien's treatment of his own languages.

Yes.  He never wrote a "Complete Grammar of Primitive Quendian" or 
anything like that!  (In fact, he never wrote one of Quenya or Sindarin, 
either, it seems!)  Mark Rosenfelder is another example. He didn't start 
with Proto-Eastern, not even with Cad'inor.  He started with Verdurian, 
which he eclectically pieced together from various European languages 
(you still can tell where most of the words came from), and worked from 
there *backwards* first to Cad'inor and then to Proto-Eastern.

> Of course, in the realm of artlangs, we'll never get a full
> picture of the genesis of the language any more than we can
> ever hope to get the full picture of the genesis of a work
> of Picasso or Jackson Pollack.  Art is subjective.

But certainly so!  While Old Albic is an attempt at "re-creating" the 
language of pre-Celtic Britain, it is entirely *subjective* and based on 
my personal notions of the matter, and also on my personal notions of a 
language that befits the kind of civilization I have in mind, which in 
turn is an expression of my personal notions of the human condition and 
the meaning of life.  Other people faced with the same starting point - 
re-creating a language of pre-Celtic Britain - would surely arrive at 
something completely different, and I am in no position to declare that 
they were "wrong".

>> Much like we can only extrapolate physics backwards to a
>> few instants after the Big Bang, we can only extrapolate
>> language backward to earlier forms that already are much
>> the same as modern languages, differing only in details
>> that could just as easily be synchronic variations across
>> modern languages as diachronic ones across one's
>> history.
>
> This IMHO is irrelevant.  Tolkien's proto-Elvish has nothing
> whatever to do with the extrapolation of any natlang group
> back to its origins.

Indeed not.  He created a stand-alone language family that was not 
related to any natlang family.  (It is different with my conlang family, 
which I fancy to be remotely related to Indo-European and Uralic, so 
Proto-IE and Proto-Uralic are of some relevance to it.)

> ---------------------------------------------------------------
>
> On 12/06/2015 02:22, Jeffrey Brown wrote:
> [snip]
>>
>> Linguists definitions:
>> 1. Descriptive: The language described as the people
>> actually speak it.
>> 2. Prescriptive: The languages described as the elite
>> state it ought to be spoken.
>
> Perhaps "perfect" state.  It's the idea, as it seems to me,
> that there is some perfect (Platonic) form of the language
> to which anyone educated ought to conform.  Such ideas lead
> to people saying things like "Cicero wrote better Latin than
> Caesar" - which IMNSHO is absolute baloney.

AMEN!

>> In other words, in order to have a *descriptive"
>> description for the language, someone (like a field
>> linguist) has to visit the people and listen to them and
>>  write down how they speak.
>
> Yep.

And the author of a fictional language usually assumes the role of a 
field linguist who explores the language of his fictional people. At 
least, that is what I do with my fictional languages, and Tolkien did 
with his!  Hence, an "out of character" grammar of a fictional language 
will be *descriptive*, while an "in character" grammar may be either 
descriptive or prescriptive, depending on which stance the fictional 
writer takes to the language.

>> But, there are no speakers of a conlang before it is
>> released! You cannot write a descriptive description
>> [see note 1] because the language is still inchoate and
>> unformed.
>
> No, no.  Tolkien wrote of the "historical background (a
> necessary thing as a constructor finds in the end, both for
> the satisfactory construction of word-form, and for the
> giving of an illusion of coherence and unity to the whole)."

Yep!

> IMO if one does not have an imaginary set of _speakers_, one
> is not likely to produce a naturalistic artlang (Of course,
> not all art aims at naturalism - far from it - but here we
> are concerned, as the subject lines says, with
> _naturalistic_ forms).  The constructor IMO is to some
> extent like a field linguist discovering how his people
> actually speak.

Just that!  Whenever I make up a naturalistic artlang, I make myself a 
picture of its speakers, be it a fully fleshed conculture as with Old 
Albic, or a simple notion such as "a transhumant shepherd clan in the 
Carpathians", as with one of my side projects.  And the role I assume is 
that of a linguist studying the language of those people.

>> Conlangers definitions:
>
>> 1. Descriptive: The language described as the conlanger
>> imagines that the people would speak it.
>
> Exactly!

Seconded.

> 2, Prescriptive: The language described
>> as the conlanger imagines the elite would say it ought to
>> be spoken.
>
> An optional extra. This, I think, is not likely unless the
> language becomes that of a dominant culture which is
> considered prestigious and certain grammarians (in your
> imaginary world) try to preserve a 'perfect" form of the
> language.

Right.  I could, in theory, write a prescriptive grammar of Old Albic, 
as the imagined culture of that language is an urban civilization where 
people writing such a thing would have existed. I could hardly do so for 
a Duniscian language (the Carpathian shepherd languages I mentioned 
above), as these languages are not prestigious or dominant, but indeed 
severely endangered in their conworld (the League of Lost Languages)!

> [snip]
>>
>> Note 1: If you are a Platonist, and believe that
>> intellectual things have a real existence, and that
>> their properties can be discovered, rather than
>> invented,
>
> No, I disagree.  I do not think that Tolkien, e.g.
> considered there was a 'Form of Quenya' or a 'Form of
> Sindarin' in a Platonic universe that he was discovering.

Nor do I think so!  In fact, he constantly changed and revised his 
languages, and never arrived at anything like a "stable form", much less 
a Platonic Form!

> What he had done AIUI was to imagine his races of elves and
> men and by 'getting the feel' of the languages they spoke to
> 'discover' the forms of that language. It was an artistic
> process.

Yep.

> [snip]
>>
>> Note 2: Once there is a community of speakers of the
>> conlang, such as Klingon or Esperanto, then it is indeed
>>  possible to write a descriptive description of the
>> language showing how Esperantists or Klingophones
>> actually speak, and how that differs from the
>> prescriptive descriptions of how they ought to speak.
>
> Interesting observations.  But Esperanto was designed as an
> artificial auxlang.  Klingon was intended not to resemble an
> naturalistic earth language, but to be an exolang.  So tho
> both have now gained a population of speakers, we have here
> a somewhat different phenomenon than that which this thread
> AIUI is about.

Yes!  Esperanto is not a fictional language.  Klingon *is* a fictional 
language, but I would not call it "naturalistic", and its real-world 
speaker community is wholly distinct from its fictional speaker community!

--
... 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