Hallo conlangers!

On Sunday 04 December 2011 17:29:11 Michael Everson wrote:

> On 2 Dec 2011, at 08:14, R A Brown wrote:
> [...]
> > But IME it is also unusual to classify auxlangs as Indo-European.
> 
> I suppose it depends on the features and the choice of the classifier. I
> classified Esperanto as Indo-European in my Alphabets of Europe project a
> decade or so ago. :-)

Hmmm.  Esperanto certainly has Indo-European origins.  It is
built from material from several Indo-European languages, and
its grammar conforms closely to the "Standard Average European"
linguistic area which has no non-Indo-European languages among
its core members; yet, many Indo-European languages remain
outside the area, and Finnish, Estonian and Hungarian have some
SAE traits, though they are hardly "typical" SAE languages.

Also, Esperanto is *built from* Indo-European material and not
*diachronically evolved* from Proto-Indo-European.  I have
likened the difference between diachronic development from and
construction out of Indo-European substance to that between a
branch of a tree and a piece of furniture crafted out of wood
from the same tree, though this comparison is rather inept,
I have to admit.

Therefore, I feel more comfortable with categorizing Esperanto
as a Standard Average European language (which is defined by
the - synchronic - presence of certain structural features) than
as an Indo-European language (which is defined by diachronic
origin).
 
> > I wrote that Volapük is a rather quirky _Euroclone_, which is somewhat
> > different.  That term has long been used in conlang circles to denote
> > any conlang, whether intended as an auxlang or not, which exhibits the
> > characteristics of 'standard average European' (SAE) natlangs.
> 
> "Characteristics" is the operative word, I suppose.

Yes.  But SAE characteristics do not make a language Indo-European!
It seems that you are confounding "Indo-European" and "Standard
Average European" here.  They are not the same.

Imagine a pre-Indo-European language surviving in a remote valley
of the Alps which is an isolate like Basque, but unlike Basque has
converged towards the Standard Average European type, such that it
has all the characteristic features of the SAE area, and works
exactly like German or French.  Would such a language be an Indo-
European language?  No.

Now imagine a group of Indo-European speakers wandering even
farther east than the Tocharians.  Their language gradually takes
up typological characteristics of Chinese, becomes isolating,
monosyllabic and tonal, a structure unlike any other known Indo-
European language; yet, the majority of its vocabulary is inherited
from Proto-Indo-European, only changed a lot by dozens of regular
sound changes.  Would such a language still be Indo-European?  Yes.

"Indo-European" is *not* defined by any set of typological
characteristics, it is defined by the *descent of the language
from Proto-Indo-European*.  I know that some 20th-century linguists
tried to define "Indo-European" by synchronic typological
characteristics - they remained a minority, and just about every
academic linguist worth his stripes today would consider such
attempts fallacious.

Esperanto, Novial and other European-based a posteriori IALs
have no diachronic development in the proper sense (other than
what happened to them in the course of their use *after* they
were released to the public by their designers, but that is not
what matters here).  Their words certainly do have origins in
Indo-European languages (and hence you could compile an
etymological dictionary of Esperanto (and it has certainly been
done)), but one can say that, by the standards of historical
linguistics, 100% of them are loanwords.  They thus do not make
these IALs Indo-European languages any more than the many English
words borrowed from Norman French make English a Romance language.

At least, that is the common viewpoint in linguistics, and it
is followed by Wikipedia.

> > The term Euroclone is sometimes considered pejorative;
> 
> It seems to me to be somewhat dismissive.

It is.  The word "clone" has negative connotations, especially
that of non-originality.  But such dismissivity is not what I
intended.  I simply used it in the meaning "Standard Average
European conlang".
 
> > but it rather depends on what the author intends.
> 
> Well, yes.

Sure.
 
> On 2 Dec 2011, at 11:29, Philip Newton wrote:
> > I thought the grammar of Tok Pisin was essentially Austronesian? That
> > would make it no more Indo-European than English is Romance.
> 
> It depends what you're looking at. "French and English are particularly
> close, since English, through extensive borrowing, is typologically closer
> to French than to other Germanic languages." But if Tok Pisin's is
> "English creole" and its etymologies lead to English being identified as
> its proto-language, and English is Indo-European...

Still, it is not customary classified as Indo-European, due to
the special circumstances of it coming about.  Unlike English,
which was handed down as an L1 from generation to generation
and gradually changed during this relay from what it was many
centuries ago to what it now is (and if you compare Modern
English with the current scholarly reconstruction of PIE
- which can be considered reasonably, though not perfectly,
accurate - you will see that the language has changed *a lot*
in almost every regard), Tok Pisin was started by people who
spoke different languages as their L1 and pieced it together
from fragments of English and other languages they had picked up
later in their lives.  There is no true diachronic continuity
- which is what makes a language a member of this or that family -
from Early Modern English to Tok Pisin, hence, Tok Pisin is not
considered a regular descendant of Early Modern English (or any
other stage of English), but an "English-based creole".

Now, if creoles are not considered lineal descendants of their
base languages, it would be perverse to ascribe the family
status denied to them to a posteriori conlangs!
 
> On 2 Dec 2011, at 14:06, Jörg Rhiemeier wrote:
> > Pidgins and creoles are usually placed in a class of their own (at least,
> > Wikipedia does so, and most other sources I am aware of), not in the
> > families of their base languages.
> 
> Wikipedia classes Tok Pisin and lots of other languages as "English
> creole". http://en.wikipedia.org/wiki/English-based_creole_languages

Which is not the same as classifying them as Germanic languages
descended from English.
 
> >> But IME it is also unusual to classify auxlangs as Indo-European.
> > 
> > Indeed.  Artificial languages are not included in natlang families.
> 
> By whom? I can think of no reason no to consider Interlingua to be both
> Romance and Indo-European.

Interlingua comes close to being a Romance language in being
based on Romance languages and modelled closely to them.  Yet,
there is no unbroken diachronic continuity from Vulgar Latin
to Interlingua.  Interlingua is not the result of gradual
diachronic changes affecting the language of a community that
once spoke Vulgar Latin, hence it is not classified as Romance
nor as Indo-European.

> > While languages such as Esperanto or Novial are made up of Indo-European
> > material exclusively, they do not descend from any IE language in a
> > "natural" way, and thus are not considered belonging to that family. 
> > Like pidgins and creoles, they are placed in a category of their own.
> 
> I don't agree. Both Slavic and Gothic translations of the Bible reflect
> Greek participles very exactly; it is debatable as to whether this usage
> was a natural part of Slavic and Gothic before those translations were
> made, though certainly in modern Russian the use of participles is very
> widespread. Might not such innovations by translators be considered
> language engineering?

To a degree, yes.  Yet, the Slavic and Gothic Bible translators
did not *invent* the languages they translated into, they merely
tried to find a balance between the grammatical rules of their
target languages and the (understandable) desire to stay as close
to the authoritative text they were translating as they could, and
thus came up with usages that would have been considered more or
less odd by the native speaker communities of the target languages
(but after the translation, would soon be considered good style
due to the authority of the Bible).  But certainly, Old Church
Slavic and Wulfilan Gothic aren't conlangs in the way Interlingua
is!  (Of course, this touches on the question where to draw the
line between "language planning" and "conlanging".)

> > Only diachronic conlangs based on natlangs, such as Brithenig, could be
> > said to belong to the base language's family, but even that is not the
> > usual practice in linguistics; for instance, Wikipedia classifies
> > Brithenig as artificial, not as Indo-European.
> 
> I'd say that the Wikipedia is wrong here.

Not really; it distinguishes between facts about the real world
and facts about fictional settings.  Brithenig is fictional; in
this world at least, there is no chain of L1 speakers of Vulgar
Latin that gradually mutates into Brithenig.  There was not even
an *idea* of this language until Andrew Smith started wondering
in the 1990s, "What would a Romance language of Britain be like,
if there had been any?", and proceeded to make up his language.

Wikipedia would be wrong if it presented Brithenig as a real
Romance language spoken in Britain, because that is not what it
is.  It is a Romance language spoken in a *fictional* version of
Britain, and that is not the same.  I think the classification
of Brithenig as a "fictional Romance language" fits the bill, but
don't drop the adjective "fictional".  The distinction Wikipedia
makes here is the same as that which it makes between real people
and fictional characters.
 
> > Unless it evolves non-SAE features, of course, but that is hard to
> > achieve in a plausible way.  Brithenig at least has initial mutations,
> > which are a non-SAE feature, but its overall structure is still pretty
> > much SAE.
> 
> Initial consonant mutation is normal in standard Tuscan Italian. See
> http://en.wikipedia.org/wiki/Tuscan_gorgia . Indeed it is normal in
> Spanish, where one has "vaca" [baka] and "la vaca" [la βaka]. No, those
> haven't been grammaticalized in the same way that they have been in
> Celtic, but it's still mutation.

Yes.  The Insular Celtic languages' initial mutations started
just that way.  The Insular Celtic languages only went a step
further and grammaticalized them.  Hence, it is not implausible
that a Romance language in Britain evolves this feature, too.
 
> >> In the auxlang world some consider that being a Euroclone is a flaw in
> >> that such a language is not truly neutral hence, for example, a_priori
> >> auxlangs.
> > 
> > Yes.  A common criticism of Esperanto and all the other "classic"
> > auxlangs is that they draw exclusively on European languages and are
> > thus eurocentric.
> 
> What's wrong with that? Perhaps evidently speakers of Sino-Tibetan aren't
> much interested in conlangs and auxlangs. What are the analogues to
> Esperanto, really, in non-Indo-European? Trade languages like Chinook or
> Swahili?

There is interest in IALs outside the western world.  Afrihili
and Frater come to my mind - both, AFAIK, not being euroclones,
even if Frater draws its vocabulary mostly from Latin.
 
> On 3 Dec 2011, at 15:42, R A Brown wrote:
> > While probably Novial is of 100% IE material, Esperanto isn't.
> 
> Nor is English, which has loanwords from most language families.

Surely, English has loanwords from many non-IE languages, but
its lexical and grammatical core is Indo-European, more precisely,
Germanic.  What matters for the classification of English is the
diachronic continuity from Proto-Indo-European via Proto-Germanic
to English.  No such continuity exists for conlangs.
 
> > Yes, and IMO it is very questionable that a Romance language derived from
> > the Vulgar Latin of Romano-Britain would have evolved initial mutations
> > in the Welsh manner. But that's another matter.
> 
> See above.
> 
> Anyway, I think that it is not axiomatic that constructed languages
> "should" not be considered to be members of language families. Certainly
> etymology counts.

Etymology counts for the distinction between a priori and a
posteriori conlangs, and for the classification of the latter
as "IE-based", "Romance-based", etc.; yet, as the diachronic
continuity is missing, they are not considered regular members
of the relevant families.  Surely, a language such as Brithenig
with its fictional diachronic development from Vulgar Latin has
a better claim at being "Romance" than Interlingua which does
not have such a development either real or fictional, but it
still is a *fictional* language, spoken in a fictional Britain.

On Sunday 04 December 2011 R A Brown wrote:

> Oh dear, things seem to be getting taken out of context and
> criticism seen where they are not intended.  It's starting
> to remind me a little bit of what I hoped I had escaped from
> when I left Auxlang.

Indeed.

>       I just would like to correct two
> statements which seem to have been taken criticisms when
> they were not intended to be so.
> 
> On 04/12/2011 16:29, Michael Everson wrote:
> [snip]
> 
> >> Yes.  A common criticism of Esperanto and all the
> >> other "classic" auxlangs is that they draw exclusively
> >> on European languages and are thus eurocentric.
> >
> > What's wrong with that?
> 
> With respect, I really do think this sort of things should
> be argued on that other list.  All Jörg did was to state a
> _fact_. He wrote that it was a common criticism. So it is.
> 
> He has not commented on whether it is a valid criticism or
> not.  Nor have I.

Right.  The question whether that criticism is justified or not
does not belong here.  That is exactly the kind of stuff for
which the Great Sundering was brought forth.

> > Perhaps evidently speakers of Sino-Tibetan aren't much
> > interested in conlangs and auxlangs. What are the
> > analogues to Esperanto, really, in non-Indo-European?
> > Trade languages like Chinook or Swahili?
> 
> Afrahili, for a start.

Yep.  So far, all non-European IALs have remained obscure, but
that holds for most *European* IALs, too.  At least 99% of all
auxlangers worldwide are evidently quite happy with Esperanto!

> But the point is that auxlangs are either a_posteriori or
> a_priori.  The a-posteriori either draw (almost) exclusively
> from European sources or attempt to be "world languages",
> drawing from a wide range of languages across the globe.
> 
> If we start evaluating the good and bad points of each
> approach, IMO we really are in Auxlang territory - and IME
> it too easily leads to flames.

Sure.  It easily leads to flames, and thus belongs to AUXLANG
and not here.  Also, I have no opinion on that matter (indeed,
I am not partial to the idea of an artificial IAL at all, and
do not expect any of them to succeed in the foreseeable future).

> [snip]
> 
> > On 3 Dec 2011, at 15:42, R A Brown wrote:
> >
> >> While probably Novial is of 100% IE material,
> >> Esperanto isn't.
> >
> > Nor is English, which has loanwords from most language
> > families.
> 
> I am very well aware that English is not!!!!
> 
> I've been speaking it for 70 years or more and I have had a
> great interest in languages for about 60 of those years.

English has borrowed many words from languages all over the world,
including non-IE ones.  And that may include more than a few
modern loans such as _mokassin_ or _chop suey_: some linguists
estimate that about a third of the vocabulary of Proto-Germanic
is borrowed from a lost non-IE substratum language of central
Europe!

But I don't see how these facts are relevant to the question how
a posteriori conlangs are to be classified.

> Once again, I am NOT criticizing Esperanto. I was merely
> commenting on a statement made by some else. I did not say
> that Esperanto was at fault in this, nor did I make any
> other comparisons.

Fine.  Enough said about this.  Auxlang criticism doesn't belong
here, so let's keep *that* can of worms closed.

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