Danny Wier <[log in to unmask]> writes:

> I'm trying to imitate Tolkien as little as possible.  The mutant races, both
> Elves and Orcs, are closer to human than those in other concultures, and can
> have offspring from humans which in turn can have children of their own (in
> other words, they're not sterile like mules are, though half-Orcs are
> pejoratively called "mules").

Well, in my conworld the Elves are even closer to human than that: they
are as close to human as they positively could be, because they ARE
humans.  Not even a mutant sub-race, just a particular ethnic group in
western Europe which has all but vanished *here* but still survives in
several others.  Though they are somewhat taller and thinner than most
people, and look similar to Tolkienian Elves.

The reason why I call them "Elves" (besides their look) is that I
imagine them to be what underlies the Celtic and Germanic mythological
traditions from which Tolkien derived his Elves.  They are also similar
to Tolkienian Elves with regard to culture and language, though neither
their culture nor their languages are simply stolen from Tolkien, of
course; both are to a large part my own invention and contain many
things not found in Tolkien's creation.  (My interpretation - entirely
fictional, of course; I don't claim that it is really what happened! -
is that Tolkien got hold of fragments of old writings about these people
and used them in his construction.)

I actually poundered the issue whether I should depart completely from
Tolkien's languages, but I have found too many things in the real world
that *could* be construed as traces of languages very similar to
Tolkien's.  Here are a few of them:

Engl. iron < Celt. *isarn- < Quend. *el-sarn "meteorite" (lit.
"star-stone")
Engl. silver < PGmc *silbr- < contamination of Quend. *kjelep- "silver"
                                           and Quend. *sil- "shine"
Engl. London < Celt. *Londinion < Quend. *Lond Din "quiet haven"

Then, there were numerous tribes called "Veneti", "Venedae" or something
similar all over western and Central Europe 2000 years ago.  These might
all have been splinters of the mysterious Bell Beaker people, whom I
imagine to have been the original "Elves".  (I remember someone
mentioned such a theory in CyBaList as well.)  Thus, the name by which
the Bell Beaker people referred to themselves could easily have been
something like *kwendi or *kwenedi.  Even the Guanches of the Canary
Islands fit into this pattern.  They were Berbers, but probably with a
European substratum which could well have been Bell Beaker folk or
descendants thereof.

> Georgian just happens to be one of my favorite languages, though I don't
> know it well.  As y'all can easily tell, many of my conlang projects involve
> large phonologies with the "holy trinity" of stops/affricates found in
> languages from Navajo to Georgian to Korean: voiced, voiceless aspirate,
> voiceless glottalized/ejective.

My own Proto-Quendian (as well as Tolkien's) comes close here, except
that the voiceless non-aspirated stops aren't glottalized (which of
course doesn't rule out that they may have been glottalized in an
earlier stage).

I also think that the glottalic theory of PIE makes sense.  The problem
of going from glottalized to voiced can be solved by an intermediate
stage:

"Glottalic" PIE          "Intermediate" PIE      "Traditional" PIE

 voiceless (aspirated) -> aspirated            -> voiceless
 glottalized           -> voiceless            -> voiced
 voiced                -> voiced               -> voiced aspirated

The last change could have happened rather late and never reached two
remote corners of the IE dialect continuum, which later became Germanic
and Armenian, respectively.  It is also possible that the original
voiced stops never became what is traditionally "voiced aspirated", but
merged with the originally glottalized grade in those branches where the
contrast between these two grades is lost (e.g. Celtic and Slavic).  A
reason to assume that the change from the intermediate stage to the
traditional system is that it apparently managed to overtake
satemization in Armenian.

In my conworld (where Quendian seems to be related to IE and
Kartvelian), the change from glottalic to intermediate might have
happened before IE and Quendian parted from each other, but after
Kartvelian began going its own way.  Or IE and Quendian "deglottalized"
independently.

> Yes, I did read your phonology; the seven vowel system of Nur-ellen is a lot
> like my nine vowel system of Tech (my other conlang-in-progress with a bunch
> of consonants but based on Afro-Asiatic and Nostratic), and also Q if you
> ignore short-long vowel distinction.  My canonical nine-vowel system:
>
> With diacritics, may not come out right with some e-mails:
> i ü u
> e ö o
> a ä å
>
> ASCII-friendly version:
> i  u" u
> e  o" o
> a  a" a.

The bottom row corresponds poorly with the top and middle row which are
both based on the pattern: front unrounded - front rounded - back
rounded.  The vowels in the bottom row are, hmm, central unrounded -
front unrounded - back rounded.  (Of these, Nur-ellen only has /a/,
while the non-low vowels are the same as in your pattern.)

> A diaeresis or umlaut above a, o and u indicate fronted vowels, while a-ring
> (a-period in ASCII) is a rounded a.  Another variation of my nine-vowel
> square replaces front rounded vowels with central unrounded vowels:
>
> i  i- u
> e  @  o
> ä  a  å
>
> This just happens to be the same nine vowel qualities of Thai and Middle
> Korean.

I also came up with this system for my own version of Pictish.

> I could also combine these two systems to make an eleven-vowel chart, and
> even add a twelfth vowel: low front rounded.

Yes, I also had that idea, and perhaps will use it some day in some
conlang.
Sort of a three-storey version of the Turkish vowel cube, if we push the
central unrounded series back a bit ;-)

Jörg.