Re: DECAL: Examples #1: Phonetic inventory examples & motivations

On Wed, Jan 12, 2005 at 06:24:35PM -0800, Sai Emrys wrote:
[...]
> First off: phonetic / phonemic inventory.
>
> Q1: What is your *phonemic* inventory? I.e., what are all of the
> discriminated phonemes in your conlang(s). (IPA / CXS / X-SAMPA)
Ebisédian's phonemic inventory is extensive but not very interesting.
The consonantal phonemes are:
	[G]
	[x]
	[N]
	[g]
	[k]
	[k_h]
	[D]
	[T]
	[n]
	[d]
	[t]
	[t_h]
	[Z]
	[S]
	[dZ]
	[tS]
	[tS_h]
	[z]
	[s]
	[l]
	[r`] or [r] (in some dialects)
	[B] or [v] (but the latter gets laughed at as "foreign")
	[P] or [f] (ditto for the latter)
	[m]
	[b]
	[p]
	[p_h]

Vowels:	[u]
	[8]
	[y]
	[o]
	[@\] or [@] (the former is preferred)
	[i]
	[A] or [O]
	[a]
	[E] or [&] (the former is preferred)

Vowels also have length, nasality, and "breathing". Taking [u] as an
example, you have [?u] (normal breathing), [wu] (smooth breathing),
[hu] (rough breathing). This applies for all vowels. Smooth breathing
for open vowels is pronounced without a distinct onset (I don't know
how to represent that in IPA).


Tatari Faran, OTOH, has a much more interesting phonemic inventory.
When designing TF, I wanted its phonetic inventory to be small and
simple yet interesting. So:

[p] [t] [k] [?]
[d]/[4] - these are the same phoneme, realized as [d] word-initially
	and [4] medially.
[b] [m] [n]
[f] [s] [h]
[dz] [ts]

(Note that there is no lateral consonant in TF.)

Vowels:
	[a] [E] [i] [O] [M] [u]
	[a:] [ej] [i:] [u:]

Glides:
	[aj] [ao] [ja] [wa] [uj]


[...]
> (Side question: CXS is the "standard" notation for this list?)
Yes.


> Q2: What are the allophones? I.e., for each phoneme, what are the
> "normal" variants that don't change meaning?
Hmm. I think I've already listed them above. It could be pointed out
that Tatari Faran stops are all *unaspirated* (hence the [t] is the
"Spanish t" rather than the word-initial "English t"). The aspirated
counterparts are understood, but regarded as foreign and "weird". The
[d]/[4] phoneme can be all pronounced as [d] or all as [4] and still
be intelligible, but of course that would sound "weird".


> Q2b: If you have any, what are the connotations / implications of the
> different allophones? E.g., do you use them for different dialects,
> registers, "accents", etc.?
They would probably occur in dialects, but so far, I haven't done much
work on dialects in either Ebisédian or Tatari Faran.


> Q3: How do your choices for the above reflect the goals of your
> language? E.g., if it's an auxlang [here!?], it's probably motivated
> by having common, strongly "universal" common-use phonetics to
> maximize learnability. So, for whatever your goals are for the
> conlang, how do they apply to the choices you made for phonetics /
> phonology?
[...]

For Ebisédian, I simply threw in all the phones I could pronounce,
including some I couldn't (at least until I made Ebisédian :-P), and
then "rationalized" it by turning labiodentals into true labials ([f]
-> [P] and [v] -> [B]). Like I said, extensive but not very
interesting. Ebisédian was in fact intended to be sortof an IAL (in
the sense of being like IPA in Ferochromon), back in the days when I
had no idea about parts of the IPA that I didn't even know could be
pronounced. :-) It was "universal" in the sense that I tried to
include all the sounds I could pronounce, but of course I didn't
realize back then that there are a LOT of sounds I couldn't pronounce.
Nevertheless, in retrospect Ebisédian's phonology turned out to be too
regular and too uninteresting because of this early decision.

Which brings me to Tatari Faran: after learning the hard way with
Ebisédian, I decided to try my hand at a much smaller phonetic
vocabulary, but with some unique twists added. So I kept the phonemes
at a minimum: instead of the 27 consonants of Ebisédian, TF only has
13 consonantal phonemes (14 phones). Initially, [l] was in there, but
later on I decided I could cut that out as well. I also threw out [g]
quite early on. I wanted to get rid of [d], but since I already had
several words that had [d] (and would sound really odd if I replaced
it with something else), I decided that [d] and [4] should be
allophonic, with one being used word-initially and the other medially.
At one point, [l] kept on coming back, but finally I put my foot down
and rejected it forever, and meticulously purged it completely from
every lexicon entry. :-)

With TF, I am quite happy that it turned out well: the phonology is
much more interesting, and has the desired small size and sound that I
was looking for. The glottal stop, for one, was an inspiration by
Hawai'ian, which also sports a minimal consonantal vocabulary.


T

--
The peace of mind---from knowing that viruses which exploit Microsoft system
vulnerabilities cannot touch Linux---is priceless. -- Frustrated system
administrator.