OP velar fricative orthography

[Rory M Larson]

> The most recent proposal from the ULCC is the voiceless double [xx] and
voiced single [x]. The reasoning being that regardless of sound quality,
everyone would recognize the [x] and pronounce the word closer to correct
than otherwise.

Examples:

xxude ‘gray’
xxuga ‘badger’
xage ‘cry’
xebe ‘shallow’
xthi ‘sore; pus’
xthabe ‘tree’

------------------------------

I'd like to add one point here.  Although I have said that a fricative in a
consonant cluster is probably not distinctive-- i.e., will not phonemically
alternate between the voiced fricative and the corresponding unvoiced
fricative--  that does not mean that it will not take one or the other of
those values.  Thus, in English, we might distinguish zip from sip, and
zing from sing, but there is only one voicing option for [alveolar
fricative] + [alveolar stop]: [st].  We have sting, but we would never have
zting, zding or sding.  Even in combination with a voiced consonant like m
or n, we have smile but not zmile, snow but not znow.

Similarly in Omaha, as we seemed to find out with the Macy speakers the
other day, the velar fricative seems to be voiceless in combination with
another consonant, even a voiced one like ledh.  So even though, for purely
orthographic purposes, we could get away with using the symbol for the
voiced velar fricative in combination with consonants, we really shouldn't,
as that will confuse learners who are not native speakers as to the proper
sound value.  Hence, if we go with the double x, or with the q, to
represent the voiceless velar fricative, then we should use that to
represent the sound in a consonant cluster as well:

  xxthi
  xxthabe'
  xxti

or

  qthi
  qthabe'
  qti

The big advantage in the system I proposed some years ago of using x-hacek
for the voiceless velar fricative and x-underdot for the voiced velar
fricative, is that it would be the least confusing and least orthographic
disruptive of all our options.  The current default system (La Flesche) of
using x for both phonemes is unacceptable for the same reasons that
c-cedilla is unacceptable: it merges two different Omaha phonemes and
thereby collapses a section of the language.  If the problem is not dealt
with orthographically, students of Omaha will ignore the difference, and
when the last of the native speakers is gone, the distinction will be lost
altogether (except for what Dorsey recorded).  Prior to that time,
linguists and educators who work with the speakers will fail to record the
difference if we do not have that as part of our own orthographic standard.

Any move we make from the La Flesche system will require some relearning
and instruction, which will be confusing to some.  If we use x for
voiceless and gh for voiced, then we have to get used to writing gaghe
instead of gaxe, and ghage instead of xage.  But the x-hacek vs. x-underdot
system would allow the original orthography to be kept, with just a bit of
"commentary".  People who are already used to spelling both phonemes with
an x will find the same x there, and can ignore the diacritics.  In
consonant clusters, we don't need to worry about diacritics; we can leave
the x alone.  But where it needs to be distinguished, a simple mark would
tell us which is meant.  This would keep everything that has been done up
until now in the Macy system intact, while allowing us to gradually work
the necessary distinction into the system.  Thus, from current

  xude
  xuga
  xage
  xebe
  xthi
  xthabe
  xti
  gaxe

we would simply shift to

  x^ude
  x^uga
  x.age
  x.ebe
  xthi
  xthabe
  xti
  gax.e

If you're used to the old system, just ignore the diacritics.  If you are a
linguist recording from a speaker, make sure you mark it.  If you are a
linguist or educator pulling up a previously recorded word like xude or
xage, then you know that it has not yet been distinguished, and you need to
track down the correct pronunciation to mark it.  This latter is an
extremely important consideration that keeps us recorders honest and
unconfused, and it will not be met if we keep a simple x in any position
that needs to be marked.  This is a problem with keeping a simple p, t, or
k for the tense stops as well.  If we run into a previously recorded word
that uses a simple t, we never know if it was really a tense t, or whether
it was an aspirated t that was carelessly recorded.  This is why I prefer
to write the tense stops double, always, so that I know that a tt was
intended to be marked tense, a tH was intended to be marked aspirated, and
a simple t outside of a consonant cluster needs to be tracked down and
determined.  The same problem of never knowing whether a sound was meant to
have a certain value, or whether it was carelessly recorded, attends any
system that proposes to keep a simple x in some non-cluster positions, be
it x vs. gh, q vs. x, or xx vs. x.  The x-hacek vs. x-underdot system would
not have that problem.  Imagine how difficult it would be to stamp out the
"c-cedilla problem" if, instead of c-cedilla, La Flesche had used s for
both s and z values!


The one great drawback to the x-hacek vs. x-underdot system is that it is
hard to represent on a computer.  Microsoft Word includes symbols like
g-hacek and so forth, but apparently nobody ever thought to put one on an x
before.  This is the only serious disadvantage I see to this system, but I
recognize that it may be a killer.

(Okay, I guess all that was more than one point!)

Rory