I had a hunch that the author of the letter-names document was
familiar with the writing system of something besides Latin (which does not
use, for instance, uu to represent w!) -- since the ms. is, I think,
associated with Salzburg, I thought I would take a look at the writing
system of Old High German. Here's what I found out (it is no doubt "well
known", but not to me):
The letter c was, as in many post-Roman European languages,
bivalent; before back vowels a, o, u it was [k]; before front vowels i, e
The letter ch was used for the affricate [kx], which had developed
in Upper German [Bavarian-Alemannic, including Austrian] dialects of OHG
from older [k].
The letter g, though normally [g], also had the value [j], though
[j] could also be represented by i or even e (between a consonant and a
[kw] was represented by qu
OHG did have a sound represented by th (possibly [ð]) but this
gradually became [d], starting in 8th century Upper German. So if this
document really does come from the area of Salzburg, nothing like [þ] may
have existed in the author's speech; though he may have had some knowledge
of places where it or [ð] was used, and spelled th.
z was used with two values; one [ts], as in modern German; the
other a variant of [s], when that sound arose from older t. So, for
instance, NHG Wasser "water" was OHG uuazzar.
Oh, yes, uu was used for [w].
This makes the forms in the ms. look like Upper German. Chozma (if
we accept the reading, or ignore possible copy errors) presumably
represents [kxosma], and shows that either a), our author's informant spoke
Gothic with a heavy Bavarian accent, or b), the author perceived the Gothic
in a way which allowed him to approximate it to his spelling standards.
OHG did have an [iu], which suggests that the spellings thyth, tyz
really do indicate a sound change, either [iu] > [y] or [iu] > [i]; I would
guess the former more likely, though y doesn't actually have this value in
OHG, where [y] did not yet exist.
z can be taken, as I guessed, to represent an [s].
geuua ought to represent [gewa]; but we may suspect here a) the
influence of OHG geba "gift"; b) that [v] was close enough to [w] in any
case to allow the author to represent it with uu. Of course we can't rule
out a sound change [v] > [w], but I am somewhat doubtful.
Doubled vowels could occasionally be used to represent length; so
it is possible that daaz, haal, really represent [da:s], [ha:l], and that
the g  sound had totally disappeared by this date, with compensatory
lengthening of the vowel. We certainly have a long vowel in iiz = [i:s].
I still guess that gaar is a mistake (in reading or copying) for geer =
[je:r]. The pronunciation g = [j] would be more justified before a front
vowel (e) than before a back vowel (a); but a copyist's error geer > gaar
could have come in by assimilation to OHG iar (iaar) "Jahr", year.
Since c = [ts], ezec could be read as a form of azets "easy". cz
also seems to occur as a very irregular spelling for [ts]; so noicz could
presumably be read as [noits], with ts being either a development or a
mishearing of [þs]. That does nothing, unfortunately, to explain the i,
which I would still guess to be the result of an error.
With reference to winja vs. wunja, the OHG method of writing [wu]
was uu; just the same as [w]. Wunja would then have to be written uunia
or uunea; depending on the type of writing, it's possible that uuinne could
be a deformation of one of the two.
The ae in uuaer is not explicable by reference to OHG.
About sauil vs. sigel (segl, sægl); as far as I can tell, the g in
the latter was never pronounced other than as [j] (i.e. [sijel]). But this
is likely true with sugil as well (=[sujil]). There is neither [g] nor [j]
in PGmc *so:wila-. Sigel looks to me like it comes from a different
source, though I wouldn't swear to it right now.
OHG is actually pretty archaic -- not as archaic as Gothic, but not
that far off either. I bet a Goth could have found himself conversing
(brokenly) in OHG after only a few weeks of exposure. I was able to read
some OHG sentences straight off, without any real knowledge of the
language, just by thinking in Gothic! As the following, from a translation
of Tatian's Diatessaron (a very old Gospel harmony):
OHG: In themo sehsten mânude gisentit uuard engil Gabriel fon gote in thie
Go: In thamma saihtsin menoþ sandiþs warþ aggilus Gabriel af guda in þizai
thero namo ist Nazareth, zi thiornûn gimahaltero gommanne, themo namo was
Joseph, fon hûse
þizos namo ist Nazaraiþ, du magaþai gawadjodai guma-mann, þizei namo was
Ïosef, af husa
Davides, inti namo thero thiornûn Maria.
Daweidis, jah namo þizos magaþais Maria.
The corresponding passage in Wulfila is: þanuh þan in menoþ
saihstin insandiþs was aggilus Gabriel fram guda in baurg Galeilaias sei
haitada Nazaraiþ du magaþai in fragibtim abin, þizei namo Iosef, us garda
Daweidis, jah namo þizos magaþais Mariam.
The constructions are a bit different (e.g., use of dative and
genitive) in the OHG, and there are some different choices of words, and
some words not found in Gothic at all (thiorna "virgin", which I think may
be related to þius, þiwi, and mahalen "betroth"); but on the whole the
resemblance, both in vocabulary and in structure, is quite close.
/\ WISTR LAG WIGS RAIHTS
\/ WRAIQS NU IST <> David Salo