Disputable rendering of J in mode for general use

[kriskowal]

From Fiona Jallings,

I found another problem – the J should be Anga, not Anca. Anca is used for the
S in “measure”, if it’s used at all in English General Use.

Here is how I’d write “measure” and “fission” (turn into a font using Dan Smith
key-mapping if it doesn’t show up right)

t]Ff6OJ - ef:%`B5^

But you could also write these with Harma, according to PE20, page 65. Thus:

t]Fd6OJ - ed;G`B5^

Also the EA could be written separately – might be easier for the auto
transliterator.

You could add in some sort of setting that catches the -sion, -sual, -sure
suffixes that always end up with the voiced palatal fricative. You may not be
able to catch the odd ones out (azure) but it’s better than nothing.

One way to deal with -s is to make them all Sa-rinci. In one of the few samples
we have of Tolkien writing in this mode (Letter 118) he even uses the sa-rince
for the final S in “Christmas” – even putting the A-tehta on top of it! I’ve
copied it below: Z78$t+D

(Tolkien had his E and I tehtar switched in this text, for some reason)

This can be a way to deal with a lot of the S’s pronounced as Z’s, though,
IIRC, this is one of the things you let people chose for themselves, correct?

BTW, PH is written with an extended Formen/Parma in two different documents
describing orthographic tengwar, so it’s a pretty safe bet to include. Unlike
the CH (of which there are at least… 5? 6?) different ways to pronounce it in
English it’s a bit easier to include in the auto-transliterator.

Phone: Q5(^

What else? Oh! Vowel digraphs/diphthongs!

Ones that end with Y/I made with Anna carrying the first vowel, ones that end
with E made with Yanta carrying the first vowel, and ones that end with W/U
made with Vala carrying the first vowel. You can also have the ones that end in
A made with Osse carrying the first vowel – but that one isn’t as prevalent,
and neither is the use of Yanta. The Y/I and W/U are definitely important to
include though.

The double vowels on the long vowel carrier, BTW – yes, I did this myself for
many years, but looking over the texts we have in this mode, it’s just not
supported. For example, in the LotR title page inscription, the word “seen” is
written this way:

8`V`V5

Now – part of this is the morpheme barrier (see-n). In other words, it’d be
with double tehtar over the following tengwa or with one vowel on a
vowel-carrier and the second on the following tengwa. Or, in the case of there
not being a following tengwa – two vowel carriers. The long vowel carrier seems
to be an alternate version of the short vowel carrier at the ends of words. I
found it most common with words ending in lone -Y’s.

[kriskowal]

With changes you’ve recommended, the new behavior is:

• In English to render j as "anga". j' renders as "anca".
• In other languages, j is always "anca".
• There is not presently any other way to express the "anca" phoneme. You’ve mentioned a few patterns that could be automagically pattern matched. I’m also interested in providing a cluster so people can reliably produce the phoneme, without the tick, if they’re performing phonetic transcription, perhaps "zh".

Anca also gets emitted by nj with the pre-nasal tilde above, regardless of language. I assume that’s also incorrect?

[kriskowal]

Captured each of these issues as independent projects.