Saturday, May 20, 2017

Keeping the semivowels Y and W

                            A-ko              E-ko                   I-ko                      O-ko                       U-ko               
I'm now inclined to think that I should keep y and w as the semivowel equivalents of i and u. For one thing, it's often difficult to determine pronunciation when i and u are used for both. How do you pronounce this?:

uia

OO-yah or WEE-ah?

So if:

y=/j/
w=/w/
q=/N/
c=/S/
j=/Z/

that only leaves x to assign a value to. Should it be a modifier of other characters, as it has been in the past, or should it stand for schwa. I think I should keep it as a modifier, and find something else for schwa. Lojban of course uses y for schwa.

I do want letter name for all but the vowel be the letter plus schwa sound, as I believe is the case in Lojban.

Any thoughts on using an apostrophe for it?
********
Late-breaking news. I just remembered that Gua\spi  [link] uses # for schwa. If I did the same, I'd still have X as a modifier for other letters. Thoughts?

Thursday, May 18, 2017

Tradeoffs

Getting the alphabet straight is frustrating. It's a matter of tradeoffs;

For now, I've decided not to use w and y (or j) for the semivowel versions of u and i, but just to use u and i, specifying that they become their semivowel equivalents when next to another vowel. That keeps two letters from going to waste. And I proposed keeping diphthongs from forming by putting a hyphen between the letters. But I could use the Loglan comma instead, as Steve Rice wrote:

If you want to break a diphthong into two separate sounds, put a comma without spaces around it between the vowels when writing them in text. The name 'Lois', for example, would be written Lo,is in Loglan (Without this comma the oi in Lois would be pronounced oy and the name would rhyme with 'Joyce'.).

So it would work the same way in Tceqli.


pai = PIE

pa,i = PAH-ee

kau = COW

ka,u = KAH-oo


I think I like that better than the hyphen.


So that keeps j from being used up that way, and permits it to represent /Z/.

And q can continue to represent /N/.

So with these "changes" (some are just reversions to earlier ideas of Tceqli or Loglan) we have:
from http://ceqli.pbworks.com/w/page/5455985/The%20Alphabet%20and%20Sounds#view=edit

PRONUNCIATION OF TCEQLI


First, an orthography note. Tceqli can be written in all lower-case letters or all upper-case letters. When writing about Tceqli in English, I use conventional English style capitalization, hence "Tceqli" is capitalized. And you can just follow English usage in capitalization if you like. And I use capital letters to show stress on some occasions in what follows on this page.

The Tceqli language uses the 26 letters of the Roman alphabet. 13 of them are called cuazim, or "leading letter" (C) in Tceqli:


B as in Boy /b/
C as in SHin /S/
D as in DuD /d/
F as in FluFF /f/
G as in Good /g/
H as in Hat /h/
J as in meaSure, French Jean-luc picard, Je /Z/
K as in KinK /k/
P as in PiP /p/
S as in So /s/
T as in ToT /t/
V as in Victory /v/
Z as in Zoo /z/

X is a special letter used only to form foreign sounds by making digraphs. For example, the English "th" sound in "thin" would be represented in Tceqli by "tx."

The other 12 are called falozim, or "following letters" (N) in Tceqli

six vaul, or "vowels" (V):

A as in fAther /a/
E as in bEt* /e/
I as in machIne /i/
O as in bOAt /o/
U as in bOOt /u/

Y as in bUt, AmericA  [schwa] /ə/

And two poifaivaul, "semivowels" (P):

U as in We, coW /w/
I as in You, boY /j/
(they become semivowels when next to other vowels. to prevent that, insert a comma between them.

fia = FYAH
fi,a = FEE-ah

And three truin, "nasals," (T)

M as in MiM /m/
N as in NooN /n/
Q as in siNG /N/

And two hlar, "laterals" 
L as in LuLL /l/
R as in RoaR (Midwestern American or Mandarin preferred, but any 'r' sound will do.) /r/

N.B: *E is a short sound, as in English bEt, rEd, lEg. Remember to keep it short at the end of a word like "bine" or "twale". Do not pronounce it as in English "hooray". That sound is a diphthong and is indicated by "ey." The sound is common in English, but is seldom found as the last phoneme of a word.  The word deybe is pronounced like English "day bed" without the "d."

Y is rarely written, and the sound is often just an allophone of /a/ in unstressed positions. It's also appended to many letters to give them their names:

"zbano" is spelled zy, by, a, ny, o.


A few of these are unconventional.  Q was chosen to represent the consonant in siNG because it had no other obvious use, and because the NG sound rarely has a symbol in any language. X is used this way in Esperanto, to make digraphs.
U and I make these diphthongs:

ai - as in frY
au - as in cow
ei - as in bAthe
oi - as in bOY
ia - as in YArd
ie- as in YEllow
io - as in YOre
iu - as in YOU
ua - as in WAter
ue - as in WEt
ui - as in WE

When a  comma is interposed, vawl are pronounced separately:

be,o - BE-o
cu,a - XU-a

Stress: If a Tceqli morpheme of more than one syllable ends with a vaul (AEIOU) or poyfayvawl (WY), the stress falls on the next-to-last vaul:

dilna - DILna

pamo - PAmo

zilau - ZIlau

somalay - soMAlay

If such a morpheme has more than three syllables, a secondary stress falls on the fourth-to-last syllable.

tuiamalu - tuyaMAlu

kauelomani - kaweloMAni

If it does not end with a vawl or poyfayvawl, it is stressed on the last vawl: 

diyan - diYAN

cawal - caWAL

femur - feMUR

felin - feLIN

And if such a morpheme has three or more syllables, a secondary stress falls on the third-from-last syllable:

piramun - piraMUN

starloremi - starloreMIN



Finally, any diphthongs that you find difficult to pronounce may be pronounced as two separate vowels, u for u, and i for i, provided that the morpheme is stressed as though the two vowels are a diphthong!

Ceilo - CEY-lo or CE-i-lo, NOT Ce-I-lo

Triphthongs are possible though rare:
iai as in YIkes!
uau as in WOW!
uei as in WAY
iei as in YAY
ioi as in YOIks!

Tceqli morphemes must begin with one or more tcuazim.  But this is an actual language, so there are times when borrowed names do not.  In this case, for the most part, we simply put an 'h' at the beginning if the name begins with a vowel or semivowel:

English > heqlizo      
Al> halzo
Obama > hobamazo

Names beginning with with lmnqr add z:

Lima > zlimazo
Russia >zrusizo
Norman > znormanzo

The names of letters in Tceqli:

A a
E e
I i
O o
U u
Y y

L hly
M hmy
N hny
Q hqy
R hry 

B by
C cy
D dy
and so forth


Note that the falozim group of letters are in violation of the rule that morphemes must begin with a cuazim. Consequently, they must be preceded by a pause or glottal stop.

All the cwazim, on the other hand, have names in the form of the letter itself followed by et, so no pause is required, though it's usually there anyway.

They are also used to spell words, of course, but they are also used as pronouns, or anaphora, to refer back to previously used words, in this way.

Djanzo pa gi kom pan. D pa falfa P. (Prounounced dy pa falfa pə.)
John was eating bread. He dropped it.

They replace da for clarity. To avoid confusion, they are always written upper-case followed by a space, of course.

They are also used for acronyms:

To hanho ze stan hu Hamerihaimzo.  HSH  Pronounced hə-sə-hə
The United States of America. (USA)

Janzo Fitsjeraldzo Kenedizo. JFK  Pronounced jet-fet-ket

And for foreign abbreviations, of course the falozim have to be used, too.

USA pronounced u-sy-a

And to name the chemical elements, both the tcuazim and falozim forms have to be used, of course:

Fe pronounced fy-e, Ag pronounced a-gy, etc.

In mathematical expression, lower-case letter symbols are pronounced the same way:

a+b=c  a plu by kwal cy
Finally, for situations where extreme clarity is called for, the military, or NATO, "phonetic" alphabet is used as is.

ICAO Phonetic Alphabet
LetterCode WordPronunciation
AAlfaAL fah
BBravoBRAH voh
CCharlieCHAR lee
DDeltaDEL tah
EEchoEKK oh
FFoxtrotFOKS trot
GGolfGolf
HHotelHO tell
IIndiaIN dee ah
JJulietJEW lee ett
KKiloKEY loh
LLimaLEE mah
MMikeMike
NNovemberNOH vem ber
OOscarOSS car
PPapaPAH pah
QQuebeckeh BECK
RRomeoROW me oh
SSierrasee AIR ah
TTangoTANG go
UUniformYOU nee form
VVictorVIK ter
WWhiskeyWISS key
XX-rayEKS ray
YYankeeYANG kee
ZZuluZOO loo 


Now — whatever shall we do with "w"?

Re-Loglanizing the Alphabet

Going back to my original inspiration, how about I use the Loglan alphabet, slightly modified:

b, d, f, h, k, l, m, n, p, r, s, t, v, z all as in English.

c= /S/ as in SHe
therefore tc = /tS/ as in CHeese
g always as in Goat
j = /Z/ as in meaSure
therfore
dj = /dZ/ as in John
q = /N/ as in baNG

So far that's the way I had it in Tjeqli version one. Now for some differences.

Like Loglan, I and U become semivowels when next to other vowels, unless set off with a hyphen:

pia = PYAH, pi-a = PEE-ah
taulo = TOW-loh, ta-ulo = ta-OO-loh

Anybody have a better idea than using the hyphen this way? I was going to use an apostrophe, but I want to reserve that to indicate stress when necessary.

That leaves x, y, and w. X can be used as I said before as a letter modifier as in
tx= /T/. w can be schwa. What can y do?
------------
Input has convinced me that y should = schwa, which leaves w as a free letter.

Dubya

No, not that Dubya. I just had an epiphany, or something. Let's change that rule I mentioned in the last post, and say that both i and u become semivowels when next to a vowel. And if you don't want them to, use a hyphen to stop it from happening:

bua - BWAH
bu-a - BOO-ah

Skia - SKYAH
Ski-a SKEE-ah

And that liberates both y and w for other uses. So let's use w for schwa. And j can indicate the palatalization of the consonant it follows:

sj = /S/ Sji selz si sjelz bai dxw sisjor.
tj = /tS/  (so the language would be Tjeqli.)
zj = /Z/
dj = /dZ/

I know this is an alarming revision, but please don't react like Sensei does in the illustration.

I'm going to go with the revisions in this and the last post unless somebody talks me out of it.

By George She's Got It, She's Really Got It!

A correspondent writes:
Rex, is there a compact list for pronounciation, first row the letter, second row the IPA character(s)? E.g. as at https://en.wikipedia.org/w/index.php...

And he's right, I need to make one, for a number of reasons. But before I do, I have a question. There's the IPA, where you spell the "ch" sound that Tceqli begins with as . IPA has a bunch of characters like that stretched-out "s" that aren't readily available on your keyboard. I've seen that same sound spelled /tS/ between two slashes like that. I assume this is a system designed for people who can't, for whatever reason, make the offbeat symbols like ʃ.  Anybody know? And know what the system's called, and where I can find a link to it?

Another question. How does this sound to everybody?:

You can symbolize just about any sound you want with Tceqli's 26 letters — 25 letters that have one phoneme each associated with them, and then the X, which is used to modify the other letters, mainly for foreign sounds. 

tx = th as in THin
hx = ch as in BaCH
ux = ü as in Über

You also have the option of replacing letter-followed-by-x with any version of that letter with a diacritic of any kind:

tx > t
hx > ħ
ux > ü (of course)

But that's sophisticated stuff for the future.

And then we have schwa, which won't be a basic Tceqli sound, but which i figure will be in the names of most Tceqli lettersa; ah, buh, shuh, duh, etc.,, and of course in foreign words and names.

The same correspondent writes:

In the Unified Latin Alphabet 2016 (ULA 2016) schwa is <y>. In IPA /y/ is a vowel and using <y> as vowel, allows easy pronounciation of international words like Yttrium.

Letter <y> = IPA /j/ would then be <j> = /j/, i.e. less deviation from IPA. Beside in IPA this is used in several Germanic and Slavic languages languages.

Then Tceqli would need something for /ʒ/. ULA 2016 uses <q>.

Then Tceqli needs something for /ŋ/. Maybe use the modifier here, i.e. Tcenxli.

I really don't want to make all those changes, but y for schwa is tempting. But if y is used for that, it can't be used as the semivowel it represents in current Tceqli. For that I'd have to use j (and then represent /Z/ with a ʒ and /dZ/ with dʒ. Or just go to Russian and use ж. Or, I could replace semivowel y with a simple i, and change the rule so that i next to another vowel behaves as a semivowel.

The trouble with this last is that you can't have a word like dia pronounced DEE-ah. It would be DYAH. One way around that would be to have the i-becomes-semivowel rule, but also have a way to short-circuit it by saying that a hyphen separates vowels and causes them to be pronounced separately. So dia would be DYAH and di-a would be DEE-ah.

Do give me your input on all of this.


Sunday, May 14, 2017

That = Ke or maybe Cto

I originally borrowed ke from Esperanto and several other languages to introduce a noun clause. In Loglan, the corresponding "that" is simply le (the) followed by po, which makes the le refer to the entire noun clause. So I'm thinking to follow the Loglan example to some extent, and have ke replaced by a the-word.


An ideal candidate would be the Russian чтоwhich would come into Tceqli as cto, pronounced like the Russian. It has the advantage of looking like it was deliberately derived from to. So

Go djan, ke zi pa kom to pan.
I know, that you past eat the bread.
I know you ate the bread.
becomes
Go djan, cto zi pa kom to pan.
I know, that you past eat the bread.
I know you ate the bread.

What do you all think?

Revision

Latest revision: Do is now only the present tense particle, for it to mean 'when,' or 'at the time of,' it must be followed by 'sa.' Analogously, the same goes for pa and fu, the past and future particles, that now mean before and after when followed by 'sa.' Does anybody see any problems with this change? I could also use ke in place of sa.

Thursday, May 11, 2017

More Alphabet Blather

I'm definitely eliminating c= /tS/  No way it isn't awkward.  I can use either x or c for /S/. If I use c, as do Loglan and Lojban, I'll  have x for use in modifying other letters to make foreign sounds. tx could be /T/ for example. The 26 letters just aren't enough to have special characters for both /dZ/ and /tS/.  Depending on which way I go, the language will be either Txeqli or Tceqli.  Your input is appreciated.

And, don't read anything much into the illustrations I use here. I pick them more or less at random for amusement purposes.

Monday, May 8, 2017

Asymmetry in the Alphabet

Ceqli has an alphabetical assymetry:

x = /S/
and
c=/tS/
but
j= /Z/
and
dj= /dZ/

Awkward. I want to eliminate c, reserving it for something else, and spell
/tS/ as tx, to match dj.

If I don't hear any good reasons not, I'll revise the material to reflect this soon. Don't let this shock you. Comment here, or on Facebook https://www.facebook.com/The-Ceqli-Language-249301371750083/