Message-Id: <m0pqae4-000058C@xiron.pc.helsinki.fi>
Date:         Tue, 12 Apr 1994 00:59:55 -0400
Reply-To:     Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Sender:       Lojban list <LOJBAN%CUVMB.bitnet@FINHUTC.hut.fi>
From:         Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Subject:      Dipthongs in Lojban
To:           Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 7318
Lines: 127

John Cowan on our shared account answers Matthew Faupel:
>Subject: Re: Diphthongs in Lojban
>
>la matius. cusku di'e
>> 2.  Is the restriction of diphthongs in normal words to ai, au, ei and oi
>>     purely historical?  How about the complete non-use of eu, ao etc.?
>
>I wasn't there when the decision was made, but I understand that it was
>basically intended for ease of use.  I personally would have permitted
>"eu" as well, but many people have trouble saying it.
>
>Loglan did not (and does not) have "'" at all; there are 25 diphthongs
>with rules about which are to be pronounced as monosyllables, which as
>disyllables, and which are variable.  Furthermore, "aa", "ee", and "oo",
>which are disyllabic, could only be used where one of the syllables
>bears the word stress.  The introduction of "'" made all 25 diphthongs
>sayable as disyllables, and distinct from any retained monosyllables.
>Monosyllabic diphthongs add nothing except brevity and greater
>cmavo/rafsi space; 4 apparently seemed like a reasonable tradeoff
>against increased pronunciation difficulty.

A more official answer.

Historically JCB did not recognize any diphthongs in the language
(reference Loglan 2, Chapter 1, Phonology).  All vowel pairs were
allowed except aa, ee and oo.  The fact that some of the vowel pairs
tended to diphthongize was noted in the pronunciation guides only that
/j/ was an allophone of i before vowels, /w/ an allophone of /u/ before
vowels, and /e/ as in /lek/ ('lake', according to JCB, though my
dictionary shows the vowel to be a 'long a' which is the Lojban
diphthong ei.  More recently, he has /used /eigh/ in 'eight' as the
example for e before vowels, contrasting this with the ei diphthong
which is the /ay/ in 'day' - reference Loglan 1, 4th edition - no one
has yet been able to tell me the difference between these two sounds
though JCB himself says that he hears and makes a difference.) an
allophone of e instead of the more normal /IPA epsilon/ as in /met/
('met') before vowels.  (JCB himself is only partially an American
English native speaker, BTW - he was born and raised in the Phillipines
by American parents, and spent many years during the Loglan development
wandering around continental Europe.)

Since vowel pairs only occurred in cmavo, there was little need for a
diphthong distinction from disyllable glides.

When GMR took place in 1979-82, the question of whether these vowel
pairs were disyllable monophthongs or monosyllable diphthongs suddenly
became important, since it became possible to have a vowel pair at the
end of a brivla, leading to questions of which was the penultimate
syllable - the one before the vowel pair, or the first vowel of the
pair.  JCB stood firmly on the fence for this one.

His continental training gave him no problem in saying things like lua,
rua, sia (vs. cia) as monophtongs.  He also personally pronounced 'ao'
as the diphthong /au/, using a much longer non-diphthong for 'au', and
still has never accepted that this is inconsistent.  On the other hand,
his 'taste testers' almost always said the words as disyllables.

JCB incorporated the previously forbidden aa, ee and oo but these HAD to
be disyllables with contrastive stress (he has never given examples of
brivla where these disyllables did not fall on the penultimate stress
boundary, so it isn't clear what TLI Loglan requires for rafsi in those
cases).  He then said that ae, au, ea, eo, eu, oa, oe, ou, were always
disyllables but did not require this contrastive stress (I cannot hear
any difference between his /eo/ and his /ei,o/ other than possibly
length which is otherwise not distinctive in Loglan).

The iV and uV pairs are listed in L1 as 'optional disyllables'.  He says
they can go either way, and implies that a Loglan word would change
stress position depending on dialect because of this.  Hence his
"mekykiu" (eye-doctor) is given in the pronunciation guides as
/MEH-kuh-kyoo/ (pg 95) because of his Francophile dialect, but he
authorizes (on pg 87) the disyllable with contrastive stress, which
would give /meh-kuh-KEE-oo/.

His first pronouncements of this sort appeared in the draft of his
Notebook 3, which we discussed collectively in the first Fairfax LogFest
in Sep. 1986 (before the split).  There was a clear consensus that this
dichotomy of stress pronunciations was unacceptable.  JCB never
responded to our comments, since he instead took offense at the fact
that I had allowed others to see and comment on his draft.  But when we
went to codify a phonology for remaking the words at the start of the
split in May 87, the need to resolve this (as well as the pronunciation
of e in and out of diphthongs.

Since our goal was pressure for change in policy and the language
definition, not a split, and we sought eventual reconciliation, we tried
for maximal commonalty with the existing phonology, while being
linguistically a bit more formal and non-American English biased.  Thus
we set up our consonant clusters on more formal linguistic rules (as JCB
had done pre-GMR, but we were a bit less restrictive).  We also codified
the de facto existence of diphthongs in the language, using descriptions
colored by Chuck Barton's description of the Loglan phonology (that
eventually was rewritten to become the Synopsis of Orthography and
Phonology - our oldest Lojban document).  At that point Tommy Whitlock
and Gary Burgess came up with the devoiced glide/rough breathing 'h'
sound based on their experience with Greek among other things, and I
came up with the comma/apostrophe contrast for voiced/devoiced glides,
though we didn't use this terminology at the time.

Because we wanted reconciliation, we kept the 4 mandatory monosyllables
as diphthongs, but moved the ao to au for linguistic accuracy.  We
permitted the other optional diphthongs to only be diphthongs in VV
cmavo, since those were firmly embedded in the language.  We realized
that the apostrophe expanded the number of cmavo available and relished
the thought of reducing the crowding of words in cmavo space (oh well
%^).  We also gave brief consideration to other diphthongs - I remember
discussing the Cockney /eu/ as the primary one considered, but kept to
the basic 4 because of the priority for historical consistency.

If we were redeveloping the language now, I might give more
consideration to other diphthongs that are highly contrasted with the
vowels we have.  But Lojban is overcrowded with vowels by the standards
of some languages as it is, so I woulf probably resist.  The only change
I would really want to make would be to use w and y for the semivowels
in diphthongs rather than fiddling with the ancient and linguistically
incorrect statement that these semivowels are allophones of /i/ and /u/
rather than parts of a diphthong.  I guess that would leave 'q' for the
schwa, or just using a hyphen, as both JCB and we considered.  (Of
course, we might have gone to using syllabic r/n for all instances of
the hyphen, another GMR change that JCB abandoned.  But we didn't
consider ourselves to be revolutionaries, just dissidents.)

lojbab
----
lojbab                           Note new address:    lojbab@access.digex.net
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA                        703-385-0273