[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[lojban] Re: required commas (was Re: Stress of cmene)
At 05:10 PM 5/5/03 -0400, Pierre Abbat wrote:
> > As to the ,=' rule, I think it's a mistake. Apostrophes are always
> > distinctive, whereas commas are not. {malauis} = {malau,is} <>
> > {malau'is}.
>
> I've prooved commas must be distinctive. ,=' is a more sane rule
> than the commas-don't-matter rule; it takes precedence.
You haven't proved that. To prove it, you have to exhibit two different words
which differ only in commas. If indeed commas matter, then {kikui,us} and
{malau,is} are not words, whereas if they don't then they are the same as
{kiku,ius} and {mala,uis}, which are words, so we haven't found a comma
minimal pair. Even if both the forms differing by commas are both words, and
have different meanings, that doesn't necessarily mean that they are
different words, as names are allowed to be polysemous.
This is a matter for the VTPFK. Nora should have some insight into the
question.
Nora didn't, but faced with apparent conflict between one rule in CLL which
says that close-comma is equivalent to an apostrophe, and another which
says that it is equivalent to omitting the comma, I checked back to see
the document from which Cowan wrote those sections of CLL, the Synopsis of
Lojban Phonology, Morphology, and Orthography, which was the first piece of
Lojban baselined in 1991. (The synopsis actually predates Lojban, having
been a non-TLI document about Loglan written by Chuck Barton in the early
80s, and jointly rewritten by the two of us in 1987-8. It was the first
published document about Lojban other than the brochure.)
Except for the muddled history of baselining decisions at LogFest that
threw out all prior baselines when approving new ones without requiring
baseline change documentation as we had practiced with the grammar, there
should be nothing in CLL that contradicts the earlier baseline (one reason
I am pushing for such careful documentation in any byfy changes.
At any rate, this conflict is clearly resolved in the historical record
(and I will be providing Robin with the Synopsis and other historical
documents for the web site, since some of them aren't there as I had thought).
For once supplication provides a documented resolution. See the two
sections in the following marked with 5 asterisks.
III.2 PERIODS, CLOSE-COMMAS,AND CONSONANTAL BUFFERS
The three special characters in Lojban need special attention. They are
all used as indicators of a division between syllables, but each has a
different pronunciation, and each is used for different reasons:
The period (.) indicates a glottal stop (defined below) or pause. In
addition to separating syllables, it indicates a separation into two words
as well, since pauses are not permitted inside of a word. . inside text
words thus indicates that two or more words have been written as a
compound, but should be spoken separately. . may be used outside of words
as an aid to readers, indicating a mandatory pause or glottal stop.
The comma (,) embedded in a word is called a closed comma. Such a comma
is written to separate syllables, but indicates that there must be NO
pause between them, in contrast to the period. Between two consonants or
between a consonant and a vowel, this is straightforward. Between two
vowels, a comma indicates some type of glide is necessary to avoid a pause
that would split the two syllables into separate words. A semicolon is
used instead of a comma after a word to indicate optional pauses for phrasing.
The apostrophe (') is also called the consonantal buffer. It is the
preferred solution to the problem of a syllable split between
vowels. Instead of using a close-comma, which forces an unspecified kind
of glide that could be misheard as a letter, the ' indicates a specific
sound, a rough breathing aspiration like a breathy English 'h' that is
represented in pronunciation guides as /h/.
We will now discuss these three in more detail, giving some examples.
A period (.) indicates a glottal stop or pause in Lojban. A glottal stop
is the catch in your throat that naturally occurs prior to the beginning
of a word (and sometimes a syllable) which starts with a vowel. Try saying:
ah /.ah:/ (the ':' indicates that the preceding vowel is to
be drawn
out or lengthened)
Did you feel the catch in your throat at the beginning?
A glottal stop is used in Lojban like a short pause, allowing a listener
to break up an otherwise unintelligible string of syllables into its
component words (parse - this term occurs often in Lojban discussions with
this meaning). Sometimes a longer pause is desirable. In such a case, a
writer will usually leave one or more spaces after the period. However,
the length of a pause is not significant in Lojban; a glottal stop is
equivalent to a longer pause, and you may insert any comfortable length
pause when you see a period.
A period may be at the beginning or end of a word in print; the pause or
glottal stop is mandatory. In Lojban, more so than in English, these
pauses are significant in determining the meaning of a string of
sounds. Thus, for example, .i is often seen as the primary indicator of
the beginning of a new sentence. Without the pause, a Lojban listener may
not recognize the sentence break. We will use periods to represent
mandatory pauses in punctuation guides as well as in actual Lojban text.
A period also may be found embedded in a word. When this happens, the
'word' is really two or more words, usually a compound cmavo composed of
smaller, related cmavo that interact in determining the grammar of a
sentence. Such a period is called a close-period. Like all other periods
in Lojban, it is pronounced as a glottal stop or pause, and it always
indicates the end of a word. It is preferable to use the short glottal
stop to a longer pause, as this conveys to the listener the interacting
relationship between the two words.
When more than two vowels occur together in a cmavo compound in Lojban,
the normal pronunciation pairs vowels from the left into syllables, as in:
meii /mei.ee/ where the 'ei' grouped together causes the
English /ei/ sound of "base". This example would appear to be the
compound of the two cmavo: mei and i. A pause (or glottal stop) must be
inserted between the /ei/ diphthong and the final /ee/ to keep each sound
distinct. This example could also have been written mei.i, using the
close-period. Since this makes reading easier, it is in fact
preferred. But, a Lojban reader who sees the cmavo compound meii knows
that /mei.ee/ is the only correct pronunciation.
The close-period can override left-wise grouping:
me.ii /meh.yee/ where the 'ii' is now grouped and pronounced as
/yee/. The word can now be seen as the compound of the two cmavo: me and
ii. The close-period thus forces a syllable and word break (and a pause
or glottal stop) to occur at a different point in the vowel string than
would otherwise occur.
In longer strings of vowels in cmavo compounds, a Lojban reader will
continue grouping vowels in pairs from the close-period. Rarely, in very
complex vowel strings, more than one close-period might be found.
Close-commas are found embedded with vocalic consonants (l, m, n, and r as
pronounced /l/, /m/, /n/, and /r/) in cmene (names) and in le'avla (Lojban
borrowings from other languages); they serve like close-periods to
indicate proper pronunciation. A comma embedded in a Lojban word
indicates a syllable break, as does a close-period. However, a
close-comma indicates that there should be no pause or glottal stop
between the two letters separated by the comma, and therefore that the
'word' truly is one word. See III.3 below for examples and notes of this
usage in normal Lojban words that have vocalic consonants. (Commas are
also used in pronunciation guides to separate syllables.)
Close-commas are required in cmene instead of close-periods, since cmene
are not permitted to have embedded pauses or glottal stops. In the song
about Old MacDonald's Farm, the vowel string which is pronounced
/eey,yai,eey,yai,ou/ in English could be Lojbanized with periods as
i.ai.i.ai.o (/i.ai.i.ai.o/). However, this would sound clipped, staccato,
and unmusical compared to the English. Embedded in vowels within a
Lojbanized name, a close-comma must be pronounced as a vowel glide
(usually similar to the /y/ or /w/ in the English pronunciation above),
since no pause or glottal stop is permitted. The close-comma can be used
instead of a close-period: i,ai,i,ai,on is a valid, if strange, Lojban
name. The close-commas require a new syllable, but prohibit a pause or
glottal stop. The pronunciation thus may be /iy,yaiy,yiy,yaiy,yon/.
*****
The use of close-commas in this way is risky to audio-visual isomorphism,
since glides can be heard by some as i and u. Some leeway is therefore
given with regard to audiovisual isomorphism in Lojbanized names to allow
them to be pronounced more like their native counterparts - the
close-comma is allowed to represent the glides or some non-Lojbanic
sound. In this special circumstance, the deviation affects only spelling
accuracy and the ability of a reader to replicate the desired
pronunciation; it does not affect the recognition of word boundaries. A
close-comma vowel glide may be pronounced as an apostrophe or consonantal
buffer (defined immediately below) as one of the options. It was this
usage, in fact, that led to the consonantal buffer as the generic solution
to such vowel glide problems while preserving isomorphism.
*****
Close-commas may also occur between consonants in cmene and in le'avla
(borrowed brivla) to indicate variant syllabication of consonant
clusters. No glide or other sound is necessary in these locations.
The consonantal buffer, represented by an apostrophe, is a third way of
breaking up vowel groupings into syllables, and is preferred in regular
Lojban cmene that are not attempting to match a non-Lojban
pronunciation. In other word forms, the consonantal buffer is the only
permitted way to split vowels - the close comma is not permitted - without
also causing a word break, as when the close-period is used to force a
glottal stop.
The apostrophe lerfu (') is pronounced as a short, breathy, 'h' - the
so-called 'rough breathing' mark of ancient Greek and other
languages. English examples of this sound may be produced by saying a
quick, unstressed, "aha", or by saying "Oh hello" quickly with no pause
between the words. The consonantal buffer causes a speaker to aspirate
(breathe out) quickly between vowels, thus preventing a glottal stop. It
is thus the Lojban antithesis of the close-period, and indicates a vowel
pair that does not separate logically into two words. If the consonantal
buffer were used instead of a comma in the above vowel string example:
i'ai'i'ai'on, it would be pronounced /i,hai,hi,hai,hon/, which preserves
the rhythm and length, if not the exact sounds, of the original
English. Further examples of the consonantal buffer will be found in
III.3 below.
The consonantal buffer is a consonant in nature, but is not treated as a
consonant or a vowel for purposes of Lojban morphology. It is included in
Lojban only to enable smooth separation between vowels, while clearly
joining the vowels in a single word.
*****
As stated above, the close-comma may be pronounced as either a glide or a
consonantal buffer when between vowels in cmene. When the consonantal
buffer is required, such as in regular Lojban words, the apostrophe only
should be used to represent the mandatory sound.
*****
So clearly the intent in the language design is that pronunciation of a
comma as an apostrophe (consonantal buffer) is ONLY permitted (and then
merely as an option for speakers of languages that need it) in Lojbanized
names (and usage has been to use an apostrophe in that situation). The
baseline should therefore either eliminate the pronunciation as an
apostrophe (if audiovisual isomorphism in names is deemed critical), or
restrict that option only to names as a dispreferred option.
Hope this helps.
lojbab
--
lojbab lojbab@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org