#560: Allen comments on #511 & 531: Help in Creole (fwd)

[Robert Corbett <bcorbett@netcom.com>]

From: Jeff ALLEN <jeff@elda.fr>

Dear Haiti Listers, 

Warning, this is a long message.

In response to:

steven white <stevelong@spyral.net> wrote:
>> Does the orthography allow for words to be spelled differently
>> by people who pronounce them differently?

Guy Antoine stated in " #531: Help in Creole: Antoine comments "
with:

(note: I'm snipping out some sections in order to avoid reposting the entire
message, although all of the content was very relevant.)

Guy Antoine <GuyAntoine@windowsonhaiti.com> wrote:
>That's an interesting issue, indeed.  I would like to know whether
>there is a universal linguistic resolution to that dilemma.  
<snip>
>For instance, to say "I like grenadine juice", one might say
>(phonetically): "M renmen ju grenadin" (with French sounding vowels
>u in ju and e in grenadin) whereas another will say: "M renmen ji
>grenadin" with the currently sanctioned i and e of Haitian Creole
>orthography.  Another example, among too many to count, would
>be "Jezi Kri" versus "Jezu Kri" (Jesus Christ).
>I have raised that question before, and I was told by some Kreyòl
>linguists that for the purpose of an official orthography, as in
>English or French, regional (or other sort of) differences must be
>ignored in favor of the way the majority of people speak the
>language.  
>I am more than willing to accept this answer...
<snip>
>in the back of my mind, the suspicion of political correctness
>also arises ...
>we are in effect obliterating a certain reality,
>that is the way a substantial number of Haitians (admittedly
>a very small minority in the totality of Haiti) speak today.
>
>On a practical level, if I want to faithfully report the words of
>CityPerson X or FrenchSpeakingMan Y ...
<snip>

Guy, what you are bringing up is very valid. It is in fact not at all an issue
reserved only for Haitian Creole, but for all languages.  For languages that
have had a standardized orthography/spelling system for a few centuries (not
taking into account the changes, mind you), there has often been a body of
literature that attempts to continually represent the way the people "really"
talk.  My wife recently read one of the books by Mark Twain (I think it was
"the adventures of Huckleberry Finn).  In that book, there are lots of
examples
of sentences that are not written according to orthographic conventions, but
rather in a transcribed way (modified English) to make it look more real --
more like true, spoken utterances.   It took her a while to get used to
figuring out what the different transcription rules were, since she learned
written English as a 2nd language, and then spoken during the 5 years we
recently lived in the United States;   
One of the best types of literature to read for transcriptions of real-life
utterances are comic strips and books.  I had a friend who did her Master's
thesis in linguistics on the written language represented in French comic
books.

One of the hardest things that linguists and education experts have to deal
with is that the "spoken language" and the "written language" are often quite
different, yet it is still necessary to teach the "written forms" in the most
economical way (that is, using the least among of forms to represent the
maximum amount of possible sounds that are associated with meaning).  This
is a
very tough thing to do.  Yet, it does not end there.  Setting up the spelling
system/orthography is only the beginning.  It is then the task of the literacy
experts and teachers to put the entire spelling system into practice with the
people.  Granted, it is "sometimes" easier to do this when methodologies have
been tried, established, revamped, etc for decades, even a century or two.  It
is also easier to do when the people learning to read and write, speak the
language as their first language (unlike in many African countries where
children speak one language at home until the age of 5 or 6 and then go to
school and learn to speak, read and write French or English). This type of
factor makes it an even more difficult task for the teachers.   Another factor
that helps is when the parents actually already know how to read and write in
the language and can also help their children or family members who are in the
learning process of the "written form".

We have all been through the education or literacy process, or else we would
not be communicating by e-mail today on this topic.   One of the important
points to remember is that any "written language" is ALWAYS a second language
to the person who learns it.   The "spoken language" is the only true "mother
tongue" or "native language".  Spoken language is "acquired" by children.  On
the other hand, the written form is "learned" at school, in classes, etc as a
2nd language.   The spoken American English of Minneapolis, Minnesota is my
"native language", but the Standard Written American English (and all variant
forms of it) is a 2nd language to me.   I learned to master it over a
period of
many years of education in elementary school, secondary school, college,
graduate school, etc, accompanied of course by practicing the use of the
written language in different areas of my everyday life for a few decades.  
But, it remains a fact that the written form of English was a 2nd language to
me, as the written form of any language is a 2nd language to anyone who learns
it.  This is a main factor in the overall discussion.  We are all 2nd language
learners of the written languages that represent our spoken mother tongues.

I will address the rest of the questions asked by Guy and Steven in my
reply to
Emmanuel Vedrine below.  I will address Emmanuel's comments as they appear
chronologically in his message, so answers to points raised by Guy and Steven
come a bit later in my reply to Emmanuel.

In  Re: #511: Help in Creole (fwd)  Vedrine replies to Allen
Emmanuel W. Vedrine <evedrine@hotmail.com> wrote:
>French does a make a distinction between "e ferme" vs. "e ouvert" whereas 
>Kreyol doesn't and that the "accent aigu /ansantegi" comes into question. 

Not quite.  I'll start with French and then explain the issue for Kreyol. In
French, the distinction between the front mid-closed oral non-rounded vowel
[e]
and the front mid-open oral non-rounded vowel (written as a backward 3 in the
International Phonetic Alphabet) has 
been disappearing in many regional areas in France. The foremost example of
this is the distinction between the future tense and the conditional mode (for
ex. "j'irai" 
pronounced (zhire) vs "j'irais" pronounced (zhirE)  that are slowly both
conflating
to the (zhire) pronunciation).  This conflation/neutralization of pronounced
forms 
is not found across 100% of the entire French population, but it has been
attested by several French dialectologists/phonologists, notably Henriette
Walter (Universite de Rennes) and Chantale Hutinet (Univ Lyon 2), and in my
college French courses 10+ years ago they told us the same thing in the
advanced classes.

The issue at hand in the case for French is not that there is a 2-way
distinction, but rather
that there is a phonemic (not just phonetic) 3-way distinction.  The 3rd
phoneme
is the front mid-closed "rounded" oral vowel (small 0 with a diagonal slash
according to the International Phonetic Alphabet).

In Haitian Kreyol, there is "not" a 3-way "phonemic" distinction for the front
mid
non-nasal oral vowels.  It is only a 2-way distinction phonemically, and
yes in
Haitian Kreyol there is a phonemic distinction between the mid-closed e and
the
mid-open e.    

The following examples of "minimal pairs" demonstrate this:

te  = past tense marker
te` = land, ground

me  = but
me` = sea

se  = it's
se` = sister

There are plenty more examples.  Consult (VALDMAN, Albert (1978) Le Créole:
Structure, Statut et Origine.  Paris. Editions Klincksieck) in the chapter on
phonology for more info.   Lots of other books and articles provide examples. 
I simply do not have access to them right now because I am on the train going
to work this morning (yes, most of my mail list messages are during train 
commuting trips).

As seen above, there are 2 different phonemes (/e/ represented
orthographically
as "e" and /E/ represented orthographically as " e` ").  In Kreyol, there is
not a 3rd phoneme for e, to the best of my knowledge. (Of course, I am not
considering the speech of bilingual speakers who might very well
import/borrow 
French words into their spoken Kreyol, while maintaining the French
pronunciation, contain certain French phonemes that do not necessarily
exist in
Kreyol).  I've already addressed a lot of these issues of phonological
adaptation in the case of Creole context bilingualism, code-mixing and
code-switching in the following papers/thesis/presentations:

ALLEN, Jeffrey. 1994. Sainte-Lucie: relexification, décréolisation,
recréolisation ou adlexification? Diplôme d'Etudes Approfondies Thesis.
Département des Sciences du Langage & Centre de Recherches Linguistiques et
Sémiologiques, Université Lyon 2.  183 pp.

ALLEN, Jeffrey. 1994. Has the adoption of words from English led to a new
phoneme in Kwéyòl? Paper presented at the Workshop on Developmental Issues in
Creole Languages and Linguistics held at the University of Westminster
(Britain), 30 April 1994. 

ALLEN, Jeffrey. 1994. L'influence adstratale des variétés anglaises sur le
kwéyòl sainte-lucien. Paper presented at the Colloque Morphologie et Syntaxe
des Langues Créoles held at the Université de Provence (France), 23-25 June
1994. 


In French, the 3rd phoneme appears in very specific contexts and is instigated
primarily by the "loi de 3 consonnes" that does not permit clusters of 3
consonants (stops) (for you linguist types, this does not include homoorganic
fricative/stop combinations followed by a liquid, such as "str") and so a
vowel
is required to break up the cluster.  It's nothing new in the phonology of
languages.  The most common vowel to be inserted is the non-accented (aigu or
grave) "e".   One example, among many, of this includes:

gouvern(e)ment (where the e is pronounced)

However, what Vedrine may be getting at is the fact that there is a
neutralization of the pronunciation of the front mid-open and mid-closed at
the
phonetic realization level, in the speech of many Haitian Kreyol speakers.  I
have worked with some Haitians whose pronunciation of /e/ and /E/ is so close
(actually pronounced somewhere between [e] (front non-rounded mid-closed) and
[E] (front non-rounded mid-open) that I have had difficulty understanding when
they speak in Kreyol.  This pronunciation can also carry over to their
pronunciation of French and impedes communication with some native French
speakers who are not aware of this phonetic realization process in Kreyol.

One of the best examples to show the neutralization of the mid-vowels is the
equivalent of the English word "after" in Kreyol.  Here are the frequency
counts of this single word (represented by 4 different orthographic
variants in
the written form) taken from the frequency word list that I have mentioned on
the Haiti Mailing list several times in the past, and which is available
publicly (in the Research section at http://www.ling.su.se/Creole/ ):

  250 aprè
1539 apre
  505 apwè
  169 apwe

All of the examples above are for the same word, but depending on social and
regional factors, the pronunciation varies, and so does the written form.  So,
Haitians, who are in the process of learning to write, _might_ tend to write
how they pronounce. This is _not_ a proven fact, but rather a point that has
been deduced from dozens of examples taken from written texts and that has
been
discussed in:

ALLEN, Jeffrey.  1999. La standardisation du créole haïtien par
l'intermédiaire
de la linguistique computationnelle. Paper presented at the 9è Colloque du
Comite International des Etudes Créoles.  Held at the Université de Provence,
Aix-en-Provence, France, 24 - 29 juin 1999.

& 

ALLEN, Jeffrey. 1998. Lexical variation in Haitian Creole and orthographic
issues for Machine Translation (MT) and Optical Character Recognition (OCR)
applications. Paper presented at the workshop on Embedded MT systems of the
Association for Machine Translation in the Americas (AMTA) conference,
Philadelphia, 28 October 1998.


In essence, writing how one pronounces leads to the multiplicity of written
variants of words throughout the language.

How can we talk about attempts to standardize and normalize the written
Haitian
Kreyol language if Haitians will all be writing differently.  It risks
becoming
a chaotic mess for people who need/want to communicate back and forth.   And
what about implementing spell-checking programs within word-processors?  I've
discussed this in length on this list before.  
I'm giving a couple of seminars on the topic of language databases during the
next month in which the issue of language standardization is one of key points
I plan to bring up.   The quality of any written language is judged upon
consistency, and consistency is based on norms and conventions that have been
established for expressing onself (in written form) in the language. Sure,
there are lots of great computer programs we can develop to automatically
convert the orthography (see messages posted by Marilyn Mason), to correct
mistakes due to poor OCR output of scanned texts (Allen & Hogan; Mason), and
spell-checking scripts (Allen & Hogan). Yet, these technologies are not the
resolution to a deeper problem.  They can only be assistance tools.  It's like
saying that I will use the spell-checker in Word 97 to teach me to write in
French or English correctly because it will stop on each word during the
checking process.  This is a backward way of trying to get someone to reach a
level of good education and literacy.  

In my experience as a trainer of authoring and translation computer systems, I
found that those writers and translators who relied too much on the computer
system actually ended up spending twice as much time getting their work done
due to their false expectations of the systems.  Those writers and translators
who realized the limits of the systems, and decided to use these systems as a
basic productivity aid, were much more successful in speeding up their work.  

It is rather more important to teach people to read and write in their
language, and then teach them how to use such computer-assisted writing
software programs, but not to use these software programs as language-teaching
tools.  Although there are language learning software programs (for other
languages) that are appropriate for the purpose of teaching someone a language
through specific exercises, it is important to note that computer-assisted
writing programs are usually not intended to be used in such a way.

These same issues are relevant for language learning and literacy programs.

>but in written form 
>it's like we were to follow a norm though I argue the Kreyol is a phonetic 
>language. It's also good for quick reference (e.g, in research, through the 

Yes and no.  All languages are composed of phonetic sounds and can be
transcribed phonetically.  Yet phonetic description is only that -- a
description of the sounds that are perceived. It is the simple
transcription of
pure "sound".  I can also phonetically 
transcribe the sound of spitting, humming, etc in the middle of my utterances 
in English or French, even though these sounds do not make up what we consider
to 
be language based on  contrast of sounds, combined with semantic meaning.

By analyzing minimal pairs to distinguish contrasting sounds at the
articulatory level, 
based on semantic meaning of the pronounced sequences of sounds, then we enter
into
the arena of phonology.  At these point, one no longer talks about phonetics,
but rather about phonemics.

Spelling systems (orthographies) are not established for people to write
phonetically (exactly how it is pronounced at the given moment), but rather
phonemically.  Some languages (Kreyol, Spanish, German, Swahili) have spelling
systems that may be more phonetically representative than others (French,
English), but in writing with such orthographies we write phonemically, not
phonetically.  This is a very important distinction to make.

A good case for this is the English word "example" (or its French equivalent
"exemple) in Kreyol.  Should a Haitian write it "egzamp" or "eksamp" or even
"ekzamp"?  Or for the English word "exist" (French "exister" and the various
conjugated forms) , is it "egziste", "eksiste", or "ekziste"?  These are cases
of words that present Haitians with a difficulty to know how to write them due
to perception on the part of the learner.  In literacy and education classes,
Haitians have been told that whenever you write "x" in French, you need to
write "ks" in Kreyol.  Well, such a "rule" may hold true phonetically at the
surface level for the devoiced consonantal sequences [ks] (as in French word
Taxi, that is written "taksi" and pronounced [taksi] in Kreyol), but one must
not forget that in French, there are also many words that contain "voiced"
consonantal sequences [gz], written also as "x" (eg., exemple, exister,
exonerer, exalter, exaucer, etc) are that these sequences should _instead_ be
represented by "gz" in Kreyol.  Cases of phonological progressive assimilation
(forward direction) (the devoiced "k" influencing the devoiced "s" and the
voiced "g" influencing the voiced "z") are important to take into
consideration. Let's even take a look at regressive assimilation that in fact
influenced the written form in French.
There is regressive (backward direction) devoicing assimilation for the word
"observer" in French where this word is in fact /opsErve/.  Yet, we do not
write the word "opserver" in French.  Look however at English equivalent where
the same word can be viewed as undergoing progressive assimilation with
/obzErv/ written as "observe".  Yet, despite these slight phonemic adaptations
of phonetic strings, most of us end up mastering, more or less, the written
forms of these words over time, because we have learned to memorize the
written
form and no longer base our writing on pronunciation.  This is what
literacy is
about.

The goal of elementary school teachers, language teachers, literacy
specialists,
etc is to help learners of a language (whether it be native speakers learning
to
write in their native language, or something learning a language that is
completely
different from their native tongue) understand that there is a phonemic level
that
goes beyond the pure "sound" level.

The problem is that when people start writing phonetically, that is, how they
perceive and/or pronounce the sounds, they write the same words several
different ways, and this leads to variation in written form of the same
semantic units in the language.

I strongly emphasize this issue, after having spent the last few years
analyzing written texts of Haitian Kreyol that have been written by
approximately 20 different groups.  The inconsistency found in the writing
(for
the same exact words) reveals that many writers produce texts according to a
phonetic perspective rather than along phonemic ones.  The result is a
significantly high inconsistency of written forms for most words in the
language, not only between different groups of writers, but also within the
same groups.  

Why?  Most likely due to a lack of adequate and practical literacy training
from the phonemic perspective. 

>In closing, the issue was about "e aksantegi" and "e" (as oral vowel in 
>Kreyol)


But both of the front mid-vowels are "oral" vowels in Kreyol.

[e] written as (e) as in the word "Kreyo`l".
[E] written as (e`) as in the words "se` , me` , and pe` ".

The nasal "en" [phonetically e~] is a completely different phoneme in Kreyol.

>question is: is this a norm to follow or will be followed by all 
>writers who writing in the vernacular language?
>

My question is should there be a reason to make Haitian Kreyol, or any other
vernacular language, an exception to the last 50 years of literacy work all
over the world.  

Lots of these issues are discussed in detail in TABOURET-KELLER, Andrée et al.
1997.  Vernacular Literacy: a re-evaluation.  Oxford: Clarendon Press.

Best,

Jeff

=================================================
Jeff ALLEN - Technical Manager/Directeur Technique
European Language Resources Association (ELRA)  &
European Language resources - Distribution Agency (ELDA) 
(Agence Europe'enne de Distribution des Ressources Linguistiques)
55, rue Brillat-Savarin
75013   Paris   FRANCE
Tel: (+33) 1.43.13.33.33 - Fax: (+33) 1.43.13.33.30
mailto:jeff@elda.fr
http://www.icp.grenet.fr/ELRA/home.html