[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

#3961: Several people try to get Corbett unconfused!

From: archim <archim@globelsud.net>
   And don't forget to take into consideration that most of us who live in Haiti
(and who read and write!) have***never*** seen any Kreyol here written in the 
manner discussed by the persons on the list. Maybe they could get subscripttions
to any of the newspapers which print in Kreyol in addition to French, and they 
would see that there are no "A"'s with little circles over them (I don't even 
know how to type that symbol/letter on the computer!), or any of the other "new"
ways. Where is this "newly-unified" used?


> Here are the propositions I have in mind.
> 1.  I've thought it was maintained that the notion of the Haitian
> spelling system in the orthographic system is phonetic.
> 2.  Further, I thought from what Michel DeGraff tells me that ONLY the
> spelling system dictated from on high by prescription.
> 3.  This would mean that both pronunciation and grammar are based on
> ACTUAL PRACTICE  by speakers of Haitian Creole.
> 4.  Of course # 3 allows for such empirical concepts as "standard" (based
> on actual empirical evidence of a standard as what the large majority
> use) and a notion of a substandard.  This seems true of every language
> and can be empirically shown in studies of actual practice.
> ========
> I think those are the fundamental propositions which lead to my confusion.
> What we have heard recently -- persuasive arguments to me -- is that
> there is significant PRONUNCIATION differences in various areas of
> Haiti.  Evidence was cited from the Cap Haitian area as well as from the
> south.

Dear Robert:
    I have been following this discussion with great interest. I have been
studying Creole for about a year, but being a scientist as well I also have
a great interest in logic. Therefore it seems to me that the following
statements are true.

1) Although the government has decreed how different words will be spelled,
they have not as yet decreed how these words shall be pronounced.
2) Further, from the previous discussion of "yon", "youn" etc., we further
see that a given vowel can be pronounced in different ways depending upon
the word in which it appears.
3) Therefore, it can not logically be maintained that the Haitian spelling
system in the new orthographic system is phonetic, precisely because it
cannot be determined how a printed word will sound when it is pronounced.
Pronunciation is highly variable and completely depends on the speaker. The
most that could be said at the present time is that there is a range of
valid pronunciations for each Haitian word.
4) Therefore, the situation is not really too much different than the
situation in the US. There is a generally approved spelling for any given
word, but speakers from different parts of our country will pronounce that
given word according to the accepted pronunciation of that word in their
part of the country.
5) No amount of legislation in either country could do much to change this
6) But of course it must also be admitted that with the new standard
orthography, the vast majority of speakers in a given area of the country
will pronounce a given word similarly. Therefore, it can probably be stated
that when any given area of the country is considered apart from all the
other areas, the Haitian spelling system in the new orthographic system is
phonetic within that area.


Michael Fitzpatrick


From: Michel DeGraff <degraff@MIT.EDU>

I very much appreciate Corbett's earnest and honest request at
clarification.  Corbett's confusion is largely unwarranted, although I
suspect that he may not be alone.  So it's important we try to get this

Some preliminary remarks: It's an unavoidable fact that linguistics, as a
science, often clashes against our common everyday notions --- sometimes
the clash is very loud.  `Simple logic' just won't do, no matter what some
list members may believe.  This is not just a problem with linguistics:
remember Copernicus and Galileo trying to show that the sun does NOT
actually turn around the earth.  And yet we STILL say: "The sun rises",
"The sun sets", etc.  It's the task of all science to go beyond our
immediate perceptions, even the science of language (even if we all seem to
know so much about the languages we speak).  Corbett is surely aware of
Bacon's, Kant's and other philosophers' warning us to not turn our everyday
intuitions into scientific assumptions.  (Such warnings also apply to those
who so firmly think they know what constitutes "cultural tendencies"
instead of racist stereotypes.)

I'll try to clarify a few points that may have led to this confusion.

> 1.  I've thought it was maintained that the notion of the Haitian 
> 	spelling system in the orthographic system is phonetic.

No, the spelling system is not phoneTic (notwithstanding repeated claims on
this list).  As I've said a few times, the HC official orthography is
morpho-phoneMic --- I seem to remember that Jeff Allen has also made
similar points in his many comprehensive postings.  Phonetics is concerned
with actual sounds and articulation in ALL their diverse realizations in
speech.  Phonemics (phoneMics, not phoneTics) has to do with what sounds
are linguistically-relevant when native speakers distinguish among the
words in their language --- phonetics is in the physical world whereas
phonemics is in the speaker's head.  Although our everyday, common intuition
tells us that HC "spelling is phonetic" and that letters represent "sounds"
in a direct one-to-one fashion, in fact they do not --- the sun does NOT
rise.  Letters in the HC orthography only APPROXIMATE sounds; in fact, they
represent a more abstract level of the phonetic reality of the language ---
this level of representation is what linguistic call "morpho-phonemics".

Let's try an example, Haitian Creole (HC) speakers make
linguistically-relevant distinctions between a non-nasalized "a" and a
nasalized "an".  Witness the HC words "sa" (= "that") vs. "san" (=
"without", "hundred").  But no such PHONEMIC distinction is made between a
non-nasalized "i" and its nasalized counterpart; that is, there is no pair
of words that are distinguished by the (non-)nasalization of "i"; this is
UN-like "sa" vs. "san".  Therefore no special symbol is required for this
nasalized "i" and such symbol does not exist in the HC official
orthography.  But note that this would not prevent a HC speaker to produce
a nasalized "i" in the word "machin" say.  The point here is that this
nasalized "i" (vs. a non-nasalized "i" as in "Mari") need not get a special
symbol.  So what happens is that one symbol "i" is going to represent both
nasalized (in "machin" for some) and non-nasalized (in "Mari) versions of
the phoneme "i".  There is no one-to-one relation here between "i" and the
"sounds" it approximates.

So, you see, HC spelling is not phoneTic, but only phoneMic.  If it were
phoneMic, then we would have to use something like the International
Phonetic Alphabet in order to represent all kinds of possible sounds, even
those that are not used by HC to distinguish between different words.

Similarly, MOST (and, I repeat, MOST) HC native speakers do not
SYSTEMATICALLY distinguish between the unrounded vowel "e", as in HC "kle"
(= English "key") and its rounded counterpart. The latter is found, say, in
French "bleu" (you must `round' your lips to pronounce the French "eu").
Whereas French phonemically distinguishes between the words for "blue" vs
"wheat" (French "bleu" vs. "ble'"), HC does not: HC "ble" in MOST dialects
is ambiguous between "blue" and "wheat".  Since French "eu" does not exist
in MOST HC dialects, the official orthography has decided not to include
any symbol for this vowel in the official orthography.  (Of course,
speakers who have "eu" in their Creole are free to round their lips and
pronounce HC "ble" (= "blue") like French "bleu" --- you say "tom-AY-to"
and I say "tom-AH-to"...)

One may decide to call such choices "prescriptive", but any morpho-phonemic
orthography has to choose a core set of dialects to work with, otherwise
the task become unwieldy.  Given that speakers can (in principle) be
multi-lingual, it would be rather un-economical if any particular
orthography had to represent pronunciation variations that are due to
influences (`accents') from other languages.  In the case of HC, the other
language that influences the pronunciation of HC is French, specially among
the (somewhat) bi-lingual speakers --- and we all know (and this list has
shown us time and again) how stupidly Francophile we Haitians can be.  But
as Delimon and Canal in this forum have made very clear, there are good
reasons why HC orthography should remain autonomous.

So far so good?  Let's keep following the strand of Corbett's confusion.

> 2.  Further, I thought from what Michel DeGraff tells me that ONLY the 
> 	spelling system dictated from on high by prescription.  

Yes, "prescription" which is largely based on observation of the CORE
dialects of HC and the desire to have an orthography that can faithfully
approximate the basic morpho-phonemics of these dialects --- the latter
stand in for what Corbett calls "the large majority use".

> 3.  This would mean that both pronunciation and grammar are based on 
> 	ACTUAL PRACTICE  by speakers of Haitian Creole.

Yes, but note, again, that "pronunciation and grammar" do NOT, and could
NOT, reduce to "spelling".  The former exist prior and independently to the
latter.  As of "grammar" (in the sense of syntax), most of it falls outside
of "spelling" altogether.  Your spell-checker will not `check' on your
syntax, will it?

> 4.  Of course # 3 allows for such empirical concepts as "standard" (based 
> 	on actual empirical evidence of a standard as what the large majority 
> 	use) and a notion of a substandard.  This seems true of every language 
> 	and can be empirically shown in studies of actual practice.


> I think those are the fundamental propositions which lead to my confusion.
> What we have heard recently -- persuasive arguments to me -- is that 
> there is significant PRONUNCIATION differences in various areas of 
> Haiti.  Evidence was cited from the Cap Haitian area as well as from the 
> south.

Yes, and this is true of ALL spoken languages.  No language is spoken
identically and with a unique single grammar across all of its speakers.

> Now, my confusion rises.  If:
> 1.  spelling is phonetics and based on how people ACTUALLY SPEAK (not how 
> 	some select intellectuals, scholars and government officials speak or 
> 	wished others spoke)

This is, again, one (just one) of the reasons for the confusion: spelling
is definitely NOT phonetics (see above).  Yet the orthography was designed
with the goal to APPROXIMATE how MOST HC speakers actually speak. Then
again, no morpho-phonemic orthography can approximate how EVERYONE
"actually speaks" (if this is unclear still, see above and, specially,

> and if
> 2.  there is pronunciation variation
> Then it would seem to logically follow that there would be differential 
> spellings depending upon who was the speaker.

Indeed.  But such "differential spellings" and their relationship to
"pronunciation" must still fall within the bounds defined by the official
orthography.  The main goal of the orthography was simply (and modestly) to
establish a systematic relationship between written text and the
morpho-phonemic level of pronunciation. The official orthography never
tried to prescribe how people speak.  The orthography only establishes
norms as to how the most common documented varieties of HC speech are to be
represented by the written text.  The structures of speech exist prior and
independently to the orthography.  I made that point earlier when replying
to a previous mis-understanding of my position by Corbett:

  "I discused the spelling rules as they have ALREADY been decided since the
  1979/1980 law.  These rules are meant to closely reflect the pronunciation.
  But the latter is a matter of FACT, which cannot (and was not meant to) be
  prescribed by the law.  Again, Creole pronunciation exists PRIOR TO, and
  INDEPENDENTLY of, the Creole orthography, and the latter was designed to
  represent the former as closely as possible."

Actually I also illustrated that very point with respect to "youn", "yonn",
etc.  Let me repeat the relevant passage:

  "As far as I know, there is no such `officialization' [regarding which of
  "youn", "yonn" must be used].  Unless there's been some recent
  development that I am still unaware of, the only thing that's been
  `officialized' is the orthography, per the 1979/1980 laws.  As it turns
  out, this orthography is flexible enough to capture the variations (or
  `allomorphy' using linguistic jargon) in "yon", "yonn", "on", "you",
  "youn", etc, and other such attested pronunciation variations among
  EXISTING Haitian Creole dialects.  By "existing dialects", I mean the
  Haitian Creole varieties spoken by Haitian native speakers across the
  country.  In fact, much work has been done to document the
  characteristics of EXISTING Haitian Creole dialects.  There's a recent
  monumental thesis on the topic by Dominique Fattier; see
  http://www.septentrion.com/ for ordering information --- do a search on
  keyword "Fattier"."

Corbett's confusion here may have to do with the distinction between the
concepts `orthography' vs. `pronunciation'/`lexicon'.  Having an official
orthography does not mean that the pronunciation and the lexicon of the
language are fixed once and for all.  The case of "youn/yonn/etc" is
particularly clear: the spellings of "youn", "yonn", "yon", etc. ALL of the
variants here do obey the official orthography, yet they belong to
different dialects (qua regional pronunciations).  This clearly shows that
the official orthography does not dictate a uniform pronunciation and

There always will exist dialectal variations in any language.  One
advantage of the HC official orthography is that it does allow an
approximation of the morpho-phonemics of the CORE dialects of HC.

> Two things do not seem to me logically to cohere:
> 1.  That pronunciation may well have strong regional features.  But, 
> 	pronunciation is not prescribed for the nation and the notion of 
> 	pronunciation is empirical.
> 2.  There is ONE approved spelling, and this ONLY is prescribed.  The 
> 	rest of the study and knowledge of Haitian linguistics is empirical.
> ==============
> Where does this seemingly simply logical puzzle go wrong?

Where Corbett's reasoning went wrong may be in the confusion between
"spelling", "pronunciation" and "lexicon".  So let me paraphrase (once
again) what I've said a few times earlier: It IS possible to have ONE
official spelling WITHOUT dictating ONE official pronunciation and ONE
lexicon (see the "youn"/"yonn"/etc example above, which I take it is
particularly clear).  What the official orthography does is give us the
means to morpho-phonemically represent a set of dialects, the set of CORE
dialects that MOST Haitians have been DOCUMENTED to speak.  This set of
dialects does include "regional features", such as "ape" vs. "ap"; "plasaj"
vs. "plasay"; "li vini" vs. "i vini"; "koyido`" vs. "korido`"; "je"
vs. "zye"; "mren" vs. "mwen"...  (Anyone remembers the "Men mren" / "Men
mwen" passage in Pe`len Te`t?  Both "mren" and "mwen" fall within the norms
of the official orthography.)  The Freeman/Laguerre dictionary gives a
rather comprehensive list of examples that show how different
pronunciations can be straightforwardly accommodated by the official
Of course, this CORE set of dialects is not without ends.  The official
orthography does exclude other dialects (e.g. the dialects with the French
"eu" phoneme).  This is for reason of economy, feasibility, demography,
identity, etc.  Jeff Allen and Francois Canal (and MANY others in print:
Dejean, Vernet, etc) have addressed some of the reasons why certain
(non-representative) dialects need not be covered by the official

One side note: Corbett may have also gotten confused by the "mambo"
discussion and by the repeated (erroneous) claims that "Mambo IS
phonetically correct".  So let me take yet another stab at clarifying this
issue.  In fact, the simple reason why "mambo" in HC is unacceptable is
that, according to the official HC spelling, "mambo" in this spelling would
be pronounced with a non-nasalized vowel "a", the same vowel in HC "dam"
("girl, lady") and "vag" (= "wave").  Such a "mambo" pronunciation is, as
far as I know, not to be found in ANY existing HC dialect.  The nasalized
vowel in the first syllable of the word is accurately represented as "an",
which corresponds to the actual pronunciation of this work by HC speakers.
In other words, the PRONUNCIATION of "mambo" in HC (per the official
orthography) violates the "actual practice" of HC speakers, as documented.


Does this help?  


It just occurred to me that there is yet another factor that may have
caused Corbett's and others' confusion.  Corbett (and others) may have been
extrapolating assumptions about English/French spelling to Haitian Creole
(HC) spelling.  English/French writers often assume that there is ONE (so
called) `correct' way to spell given words, even when native speakers vary
in their pronunciations of the word.  Witness the UNIQUE spelling "tomato"
and the pronunciation "to-MAY-to" in the U.S. vs. the pronunciation
"to-MAH-to" in the U.K.  

The crucial point here is that the HC official orthography does not even
try to establish such unique spellings.  What the HC official orthography
does is this and ONLY this: Given some documented pronunciation(s) for a
given word and its variants ("yonn", "youn", etc.; or "soti" and "so`ti";
or "mwen" and "mren"), the official orthography rules tell us how to
approximate these pronunciations with sequences of letters.

In fact, there is NO way the HC official orthography could establish
UNIFORM spellings for the ENTIRE set of HC words.  The COMPLETE set of
official HC orthography rules can be summarized in less than 10 pages (see
for example the relevant web pages of http://www.windowsonhaiti.com ).  And
I assume that we all know that HC has many more words than could fit in 10
pages.  So not only is it LOGICALLY impossible for the HC official
orthography to prescribe UNIQUE pronunciations for given HC words, but it
is also PHYSICALLY impossible to do so. Given dialectal variations, it is
thus expected that many words will receive "differential spellings" (in
Corbett's terminology).  In fact, given a morpho-phonemic orthography that
tries to stay close to an abstract representation of pronunciation AND
given dialectal variation, "differential spellings" become the `norm' ---
unlike the English "tomato" case.

To recap: The set of "differential" pronunciations in HC core dialects is a
matter of documentable "actual practice".  What the official orthography
does is to PRESCRIBE rules whereby one can systematically `translate' (most
of) the words from these documented core dialects into written text, and

As Corbett correctly stresses, the set of (core) pronunciations constitutes
a matter of fact: they are empirical.  But the orthographic rules (i.e. the
phoneme-to-letter and letter-to-phoneme rules) constitutes an ARBITRARY
MAN-MADE ARTIFACT the goal of which is to provide an economical, mechanical
and logical (if imperfect and limited) way to go from speech to written
text, and vice-versa.  As I discussed earlier, rules of spoken language and
rules of reading/writing do not belong to the same domain.  Rules for
reading and writing (i.e. literacy), unlike spontaneous speech, are the
products of planned and conscious human intervention, and the two sets of
rules (literacy vs. language) evolved via different psychological and
social processes.

In coming up with the HC official orthography, the linguists involved had
to make some arbitrary (but scientifically, pedagogically and
sociologically motivated) decisions as to what to represent, and how.  But
one thing which they did NOT intend to, and could NOT, enforce is a UNIFORM
pronunciation for ALL HC words.  I yet have to read an OFFICIAL document
from the Haitian state that legislates which of the "yonn"/"youn"/etc
variants is correct.  Neither do I know of any OFFICIAL document that
forces speakers to say "M ap vini" vs. "M ape vini".  Again, such
widespread variants (unlike others which were deemed non- or less
representative) can all be straightforwardly represented within the norms
of the official orthography.

Again, given various feasibility, pedagogical and sociological factors, the
official orthography does assume that there are CORE dialects that must be
representable and represented, and others (the NON-core dialects) that need
not be.  This is not surprising: in order to work efficiently and to be of
optimal service, any man-made symbolic system that is intended for massive,
popular usage has to put some limits on what it will be able to represent,
otherwise it becomes intractable and impractical.  So far, there is
widespread agreement among well-informed linguists and educators that the
HC official orthography IS tractable, practical, logical and efficient for
the vast majority of its target users.

I hope that this extra bit will clarify any left-over mis-understanding.


MIT Linguistics & Philosophy, 77 Massachusetts Ave, Cambridge MA 02139-4307
degraff@MIT.EDU        http://web.mit.edu/linguistics/www/degraff.home.html