28

01h2LexDiffFin

/renmen/ "to love" which reflects /er/ reduced to /r/ before vowels as it does
productively in any variety of Québécois French:<raller> "to go again, to return",
<rabrier> "to cover (again/up)", <rajouter> "to add on", <ressayer> "to try", <rôter>
"to take off", etc.

The (AA) difference builds crucially on the fact that Saint-Thomas French has more
than two exceptions to the /re/ generalization. That's conveniently chopping my
Bonventure County evidence which came right after the evidence from Saint-Thomas
French.

If you think Saint-Barth is too close to the creole area, then let's take the
koine variety of Bonaventure County which is enclaved between the non-creole
non-koine Acadian dialect area and the non-creole koine varieties of Laurentian
(we already talked about), both these surrounding areas maintaining productive
schwa-insertion. In Bonaventure French, schwa-insertion in front of /r-/ is just
as exceptionnal as in Saint-Barth French and Haitian Creole.

Schwa-insertion in front of /r-/, your /er/ variant, is so exceptional in Bonaventure
County French I can't come up with a single example. This "indubitably strengthens
my case" that Bonaventure County French must have gone "trough an affixless stage"
because it beats both Saint-Thomas French and HC on exceptionlessness.

>rejè vs. *rijè (cf. French rejet `reject')

Interesting example. Why wasn't final /t/ preserved, as it is in most Haitian
words? [...]

Even in Québécois, this /-t/ is unevenly preserved.

Agreed. But this would still leave unexplained the mid-low quality of the final
vowel: Haitian mid-low vowels normally are the result of French mid-low allophones
of vowels closed by /r/, which became phonemes in Haitian after the fall of
syllable-final /r/; otherwise the distinction between mid-low and mid-high vowels
was neutralized (cf. Haitian /pe/ from PAIX and /pè/ from PèRE).

In short, I stand by my earlier claim that /rejè/ is a French loanword in Haitian.

There are seven full pages of examples with word-final mid-low /è/ in Freeman's
reverse dictionary to show that French /er#/ is not the only source for HC /è#/. How
much full of fuddle duddle can you be?

The simplest hypothesis is to see the e-schwa/i-schwa alternation as the outcome
of competing changes. Ever heard of lexical diffusion?

I am familiar with the theory.

It doesn't look like it. If you were, you'd be using it.

But why should the i-schwa variant *only* be found in a few *basic* lexical items?

The "few *basic* lexical items" you came up with are:sila, vini, tini. I have
shown on comparative evidence that /vini/ and /tini/ in koine-derived varieties of
French are not items reflecting an earlier i-schwa. The remaining item /sila-a/
"this, this one" [Sylvain 1936:58 knows no /sila/, only /sila-a/] most certainly
reflects Québécois French /sula-la/ "this one" which Fournier (1981) shows to be the
agglutinated demonstrative <celui-là> followed by the determiner <là>. [I had missed
that one, sorry, but I'll correct the final version of the paper accordingly.]This
leaves us with the following question:Why should Goyette's i-schwa variant be found
in *any* lexical item of HC?

Why, furthermore, does Haitian Creole have so many different reflexes for a number
of French phonemes?

With all the i-schwas gone, I only see e-schwas left. By the way, I have shown to
some extent in Wittmann (1996; with more details forthcoming) how the phonological
systems of the colonial "Frenches" vary geographically on a continuum from the utmost
North trough the Americas to the utmost South in the Indian Ocean.

28

Printed for Henri Wittmann <hwittman@cgocable.ca>

29

01h2LexDiffFin

Lexical diffusion can explain the odd irregularity in laws of sound change, but not
the huge number of such "irregularities" found in Haitian. I therefore stand by my
earlier claim that this can only be accounted for by massive and continuous French
influence.

This just leaves one pitfall in Goyette's scenario unattended:What would be the
exact variety of French Haitian Creole borrowed "irregular" /re/ from?

Standard French seems to be the only variety around with enough muscle to
debasilectalize "regular" affixless varieties or "regular" varieties with /er/ to
"irregular" varieties generalizing /re/. If Standard French is indeed the source:
(1) Why is it that Québec where debasilectalization is so much stronger than in Haiti
failed to convert to Goyette's /re/? (2) If HC adopted the /re/ rule from Standard
French, why aren't there in HC any /ré-/ reflexes before vowels such as in Standard
French <réactiver>, <réajuster>, <réécrire>, <réentendre>, <réessayer>, etc.? (3)
How do we explain that examples with /re/ from HC and other French creoles are
attested BEFORE the "correct pronunciation" of the Paris standard changed to /re/
considering that in the second half of the 19th century the (in Nisard and Bauche)
attested non-standard varieties of Paris French had /er/? (4) Why should an
anticlerically orientated "revolutionary" etymological revial movement of the French
Republic (= an ideology reviving for political reasons procunciations which are
historically dead) have any linguistic impact either in Québec or in Haiti?

APPENDIX 7:THE "SONTAIENT" SYNDROME

Some subscribers on this list seem to be worried on different counts re the pursual
of the debate on lexical diffusion. There is nothing to worry about, I just happened
to be away on a lecture tour. Color posters of the event in Ottawa are supposed to
be available at this site:

http://www.carleton.ca/fls/pastevents.htm

The title of the lecture though had been changed to:

"Syntactic frills of Canadian French in a comparative historical perspective:The
'sontaient' syndrome."

Stephane Goyette had been invited to come and heckle, but, though in town, he was
unfortunately unable to come.

I'm now back. Thanks for the concerned inquiries.

APPENDIX 8:REPLYING TO STEPHANE GOYETTE III

Stephane Goyette wrote (26/02/01):

No wonder <pied> generalized to generic "plant, tree" and to derived uses in
lexical compounding for a number of varieties of "Frenches". Even the non-creole
French documented in Highfield has /pyé/ "tree, bush, plant" as well as /pyé de
bwa/ "tree".

But this non-creole French is creole-influenced.

Show me one other creole feature in Saint-Thomas French that I wouldn't be able to
document in another variety of French.

Otherwise no variety of overseas French has "pied-de-bois" or similar such compounds
TO THE EXCLUSION of the monomorphemic term "arbre". Consultation of the ALF (Atlas
Linguistique de la France, in case Professor Wittman doesn't know...) shows that
forms related to ARBRE are quasi-universal throughout France as the word for "tree".

It's the "quasi" that's interesting. As I read this, I just happen to have on my
desk Patrice Brasseur's edition of Louis Bizeul's Dictionnaire Patois du Canton de
Blain (around 1850). On page 127, I see <piétaud> translated by a bimorphemic "jeune

29

Printed for Henri Wittmann <hwittman@cgocable.ca>

30

01h2LexDiffFin

pied d'arbre" for "young tree". Why, in 1850, would Bizeul be using a bimorphemic
<pied d'arbre> instead of monomorphemic <arbre>? Because, at that time, everywhere
in France <arbre> /arb/ "tree" was still homophonous with <herbe> /arb/ "grass".
This is still so in ALL varieties of basilectal Québécois French:/pyédarb/ is used
whenever it's crucial to convey that /arb/ doesn't refer to the grass of your lawn.

I therefore concluded that this is a creole innovation. That the botanical usage of
"pied" in French served as a model for /pye-bwa/ is something I do not doubt. But
for such a compound to be used INSTEAD of "arbre" definitely ranks as a *creole
innovation*, rather than an inheritance from dialectal French pure and simple.

Since (a) <pied> is a noun that can mean "tree", (b) <bois> is a noun meaning "trees,
forest" and (c) "N de N" is a productive model for creating semantically transparent
compounds in French, there should be a way testing whether <pied de bois> is a
possible word in the lexicon of native speakers of French. Here is what I got in
Trois-Rivières with a quick survey à la Labov as possible interpretations for <pied
de bois> (number of votes for each alternative within parentheses):

(A) = <pied d'arbre> "tree" (4).

(B) = <pied d'arbre de forêt, pieds d'arbre d'un bois> "forest tree(s)" (3).
[According to one source, in the old days, "pieds de bois" were the forest trees
marked for felling the next day determining the work load for the day.]

(C) = <pied d'arbre, tronc d'arbre (avec la racine)> "tree trunk" (2).

(D) = <souche d'arbre> "stump of a tree" (1).

(E) = <pied de reboisement> "young tree grown for reforesting" (1).

(F) = <pied de bois> "foot prosthesis" (1). [On the model of <jambe de bois> "wooden
leg".]

As can be seen, the transparency of "pied de bois" doesn't leave native speakers of
non-creole French indifferent which goes to show that "pied de bois" cannot
"definitely rank as a *creole innovation*".

Someone on this list has suggested that Goyette's data on French are "unreliable".
For the benefit of the list owners and all those non-natives of French who
unwittingly but in good faith had recourse to any such data, I'd publicly submit
the following question:To what extent is it likely that a native speaker of
non-standard and standard varieties of French with a PhD in linguistics (everybody
I asked assures me this is actually so) would not know the botanical extensions in
the meaning of the French word <pied> "foot" and derived lexical compounds of the
type "pied d(e) X"?

I'd publicly submit the following question:To what extent is it likely that
someone who has claimed in print that people who believe creoles to exist as a
separate language type are racists would be able to coolly evaluate linguistic
evidence pointing to creoles indeed existing as a separate language type?

Where did I call you or anyone on this list a "racist"? You might as well apologize
on the spot.

Though I would subscribe to to Salikoko Mufwene's (04/02/01):

Yes, I am saying that there is an embarrassing racial bias that the field of
creolistics should feel guilty of (all of us regardless, of color, and in some of my
earlier publications too). Moreover, the accounts smack of racism, though I have no
evidence for accusing any of the creolists themselves of racism. Positions can be
racist while their proponents need not be. It's a matter of what is implied by what
we say.

and I probably said so somewhere. As it is, "ditherings of unconscious albocentrism"
affected the first research project I ever wrote up, around 1959 or so. The
substrate explanation for the frills of creole French seemed to be the only
reasonably attractive explanation and I would no doubt have persisted in that belief

30

Printed for Henri Wittmann <hwittman@cgocable.ca>