IMAGE 01h2LexDiffFin01.gif



rather detailed description of Magoua (and its relation to other koines and creoles)
he ostensibly claims here not to know about. Given the eclecticism of his knowledge
of "non-standard varieties of French spoken overseas" (more on that under 9.6), I
might be forgiven if I use the pedagogics of symptomatology in relating Magoua to
other "colonial varieties of French".



IMAGE 01h2LexDiffFin146.gif
IMAGE 01h2LexDiffFin06.gif

Koine French varieties also don't have any plural endings, pronominal cases
(pronouns such as mouén, toué, etc. are invariable)[...]

Interestingly, while "mouén" is claimed to be invariable according to case,
Wittmann's own data abounds with cases of not only "mouén", but also "j", which,
given that it is translated as a 1sg pronoun can hardly be related to anything else
than Standard French "je". Exactly what does Wittmann mean when saying that
"pronouns such as mouén [...] are invariable"?


IMAGE 01h2LexDiffFin16.gif

(16)j-à-pong là-hàch j-à-kup là-branch
"I cut the branch with the ax."

However, note that this sentence contains two subject occurrences, something that
disqualifies it from the status of a serial verb constructions in many (perhaps
most) analyses. [...]

This is a very neat attempt to demolish something out of context. Where did I say
that /j/ translated as "a 1sg pronoun" or that it occurred as "subject occurrences"?
What is it exactly what I mean when I say discussing examples (1a) to (6b):

IMAGE 01h2LexDiffFin41.gif
IMAGE 01h2LexDiffFin41.gif

The erstwhile "disjunct" pronoun (mwén, toué, lui, èl, nouzot, vouzot, euzot)
succeeds the erstwhile "conjunct" pronoun of earlier French in all bare DET
positions, that is to say in all positions other nouns do, whereas the erstwhile
"conjunct" pronouns [to wit:the earstwhile pronouns corresponding to Standard
French je, tu, il, elle, on, ils] succeed in (non-argument) agreement positions to
the erstwhile "bound" suffixes (such as -e/-s, -(e)s, -e/-t, -ons, -ez, -ont/-ent)
of earlier French.

I think I thought I was referring to something quite consensual in Romance
linguistics, something you ought to know about, to wit that preverbal predicate
markers of the type of /j-/ or /on-/ as in (29a):

(29)a(mouén) j-manj, (nouzot) on-manj, etc.

had been replacing conjugational suffixes of earlier Latin such as /-o/ or /-amus/ in

(29)b(ego) am-o, (nos) am-amus, etc.

This is not just a syntactic frill of Magoua, it's the quirk that defines
syntactically any variety of koine French (what goes with the French as "français
populaire", "français avancé" or with the German School under "néo-français",
"Neo-French", "Neueres Französisch", etc.). A very partial bibliography of sources
besides myself you might want to turn to would be (from Wittmann 1998):Meyer-Lübke
(1894, 1909), Richter (1911, 1933), Bally (1909, 1913, 1932), Rohlfs (1928), Meillet
(1921), von Wartburg (1943), Queneau (1947, 1959), Tesnière (1959), Martinet (1962),
Sauvageot (1962), Weinrich (1962), Rothe (1965, 1966), Pulgram (1967), Baldinger
(1968), Grafström (1969), Söll (1969a, b), Wandruszka (1980), Hausmann (1975, 1979,
1980), Ernst (1980), Schmitt (1980), Wüest (1985), Harris (1976, 1978), Ashby (1977,
1980, 1982, 1988), Lambrecht (1980, 1981, 1986, 1988), G. Sankoff (1982), Ossipov
(1990), Roberge (1990), Nadasdi (1995). Since this is the defining frill of koine
French (or "Neo-French", if you prefer), it is present in all "non-standard varieties
of French spoken overseas" you claim to be familiar with, except Acadian and the
Standard French taught in schools where (29a) translates as in (29c) and (29d):

IMAGE 01h2LexDiffFin01.gif


Printed for Henri Wittmann <>

IMAGE 01h2LexDiffFin01.gif



(29)cje manj, je manj-on, etc. (Acadian)

(29)dje mang-e, nous mang-eons, etc. (FS)

It has reflexes in the Indian Ocean you know well about, the "mysterious" /i/ of
Seychelles and Reunionese creoles. The categorial downgrading of erstwhile pronouns
in subject doubling positions is a cross-linguistically well-documented phenomenon,
even in the English-based pidgins and creoles (cf. Tok Pisin /i-/ from erstwhile
pronominal <he> in English).

In the most basilectal varieties of koine French (which for socio-historical reasons
mostly survive in the Americas), clitic agreement is symptomatic of a more
fundamental typological change:a complex of frills collectively characterizing an
evolutionary jump. That is what I call here the "sontaient" syndrome, a clear case
of "resyntactification":

(30)aErstwhile "lexical" pronouns in syntactic positions doubling subjects are
categorically downgraded to inflectional items marking subject agreement.

(30)bErstwhile "conjugational" copula paradigms are regularized freeing an (30ba)
erstwhile "present tense" stem and an (30bb) erstwhile "past tense" affix stem for
recovery in syntactically reanalyzed positions or functions. [In koine French, this
is the paradigm proposed in (12):<je_suis tait> /ch te/, ..., <son tait> /son te/,
something I suggested for the first time in 1972 on the basis of orthographies of the
type <est tes> for <était> "was". The reanalysis of <était> as /son te/ is rendered
in the literature from the 19th century on as <sontaient>, i.e. the standard French
3PL <sont> "are" plus the regular 3PL suffix <-aient>.]

(30)cInflectional items of type (30ba) compete with inflectional items of type
(30)a as "grammaticalized" subject agreement markers. [These are the agglutinated
chu-/j- <je_suis> type clitics competing with j-/ch- <je> type clitics in the Magoua

Erstwhile pronominal copulas as agreement clitics or affixes, again, are well-known
cross-linguistically documented frills of syntax. It's a feature of all the
"sontaient" dialects of koine French, to wit all the varieties of non-standard French
you happened to look at, and extensively documented in the literature. Since you
don't accept creole samples in etymologizing garb, you'd be able to see through the
garb of <sontaient> too, especially since it had been analysed in detail as a
"creolism" in the paper you extrapolated your Magoua data from.

Another thing "I apparently need to repeat once more" to judge from the insinuations
is the answer to the question:Who speaks Magoua?

The ethnic Magouas are the descendants of 17th century "coureurs de bois" type
squatters in the Trois-Rivières area of Québec. Though only one Magoua village
subsists in modern days and the ethnonym is highly stigmatized, Magoua speech has
become everybody's basilect in Trois-Rivières, the in-group type of speech even the
most educated must resort to in informal family reunions. The area has the highest
illiteracy rate in Québec and relevant research has shown that lack of acquisitional
models for a conjugated copula is the most important hurdle on the way to literacy.
There are three features of speech highly stigmatized that are undergoing
"regularization" on the model of mesolectal general Québec French:(a) The 3PL son-
is regularized to a son-/i- alternation according to context though; again, relevant
research with subjects from literacy campaigns show that this does not restore a
"conjugated" copula in any sense;(b) The imperative and the 2PL present
habitual-stative are regularized on the model of long forms with -é, the latter
though in the present tense only, not with the imperfective -e;(c) The homophony of
final vowels in manjé/fini, peú/pu, fó/fou is resolved by a rule of final vowel
raising affecting /i u ou/.



IMAGE 01h2LexDiffFin16.gif
IMAGE 01h2LexDiffFin16.gif

His case for gender seems to be based entirely on the "article system" of
historical French as it survives in modern Standard French. In Parkvall
(2000:78), on the other hand, he bases his conclusion which goes to admit that the

IMAGE 01h2LexDiffFin01.gif


Printed for Henri Wittmann <>

IMAGE 01h2LexDiffFin156.gif



post-nominal article /-la/ of creole French cannot be substrate-induced on
Fournier (1998; with further references extrapolated from there) where alternate
strategies for definitenes are said to have developed because the etymological
"article system" did NOT survive in koine French.

Though I don't know anything about Magoua except for what Wittmann himself has
taught us (since there seems to be no other source to turn to), I apparently need to
repeat once more that I am familiar with non-standard varieties of French spoken
overseas. In virtually all of these, gender is manifested (albeit less so than in
modern standard French) not only in the article system, but also in agreement and in
pronoun usage (even were, as in many Cajun dialects, "ça" is possible in addition to
derivates of "il" and "elle").

Though the basically Greenbergian idea that articles grammaticalize to noun class
clitics or affixes as new articles emerge from demonstratives in a cycle of
categorial downgrading alternating with compensatory morphological renewal has
become standard fare in historical linguistics (cf. Wittmann 1983), Parkvall in
his extensive treatment of noun class phenomena (2000:81-83) is either unaware of
this or such input has been weeded out in a cafeteria-principled way.

I am not at all blissfully unaware of that. It's just that I fail to see the
significance of it here.

The "the significance of it here" is fully documented in Fournier (1998, with
reference to earlier work, Wittmann & Fournier 1982, Wittmann 1983, etc.) from which
Parkvall extrapolates so extensively on page 78 of Out of Africa. Why divorcing so
ostensibly article system renewal of pp. 78-79 from article system decline on pp.
81-83 to the point of styling the latter as "Reinterpretation of morpheme and lexical
category boundaries"? After all, all that was achieved on pp. 78-79 and 81-83 is
demonstrating in reversed order that an erstwile prenominal article system of earlier
French agglutinated in creole French whereas a new postnominal article system arose
in replacement from an erstwhile postnominal demonstrative, likewise of earlier

The 168 African languages suspected of subtrate induction belong all to the
Niger-Congo family of languages, except for 121 and 122 (Hausa and Margi, and they
don't loom very large in the argumentation). Greenberg's seminal work taught us this
about Niger-Congo:

(31)aAn erstwhile prenominal article system was downgraded to a morphologically
rich noun class system pairing into 9 genders on a continuum that ranges today from
"fully functionally class markers" to class markers "in a state of attrition" and
"fossilised class markers".

(31)bNew determiner systems developed out of "demonstratives" and show either "DEM
N order" or alternatively "N DEM". Alternatively, definiteness may be expressed by
class concord marked on the verb in varieties with "fully functionally class

What is shown here in quotation marks are things Parkvall admittedly knows about.
Why does he so ostensibly "fail to see the significance" of a relation between the
two? Since the agglutinated articles of creole French are equated to "fossilised
class markers" in Niger-Congo languages, why doesn't he explore the "material" he so
explicitly read about in Fournier (1998):The withering of the prenominal article
system of Old French subsisting on a koine-creole "article agglutination" continuum
ranging from "fully functionally class markers" to class markers "in a state of
attrition" and "fossilised class markers"? Why hide from readers who mostly ignore
the history of either Niger-Congo or Gallo-Romance that the fossilized noun class
system of creole French out of a fully functionally three-class noun gender system is
a much more natural explanation than suspecting admittedly unidentifiable substrate
factors possibly relating to a 18-class noun gender system "out of Africa"? Where is
ignored that the shift from prenominal articles to noun class markers is a change
known to affect all varieties of modern French, including Standard French (cf. Harris
1980 discussed in Fournier 1998)?

IMAGE 01h2LexDiffFin01.gif


Printed for Henri Wittmann <>

[made with GoClick]