Studies of both Canadian (e.g., Laberge,
1977; Thibault, 1991) and European
French (e.g., Coveney, 2000) show that
definite on has almost entirely supplanted the standard
first-person plural subject clitic nous in informal
French.1
There is variation across Acadian
French varieties as to whether the subject “pronouns”
constitutes syntactic subjects or verbal affixes (cf., King & Nadasdi, 1997; Balcom
& Beaulieu, 1998); we ignore this issue here.
Indeed, it has been argued that for isolated varieties of French that
have had little or no contact with the standard, the
nous
variant has probably never existed as a subject clitic (see Coveney,
2000:453 for a discussion). This is most
likely the case for vernacular Atlantic Canada Acadian French. Our own
studies of varieties spoken in the provinces of Prince Edward Island
and Newfoundland, along with those of Flikeid (
1989) for varieties in the province of Nova Scotia,
reveal no instances of subject
nous in either formal or
informal style in large sociolinguistic corpora. In Acadian
communities,
on is in competition with the traditional
vernacular variant,
je…ons, as shown in (1).
2 All data cited come from the 1987
sociolinguistic interview corpus for Prince Edward Island French
constructed under the direction of Ruth King.
In this article, we examine first-person plural pronominal usage in
two varieties of Prince Edward Island French, those of the communities
Abram-Village and Saint-Louis, to see whether or not on is in
the process of replacing the traditional variant, thereby bringing
these Acadian varieties in line with other varieties of Canadian
French. We will consider the distribution of the two variants across
social categories and also across individual speakers.
BACKGROUND
According to Brunot (1967:335),
je…ons began to be replaced in French in the 16th
century. Although it is not to our knowledge attested in (earlier)
Quebec French, the geographical distribution of je…ons
actually remained widespread in France up to the late 19th century.
Flikeid and Péronnet (1989) reported
that the Atlas linguistique de la France (1902–1910)
shows that most northern varieties still retained
je…ons (with o(z) occurring in Picardy and
on occurring in the northwest) at the turn of the 20th
century.3
Lodge (1998:114) showed je…ons remained in
use in pre-Revolutionary Paris; further, the data in the Dialogues
révolutionnaires suggest that it was widely used by members
of the lower classes (both cited by Coveney,
2000).
In Atlantic Canada, recent studies have found
that
je…ons, typically considered one of the most
salient markers of Acadian French, remains the most frequent variant
only in those Acadian communities that exhibit low normative pressure
from external varieties, particularly, francophone communities in the
provinces of Nova Scotia and Newfoundland (
Flikeid
& Péronnet, 1989;
King &
Nadasdi, 1999). In some areas of Atlantic Canada, such as
northeastern New Brunswick, it has disappeared altogether (
Beaulieu & Cichocki, 2003).
In our survey of the literature, we noted a link between the survival
of je…ons and the survival of the third-person plural
traditional variant ils…ont (e.g., ils parlont
“they speak” vs. the standard ils parlent, where
the ending is phonetically null).4
Like
je…ons, ils…ont is an example of
archaic usage surviving in Acadian, appearing in France as early as the
13th century. According to Nyrop (vol. 2, no. 61),
ils…ont was in widespread usage in the center-west of
France at the time of Acadian emigration from that area.
In
the most conservative Acadian communities (e.g.,
L'Anse-à-Canards, Newfoundland), usage of the first- and
third-person traditional variants approaches categorality (
King, 1989). However, Flikeid and Péronnet
(
1989) reported the same proportion of
je…ons and
ils…ont in one of five Nova
Scotia communities they studied, but somewhat more
ils…ont than
je…ons in the remaining
four communities (e.g., 72%
ils…ont, but 56%
je…ons in Baie Ste-Marie, Nova Scotia). In communities
where
ils…ont is marginal, as in northeastern New
Brunswick (
Beaulieu & Cichocki, 2003) and
in French Louisiana (
Dubois, King, & Nadasdi,
2003),
je…ons is not attested.
METHODOLOGY
Our data come from a large corpus of sociolinguistic interviews
recorded in 1987 with 44 residents of two small villages in Prince
Edward Island. Although French is a minority language in Prince Edward
Island, there is one region, Évangéline, that constitutes
a French enclave, wherein French is the majority language in a number
of small villages. The French of one of those villages, Abram-Village,
was chosen for study. There the situation is one of stable language
maintenance, with considerable institutional support for French, with
most services, such as shops, church, and the post office provided in
French, along with French first-language education. The second village
chosen, Saint-Louis, is located in Tignish region, where the situation
is one of language shift, with French largely restricted to the speech
of middle-aged and older residents, and to high-school students
enrolled in “French immersion” programs aimed at
second-language learners of French. In each community, local residents
born and raised in the community conducted sociolinguistic interviews
of at least 90 minutes' duration. For the present study,
interviews for 22 individuals, representative of the age range of the
larger sample and representing both sexes, were analyzed. The interview
data for these 22 speakers provided 4572 tokens of first-person plural
pronouns with definite reference. Following standard sociolinguistic
methodology, cases of indefinite on (e.g., on sait
jamais … ‘one never knows’), which do not
represent instances of the variable, along with ambiguous cases with
respect to the definite/indefinite distinction, were eliminated
from the data set.
VARIATION AT THE LEVEL OF THE COMMUNITY
Table 1 shows the distribution of the two
first-person plural pronominal variants by community. We see that use
of je…ons is limited in Abram-Village, where the vast
majority of tokens contain on, but it has wide currency in
Saint-Louis; although it must be noted that nearly a quarter of all
Saint-Louis tokens do involve on. When we compare these
results to those for a comparable number of tokens for the third-person
plural variable (n = 4892), we find that the traditional
variant (ils…ont) occurs with 78% frequency in
Abram-Village and 83% in Saint-Louis.5
King & Nadasdi (1996)
provided a full discussion of the ils…ont variable in
Prince Edward Island Acadian, and Dubois, King, & Nadasdi (2003) provided a comparison of Atlantic Canada
Acadian and Cajun usage.
Thus, these results are in line with
those of earlier studies:
ils…ont occurs with close to
the same frequency as
je…ons in Saint-Louis, but is far
more frequent than
je…ons in Abram-Village. We will
consider why such a difference between the two traditional variants
might arise.
Distribution of je…ons versus on by
community
In the quantitative analysis of the variable, a number of social
factors were taken into consideration that have been shown to condition
variation in both majority and minority speech communities, including
age, sex, community, language use, French language education, and
speakers' position in the linguistic marketplace (cf., King & Nadasdi, 1996; Mougeon & Nadasdi, 1998). Given the different
status of French in the two communities, it was necessary to perform
separate multivariate analyses for each community, which was achieved
in a number of independent runs of goldvarb-2. Table 2 presents the results for Abram-Village. For
this community, robust results were obtained for position in the
linguistic marketplace, that is, with “how speakers'
economic activity, taken in its widest sense, requires or is
necessarily associated with competence in the legitimized (or standard,
elite, educated, etc.) language” (Sankoff & Laberge, 1978, following Bourdieu). Those speakers with
higher marketplace scores tended to use more on and those with
lower scores used more je…ons.
Overall results for Abram-Village
The Saint-Louis results are somewhat different, as there is little
division among speakers in terms of marketplace. Because, with very few
exceptions, only students enrolled in French immersion classes can be
said to be influenced by nonvernacular French, the marketplace variable
was not included in the analysis of the Saint-Louis data.6
As we have not included data for French
immersion students in this analysis, there are no younger speakers
(i.e., between 15 and 21 years of age) in the Saint-Louis sample used
here. We have found in other research that these young Saint-Louis
residents do not control vernacular variants (King,
2000).
Here, the amount of French language education
emerged as significant, that is, speakers with higher levels of
education used the
on variant more frequently. These results
are presented in
Table 3.
Overall results for Saint-Louis
These results are not surprising and mirror the findings of previous
work on a number of other morphosyntactic variables in the two
communities (cf., King & Nadasdi, 1996).
The vernacular variant has much greater currency in Saint-Louis, where
external varieties of French have had significantly less impact, as
noted previously.
Of the remaining social variables, the most interesting is the
differing contributions of speaker sex in the two communities. In
Abram-Village, men more so than women favor the traditional variant,
but in Saint-Louis, the star tradition-bearers are women: fully 94% of
their tokens involve je…ons. This difference is
understandable when one considers the sociolinguistic structure of each
community. Following Eckert and McConnell-Ginet (1992), we interpret the sociolinguistic behavior of
women (and men) within the local context, examining the roles of women
in the two communities. In Abram-Village, women are employed in a
variety of occupations and take part in voluntary activities in which
normative French has status (e.g., school secretary, night school
teacher, member of the community theatrical group, etc.). In
Saint-Louis, most women do not work outside the home, and, if they do,
they work at jobs that are conducted in English. Although the
Saint-Louis women are clearly fluent speakers of the Acadian variety,
both participant observation and self-report data indicate that they
have less contact with external varieties of French than do all the
other members of our sample.7
These
results are in line with the results of King & Nadasdi (1999), where the Saint-Louis women stood out from
the rest of our consultants, in that case, as the most frequent and
most innovative codeswitchers.
Not surprisingly, they make
almost exclusive use of the vernacular variant.
Although only 18% of the Abram-Village tokens contained the
vernacular variant je…ons, both the overall results and
the quantitative results by social category mask the fact that what
seems to be a declining vernacular variant is not simply fading away.
In the following section we examine variation in the usage of
individual speakers in this community.
VARIATION AT THE LEVEL OF THE INDIVIDUAL
Close inspection of the interviews shows that some Abram-Village
speakers have near-categorial or categorical use of on. For
example, one resident, Elizabeth,8
All
names are pseudonyms.
a 35-year-old secretary at the local
French-medium high school, used 99%
on, and another, Michel, a
21-year-old accountant with many contacts outside the community and
outside of Évangéline, had only
on tokens. Their
avoidance of
je…ons conforms to their high marketplace
ranking. For them,
je…ons is stigmatized. However, when
we examine the data for Abram-Village speakers with mixed usage, we
find that a discourse-level constraint is operative in several cases.
For such speakers,
je…ons predominates in narrative
contexts, whereas
on predominates in nonnarrative contexts.
To illustrate this difference, we turn to some of the Abram-Village
individual interviews in detail, beginning with Marie, who was 81 at
the time. Marie had gone to school up to grade seven, not atypical for
her generation. In her younger days, she had cleaned other
people's houses for a living; in 1987, at the time of the
interview, she lived in the local seniors' home. Her interview
lasted for 90 minutes and covers a wide variety of topics concerning
habitual aspects of life in her community, including her family
history, childhood games, school, celebration of Christmas and other
calendar customs in the old days, her early years of marriage, and,
finally, life at the home. The interview contains 159 on
tokens and 27 je…ons tokens. Twenty-one, or 83%, of
Marie's je…ons tokens occur in the six narratives
that she recounted during the interview, five of which were quite
short. The numerical results are displayed in Table
4.
Marie's wedding day narrative is reproduced next. Following
Labov and Waletsky (1967) and Labov (1997), each clause is identified by structural type
(i.e., abstract, orientation, complicating action, coda, or
evaluation).
Marie's Wedding Day Narrative
Marie's Wedding Day Narrative – Translation
The wedding day narrative involves going to the church, finding the
priest absent, waiting for hours until he returns from performing last
rites, getting married at last, and then going home for the wedding
meal and dance. This narrative contains ten instances of the variable,
all with the je…ons variant; both orientation and
complicating action clauses are involved. Je…ons is the
only variant that occurs in five of Marie's six narratives. (In
the sixth, there are three orientation clauses with on.) In
contrast, when Marie is talking outside the narrative context, the vast
majority of tokens contain on (156/162), as shown next in
two other excerpts, one about dating when she was young and the other
about cardplaying at the home.
Next, we turn to Daniel, 45 years old at the time of the interview
and the owner of his own construction company. He completed grade 12 at
the local high school, and at the time of the interview lived with his
wife, Louise, the company bookkeeper, and their children in their own
home. Daniel's interview likewise lasts 90 minutes and, like
Marie's, covers a wide range of topics. It contains nine
narratives. Quantitative results are shown in Table
5. Altogether, Daniel has 224 on tokens and 35
je…ons tokens.
Twenty-five of the je…ons tokens occur in narratives.
In the narrative reproduced next, Daniel starts talking about the time
he and his friends decided to play a trick on a neighbor one Halloween.
They went to the neighbor's house around 9:30 pm, but the
owners caught sight of them so they went back around two in the morning
and played the trick. The next morning, the police were called and went
out hunting for the culprits. Although Daniel and his friends
didn't get caught, he got the scare of his life.
Daniel's Halloween Tricks Narrative
Daniel's Halloween Tricks Narrative –
Translation
In the first orientation section, there is a case of indefinite
on, on faisait (line 3), not an instance of the
variable. Beginning with line 5, there is a series of sixteen
je…ons tokens, contained in orientation, complicating
action, and evaluation clauses. After a second evaluation in line 34
(J'avais jamais eu si peur/‘I had never been so
afraid’), Daniel sums up, moving out of the narrative in lines 35
(les polices, le lendemain matin, ils étiont dans les
chemins/‘the police, the next morning, they were
cruising around’) and line 36 (puis ils cherchiont les jeunes
qu'aviont fait ça/‘and they were looking for
the young people who had done it’). In line 37, he says, oh,
je sais pas, on faisait…/‘oh, I don't know,
we used to…’ and the interviewer overlaps with him with a
question: Vous avez fait prendre?/‘You got
caught?’. Daniel then returns to a second frame-out (cf., Galloway Young, 1987), the coda, in line 38:
J'avons pas fait prendre/‘We weren't
caught’. This on faisait/j'avons pas fait
sequence is interpretable in terms of the interruption: the question
momentarily returns him to the narrative frame. In the text that
follows, Daniel clearly has moved on, and there are a number of
on tokens immediately following the narrative.
In another lengthy narrative, which likewise recounts events which
took place when Daniel was a child, there are five on tokens
and eleven je…ons tokens. The beginning and end of the
narrative are found below. The on tokens occur at the
beginning, in orientation clauses in lines 1 and 2 (Moi puis mon
frère, on avait été voir pour une
brouette/‘Me and my brother, we went to see about a
wheelbarrow’; On avait été su mon
menoncle/‘We went to my uncle's’) and in a
following evaluation clause in line 3 (puis c'est pour dire
comment honteux qu'on était/‘and it's
to give you an idea of how shy we were’). These are followed by
eleven je…ons tokens, contained in complicating action
clauses, such as in line 28 (puis je nous emmenons en courant par
chez nous/‘and we ran home’). The final instance
of on occurs in line 30, in an evaluation/coda
(C'est de même qu'on était, oui, on
était pas mal honteux/‘That's how we were,
yes, we were pretty shy’). The variable usage displayed may be
interpreted in terms of Daniel's framing in and out of
narrative.
Daniel's Borrowing the Wheelbarrow Narrative,
Excerpt
Daniel's Borrowing the Wheelbarrow Narrative,
Excerpt – Translation
On the other hand, outside the narrative frame, Daniel uses
on almost exclusively, such as when telling of his and his
wife's settling in Abram-Village, where he'd been born and
raised, after their marriage, excerpted next.9
Although this excerpt does involve the temporal sequencing
of past events, it constitutes a report rather than a narrative, as no
evaluation expressing the relevance of this account to the present
discourse is provided (see Fleischman, 1990;
Labov, 1972; Labov &
Waletsky, 1967).
Extract 3: Settling Down in Abram-Village
We argue that Daniel and Marie's interviews reveal a pattern,
whereby je…ons serves as a marker for the performance
of narratives of personal experience. An objection that might be raised
to this interpretation is that what we are calling narrative versus
nonnarrative context might be more accurately considered as a formal
versus informal contrast, recalling that in Labovian methodology, the
elicitation of narratives of personal experience is a technique for
getting at the interviewee's informal style. However, there are a
number of arguments against this analysis of our data. First, the
contrast obtains even when we consider highly reduced narratives,
consisting of only three or four narrative clauses. Second, one good
indicator of level of formality in the Acadian context is the frequency
with which speakers use words (which they view as) of English origin.
In the interviews on which this analysis is based, we detect no
difference in such frequency in narrative versus nonnarrative contexts
in the insider interviews that serve as our data base.
In Verbal Art as Performance, Bauman (1977:17) discussed the culturally conventionalized
means by which members of a speech community key the performance frame.
The communicative resources employed may include, among others,
figurative language, special paralinguistic features, special formulae,
and special codes. Special codes may involve “one or another
linguistic level or features”. In the present instance, use of a
particular grammatical variant keys the performance of a specific
verbal genre, narration. Je…ons is available for such
use, as it is, in Cornips and Corrigan's (in
press) terms, a low-level morphosyntactic variable, open to social
(re)evaluation in a way that high-level variables (e.g., passivity) might
not be. In our corpus, narration usually involves narratives of personal
experience, but also includes family and community narratives.10
Butler (1991) found
that these three types of narratives predominated in his corpus-based study
of narratives in a Franco-Acadian community in the mid-1980s.
Because the variable at issue involves first-person plural reference,
unsurprisingly it occurs most frequently in narratives of personal
experience.
What of other Abram-Village residents who show mixed usage? In terms
of the presence or absence of discursive constraints, two other types
of speakers emerge. First, there are speakers who use a significant
number of je…ons tokens overall. This is the case of
Carole, an 18-year-old high school senior, who uses the traditional
variant 80% of the time. Her interview contains two narratives, both of
which have je…ons, not on.11
Like many younger speakers, Carole does not produce as
many narratives as do older residents. However, her high frequency of
je…ons usage across discourse genres is our primary
interest.
That Carole, an educated young woman, would have
such a high frequency of use of the vernacular variant is not
surprising. As Flikeid (
1992) argued,
education level alone gives us an incomplete picture of Acadians'
relative sensitivity to the relationship between standard and
vernacular usage. It does not take into account the various ways in
which one is exposed to standard varieties, for example, through
contact with outsiders, through jobs that may be interpreted as low in
terms of one's socioeconomic status, but that bring one into
contact with, or require, use of the standard or legitimized language.
We suggest that Carole's age indicates a lack of life experience
needed to recognize the sociosymbolic meaning(s) of the two
variants.
Second, there are speakers for whom the narrative/nonnarrative
split is not as clear as in the case of Marie and Daniel, but who do
tend to use je…ons in narrative. This is the case for
Sylvie, a 42-year-old letter carrier and amateur actor, who uses this
variant 23% of the time. Her interview contains eleven narratives and
her use of the variable is shown in Table 6.
Here we see a high proportion of on tokens overall, but with
je…ons is found somewhat more frequently than
on in narratives. However, turning to the individual
narratives, we find that two of the eleven have only on
tokens, ten such tokens in total. In the remaining nine narratives,
there are fourteen on tokens and thirty-four
je…ons tokens. In these narratives all but one clause
containing complicating action have je…ons.12
The tendency toward use of
je…ons in clauses involving complicating action
suggests that such usage may involve a foregrounding strategy. Hopper
(1979:213) noted that it is “a
universal of narrative discourse that in any extended text an overt
distinction is made between the language of the actual story line and
the language of supporting material which does not itself narrate the
main events,” referring to the former as foreground and the
latter as background. We thank Jen Smith for pointing us toward this
work.
is limited to orientation and evaluation
clauses (sometimes evaluation and coda clauses combined) near the
beginning or end of the narrative. Thus, in the majority of
Sylvie's narratives, the pattern is similar to what was found for
Daniel. One might ask why she is more “advanced” than
Daniel, an Abram-Village resident of around the same age. More
specifically, why has
on “taken over” in two of
her narratives? The answer most likely lies in Sylvie's
participation in voluntary activities involving frequent contact with
speakers of external varieties. As an actor in local summer theatrical
productions, she has significant contact with francophone tourists,
particularly tourists from Quebec, for whom
on is the only
informal variant (cf.,
Laberge, 1977). These
results, then, show that a particular declining variant is not simply
fading away. Some Abram-Village speakers have abandoned the traditional
variant
je…ons, while others continue to use it quite
frequently. But for some mixed-use Abram-Village speakers, it has taken
on a new function in discourse, that of indexing the performance of
narratives of personal experience, a function observable in a
stylistically rich database.
THE FATE OF JE…ONS
In general, then, contact with other varieties leads certain Acadian
French forms to be viewed as substandard. Abram-Village speakers make
less frequent use of the vernacular variant, and consequently, use
on to a greater extent. However, participating in an education
system that extols the virtues of the standard variety results not in
widespread use of the standard variant nous, but, rather, in
the stigmatization of the traditional variant, je…ons,
and rise of the colloquial French on. Saint-Louis speakers, on
the other hand, have had little contact with other varieties; they do
not appear to consider je…ons to be stigmatized and
therefore make frequent use of this form. But stigmatization is only
part of the je…ons story. For some Abram-Village
speakers, rather than being stigmatized, je…ons has
instead become associated with a particular discourse function, that of
indexing the performance of oral narrative.
We noted previously that our results confirm a general tendency in
Acadian varieties for the first-person vernacular variant
je…ons to be less well preserved than the third-person
form, ils…ont. It is safe to say that
ils…ont (76% of occurrences of the third-person plural
variable) has not undergone the fate of je…ons (18% of
occurrences of the first-person plural variable) in Abram-Village. We
suggest that this discrepancy occurs because the contrast between the
vernacular and nonvernacular forms is more marked in the first-person
plural than in the third-person plural. A comparison of the two cases
shows that what differentiates the two variables is that the
first-person variable involves a change in pronominal clitic along with
loss of a suffix on the verb (e.g., je parlons vs. on
parle), while the third-person case involves (uniquely) loss of
the suffix (e.g., ils parlont vs. ils parlent).
Further, the retention of the -ont suffix in the third-person
plural may be reinforced by the occurrence in the standard variety of
high-frequency irregular verbs with -ont, such as
aller “to go” – ils vont,
être “to be” – ils sont, and
avoir “to have” – ils ont. The fact
that the first-person plural vernacular variant is highly distinctive,
providing a marked contrast with the on variant, seems to be
the motivation for its strong association with local norms and its
reevaluation, either as a performance key or as a stigmatized
variant.
