A sociolinguistic study of verb morphology in Montreal French:
	     reanalyzing the alternation between avoir and être

		Pierette Thibault, Université de Montréal
		  Gillian Sankoff, U. of Pennsylvania

The major domains of usage of avoir  'to have' and être 'to be' as
auxiliaries in French, as in many other languages, link the former to
compound tense constructions and the latter to passive constructions (cf.
Benveniste 1966:205), as shown in examples (1a) and (1b).
	(1a) J'ai apprécié la viande 'I enjoyed the meat' (past meaning)
	(1b) La viande est appréciée 'The meat is enjoyed' (passive meaning)

When used with transitive verbs -- two argument verbs -- the distinction
between the two is aspectual as shown in examples (2b), an unaccusative
construction, and (2c), a resultative:
	(2a) Le charbon trop chaud a brûlé la viande
	       'The over-heated charcoal burned the meat'
	(2b) La viande a brûlé
	      'The meat burned' [focus on the action, punctual aspect]
	(2c) La viande est brûlée
	     'The meat is burned' [focus on the result, nonpunctual aspect]

With intransitive verbs -- one argument verbs -- such a distinction does
not apply in standard French. Some are constructed with avoir: marcher  'to
walk', ramper 'to crawl', voler 'to fly', plonger 'to dive' etc., while
others are constructed with être: aller 'to go', arriver 'to arrive',
partir 'to leave', monter 'to go up' and a few others. A similar prescribed
distribution applies to stative verbs: progresser 'to improve' among others
is constructed with avoir, while rester 'to stay' or 'to remain' is
constructed with être.

In Québécois French, however, both avoir and être  are used freely with a
number of intransitive verbs. Sentences (3) and (4), spoken by the same
speaker, describing the same sequence of events within the space of a
minute, illustrates the alternation.
(3) J'ai rentré à cinq heures, j'ai été opérée le lendemain matin à dix
heures et demie.
	(4) Je suis rentrée juste la veille de l'opération à cinq heures,
j'ai été opérée le lendemain matin à dix. (Germaine C., 1984)
		"I went in at 5 o'clock, I was operated on the next morning
at 10 (10:30).'

In a paper first published in 1977, we argued that this distribution,
albeit partially semantically motivated, was also to a considerable extent
arbitrary. In Montreal French, we found that the distribution of avoir and
être was both lexically and socially motivated. Our study analyzed the
alternation between avoir and être with intransitive and stative verbs in
contexts where there was a clear indication of a focus on the action as in
(3) and (4), thus ruling out the aspectual distinction illustrated between
(2b) and (2c). The data were drawn from recorded interviews with a socially
stratified sample of 120 Montreal-French speakers (the 1971
Sankoff-Cedergren corpus).

 Alternative interpretations of the variation have been proposed French
(Blanche-Benveniste 1977; Kayne 1993 and others).  First, it has been
suggested that some speakers (namely the working class speakers) express an
aspectual opposition that does not exist in standard French. Second, since
some of the verbs in question also have transitive and unaccusative uses,
as in (5a) and (5b), the unaccusative/passive opposition may have been
transformed into marking a focus on the action vs. a focus on the result of
the action as in (6a) and (6b).
	(5a) Ils ont sorti la télévision en cinquante.
	       'They came out with TV in 1950'.
	(5b)  La télévision a sorti en cinquante.
	       'TV came out in 1950.'
	(6a) Le chien a sorti. 'The dog went out'
	(6b) Le chien est sorti. 'The dog is out'
A third line of argument stems from a class of standard prescribed uses of
être : the reflexive as in (7)  and the pronominal uses of the verbs, as in
(8). According to this view, the reflexives, behaving like transitives,
would be more likely to show avoir, whereas the pronominals, as in (8)
would use être.
	(7) Comme moi, je m'ai donnée bien jeune (Réjeanne C., 1971)
		'Like me, I gave myself away very young'
	(8) Je me suis levée, il était neuf heures. (Eugénie G., 1971)
	      'I got up, it was 9 o'clock.'

Our speakers, however, also use être  with pronominals, as in (9).
	(9) Ah bien je m'ai levé, j'ai été aux toilettes, j'ai mangé des
céréales. (Pierre G., 1984)
		'Oh well I got up, I went to the bathroom, I ate some cereal.'

Almost twenty years after our first study of the avoir/être alternation in
Montreal-French, we decided to undertake a reanalysis for the following
reasons:
1. In 1984, half of the initial sample of speakers were reinterviewed.
Since our initial study only included half of the 1971 speakers, we wanted
to take the opportunity to double the amount of data and test the influence
of more quantitatively manageable factors, including
those stemming from more recent inquiries in the field.
2. Such comparative data allowed us to see whether speakers modified their
use of the auxiliaries as they grew older.

Our data provide counterexamples to all three of the proposals that suggest
categorical reanalysis, e.g. (3), (4) and (9) above.  As in our 1977
analysis, our results show that the social differentiation is by no means
straightforward.  Although there is a tendency for the non-standard use of
avoir to be associated with working class speakers, this applies
differentially to the various verbs, and the social class based pattern is
far from categorical. A semantically driven association between auxiliary
use and verb meaning is weakly motivated at best. Rather, there seems to be
a lexically-focused distribution, differing from the standard pattern in a
nuber of details.  Lastly, we see little change over time, and no
indication that one competing form is replacing the other.

References

Benveniste, Emile. 1966. Problèmes de linguistique générale I. Paris,
Gallimard.

Blanche-Benveniste, Claire. 1977. L'un chasse l'autre: le domaine des
auxiliaires. Recherches sur le français parlé. Groupe aixois de recherche
en syntaxe. Aix, Université de Provence, pp.100-148.

Kayne, Richard. 1993. Toward a modular theory of auxiliary selection.
Studia Linguistica 47.

Sankoff, G., & P. Thibault. 1977. L'alternance entre les auxiliaires avoir
et être en français parlé à Montréal. Langue française 34: 81-108. [revised
version in G. Sankoff, The Social Life of Language, University of
Pennsylvania Press, 1980, pp.311-345].