McGILL LAW JOURNAL
REVUE DE DROIT DE McGILL
Volume 34
Montreal
1989
No 4
Education and Linguistic Security in the Charter
Denise Raume* and Leslie Green** t
The authors provide an interpretive frame-
work for minority language education rights
as guaranteed in section 23 of the Canadian
Charter of Rights and Freedoms. They argue
that the purpose of such rights is to protect
linguistic security. Attending to that value
and to the text of the Charter, they seek to
explain the nature and ground of the limi-
tation which confines application of the right
to circumstances in which numbers warrant.
In doing so, they critically discuss a number
of judgments bearing on the content of the
right, the relevance of cost in securing the
right, and the appropriate judicial remedies
for enforcing it.
Les auteurs tentent d’6tablir un cadre d’in-
terpr6tation du droit fA l’ducation dans la
langue de Ia minorit6 garanti par l’article 23
de la Charte canadienne des droits et liberts.
Selon eux, ces droits ont pour but la sfiret6
linguistique de ]a minorit6. Consid6rant tant
le but que le texte de ces dispositions de la
Charte les auteurs tentent d’expliquer ]a na-
ture et le fondement de la restriction qui sti-
pule qu’on ne peut exercer ce droit que 1A ofi
le nombre lejustifie. Ils 6tudient et critiquent
la jurisprudence en particulier en ce qui a
trait au contenu de ce droit, A la pertinence
de tenir compte des cofits et aux recours ap-
propri6s pour faire respecter ce droit.
*Associate Professor, Faculty of Law, University of Toronto.
-Associate Professor, Osgoode Hall Law School and Department of Philosophy, York
University.
tWe are grateful to the Human Rights Law Fund of the Department of Justice and the
Connaught Programme in Legal Theory and Public Policy at the University of Toronto for
their financial support of this project. The opinions expressed are solely those of the authors.
McGill Law Joumal 1989
Revue de droit de McGill
REVUE DE DROIT DE McGILL
[Vol. 34
Synopsis
I.
The Value of Linguistic Security
II. Rough Equality of Instruction
III. Why Numbers Warrant
IV. Numbers and the Sliding Scale
V. Numbers and Cost
A. Equivalent Cost
B. Reasonable Cost
C. Efficacy and Cost
VI. Where Numbers Warrant
VII. Control of Facilities
VIII. Remedial Problems
The Canadian constitution protects the use of minority languages in
two distinct ways. General guarantees of freedom of expression, freedom
of association, non-discrimination and natural justice create a regime of
linguistic tolerance in which no language is a ground of social liability.
Beyond these consequential protections, two language groups are protected
directly. French and English-speaking citizens enjoy further rights, even
where they are in the minority, to use their languages in some courts and
legislatures, to have legislation enacted in their languages, to receive federal
government services in those languages, and to have their children educated
in their mother tongue. The educational rights in s. 23 of the Canadian
Charter of Rights and Freedoms1 are the concern of this article. They are
perhaps the most important aspect of the Canadian regime of official lan-
guages, but have not yet been provided with a clear analytical and normative
framework. Through a purposive analysis of the language rights provisions
we seek to sketch the main elements of such an approach.
‘Part I of the Constitution Act, 1982, being Schedule B of the Canada Act 1982 (U.K.), 1982,
c. 11.
1989]
EDUCATION AND LINGUISTIC SECURITY
or should –
In particular, we shall pay special attention to a feature of the education
rights which is at once intuitively plausible and conceptually puzzling. They
are subject to certain limitations with respect to numbers: the right of parents
to have their children educated in the minority language applies only where
numbers warrant. The numbers proviso is a plausible feature of such rights
because no one thinks that the constitution does –
oblige
provinces to set up a school system for the sake of a single child. It is
puzzling, however, because most familiar human rights are not subject to
a numbers constraint. It would be a very odd right to freedom of expression,
for example, that applied only to large bodies of opinion. It is secured even
for a single person. Why then are language rights, or at least some of them,
different? We hope to illuminate this issue by appeal to the value of linguistic
security, which we take to provide the best justification for language rights.
It is uncontroversial 2 that the minority language education provisions of
the Charter are remedial in character. It is our argument that those remedies
should be understood to have the aim of promoting the linguistic security
of official language minorities, and that some aspects of linguistic security
involve collective goods whose benefits are only available in groups.
I. The Value of Linguistic Security
Two mischievous notions about language rights have some currency in
Canada. The first is that language rights are a mere product of political
compromise and have no foundation in principle. The second contradicts
the first. According to it, language rights are founded on the principle of
survival: governments have a duty to ensure that minority languages con-
tinue into the future. These are not politically innocent notions, for each
has implications for the way in which language rights should be interpreted
and the weight they should be given. But they are both founded on mistakes.
The first confuses the genesis of constitutional rights with their justi-
fication. All rights entrenched in positive law have a particular form that
attempts to make concrete certain abstract values which the law prizes.
Every constitutional right thus marks a kind of compromise between com-
peting interpretations of the values it protects; every one strikes some bal-
ance between legislative sovereignty and minority protections; every one
can be protected only by a combination of non-interference and positive
action on the part of government. Because these are features of all consti-
tutional rights, they do not distinguish language rights from the rest and
therefore provide no ground for interpreting them differently. That is why
the Supreme Court of Canada, to whom this first mistake is due, has not
been able to draw the proposed distinction between “compromise- and prin-
2A.G. Quebec v. Quebec Association of Protestant School Boards, [1984] 2 S.C.R. 66.
McGILL LAW JOURNAL
[Vol. 34
ciple-based” rights in a consistent and persuasive way.3 Such truth as there
is in the idea amounts to this: the courts must give effect to the terms of a
constitutional agreement without, under the guise of interpretation, amend-
ing them. That claim is as harmless as it is sound. It does nothing to show
what those terms are, nor how courts should proceed when they are equiv-
ocal.4 Thus, the claim that they originate in a compromise does not in fact
justify the Court’s recent policy of reading some language rights restrictively.
The second view, according to which minority language rights are rooted
in the principle of survival,5 makes a different error. It confuses the justi-
fication of a right with the likely by-product of its exercise. Minority lan-
guages are under threat from a variety of sources, but they die out for a
common reason: they are abandoned by their speakers. Language rights aim
to protect speakers from certain pressures to abandon their languages. When
linguistic choices are made in a secure environment, roughly, one without
unfair pressure to conform to majority practices, they will in fact typically
lead to a higher rate of survival. Does it follow, then, that the aim of language
rights is to protect the endangered species of the linguistic world?
We can test that hypothesis by considering some policies aimed at
ensuring language survival. Suppose, for example, that one of the majority
English provinces required all French speakers to send their children to
French schools and denied them access to English instruction. Or suppose
that by residential zoning it attempted to reduce exogamy among declining
minorities. Set aside the question of whether these measures would violate
other rights, and let us ask simply whether as far as language goes, they are
aimed in the right direction. Could they be said to take at least one step
towards justice? On reflection that seems dubious. The problem is not simply
3Compare: MacDonald v. Montreal (City o), [1986] 1 S.C.R. 460, 27 D.L.R. (4th) 321, 67
N.R. 1, 25 C.C.C. (3d) 481; SocitO des Acadiens du Nouveau Brunswick Inc. v. Minority
Language School Board No. 50, [1986] 1 S.C.R. 549, 27 D.L.R. (4th) 406, 66 N.R. 173 [here-
inafter SocitO des Acadiens cited to S.C.R.]; Ford v. A.G. Quebec, [ 1988] 2 S.C.R. 712 at 751,
54 D.L.R. (4th) 577 [hereinafter Ford cited to S.C.R.]. In Ford, the Supreme Court reinterprets
Socit6 des Acadiens as holding that the distinctive basis of language rights is that they impose
positive duties on government. It is unclear whether this is meant to supplant or supplement
the distinction drawn in the earlier cases between rights based on ‘compromise’ and those
based on ‘principle’.
4This argument is more fully made in our paper “Second-Class Rights?: Principle and Com-
promise in the Charter”, forthcoming. For a brief statement of the position see L. Green, “Are
Language Rights Fundamental?” (1987) 25 Osgoode Hall L.J. 639.
5This is, explicitly or implicitly, the view of most of the commentators on language rights
in Canada. See for example, J.E. Magnet, “Collective Rights, Cultural Autonomy and the
Canadian State” (1986) 32 McGill L.J. 170 at 184; Andr6 Bradn, “Language Rights”, at 21,
Pierre Foucher, “Language Rights and Education”, at 257, and Emmanuel Didier, “The Private
Law of Language”, at 327, all in Michel Bastarache, ed., Language Rights in Canada, (Montreal:
Y. Blais, 1987).
1989]
EDUCATION AND LINGUISTIC SECURITY
that language rights and other liberties are here in conflict, but that moral
rights to language use are themselves violated by the policies in question.
Prohibiting the minorities from learning the majority language and banning
minority-language instruction offend common principles: they attack lin-
guistic security by creating unfair pressures to conform. These pressures do
not become acceptable when they are inflicted on a minority within the
minority community itself. Draconian measures to promote minority lan-
guages may evince a kind of concern for the health of the languages, but
they do not give appropriate concern for the interests of their speakers.
That security and not survival is the root value is suggested by con-
sidering the importance of language. Apart from its instrumental value in
communication, language is also an important marker of identity. Those
who wish to use minority languages do so partly as an expression of be-
longing to and identifying with a community. But language use has this
valuable expressive dimension only if rooted in a free and fair context.
Those who are forced to use a particular language cannot be thought thereby
to express their identity. That does not mean that language must be con-
sciously chosen. Language is only partially a realm of free choice. Children
have a mother tongue long before they develop the capacity for reflective
and informed choice about ethnic identification, and parents typically trans-
mit their mother tongues as a matter of course. But these normal processes
of social development contribute value to their outcomes only in circum-
stances which are fair and unbiased. Thus, while facilitating minority lan-
guage education and requiring it both promote the survival of minority
languages, this equivalence in consequences does not establish an equiva-
lence in aim. The point of language rights is to give speakers a secure en-
vironment in which to make choices about language use, and in which ethnic
identification can have positive value.
The confusion of survival and security is easily made, for the conditions
threatening security also make survival less likely. Evidence of assimilation
and decline among the francophone minorities 6 made it clear that the lack
of adequate protection in the 1867 constitution 7 had exacerbated their de-
6This was first recognized by the francophone communities themselves when various pro-
vincial governments revoked the freedom that officially or de facto denominational schools
had previously had to use French as the language of instruction. See Pierre Foucher, Consti-
tutional Language Rights of Official-Language Minorities in Canada (Ottawa: Ministry of
Supply and Services, 1985), for a survey of the history. See Les heritiers de Lord Durham
(Ottawa: FEderation des francophones hors Quebec, 1977), translated as The Heirs of Lord
Durham (Ottawa: Bums and MacEachern, Ltd., 1978) for the views of La Fdration des
Francophones hors Quebec. See the Report of the Royal Commission on Bilingualism and
Biculturalism, (Ottawa: Queen’s Printer, 1968), especially Vol. 2, for the beginning of serious
efforts at constitutional reform in this regard.
7Constitution Act, 1867 (U.K) 30 & 31 Vict., c. 3.
REVUE DE DROIT DE McGILL
[Vol. 34
mographic fragility, and the desire to remedy this was a driving force of the
language rights provisions. Nonetheless, the decline of the minorities is a
symptom and not itself a disease. It is presumptive evidence that there is
strong and potentially unfair social pressure to abandon their language. But
this evidence is rebuttable. It is possible (though not probable under normal
circumstances) that even in a completely secure environment, some mem-
bers of minority language groups would still make free and informed de-
cisions to integrate with a majority community. The need to identify with
a community may be deeply rooted in human nature, but we know that
there is nonetheless much flexibility regarding the community with which
one identifies.
These considerations suggest that it is not the survival of languages but
the security of their speakers that justifies language rights. To have linguistic
security in the fullest sense is to have the opportunity, without serious im-
pediments, to live a full life in a community of people who share one’s
language. This opportunity is taken for granted by those in linguistically
homogeneous societies and by those who speak the majority language in
multilingual societies. Through sheer numbers they enjoy defacto linguistic
security without need for special legal protections. No doors are closed, and
no aspects of human fulfillment are unavailable on account of language.
Abandoning one’s mother tongue (oneself or on behalf of one’s children) is
of course a conceivable option for them, but not one to which they are
driven by force of social circumstance and not one which will even be
considered in the normal course of life. It is otherwise for members of
linguistic minorities.8 Without special protections, minority language speak-
ers are inevitably placed under strong pressures to abandon their mother
tongue. Because of its central role in every aspect of human co-operation,
people share a common interest in communicating with others. To be ex-
cluded from this is to be denied most of what is valued in life. The more
restricted the existence available in one’s mother tongue the more rational
it becomes to take up the language that offers greater opportunities. This
does not mean that the minority language speakers do not value their lan-
guage or communities, any more than the decision of hold-up victims to
part with their wallets means that they do not value their money. It means
8Linguistic security is, of course, a universal good. This raises the question of why only two
such communities should now enjoy the special protections of the official language regime.
The point cannot be argued here, but it seems to us that the defense lies in the fact that the
French and English are the nation’s largest and most deeply established communities. If a
sound argument can be made that other language groups also deserve special protections beyond
what they now enjoy under the regime of linguistic tolerance, this would not weaken the claims
of the French and English minorities. Moreover, one might have regional as well as national
official languages. For convenience, however, our discussion is limited to the present official
languages.
1989]
EDUCATION AND LINGUISTIC SECURITY
simply that there are some burdens that outweigh it, and some costs that
it is unjust to expect them to bear.
Part of the environment that ensures linguistic security must be pro-
vided by the members of a particular language community themselves. Ac-
tivities that require the participation of many people depend upon their
choices converging. To the extent that linguistic security requires a social
environment that includes such activities, the community is dependent
upon its own efforts. The larger the community, the greater the range of
options its members will have simply because it is more likely that there
will be sufficient numbers with common interests. If a community is very
small, much of this will not be feasible and could not be promoted without
grave risk to other values. For example, it would be illegitimate for gov-
ernment to require minority language speakers to have more children to
increase the size of their community, or to force non-native speakers of that
language to participate in certain cultural activities to make their provision
feasible. But provided there is at least a small core of willing native speakers,
the government can act in ways that support and facilitate their ambitions.
This support takes two forms. First, a secure environment is partly
established by those general rights which establish a regime of linguistic
tolerance: freedom of expression, association, non-discrimination, etc. These
ensure that one’s language is not made a ground of liability. In the case of
very small linguistic communities, there may be little more that can be
done. But when groups have a standing and vitality of the French and
English communities in Canada, governments have the capacity and duty
to do more. They may provide for services, counteract unfair bias, and
generally facilitate their activities in ways that will have tangible and ben-
eficial consequences. These facilitative actions may intervene quite vigor-
ously in the customary linguistic order, without threatening security in the
way the compulsory measures discussed above would. The role of govern-
ment in protecting linguistic security is thus easily explained. The familiar
official language rights serve the interests of linguistic security by facilitating
participation in activities under government control. Participation in po-
litical life involves communication with officials. A community that could
not participate in the political life of its country would be severely handi-
capped, and, if participation must be on the majority’s terms, then the
incentive to assimilate is obvious. Similarly, the denial of government ser-
vices, whether the court system or the kind of everyday help and advice
that many government departments provide, turns the use of one’s mother
tongue into a handicap and sometimes even a source of shame. But, unlike
ethnic groups, government has no mother tongue of its own. The choice of
its working languages is a matter over which the government has complete
control. Participation can therefore be guaranteed in one’s own language
McGILL LAW JOURNAL
[Vol. 34
without sacrificing the legitimate interests of others. How does education
fit into the emerging picture?
The system of education, particularly at the primary and secondary
levels, makes major contributions to the security of one’s linguistic envi-
ronment. Provision for minority language education is a complex good with
many different facets. For convenience, we distinguish two main aspects.
There are powerful individual benefits of children being able to learn in their
mother-tongue: it is easier to master other subjects when one knows the
language and feels socially at ease in the class-room. It also opens doors to
participation in one’s community and fosters a positive attitude towards it.
The absence of minority language education is quite obviously a powerful
assimilative force. Children grow up with a grasp of their mother tongue
which is inadequate for the kind of adult pursuits which require strong
communication skills. In such circumstances it is hardly surprising that
people abandon their first language and do not teach it to their children.
Before long, such a community ceases to be viable and its language, if it
persists at all, has merely folkloric status.
Education cannot however be fully understood as an individual good.9
Minority language instruction benefits the linguistic group as well. It has
collective benefits which flow from the language being a vehicle of instruc-
tion. For example, it provides and renews cultural capital. This is true at
the level of both “high” and “popular” culture: the productive and appre-
ciative capacities must be nurtured and trained through a comprehensive
education. Musicians, writers, artists obviously depend on and draw on
common cultural capital in representing and contesting the life of the com-
munity. But even folk and oral traditions, sporting culture, etc., all draw on
a stock of common forms and images. In modern societies this capital is
largely controlled by the educational system.
Other direct collective benefits are more instrumental: the education
system provides jobs for members of the minority community. There are
also indirect collective benefits which flow from the existence and admin-
istration of minority-language instruction. 10 For one thing, a community
9p. Foucher, supra, note 5, emphasizes the individual benefits to highlight the individual’s
standing to enforce the right. There is however no good reason to tie the nature of a right to
the issue of who may sue for its protection. The primary beneficiaries of minority language
education rights are children, but control over the right is normally exercised by their parents.
10See Rapport de la Commission minist~rielle sur l’Education secondaire en languefrancaise/
Report of the Ministerial Commission on French Language Secondary Education (Toronto:
Queen’s Printer, 1972) at 13 (The Symons Report): “The school occupies a central role in the
cultural life of the community… . The French language schools must truly be community
schools and easily accessible to the general population of the linguistic group they exist to
serve….”
1989]
EDUCATION AND LINGUISTIC SECURITY
with public institutions will have greater visibility and status. More im-
portantly, an educational facility such as a neighbourhood school is an im-
portant focus of social and cultural activities for the community, especially
in smaller towns. And managing a school system by electing trustees, hiring
teachers, setting policy, etc. are all important parts of the political life of
such communities and contribute to their richness and vitality.
These are only some of the ways in which minority-language education
enters the collective life of the community. Many of them exhibit interesting
structural features. Some collective benefits are public goods in the econo-
mists’ sense: none can be excluded from their benefits and they do not
diminish with consumption. This is clearly the case with respect to the
diffuse effects of a minority language education system on the security, status,
and vitality of the community. And, where publicly funded education is the
norm, it is true of educational options themselves: they become available
to any parents who wish to take advantage of them.” Moreover, the exist-
ence of these schools makes the entire community more vital in diffuse ways
which generate benefits even for those who do not directly participate in
its activities. For example, the increased use of minority languages obviously
increases the instrumental value of being able to speak them, and this benefit
accrues to all.
But minority-language instruction has further collective benefits which,
though excludable, are social and non-rival. 12 Where these flow from the
inherent value of participating with others in some social activity, we call
them participatory goods. A school plays a significant role in fostering human
relationships, teaching co-operation, and imparting other social skills in a
way that could not be achieved under a system of private individual tuition.
Public education is the central means by which children are introduced to
and can participate in the cultural traditions of their community. Manage-
ment and control of an education system, similarly, provides a forum in
which parents can exercise and develop skills of self-government. In all these
ways, minority language education has a significant social role. Let us now
see to what extent these abstract notions can illuminate one of the darker
corners of the Charter.
“We set aside the presence of some rivalness in consumption due to ‘crowding’ which in
any case usually affects majority schools more than minority ones. For the general analytic
framework see M. Peston, Public Goods and the Public Sector (London: Macmillan, 1972). On
norm-dependent public goods see note 12, infra.
12See D. Raume, “Individuals, Groups, and Rights to Public Goods” (1988) 38 U.T.L.J. 1.
REVUE DE DROIT DE McGILL
[Vol. 34
II. Rough Equality of Instruction
The language rights provisions are quite complex. Section 23 does not
simply announce that official language minorities have the right to education
in their own language, in the way, for example, that s. 2 succinctly declares
a right to freedom of expression. Rather, it provides to parents falling in
certain precisely drawn categories,13 a right to have their children receive
“primary and secondary school instruction” in the minority language of the
province, and limits the application of that right to areas where numbers
warrant. But what does such “instruction” amount to? How should it be
understood in light of its remedial purpose?
Generally speaking, the courts have endorsed the view that the right
to minority language instruction means the right to minority first-language
instruction that is of roughly equal quality to that of the majority,’4 thus
interpreting “instruction” to mean education in which the minority language
is the vehicle and not just the object of study. This is obviously the correct
approach from the point of view of fostering linguistic security. The pres-
sures to abandon one’s language take the form of comparative incentives:
the greater benefits of speaking one language rather than another. For that
reason, the relative standing of the majority and minority educational sys-
tems bears on the success of the latter in fostering linguistic security. A
13S. 23. (1): Citizens of Canada
(a) whose first language learned and still understood is that of the English or French
linguistic minority population of the province in which they reside, or
(b) who have received their primary school instruction in Canada in English or
French and reside in a province where the language in which they received that
instruction is the language of the English or French linguistic minority population
of the province,
have the right to have their children receive primary and secondary school instruction in that
language in that province.
(2) Citizens of Canada of whom any child has received or is receiving primary or secondary
school instruction in English or French in Canada, have the right to have all their children
receive primary and secondary school instruction in the same language.
“4Reference re Education Act of Ontario and Minority Language Education Rights (1984),
47 O.R. (2d) 1 at 43, 10 D.L.R. (4th) 491 (Ont. C.A.) [hereinafter Ontario Reference cited to
O.R.]: “The quality of education to be provided to the minority is to be on a basis of equality
with the majority.” See also: Marchand v. Simcoe County Board of Education (1986), 55 O.R.
(2d) 638 at 660, 29 D.L.R. (4th) 596 (Ont. H.C.) [hereinafter Marchand cited to D.L.R.]; Mah
v. The Queen in Right of Alberta (1987), 80 A.R. 161 at 176, 180, 42 D.L.R. (4th) 514 (Alta
C.A.) [hereinafter Mah cited to D.L.R.] (leave to appeal granted by the Supreme Court of
Canada, December 7, 1987); Commission des Ecoles Fransaskoises Inc. v. Saskatchewan, [ 1988]
3 W.W.R. 354 at 360, 48 D.L.R. (4th) 315, 64 Sask. R. 123 (Sask. Q.B.) [hereinafter Commission
des tcole Fransaskoises cited to W.W.R.]; Reference Re Minority Language Educational Rights
(PE.I.) (1988), 69 Nfld. & RE.I.R. 236 at 256, 211 A.PR. 236 (S.C.A.D.) [hereinafter RE.L
Reference cited to Nfld. & RE.I.R.]; Lavoie v. Attorney General of Nova Scotia (March 29,
1989) S.C.N.S.A.D., unreported, at 31.
1989]
EDUCATION AND LINGUISTIC SECURITY
substantially inferior system would do nothing to advance, and might even
impair, the security of the group. First, for instrumental reasons, minority
language parents will rarely sacrifice the quality of their children’s education
to its linguistic character. They may accept certain trade-offs in the range
or nature of resources, but not a fundamental difference in quality. Second,
an inferior system may express profound disregard for the status of minority-
language citizens and would therefore be discriminatory. Education rights
can improve the situation of the minority only if they offer a reasonably
attractive option. For a variety of reasons, including especially the fact that
the minority is always smaller than the majority, the options cannot be
identical ones. Since nothing can change the general social matrix of a region
so that all diffuse benefits are equalized as between minority and majority,
the standard of equality is necessarily a rough one. 15 But to the extent that
education can improve matters in such a situation, it should aim to do so
by making the options as nearly equal as feasible with respect to both the
quality and range of benefits that the education system provides.
Despite the apparent judicial acceptance of the idea of rough equality,16
the courts have given little consideration to its nature. Apart from the ob-
vious point that it does not require identical facilities to those enjoyed by
the majority, 17 their comments have been confined to refusing to consider
the equality provisions of s. 15 of the Charter as applicable to these cases. 18
The courts are perfectly correct to say that a requirement of equality flows
from s. 23 itself, without need to resort to s. 15. For the reason given above,
it is impossible to suppose that the Charter leaves the provinces free to
provide an inferior education system for the minority community. But that
there is no need to resort to s. 15 does not mean that one cannot resort to
it, nor that its concept of equality is not already instinct in s. 23. There is
15And it is easier to apply to some dimensions, e.g., funding, than to others. In view of the
historical underfunding of minority denominational schools in some provinces, careful scrutiny
of the level of provision for linguistic minorities would be a prudent policy. See J.E. Magnet,
“Minority-Language Educational Rights” (1982), 4 Sup. Ct. L. Rev. 195 at 213-14.
t6The most concrete determination of what equality requires has been in Marchand, supra,
note 14, in which Sirois, J. held that the school board’s failure to provide adequate shop facilities.
for the local French high school was unacceptable. Lawyers for the Attorney General of Ontario
argued that the minority community deserved a high school, but not one with its own shop
facilities. These arguments could not withstand evidence presented to the Court detailing the
disadvantages suffered by minority students who wished to take a course common in majority
language schools. See note 38, infra.
17Mahe, supra, note 14 at 546; Commission des Ecoles Fransaskoises, supra, note 14 at 368;
PE.L Reference, supra, note 14 at 256.
‘8One exception is the Appeal Division of the Nova Scotia Supreme Court in Lavoie, supra,
note 14. However, despite accepting the relevance of s. 15, the Court held that there was no
denial of equality in the fact that the province operated several anglophone schools with less
than 68 pupils but refused to establish a school for 50 francophone students. No argument was
offered in support of this conclusion.
McGILL LAW JOURNAL
[Vol. 34
no more need to suppose that s. 23 equality and s. 15 equality are different
concepts than there is to suppose that two different conceptions of equality
are invoked in s. 15 and s. 28, or two different concepts of fairness in s. 7
and s. 11 (d). Without a fuller specification of the putative distinction be-
tween s. 23 equality and s. 15 equality little more can be said.
The principle of rough equality bears on the acceptability of immersion
programmes as a way of meeting a province’s s.23 obligations. Immersion
programmes progressively introduce children to a second language by teach-
ing them various subjects in that language. The objection to such pro-
grammes as facilities for the minority community is thus not based on the
claim that they are not a form of instruction, for unlike some second-
language programmes they do not treat the minority language merely as a
subject of study. Rather, the objection is that integration of first and second-
language speakers provides an unattractive option. Immersion schooling
can be a stimulating method of second language training but it is stultifying
when used as the medium of instruction for native speakers. 19
The limits of immersion training were recognized in the JE.I Refer-
ence,20 although the Court also held that the province is free to offer such
instruction to groups too small to qualify under s. 23. However, in the
Commission des Ecoles Fransaskoises case, Wimmer, J. seemed to hold that
if there are insufficient students to justify the provision of a separate “fa-
cility”, francophone students can be combined with others, even if those
others are less fluent in French.21 Since his main point was to reject a claim
that s. 23 schools should be limited to francophone children, this may not
have been an endorsement of immersion schooling. It is clear that a child’s
ability to speak the minority language is not a condition for eligibility under
s. 23 and thus cannot be used as a ground of exclusion. However, having
to admit eligible non-minority language speakers is very different from com-
bining eligible children with large numbers of ineligible, immersion students.
Some kind of remedial programme may have to be devised for eligible
19The trial judge in Societe des Acadiens du Nouveau Brunswick Inc. v. Minority Language
School Board No. 50 (1983), 48 N.B.R. (2d) 361, 126 A.P.R. 361 (N.B.Q.B.), [hereinafterSocite
desAcadiens] heard and accepted evidence of the detrimental impact of this form of education
on native francophones. Even among those who recognize that s. 23 provides rights to minority-
language as a vehicle and not merely subject of instruction, the second-language model dies
hard. In Mahe, supra, note 14 at 535, for example, Kerans, J.A. said: “The right to receive,
out of public funds, ‘minority language instruction’ can only mean … the right to become
sufficiently fluent in that language .. ” It is somewhat odd to think that native speakers of,
e.g., English in Montreal have the right only in order to become fluent in English.
2oSupra, note 14. Without referring to immersion instruction, Kerans, J.A. in Mahe, supra,
note 14 at 535, also seems to have acknowledged that what he called “effective” instruction
should be directed at making the children full participants in the minority language community.
21Supra, note 14 at 367-68.
1989]
EDUCATION AND LINGUISTIC SECURITY
children who are not already fluent, but their numbers are likely to be small
enough that their presence will not have an adverse impact on the linguistic
environment. However, to admit any child merely at the request of the
parents risks altering the balance to favour an immersion atmosphere. Since
that would make the option of minority language instruction substantially
less attractive, it would not adequately promote linguistic security. The con-
test between majority demands for second-language instruction and mi-
nority rights to first-language instruction is settled in that only the latter are
entrenched by the Charter.
Further reasons for rejecting immersion can be gleaned from Whit-
tington v. Saanich Sch. Dist. 6322 in which it was held that having begun
education in an immersion programme does not entitle a child to so con-
tinue. The decision establishes that members of the majority who wish to
be educated in the minority language were not the intended beneficiaries of
s. 23. However, the reasoning in the case can be generalized. The notion of
“instruction in English or French” is used both to define the right and to
designate two grounds of entitlement: para. 23 (1)(b) and subsection (2).
There is no reason to think that it bears a different meaning in these closely
adjacent contexts. Thus, the reasons supporting the view that immersion
schooling cannot ground entitlement also support the view that provision
of immersion schooling is not sufficient to fulfill the provinces’ obligation
to provide minority language instruction to those so entitled.
III. Why Numbers Warrant
We turn now from the nature of instruction to the limits on that right.
The most puzzling feature of s. 23 is surely the numbers proviso. This is
indeed an unusual provision to find in a human rights law, but its role
becomes intelligible if we consider the unusual features of the interest which
it protects. Minority language rights have a collective dimension; they are
group rights, secured for individuals but for the sake of interests which are
partly indivisible from those of others. Numbers are relevant to minority
language instruction rights at two levels. First, the diffuse cultural benefits
which find their source in education constitute a public good for the entire
community and require the participation of many to sustain it. Second, the
more immediate benefits of minority language education constitute a shared
good for the children who participate. Given the requirements of effective
pedagogy, the participation of many is necessary to achieve the shared goods
of education. Since neither of these goods can be enjoyed as an individual,
neither can be understood as a matter of individual right in the classical
sense. This is not to say that an individual may not sue under s. 23. Rather,
22(1987), 16 B.C.L.R. (2d) 255, 44 D.L.R. (4th) 128 (B.C.S.C.).
REVUE DE DROIT DE McGILL
[Vol. 34
it means that the justification for protecting minority language rights does
not rest solely in the interests of a single child taken individually, but in his
or her interests as a member of a linguistic community.
To have a right is to have an interest sufficiently urgent to warrant
holding others duty-bound. In view of the social dimension of education,
some of those interests give rise to duties only when shared by a group.
There is obviously nothing that can be done for a single family, since all
their linguistic contacts will be with majority language speakers. Life in their
own linguistic milieu is already impossible for them and no institution could
provide for them the indirect collective benefits that schools normally con-
tribute to their communities. Thus, with respect to the collective dimension
of linguistic security, there is a background notion of a minimally viable
language group. The size of that group will vary with activity and context.
It cannot be captured by some fixed proportion of the whole community,
since the significance and possibilities of group life in five percent of a
metropolitan area obviously differ from those in five percent of a rural one. 23
The numbers to which the proviso refers are therefore unavoidably context-
dependent.
In respect of the immediate benefits of education, the numbers question
is largely one of determining what is pedagogically effective. As argued
above, education has a social dimension. The acquisition of social skills
and a culture itself mandates that children be taught in groups sufficiently
large to reap these benefits. Consideration of the nature and purpose of
minority language education itself establishes that schools are not to be
provided for one or two children, and that with larger groups richer and
more diverse forms of educational experience become possible. The mini-
mum size of group for a given kind of minority instructional facility is
primarily a pedagogical question, and one in principle open to evidence.
Indeed, the considerations here are in fact no different from those which
bear on majority language instruction: a school is never provided for a
handful of children, nor a science laboratory for two students. We must also
bear in mind the remedial nature of s. 23, however. The minority com-
munities outside Qu6bec are weaker than they would have been had an
educational system not been largely denied them for most of the last century.
To insist on a full and vibrant community before qualifying for a school
would only compound inherited injustice. To redress these inherited biases,
the thresholds should be set at the lowest level, Le., not at optimal class
sizes, but at minimal ones.
23This need not exclude a small set of reasonably uniform standards, perhaps varying as
between rural and urban areas, provided they are sensitive to activity and context.
1989]
EDUCATION AND LINGUISTIC SECURITY
This explains why some sort of numerical threshold is relevant in ar-
guing for the provision of minority language education. Further exploration
requires consideration of the impact of rising numbers on entitlement to
different kinds of facilities. To this we now turn.
IV. Numbers and the Sliding Scale
We shall argue that an appropriate understanding of the social character
of education supports a “sliding scale” interpretation of s. 23 according to
which the nature of the facilities to which one is entitled depends upon the
size of the group that may potentially benefit. We must reject the idea of a
single numerical threshold, both because no single number is appropriate
for the entire range of possible educational services and because any single
number likely to be chosen would be so high as to prevent education from
advancing linguistic security. However, the notion of a sliding scale is am-
biguous between two quite different approaches. We shall distinguish be-
tween a scale which consists of a series of ordered thresholds and one which
slides continuously, and argue that the former better serves the purpose of
fostering linguistic security. But first, we must remove an obstacle in the
path of either version.
In Mah6, the Alberta Court of Appeal argued that the sliding scale
approach cannot give a satisfactory account of the relationship between
paras (3)(a) and (b) of s. 23. Subsection (3) states:
The right of citizens of Canada under subsections (1) and (2) to have their
children receive primary and secondary school instruction in the language of
the English or French linguistic minority population of a province
(a) applies wherever in the province the number of children of citizens
who have such a right is sufficient to warrant the provision to them out of
public funds of minority language instruction; and
(b) includes, where the number of those children so warrants, the right
to have them receive that instruction in minority language educational fa-
cilities provided out of public funds.
After describing the sliding scale approach as the idea that “…a new right
arriv[es] at each new increase in the numbers of students”, Kerans, J.A.
argued that this “…renders s. 23(3)(a) superfluous, because, if all possible
situations fall within the scope of the second paragraph, it protects also the
right expressed in the first.”’24 This led the Court to interpret paras (a) and
(b) as encompassing, respectively, two distinct dimensions of education:
teaching and governance. 25 However, that it renders para. (3)(a) superfluous
is an objection to the sliding scale approach only if paras (a) and (b) are
24Mah, supra, note 14 at 537.
25Ibid. at 537.
McGILL LAW JOURNAL
[Vol. 34
correctly construed as two watertight compartments defining independent
rights.
This “two rights’ analysis (originally attributable to the Ontario Court
of Appeal in the Ontario Reference26) is rapidly becoming the standard
analysis of the structure of s. 23. On this view, para. 23(3)(a) grants a right
to “instruction”, whereas para. 23(3)(b) grants a distinct right, subject to
distinct qualifications, to “minority language [educational] facilities”. 2 7 This
suggests that whatever is included in facilities must be excluded from in-
struction and that the task of the court is to decide into which category each
concrete claim fits before applying the appropriate numbers test. If the con-
tent of para. (b) must be different from that of para. (a), any interpretation
of “educational facilities” which also covers some of the content of “in-
struction” is unacceptable. Kerans, J.A. understood the sliding scale ap-
proach as an interpretation of the concept of “educational facilities”. Since
the scale includes the full range of educational services available in different
circumstances, from the most modest to the most ambitious, para. (3)(a)
becomes superfluous: “instruction” is simply the lowest end of the “facil-
ities” scale.
This interpretation fails because it reads the parts of the section in
isolation from one another rather than as parts of a coherent whole. It also
overlooks the fact that the right granted by the section is already stated in
subsections (1) and (2), and not initially defined by subsection (3). Subsec-
tions (1) and (2) not only define the eligibility criteria, they also create the
right, which is to have one’s children “receive primary and secondary school
instruction” in the minority language of the province. This is emphasized
by the opening line of subsection (3) which refers to the right under sub-
sections (1) and (2). Subsection (3) is thus not an independent source of
any rights; it merely performs a clarificatory or scope-limiting role by spec-
ifying 1) what the right to receive instruction includes, i.e., the right to
minority language educational facilities provided out of public funds, and
26Supra, note 14.
27Ontario Reference, supra, note 14 at 29. This has been followed by the Court of Appeal of
Alberta in Mah , supra, note 14 at 536, which was in turn quoted approvingly by the Sas-
katchewan Court of Queen’s Bench in Commission des Ecoles Fransaskoises, supra, note 14
at 366. Lavoie, supra, note 14, is also consistent with this approach. Most commentators have
also adopted a similar analysis. See P Foucher, “Les droits scolaires des acadiens et la charte”
(1984) 33 U.N.B.L.J. 97, J.E. Magnet, “Minority-Language Educational Rights”, supra, note
15, and D. Proulx, “La prdcarit6 des droits linguistiques scolaires ou les singuli~res difficult6s
de mise en oeuvre de l’article 23 de la Charte canadienne des droits et libertfs” (1983) 14
R.G.D. 335.
1989]
EDUCATION AND LINGUISTIC SECURITY
2) where in the province the right applies. 28 Shorn of the eligibility criteria,
the substance of the right therefore amounts to this:
Eligible citizens have the right to have their children receive minority language
primary or secondary school instruction, which includes, where the number
of children warrants, the provision of minority language educational facilities
out of public funds, and which applies wherever in the province the number
of children is sufficient to warrant its provision out of public funds.
On this analysis there is only one right provided by the section: the right
to minority language instruction.29 Like any right, 30 it may require different
things in different circumstances, some of which are determined by the
specificatory clauses of s. 23.
It is therefore clear that the “two rights” theory suggests a misleading
view of the relationship between paras (a) and (b). It has generally been
assumed that the “right” to educational facilities allegedly contained in para.
(b) guarantees something more than the “right” to instruction attributed to
para. (a). “Instruction” and “educational facilities” are both general con-
cepts, but the fact that s. 23 says that the former includes the latter means
that the Charter regards “instruction” as of at least equal breadth to “ed-
ucational facilities”. Rather than supporting a sharp distinction between the
two concepts, a natural reading of s. 23 shows that they must be connected.
Indeed, the courts have recognized this to a certain extent in noting that
instruction can be provided only if there are some facilities in and through
which to do so. 31 No instruction is possible without at least an instructor,
teaching materials, and a room in which to teach. But how can that manifest
truth be reconciled with the view that locates all rights to facilities in para.
(3)(b) and subjects them to a separate numbers test? The “two rights” theory
must be abandoned.
If anything, one might stand the Mah6 argument on its head and say
that para. (a) subsumes para. (b). But, of course, the fact that “instruction”
includes “facilities” does not make the express mention of facilities redun-
dant. As noted by Kerans, J.A. in Mah6, “instruction” is not a legal term
of art with a precise meaning. 32 The right to “instruction” has no consti-
tutional tradition either in Canada or in most other western nations. Un-
28Here, we are mainly concerned with the relationship between the concepts of “instruction”
and “educational facilities”. Below we return to the issue of the geographical dimension to the
applicability of the right.
29More accurately, the right to have one’s children receive such instruction, but the abbre-
30See J. Raz, The Morality of Freedom (Oxford: Clarendon Press, 1986) at 171.
31See Ontario Reference, supra, note 14 at 37; PE.L Reference, supra, note 14 at 260-61;
32Supra, note 14 at 533.
viated formulation is more convenient.
Lavoie, supra, note 14 at 18.
REVUE DE DROIT DE McGILL
[Vol. 34
specified, it could have been taken to mean anything from a weekly class
in the minority language to everything normally included in a comprehen-
sive educational system. It was therefore sensible to draft the section so as
to signal clearly that the more generous end of the spectrum was meant to
be included. For example, para. (b) indicates that in some circumstances
separate schools for the minority are warranted and thus pushes the courts
toward a broad rather than a narrow interpretation of the right to instruction.
Since the numbers proviso suffices to ensure that more comprehensive fa-
cilities are guaranteed only when appropriate, there is no need and no ground
for a conceptual barrier between paras (a) and (b).
Once it is understood that s. 23 contains only one right, the supposed
textual obstacle to the sliding scale crumbles and it becomes clear that the
sliding scale makes sense of the section read as a whole. The social character
of education dictates that what is required by instruction varies with the
number of children available, and the inclusion of minority language ed-
ucational facilities simply expands the potential scope of interpretation.
“Instruction” covers a wide variety of arrangements, but its social dimen-
sion and certain obvious requirements of sound pedagogy mean that some
facilities cluster together in natural ways. Instruction is a lumpy good, not
a continuous one. There is no point in providing a class-room to students
who have no teacher. But on the best understanding of a successful edu-
cation, at least a small group is required to achieve the social benefits of
education.
Many educational projects are organized around group participation,
and quite generally the presence of other children is a stimulus to learning.
Similarly, we unite in a single school a number of classes and teachers,
enabling children to interact with others of various ages in a common setting
which deepens the social dimension of the educational process. Greater
numbers make possible many curricular and extra-curricular activities im-
portant to a well-rounded education. Specialized subjects, optional courses,
group activities such as drama societies and team sports, and even the
availability of a range of options all require enough students that those
interested will be able to find like-minded colleagues. Greater numbers also
make feasible the provision of specialized facilities which enrich education,
but which may be warranted only if they are in sufficient use throughout
the day. The more sophisticated these activities become and the more im-
portant the role they play in education (which typically increases in the
transition from elementary to secondary school), the higher the numerical
threshold. These considerations indicate roughly the numerical threshold
for a homogeneous minority language school. Of course, the value to the
community of homogeneous schools should not be underestimated. Their
impact on the self-respect of both students and other members of the com-
19891
EDUCATION AND LINGUISTIC SECURITY
munity is also an important contribution to a linguistically secure environ-
ment. But the net value of homogeneity decreases as educational
opportunities become narrowly confined. It may be better to give minority
language children access to a fuller range of activities in a mixed school,
even if not in their own language, than to deprive them of the opportunity
altogether. At some point, a homogeneous school may even become self-
defeating. 33
Thus, the concept of instruction encompasses various kinds of services
suitable to different numbers of children. No single numerical level could
be appropriate to all of these. What would be the likely result of choosing
a single threshold? Because the level appropriate to the creation of a class
would not justify the provision of the more ambitious facilities, the choice
of a single threshold would probably be set higher, say, at the level appro-
priate for a school. But this in turn would exclude many children from the
protection of the provision even though they and their community would
greatly benefit from something less. Far from fostering linguistic security,
this would be a retrograde step. Most provinces already allow the creation
of individual minority language classes within majority language schools.
Section 23 cannot be understood to permit the provinces to reduce their
level of services. 34 Moreover, the choice of that threshold would also incline
the courts against interpreting s. 23 to include facilities like educational
television 35 because, while a single threshold would entail that a group reach-
ing it is entitled to all the benefits of s. 23, the numbers sufficient for a
school would not justify the provision of mass media. In any event, a single
threshold is inconsistent with the view espoused in many cases that the
satisfaction of para. 23(3)(b) requires greater numbers than para. 23(3)(a).
What is needed, therefore, is something like a sliding-scale with the
type and level of services related to the number of children involved. Be-
cause education is a lumpy good, however, this should not be interpreted
as a continuous, increasing function of numbers. There are limits on the
extent to which the scale may slide. This seems to have been recognized by
Sirois, J. in Marchand.36 He rejected the government’s argument that there
were enough minority language students in Penetanguishene to justify the
33For this reason, the appellants’ concession in Lavoie, supra, note 14 at 25, that sometimes
less than “ideal” arrangements must be accepted is misguided. The quality of arrangements
can only be assessed by also taking into account the range of experience available to the pupils,
not only their isolation from majority language students.
34A.G. Quebec v. Quebec Association of Protestant School Boards, supra, note 2.
35The possibility that such services may be included in s. 23 is suggested by J. Magnet,
“Minority Language Educational Rights”, supra, note 15.
36Supra, note 14.
McGILL LAW JOURNAL
[Vol. 34
provision of a high school, but not enough for one with shop facilities. 37
The Ontario government defended its position by relying on an argument
which would permit a very finely graduated scale on which every type of
educational equipment and facility could be related to numbers. On this
logic there would be endless thresholds for the minority to cross: one for a
class, a higher one for a school, a still higher one for a school with a shop,
yet another for a shop with a drill-press, and why not another for a drill-
press with drill-bits, and so forth?38 The vision of a right to educational
resources as an indefinitely articulated, numbers-related scale is fundamen-
tally flawed. It ignores the need for rough equality of resources between
majority and minority schools if linguistic security is to be protected at all.
Within this constraint the instruction atom may be split, but not indefinitely.
Pedagogical considerations determine which groups of facilities are indi-
visible (e.g., no shop without a lathe), considerations of rough equality and
adequate use determine which facilities are essential in which schools (e.g.,
no school without a shop).39 Because instructional facilities have only lim-
ited divisibility, the sliding scale is not continuous, but is a series of ordered
thresholds.
V. Numbers and Cost
The above arguments about why and how numbers are relevant contrast
with a view according to which they are a measure of the cost of providing
minority language facilities. In Mah6, Kerans, J.A. said: “Numbers, in my
view, are relevant only as the criterion for reasonable cost.”’40 Although some
in Commission des Ltcoles Fransaskoises, supra, note 14 at 368.
37Ibid at 613. On the other hand, the Court of Queen’s Bench implicitly accepted this notion
38Lest this be thought sheer fantasy, the reader would do well to consider the facts in Mar-
chand. The French students allowed access to the shop facilities in the English school, “were
permitted to use one room in which they could not touch anything, in which they had to bring
their own defective equipment. This applies to both the technical shops and the home eco-
nomics training.” Ibid. at 604.
39We noted above that this issue is connected to the concern for ensuring that minority
language education is of roughly equal quality to that provided for the majority. In Marchand,
supra, note 14, the minority students were severely disadvantaged in being required to use the
shop facilities at another school. Of course shared facilities may sometimes be acceptable,
especially as a temporary expedient or when the combined numbers of minority and majority
students are still very small. Nonetheless, a sharing system must always ensure that the minority
is not unfairly treated. See note 38, supra.
4OMah, supra, note 14 at 541-42.
1989]
EDUCATION AND LINGUISTIC SECURITY
courts have mentioned cost as a relevant factor,4 1 no one else has suggested
that it is the only way to understand the numbers limitation, 42 nor even
that it is the most important element in it. So far, the courts have had few
serious opportunities to explore the concrete implications of taking cost into
account. A closer look suggests that it is both unwise and unnecessary to
do so.
The idea that the numbers requirement amounts mainly to a cost con-
straint is based on two things: on the plausible intuition that it would be
too expensive to provide schools or classes for very small numbers of pupils,
and on the reference to “public funds”. On these foundations some have
sought to erect the thesis that minority language education rights are subject
to general considerations of economy. Kerans, J.A., again, stated that “the
reason for the limit is not to burden a province with substantial extra cost.”’43
Thus, educational rights are guaranteed only where it is not too expensive
to do so. And, since this is no guarantee at all unless the courts can test the
constitutionality of the provinces’ criteria of expense, it must be conjoined
with fairly robust review powers, reaching even to details of the provincial
budget.
The twin foundations mentioned above are, in our view, too slight to
bear the weight of such a conclusion. First, there is in fact no reference
anywhere in s. 23 to the concept of cost at all. The only restriction is defined
in terms of number of children. The mention of “public funds” merely
secures the entitlement at public expense, provided that the numbers test
is met. Second, even if cost were involved, it could not be, pace Kerans,
J.A., as a measure of burden on the public purse. As the number of eligible
children increases, the total cost of providing instruction will in fact increase.
But whatever the numbers test means, it is certain that it provides more
entitlements, not fewer, as the size of minority groups increases. Its guar-
antees therefore expand as the burden on the public purse grows. In other
words, only those measures of cost which vary inversely with numbers are
4 1E.g., PE.I. Reference, supra, note 14 at 256; Commission des Ecoles Fransaskoises, supra,
note 14 at 368; Lavoie, supra, note 14 at 24. In the latter case, however, the Appeal Division
gave no consideration to cost in determining whether a right to some form of instruction
existed, while rejecting the claim for a separate school wholly on the ground of cost. This may
indicate that the Court did not view cost as relevant to the interpretation of the numbers
proviso since the right to instruction is also subject to a numbers constraint. However, it is
difficult to see what other textual basis there is for using cost as a factor.
421n the trial judgment in Mah, (1985) 22 D.L.R. (4th) 24 at 47, 39 Alta. L.R. (2d) 215 at
240, Purvis, J. suggested without argument that apart from cost, whether numbers warrant
could also depend on: the difficulties of transport, the age of the children, the possibility of
providing residential accommodation for students, and the impact such arrangements would
have on the social and other development of the children.
43Supra, note 14 at 541-42.
REVUE DE DROIT DE McGILL
[Vol. 34
admissible as interpretations of the numbers proviso. This excludes all costs
which vary directly with numbers. For example, books in the minority
language will frequently be more expensive than majority language ones.
But since the books needed will increase with the number of students, that
expense rises with numbers. It would make no sense to deny such a claim
on the ground that it requires an expense which the numbers are insufficient
to justify. Cost can therefore be relevant only as an average or proportionate
index. This in turn admits of two interpretations.
A. Equivalent Cost
The strongest view would be that minority language instruction is war-
ranted provided it costs no more than would majority language instruction
for those same children, taking the average majority-language expenditure
as the appropriate baseline. Under what circumstances would this permit,
for example, a minority language school? If, within a certain area, there is
a homogeneous minority-language neighbourhood, its local school would
already be filled with minority-language children. Suppose also that they
are taught by minority-language teachers but, until now, the majority lan-
guage has been used. If a claim were made on behalf of these students for
minority language instruction, the only change that would be necessary
would be in the language of instruction. From the point of view of cost, it
would make no difference which language is used. However, without even
considering the Charter guarantee, the only reasons for not allowing these
schools to be conducted in the minority language are intolerance or mis-
guided views about what is best for the minority community. Groups like
this would succeed under the Charter, but they would have just as valid a
claim on the public purse as speakers of the majority language even without
it (although this claim would not be guaranteed recognition by the province).
Section 23 would thus be interpreted as a bare anti-discrimination provision
protecting against intolerance and misguided paternalism. This is important,
but it is not all the Charter was meant to do.
Cases like this are, of course, vanishingly rare. Neighbourhoods are not
homogeneous, and an educational system is already in place, distributing
students, teachers, and resources in a particular way. Thus, some additional
costs are inevitable. First, there are the costs of transition from a system
which is not designed with minority language rights in mind to one which
is. These include new capital costs, costs of planning and reorganization,
redistributing students and resources among schools, hiring new teachers,
etc. Second, there are the continuing costs of running the new system. The
very division of schools will itself have costs as some economies of scale
become unavailable. This applies also to the minority language schools and
19891
EDUCATION AND LINGUISTIC SECURITY
classes themselves: being smaller they are less likely to be used to optimal
capacity. In addition, there are certain to be new costs in transportation and
perhaps accommodation where the minority language population is un-
evenly spread. Children who were previously able to walk to the local ma-
jority language school may have to be bused if enough children are to be
gathered to create a minority language school.
These are all additional costs which, in per capita terms, decrease as
the number of students using the facility rises. However, if services can be
denied on the ground that it costs more to educate students in the minority
language than it would in the majority language, few groups would be large
enough to succeed. A class would be created only where there are enough
minority students, without busing, to fill an average size class and where a
minority language teacher is available for reassignment. In other words, this
interpretation would allow anglophone provincial governments to reduce
their existing level of service. Since it has been accepted that s. 23 is re-
medial, 44 this interpretation must be rejected.
B. Reasonable Cost
It might be argued that precisely because certain expenses are inevitable
they should be regarded as foreseen by the Charter and thus deemed rea-
sonable. On this view, the cost criterion limits rights only when the ex-
penditure exceeds reasonable transition and continuing costs. The burden
of establishing reasonableness would fall on the plaintiff.45 Is this an im-
provement over the “equivalent cost” interpretation?
What is the threshold of reasonability and how are the courts to de-
termine it? Is it to be defined as additional per capita cost? Or as some
proportion of the provincial education budget? The latter would require a
consideration of the entire education system and all potential claims for
minority language instruction across the province, making it difficult if not
impossible for plaintiffs to prove that a school is warranted in their neigh-
bourhood. A test of per capita expenditure can be applied at the local level
and allows cost to be tied to numbers in the right way. This would require
the courts to determine how much it costs to educate a student in the existing
system, compare this with the expense under the plaintiff’s proposal, and
4″A.G. Quebec v. Quebec Association of Protestant School Boards, supra, note 2.
45Maht, supra, note 14 at 543.
McGILL LAW JOURNAL
[Vol. 34
decide what level of additional expenditure is reasonable. 46 If this is not to
be a wholly ad hoc assessment, however, the courts will need some criteria
for deciding how much additional expense is too much. What tests are
appropriate in judging this?
Whether something is unreasonably expensive can be assessed only by
considering its costs and benefits in light of the alternative uses to which
that money might be put. But that general direction settles almost nothing.
Should the courts, for example, consider only how the money might oth-
erwise be spent within the local education system, or should their calcu-
lations range province-wide? Is it legitimate for them to consider alternative
expenditures outside education on, say, health care or swimming pools?
Perhaps twice as many people might benefit from the same expenditure on
swimming pools for those schools which do not yet have one, or a new
community skating rink, as would benefit from a minority school. Must the
courts develop standards for weighing and comparing these benefits? Should
they take into account the diffuse benefits of either project to those who do
not directly participate?
There are two obvious objections to pursuing this line. First, to require
plaintiffs to prove that their claim represents as good a use of public money
as any other possible use would impose an impossible burden. 47 It is im-
plausible to assume that a provision which was meant to improve the sit-
uation of minorities can bring such burdens. Second, even if evidence could
be garnered, the courts would effectively be writing the provincial budget
in deciding the issue. Nor could this problem be avoided by confining the
comparisons more narrowly within the provincial or local education system,
for the court would first have to decide whether the existing budget for
education, at either level, was adequate.
Could courts avoid these budgetary decisions –
and the questions of
institutional role and political responsibility that they raise – by narrowing
their scrutiny? Perhaps they would say that the provinces have full discretion
46This is bound to be a very complicated matter. For example, the courts would have to
decide how to weigh the additional expense of building new schools where necessary. It would
be unfair to divide this by the existing number of eligible students as part of the per capita
additional expense, for this does not take account of the future use of the school. Should these
start-up costs therefore be discounted or amortized in some way? At what rate? For an example
of some of the problems one might expect in the application of this approach, see the third
trial judgment in Lavoie (1988), 90 N.S.R. (2d) 16. Although Hallett, J. held that the government
was wrong to include the capital cost of renovation in its assessment of the additional per
capita cost of providing a minority language school, he nevertheless found that this capital
expenditure was not warranted for the sake of fifty children.
1n Lavoie, supra, note 14 at 23-24, the Court held that although the plaintiff has the initial
burden of adducing evidence to support his or her claim, evidence of cost should be presented
by the province to whom it is more readily available.
47
1989]
EDUCATION AND LINGUISTIC SECURITY
to decide how much extra expense is unreasonable, subject only to a test
of rationality and good faith. That argument was rightly rejected in Lavoie.48
It would, again, turn s. 23 into a bare anti-discrimination provision. Since
there are always many good uses to which the government can put public
funds, the province will almost always be able to make a plausible argument
that money which minority parents want devoted to the minority language
school system would be better spent elsewhere.
Fortunately, these difficulties need not be faced. To begin with, the text
in fact gives little support to the view that cost is the engine of the numbers
limitation. The cost of providing some facility depends in part on the level
of demand for it. The proviso, however, specifies that the number of children
of eligible parents is the criterion, not the number demanding access. If the
courts were meant to compare the benefits of spending on minority edu-
cation to other possible uses, the value of the former would vary according
to the number who will actually attend the school. But s. 23 expressly refers
to a potentially larger group for the purposes of measuring numbers. This
seems to have been misunderstood in the PE.L Reference and in Lavoie.
In the former, the interveners wanted the government to plan services ac-
cording to statistical data about the prevalence of francophone families. 49
The Court responded that it is reasonable to require evidence of actual
demand before committing resources:
It would be illogical to provide minority language instructions without some
evidence of demand. It would be illogical and imprudent to provide a teacher,
a classroom, equipment or a school facility without knowing if there existed
a demand for the services offered. The cost of education is very high and it is
never funded to the extent demanded by teachers or parents. To waste money
on an insufficient demand would do an injustice to the whole school system.50
In Lavoie, the Court refused to order the establishment of a separate school
because only fifty pupils had registered, even though there was evidence of
as many as four hundred and twenty-nine eligible students in the area. 51
Both these judgments wrongly read the numbers proviso of s. 23 as if
it were the limitation in s. 20 which requires that minority language gov-
ernment services be provided whenever there is “significant demand”. If a
demand constraint were intended in s. 23, we would expect to find the same
48Supra, note 14 at 17. It is of interest, however, that the trial judge in Lavoie, supra, note
46 at 22-23, thought that alternative uses of the money were relevant, and largely for this reason
decided that the legislature was better placed to decide. Surprisingly, Hallett, J. rather quickly
dismissed the argument that the courts had the advantage of “not being concerned about
political pressure.”
49PE.L Reference, supra, note 14 at 250-51.
5OIbid. at 251.
51Lavoie, supra, note 14 at 24.
REVUE DE DROIT DE McGILL
[Vol. 34
language here. The clear difference of language between the two limitations
has a plausible justification. For various reasons, minority parents might
well be reluctant to exercise their right, especially in the early years of the
system. 52 There may be uncertainty about the nature and quality of the
facilities offered, and families may be reluctant to change their practices,
particularly if it involves dividing their children among different schools.
By using only the number of children of parents who have the right as its
test, s. 23 requires the provinces to create classes and schools even though
they may for a time be under-attended, in order to make the facilities avail-
able for use by the minorities if they choose. The education provisions
promote linguistic security by facilitating parental choice of language of
education; the extent to which the option is taken up is conditioned in large
part by the character and quality of the education offered. But no effect can
precede its cause, so there is no reason to expect or require that vigorous
exercise of s. 23 rights must precede the availability of institutions within
which those rights can in fact be exercised.
C. Efficacy and Cost
The above considerations strongly suggest that the numbers criterion
is not primarily a measure of cost. The contrary argument seems mainly to
be based on the desire to avoid the repugnant conclusion that very small
groups might be entitled to all the accoutrements of a normal school system,
which would, of course, be excessively expensive. However, the repugnant
conclusion does not follow at all, quite apart from any consideration of cost.
As we have argued above, the nature of minority education itself dictates
that certain facilities have a built-in numerical threshold. Below that thresh-
old minority language children would not get the kind of education that the
principle of rough equality requires. These pedagogical criteria are them-
selves sufficient to ensure against wasteful expenditure on minority facilities,
and to give cost further weight, in the form of an independent constraint
on rights, would involve the courts in budgetary issues and undermine the
purposes of the provisions.
Minority language education will often, perhaps always, be more ex-
pensive than a unilingual education system. But to regard this as an objec-
tion is to treat the educational systems already in place before 1982 as setting
a normatively relevant baseline. Minority language instruction is then to
be “added” to this system only if it is not unreasonably expensive. 53 But
52This will be especially so if, as so often, the school board or province has demonstrated
its resistance to providing facilities.
53This attitude was amply demonstrated in Lavoie by the school board’s argument that the
facilities suggested by the minority parents could not be made available because the board had
already promised them to various constituencies within the anglophone school system.
1989]
EDUCATION AND LINGUISTIC SECURITY
why should a system, recognized as deficient and for which constitutional
remedies were provided, now be taken as a baseline? This is simply to reason
as though the Charter never happened. In fact, it establishes a new baseline
and requires us to assess the overall provision of education in that light.
Necessary redistribution of educational resources is therefore wrongly de-
scribed as “extra” spending on the minority community; it is their due under
the constitutional obligations entrenched by the regime of official languages.
The sufficiency of numbers should be determined according to the require-
ments of establishing effective learning environments which promote lin-
guistic security. That will remain a complex and, at the margins,
controversial matter. Unlike cost constraints, however, it will be one fully
within the purposes of the education rights provisions.
VI. Where Numbers Warrant
So far we have considered how many pupils are necessary for the ap-
plication of s. 23. We now turn to the geographical dimension of the test.
Paragraph (3)(a) provides that the right “applies wherever in the province
the number of children … is sufficient to warrant the provision to them out
of public funds of minority language instruction”. This introduces further
complexity. Although a linguistic community is often thought of as having
territorial boundaries, in a modem and mobile society these boundaries are
quite porous. Some families may temporarily have to move to an area in
which their language group is very small. The cost of doing so should not
include permanently cutting off their children from the community to which
they hope to return. Other more stable groups are diffuse. For example,
there is a fairly large minority language community in Toronto, but there
is no “French Quarter”. Schools, however, usually have territorially defined
catchment areas. The difficulty is that these areas generally antedate the
constitutional guarantee of minority language education rights. Thus, “wher-
ever in the province” cannot mean “wherever within the jurisdiction of
existing school boards” since a community may overlap those boards. 54
Moreover, just as school jurisdictions are flexible, children are mobile, so
transportation facilities must also fall under the scrutiny of the constitutional
requirements. An assessment of numbers in a small town may conclude
that a school or even a class is not justified, but if we expand the area and
consider the number of children within busing distance, there may well be
enough. Provinces must therefore take into account available means of
54Ontario Reference, supra, note 14 at 32-33; Commission des Ecoles Fransaskoises, supra,
note 14 at 370-71; RE.L Reference, supra, note 14 at 250.
McGILL LAW JOURNAL
[Vol. 34
transportation when designing districts in which minority language instruc-
tion is to be provided. 55
Let us provisionally say that instruction must be provided wherever
sufficient children can be gathered through reasonable means. Reasonable
means would clearly include busing. What then of those parents who live
too far from a required area for their children even to be bused? Are they
simply out of luck with respect to the application of their rights? The as-
sumption is unnecessary. Provided there is at least one centre of minority
language population large enough to support a school, it may be possible
to accommodate all eligible minority language students in the province in
that (or in one of those) school(s). On this view, para. 23(3)(a) does not
mean that instruction need only be offered to those children who live in an
area where there are enough others to warrant its provision, but that in-
struction need only be offered in those locations and made available to all
children who qualify under s. 23. This could be done by transportation, or
even by subsidizing lodging, for those students who qualify but live too far
away.56
We can now see why the eligibility criteria for the right are themselves
defined individualistically and not by reference to numbers. Every parent
covered by subsections (1) and (2) has the right to have his or her children
receive minority language instruction. This right, however, can be exercised
only in those locations where sufficient children can be gathered through
reasonable means. In principle, then, the right is inert only when a province
has no location in which provision of instruction is warranted. 57 Thus, no
eligible child is automatically excluded from the minority educational sys-
tem, although only some of those children are guaranteed local instruction.
55This was taken for granted in Lavoie, supra, note 14, the only case so far to address the
issue of whether the provision of a minority language school was warranted in a given area.
However, for a possible endorsement of the view that the province is under no obligation to
fund transportation, see Chaddock v. School District of Mystery Lake No. 2355 and Manitoba
(1986), 31 D.L.R. (4th) 82, [1986] 5 W.W.R. 673, 43 Man. R. (2d) 81 (C.A.), leave to appeal
to the Supreme Court of Canada refused, January 15, 1987. Although s. 23 was not raised in
this case, O’Sullivan, J.A., for the Court, opined that neither the framers of the Charter, nor
the Public Schools Act intended that French and English instruction must be made available
to every child on exactly the same basis, and thus denied the plaintiffs’ claim that they were
entitled to have their children bused at public expense to the only French elementary school
in Thompson, Manitoba.
56This has also been suggested by Magnet in “Minority-Language Educational Rights”, supra,
note 15 at 213, “Language Rights: Myth and Reality” (1981) 12 R.G.D. 261 at 269, and in
“Les 6coles et la Constitution” (1983) 24 C. de D. 145 at 154, and would seem to be required
by Foucher’s emphasis on accommodating each child of eligible parents: “Language Rights and
Education”, supra, note 5 at 272-74.
57That is not presently the case anywhere. See Foucher, Constitutional Language Rights of
Official-Language Minorities in Canada, supra, note 6, for recent data.
1989]
EDUCATION AND LINGUISTIC SECURITY
Naturally, exercise of the right will be more burdensome, and thus less
valuable, where numbers do not warrant local provision. In such cases,
parents must decide whether the benefits are worth the burdens they impose
on their children.
Is this a desirable situation? The trial judge in Mah did not think so.
He stated: “Children are not by s. 23 the pawns of their parents in some
form of cultural or linguistic conflict. Section 23 rights are limited by the
children’s rights to a reasonably normal childhood without the intervention
’58 Setting aside the of-
of unusual compulsory transportation obligations.
fensive supposition about the motivation of French parents, three points
are relevant. First, s. 23 rights do not impose “compulsory transportation
obligations”. No parent is obligated to send any child to a minority language
school. Second, s. 23 rights are not limited by children’s rights to a reasonably
normal childhood. Like all Charter rights, they are subject only to such
limits as meet the test set out in s. 1, including the requirement that the
limits be established by law. That is not now the case with respect to the
putative right to a reasonably “normal” childhood. Finally, it is not, in any
case, obvious that having to ride a bus to school is more offensive to the
ideal of a normal childhood than is being educated in a foreign language.
The objection is thus without merit.
The decision whether to exercise these rights must rest with parents for
an evident reason: they are best placed and most reliably motivated to take
that decision in the interests of their children. The duty of the provinces is
not to usurp the decision, but to facilitate it by providing services on fair
terms and easing the costs of using them. This may be done in various ways.
Local concentrations may warrant schools, dispersed groups only classes,
transportation or accommodation. Given the importance of education in
making it possible for a child to participate fully in the minority language
community, one should not lightly disregard a parent’s desire to keep this
door open. Boundaries are thus controlled by the notion of linguistic se-
curity. Long absences from home may promote linguistic security in one
way, by giving schooling, but harm it in others, by decreasing the time spent
with family. Time spent on buses is time away from ancillary and sometimes
even primary aspects of education. Once again, as in the case of cost, it is
not competing considerations which define the substance of the right, but
the protection of linguistic security itself.
58Mahe, supra, note 42 at 240 (Alta L.R.).
REVUE DE DROIT DE McGILL
[Vol. 34
VII. Control of Facilities
Following the Ontario Reference,59 it has been accepted in Alberta,
Saskatchewan, Nova Scotia, and more cautiously in Prince Edward Island, 60
that s. 23 includes some sort of managerial control by the parents of eligible
children. (It should be noted that even before the Charter, New Brunswick
and Qu6bec already had in place educational systems giving the minority
extensive managerial control. 61) The Ontario Court of Appeal held that,
overall, the wording of the text was more consistent with the requirement
of parental management. The Court held, first, that the deliberate distinction
drawn between facilities and instruction must mean that more is involved
in the former. Second, it relied on the fact that the French text of s. 23 refers
to “6tablissements d’enseignement de la minorite linguistique” which seems
to contemplate that these facilities must “appertain to or be those of the
linguistic minority”. 62 Only if that community controls its schools can they
be the schools of the minority. In Mah, the Court of Appeal of Alberta
held that para. (3)(b) must be regarded as a cogent extension of para. (3)(a),
and that only governance is sufficiently important in this regard to justify
constitutional protection.63
This may be the most serious misunderstanding arising out of the “two
rights” view. The Ontario Court of Appeal treated all control rights as having
their source in para. (3)(b), as distinct from para. (3)(a), and as being subject
to a potentially higher numbers threshold. This implies that there may be
minority parents who have instruction rights but who have no control rights
at all. Similarly, in Lavoie, once it was decided that the appellants were not
entitled to a separate school, the Court gave no further consideration to
their right to some form of managerial control of the programme of instruc-
tion to which it held they were entitled. The Alberta Court of Appeal rea-
soned in Mah6 that since paras (a) and (b) must be different, and since
assigning instruction to para. (a) and control to para. (b) would constitute
a difference between them, then that is what the difference in fact is. The
argument is fallacious because there are other ways of marking the difference
which better accord with the text.
In our view, para. (3)(b) sets a threshold only for mandatory separate
facilities. Shared or common facilities are permissible below that threshold,
5 9Supra, note 14.
6Mah , supra, note 14; Commission des tcoles Fransaskoises, supra, note 14; Lavoie, supra,
61See Foucher, Constitutional Language Rights of Official-Language Minorities in Canada,
note 14; RE.L Reference, supra, note 14.
supra, note 6, for details.
62 Ontario Reference, supra, note 14 at 37-38.
63Supra, note 14 at 536.
1989]
EDUCATION AND LINGUISTIC SECURITY
though pedagogical considerations set the limit at the basic instructional
unit, i.e., the class-room. Supposing the Ontario Court to be correct in its
argument for management rights under para. (3)(b), does it follow then that
there are no management rights apart from para. (3)(b)? Is the minority
entitled to a share in the direction of instruction only if it is large enough
to warrant separate facilities? Once the “two rights” view is rejected, that
conclusion does not follow.
The argument for control rights is not in our view mainly established
by the wording of the section (though it is consistent with it), but by its
purpose in fostering linguistic security, as much of the Ontario Court’s ar-
gument implicitly recognizes. From this perspective, control is as important
to small groups as it is to large ones. First, the history of minority language
education in Canada strongly supports the view that the minorities cannot
rely on majority language school boards for fair consideration of their special
rights, and the Charter was drafted with the history of arbitrary and unfair
policies in mind.
Second, as we suggested above, an educational system is a participatory
good for a community, and self-government has intrinsic value. Linguistic
security requires that the minority education system provide, to the extent
possible, the same kind of collective benefits that the majority system pro-
vides. Parents’ sense of participation in the future of their children and,
through them, their society, is an important by-product of an educational
system incorporating parental control. It is important that the minority be
able to share in this contribution to the future.
Third, there are good strategic reasons for preferring minority control
to a situation in which the courts have continually to supervise the decisions
of majority boards as they effect the minorities. Consider the lessons of
Marchand,64 decided before Ontario had enacted a system of minority con-
trol. The Court was asked to decide whether the majority-controlled board’s
refusal to provide industrial arts facilities for the French high school violated
s. 23. Even after the Court decided that it did violate s. 23, the board
continued its obstruction and, rather than implement a proposal put forward
by the French Language Education Council (“FLEC”), (created in the mean-
time under the Education Amendment Act, 1986, (No. 2)65) the board pre-
sented a motion to the Court asking for clarification of the type and extent
of the facilities necessary to satisfy the Court’s first decision. 66 It was only
because the new Ontario structure had been implemented in the meantime
and gave exclusive jurisdiction over such matters to the FLEC that the Court
“Supra, note 14.
6SS.O. 1986, c. 29.
66Marchand v. Simncoe County Board of Education (No.2) (1987), 61 O.R. (2d) 651 (H.C.).
McGILL LAW JOURNAL
[Vol. 34
was able simply to decide that the FLEC’s proposal was consistent with s.
23 and should therefore be implemented.
Without some form of managerial control by the minority themselves,
each unfavourable decision of a school board – whether to provide a class,
or a school, how to organize busing, or the extent of the facilities to be
provided –
could become the subject of litigation. In order to avoid ob-
structionist tactics by the boards, the courts would then have to make very
detailed orders defining the required facilities. (Such obstruction is, sadly,
already well documented. 67) In short, the courts would have to decide, in
minute detail, the shape of minority language education for each province.
Quite apart from the obvious burden this would impose on the courts, it
would reduce the effectiveness of s. 23 in fostering linguistic security because
it would bleed the minority community of time, energy, and resources re-
quired by constant litigation. That is a recipe for frustrating rather than
promoting the remedial intent of the Charter.
Interestingly, the P.E.I. and Alberta courts have both held that even
parents who have no entitlement under para. (3)(b) nonetheless have some
right of participation, though it falls short of exclusive control. In Mahg,
the Court held that the concept of “instruction” itself might require “close
ties to parents and local francophone institutions; perhaps some involve-
ment in the affairs of the school by local s. 23 persons”. 68 And in the PE.L
Reference, the courts found that the right of parents to participate in pro-
gramme development and delivery is already implicit in para. (3)(a). This
will require that the minority community be associated with both the school
boards, as trustees, and with the Ministry of Education, as staff.69 This seems
correct, but it is quite inconsistent with the procrustean “two rights” view
which assigns different aspects of education to the two subsections.
The right to control is properly based on the remedial purpose of s. 23
in promoting linguistic security. It inheres in the notion of instruction itself
and arises before the entitlement to separate facilities, though in that case
necessarily in a truncated form. If this is so does it follow that the second
reference to the numbers proviso in para. (3)(b) is redundant after all? The
Ontario Court of Appeal reasoned thus:
[B]oth paras. (3)(a) and 3(b) refer to the “numbers warrant” test. The repetition
in para. (3)(b), even though in slightly different terms, would not be necessary
unless the facilities there referred to are different from those included in the
providing of instruction. It would appear, further, that a different numbers test
67See the history recounted in the Ontario Reference, supra, note 14 at 40-41.
68Supra, note 14 at 535.
69Supra, note 14 at 259.
1989]
EDUCATION AND LINGUISTIC SECURITY
might apply. Logically a larger number would be required for para. (3)(b) than
for para. (3)(a).70
But what can this mean? That a certain number of children are needed for
a class but a higher number of children are needed before those parents are
entitled to any share in the management of that class? If there were a higher
number of children, the facilities to which they were entitled would be
greater and thus different rights of control would be appropriate. Why should
the numerical threshold for rights to control of one kind of facility be greater
than the threshold needed for the provision of that facility?
The Ontario Court’s objection confuses redundancy with specification.
The numbers test is repeated simply for the reason the Court gives: in some
circumstances parents are entitled to have their children receive education
in the minority language; in other circumstances that includes the right to
receive that education in separate facilities. The threshold for separate fa-
cilities may be higher than the threshold for instruction; but that does not
show that the threshold for management rights is higher still nor that those
in shared facilities are not entitled to some share of control. Given that the
clarifying point about educational facilities was placed in a separate para-
graph, the reinclusion of the numbers proviso makes it absolutely clear that
similar considerations are relevant here also. To think that the repetition of
“numbers” must either be redundant or else refer to different numbers is
like thinking that the second occurrence of “public funds” is either redun-
dant or refers to different funds.
In our view, management rights are instinct in the general notion of
“instruction” understood in light of the remedial purpose of protecting the
linguistic security of the minorities. But since the sort of facilities provided
depends on numbers of eligible children, the sort of management structures
will necessarily vary. One should not have a school board if there are no
schools for it to govern. One should not have exclusive control over shared
facilities. But that leaves open many important alternatives. Which of these
possible management structures are required under the Charter?
This should not be answered entirely according to the criteria used for
majority language schools. We are not confined to asking whether there are
enough students within an area of the size of a typical school board to justify
a management structure. It may be the case, for example, that a city has
been divided into several school districts, but that only if the city is con-
sidered as a whole are there enough schools to justify creating an autono-
mous or semi-autonomous managerial body. A holistic interpretation of s.
23, integrating the right and its conditions, makes the guarantee of parental
control subject to the same geographical dimension as the guarantee of
70Supra, note 14 at 38.
REVUE DE DROIT DE McGILL
[Vol. 34
instruction. In other words, the right to manage applies in those locations
in the province where numbers warrant. Just as the provision of instruction
may require transporting students across pre-existing school districts, the
provision of management rights may require bringing together their parents
through the legal definition of an appropriate management area. Indeed, in
some cases, e.g., in Prince Edward Island, it might even be appropriate to
have province-wide management structures for minority language facilities.
It is useful to bear in mind that the difficulties in assembling parents from
a wider area than is now standard are fewer than in assembling their children
daily for school. Thus, the geographical limitations are no more stringent
when applied to management than they are when applied to instruction.
It is therefore possible to create new types of facilities appropriate to
the conditions that exist in the minority language sector. The government
of Ontario, for example, attempts this in The Education Amendment Act
(No.2), 198671 which constitutes trustees representing the minority language
group as French Language Educational Councils responsible for the man-
agement of minority language schools within a given district. Of course, the
management structures available to the minority must not violate any other
constitutional right, including the right to equality. But there is no reason
why one kind of managerial structure will be appropriate to all provinces,
or throughout a single province. This kind of flexibility seems preferable to
assuming on the rigid “two rights” theory72 that there is a gulf between the
guarantee of instruction and that of facilities and that the latter refers roughly
to existing school boards and is subject to a high numbers threshold. On
that view, parents who live in cities with a relatively large minority language
community are entitled to something like a school board, while parents who
live in rural areas or small towns where the best that can be done is to
provide one or two classes in the local school are entitled to nothing beyond
whatever influence they can muster on a local board on which they will
always be outnumbered.
VIII. Remedial Problems
The last issue –
and we can only briefly touch on it here –
is the
appropriate remedial posture of the courts with respect to education rights.
While rights cannot be reduced to or defined in terms of the remedies they
bring, the character of those remedies is clearly of the greatest moment in
7’Supra, note 65.
72Mah, supra, note 14; Commission des Ecoles Fransaskoises, supra, note 14. The same
tendency is evident in the Ontario Reference, supra, note 14, despite the fact that its own
endorsement of the government White Paper proposals implicitly recognizes the possibility of
alternative structures.
1989]
EDUCATION AND LINGUISTIC SECURITY
assessing the value of such rights. Without effective enforcement the most
generous of constitutional regimes is idle.
The special remedial problems involving s. 23 flow from the fact that
it creates rights to government services and does so by giving the provinces
a duty to use their legislative powers over education in a particular way.
Minority-language education rights may thus be infringed, not only by gov-
ernment action, but also by inaction. One might put this by saying that
these are positive rights and not merely negative ones. But that contrast
should not be overdrawn: most human rights require positive government
action to some extent. Mere restraint is sometimes enough to ensure that
a government does not interfere with rights, but the government’s duties are
not discharged in this way. They must not only refrain from interfering with
rights; they must also protect them from interference. 73
A more helpful distinction turns on the technique by which these rights
are protected. Section 23 is an example of a duty-imposing constitutional
rule rather than a power-conferring or validity-establishing rule.74 It does
not limit provincial jurisdiction to legislate with respect to education in the
same way that, for example, the division of powers limits the provinces’
capacity to legislate with respect to banking. In the latter case, the consti-
tution deprives the provinces of a power and any attempt by them to exercise
it is ultra vires and the purported legislation null and void. In contrast, s.
23 does not deprive the provinces of legislative powers, but rather directs
them to use their existing powers over education in a certain way: it gives
them a constitutional duty to provide minority language instruction. Where
a province is in breach of this duty, individuals are entitled to apply to the
courts who must do what they can to rectify the situation. 75 But what should
the courts do?
The traditional remedy of invalidating legislation will have little scope
here. One can, after all, only invalidate legislative action, not inaction. The
73Consider, for example, the right to security of the person. The government has a duty not
only not to interfere with this itself, but also to protect citizens from such interference on the
part of others. And the government can violate the right to vote not only by prohibiting some
citizens from voting, but by failing to create institutions which give citizens the power to vote.
The notion that all constitutional rights are merely rights to non-interference results from an
inappropriate extension of the distinction between not harming and helping from the realm
of private to the realm of public morality. See H. Shue, Basic Rights (Princeton: Princeton
University Press, 1980) at 35-40.
74For a more general discussion of the difference between these two kinds of constitutional
provisions, see, D. R~aume, “Language Rights, Remedies, and the Rule of Law” (1988) 1 Can.
J. of L. and Jurisprudence 35.
75Reference re Manitoba Language Rights [1985] 1 S.C.R. 721, 59 N.R. 321, 19 D.L.R. (4th)
McGILL LAW JOURNAL
[Vol. 34
cases thus far litigated 76 suggest that neither the courts nor counsel have
fully grasped the implications of this. Many existing statutes give school
boards or officials a discretion wide enough to encompass decisions which
would violate the Charter. Should these be invalidated? That would leave
the relevant official without the power even to make permissible decisions.
For example, the plaintiffs in Mah asked the Court to invalidate the pro-
vision of the SchoolAct 77 which gave the Minister discretion to create school
districts but did not require the establishment of francophone districts in
appropriate circumstances. But what purpose would invalidation of the Act
serve? No doubt it would put a diffuse sort of political pressure on the
government, but that may backfire and it is in any case purchased at the
cost of serious disruption of the entire school system. Further, invalidation
prevents the Minister from acceding to the plaintiffs’ request for a new
district until new legislation is passed, and that is counter-productive. Sim-
ilarly, the request of the plaintiffs in Lavoie that the Acadian Schools
Amendment 78 be invalidated because the criteria for the designation of the
area served by an Acadian school did not match the s. 23 eligibility criteria
was self-defeating. It would have prevented the establishment of any Aca-
dian schools, except through court order, until new legislation was passed.
There is no more logic in invalidating an official’s power in these circum-
stances than there is in invalidating a power to create hospitals because it
does not include the power to create minority language schools.
Though they have instinctively realized that invalidation is inappro-
priate in these cases, the courts have not yet developed adequate alternatives.
The judgements in Lavoie illustrate some of the problems. Before passage
of the Charter, Nova Scotia amended its Education Act 79 to permit the
Minister to designate French-language “Acadian schools” and areas served
by them, based on the number of children of French mother-tongue in the
district. The Charter, however, grants rights on the basis of mother-tongue
of the parents (which, in view of continuing assimilation, typically designates
a larger group). The trial judge held, mysteriously, that the legislation was
to the Charter80
both “inconsistent” with and yet also “complementary’
The Appeal Division ruled the legislation consistent with the Charter be-
cause, unlike Bill 101,8 1 it did not “directly collide” with s. 23 and could
76See especially the legislation which was challenged in Mah , supra, note 14 and in the RE.L
Reference, supra, note 14.
c. 20.
77R.S.A. 1980, c. S-3, as am.
‘7An Act to Amend Chapter 81 of the Revised Statutes, 1967, the Education Act, S.N.S. 1981,
79R.S.N.S. 1967, c. 81, as am.
8OLavoie, supra, note 46 at 27.
81Charter of the French Language, R.S.Q. 1977, c. C-I 1, as am.
1989]
EDUCATION AND LINGUISTIC SECURITY
therefore coexist with it.82 Yet these words were used by the Supreme Court
to signify that Bill 101 was such a great violation of Charter rights that it
could not be saved under s. 1 and need not therefore even be tested against
it. The Supreme Court did not say that any violation short of such direct
collision could coexist with the Charter, but only that these would have to
be -tested against s. 1 in the usual way. Misunderstanding this, the Appeal
Court opined that so far from detracting from the rights under the Charter,
the Acadian Schools Amendment had the effect of “magnifying” them.83 No
doubt this came as a surprise to the Acadians. It seems odd that a govern-
ment qan magnify one’s rights-by a combination of inaction and inadequate
provision. After all, the Acadian Schools Amendment is not what Nova
Scotia has in addition to a regime implementing the Charter, it is what it
has instead of a regime implementing the Charter.
It is clear that to provide for the creation of francophone schools from
which some children eligible under s. 23 would be excluded does not fulfil
the constitutional duty to provide instruction for the class defined by s. 23.
What then will the courts say when a parent whose child is excluded under
the statute but is eligible under the Charter sues to have the child admitted
to an Acadian school? To order the child to be admitted is to recognize that
the legislation is defective even if not invalid. So we need a more subtle
vocabulary to describe the kind of unconstitutionality involved here: there
has been a breach of constitutional duty even though there has been no
invalid exercise of powers.
Some flexibility can be introduced, perhaps, by avoiding the all-or-
nothing character of invalidation and allowing instead that a statute or
decision can be. declared ineffective to the extent that it conflicts with the
requirements of s. 23. For example, one of the issues in the Commission
des Ecoles Fransaskoises case 84 concerned a regulation allowing the Minister
to refuse to recommend the designation of a school as francophone if the
programme could not be sustained for more than three consecutive years
or if adequate provision could not be made for anglophone students. The
regulation is inconsistent with s. 23 to the extent that it permits the Minister
to refuse to designate a school when s. 23 would require it. The Court, rather
than using the language of invalidity, declared the provision of no force and
effect to the extent of the inconsistency, thus preventing the Minister from
82Supra, note 14 at 28. The Court of Appeal in Mah, supra, note 14 followed a similar
approach arguing that the legislation which failed to recognize all the rights granted by s. 23
was supplementary to the Charter, not contradictory. The point is that contradicting s. 23 is
not the only way to violate it. Ignoring its requirements is also a violation of duty.
83Lavoie, supra, note 14 at 27.
“aSupra, note 14.
REVUE DE DROIT DE McGILL
[Vol. 34
basing decisions on unconstitutional grounds while not excluding valid uses
of his powers.
However, introducing remedial flexibility in this way or by the inter-
pretative technique of “reading down” over-wide discretion is powerless
against inaction. In the Commission des Ecoles Fransaskoises case the Court
declared that certain provisions of the Education Act 85 are of no force and
effect to the extent that they fail to recognize the managerial rights of the
minority. What consequence could this declaration have? It makes known
the Court’s view that the legislation is deficient, but does nothing to fill the
gap. It has no immediate effect, since no particular decision or act is rendered
ineffective by it. The most this can accomplish is to allow the courts to deny
effect to any later decision made by officials which actively interferes with
the ability of the minority to manage their own schools. This shows that in
the face of legislative inactivity, mandatory orders may ultimately be the
only effective remedy.86 Merely to refuse to give effect to improper decisions
will not itself ensure that the correct decisions will be made. Such remedies
are certainly available under the Charter, provided only that they are “ap-
propriate and just in the circumstances”. 87 The failure of invalidation sug-
gests that they are appropriate, at least in the sense of efficacious and
proportionate to the violation of rights. Are they, in these circumstances,
also just remedies? Or would their use raise some other general concern,
such as the appropriate division of labour between courts and legislatures?
An educational system is a complex and expensive regime of physical
resources, human skills, and organization. There is no doubt that legislatures
and administrators are in the best position to create such a complex struc-
ture, and that courts have a lively awareness of their own limitations in this
respect. They rightly hesitate to draft comprehensive blueprints for imple-
menting s. 23, not only because of their limited competence, but also because
there is no single scheme which is required by the Charter.88 There is room
for legitimate variation among the provinces in accordance with local needs
and circumstances. In some cases, these variations will reflect the different
responsibilities to religious minorities which must be accommodated by the
minority language system, in others matters of convenience or efficacy. This
will inevitably lead to differences among the structures ultimately adopted
across Canada.
85R.S.S. 1978, c. E-0.1 (Supp.) as am.
86For a similar argument, see Proulx, supra, note 27.
87S. 24(1): Anyone whose rights or freedoms, as guaranteed by this Charter, have been infr-
inged or denied may apply to a court of competent jurisdiction to obtain such remedy as the
court considers appropriate and just in the circumstances.
88Ontario Reference, supra, note 14 at 43; Mahe, supra, note 14 at 533-35, 544; Commission
des Ecoles Fransaskoises, supra, note 14 at 361; PE.L Reference, supra, note 14 at 261.
1989]
EDUCATION AND LINGUISTIC SECURITY
There are, however, at least two remedial options between declarations
of ineffectiveness and comprehensive educational planning by the courts.
With respect to small-scale issues, the courts can provide for some imme-
diate needs of litigants rather than leaving them subject to the vagaries of
the legislative timetable. For example, assuming adequate evidence were
presented, a situation like that giving rise to the PE.L Reference could be
met with an order requiring the establishment of a school for all eligible
children. This would ensure that at least these children get the education
to which they are entitled. To make them wait until the government designs
a minority language educational system is bound to have an adverse impact
on their education. Provided that the court is presented with a reasonable
and constitutional interim proposal, it should be implemented through man-
datory orders. The courts need not be satisfied that the proposal is the best
available one; they need only be confident that it will fulfill the essential
needs of the litigants without closing any option which the government
might legitimately wish to pursue. This does not irrevocably commit the
legislature to a particular approach to the organization of the system. Indeed,
the court could make it explicit in the order that it is effective only until
the government makes provision for the relevant situation. In this way the
ultimate power to determine the contours of the province’s system would
rest with the government, but parents and their children would not be re-
quired to wait for the benefits to which they are immediately entitled under
s. 23.
Further, there may remain cases of complete and continued government
inaction. It is impossible for the courts to refuse to intervene in such sit-
uations without completely nullifying the aims of s. 23. There are two al-
ternatives. The courts could do their best to implement s. 23 on a piecemeal
basis in response to claims raised by parents. But this would be a lengthy
process and would tax the minority community with the need to litigate
every complaint. Further, as more and more claims are adjudicated, the
courts will inevitably face the necessity for more systematic planning and
will therefore have to make the kind of judgments about which they now
seem so hesitant. A better option may be to issue mandatory orders requiring
the Minister of Education to design and establish a system for the imple-
mentation of s. 23. This would leave it to political officials to determine
the shape of the system, subject, of course, to judicial review to ensure that
whatever is adopted does in fact conform to the Charter. An appropriate
time limit should be included in the order. This would put an end to gov-
ernmental inactivity and ultimately establish a mechanism that should not
require constant judicial supervision.
A final consideration suggests that mandatory orders are not merely
permitted but required by justice. Time is running out for many minority
McGILL LAW JOURNAL
[Vol. 34
at least at the outset –
groups. The courts may well prefer to assume –
that no province wishes to deny minority rights, and that the courts need
only declare the inadequacies of the existing legislation to induce the pro-
vince to amend it. This may be the correct attitude in the first instance. But
Canada is already seven years into the Charter era: that is long enough for
a whole cohort of minority language children to have passed through the
elementary school system. A recent study shows that, outside Quebec, fa-
cilities are still unavailable for half of the minority language students who
are constitutionally entitled to them.89 This should be understood in light
of the fact that, unlike s. 15, s. 23 had no time-delay fuse; it was expected
and intended to have immediate effect. 90 This seems to have been ignored
by the courts in Mah and Lavoie who said that the inadequate statutes,
having been passed before the Charter came into force, could not have been
intended to implement s. 23 and therefore cannot now be faulted for failing
to do so.91 They may be correct in holding that those statutes are not, for
that reason, invalid. But that leaves many other remedial options. Does
justice permit a government to evade its constitutional duties by the ex-
pedient of not trying to fulfil them? Is it not significant that the provinces
have already had years to exercise their powers and, where necessary, amend
or supplement their legislation as required by the Charter? To the extent
that provincial regimes remain inadequate there has already been a depri-
vation of rights for a considerable period: another generation has been lost.
The intended beneficiaries cannot fairly be expected to wait any longer. Not
only their own education, but the linguistic security of their whole com-
munities is at stake.
89Ang~line Martel, Les droits scolaires des minorit~s de langues officielles au Canada; de
l’instruction d la gestion (Ottawa: Commission des langues officielles, 1989): Forthcoming.
9As Sirois, J. noticed in Marchand, supra, note 14.
91Mah, supra, note 14, at 544; Lavoie, trial judgement, supra, note 46.
