McGill Law Journal ~ Revue de droit de McGill
RANDS LEGAL REPUBLICANISM
David Dyzenhaus*
Justice Rands judgment in Roncarelli v.
Duplessis is best understood in light of recent
political and legal theory that argues for the
importance of the republican ideal of non-
domination for in it he sets out an account of
the rule of law that gives clear expression to
that ideal, one founded in a more basic ideal of
respect for persons. As Rand understood things,
Roncarelli was a member of a disliked minority,
who was singled out for persecution when he
had done nothing more than exercise his rights
as a free and equal subject of the law. Those
who singled him out for persecution sought to
achieve their ends through law.
The author argues that since government
under law is valuable because it helps to secure
non-domination (the rule of law rather than the
arbitrary rule of men), to use law to single out
an individual for domination is, as Duplessis
discovered, rather a complex business. No mat-
ter ones grip on power, one might find that
ones ends simply do not count as public ends
within a system of public law because such a
system is predicated on respect for the persons
who are subject to its authority.
limportance de
Le jugement du juge Rand dans laffaire
Roncarelli c. Duplessis se comprend le mieux
la lumire de la thorie politique et juridique
rcente qui dfend
lidal
rpublicain de non-domination . En effet, lap-
proche de Rand quant la primaut du droit
exprime bien cet idal, qui est son tour bas
sur lidal plus fondamental du respect des
personnes. Selon Rand, Roncarelli tait membre
dune minorit qui ntait pas apprcie, il a t
point du doigt et perscut alors quil navait
rien fait dautre que dexercer ses droits en tant
que sujet libre et gal de la loi. Ceux qui lont
perscut ont tent datteindre leurs objectifs
par lentremise du droit.
Toutefois, comme la appris Duplessis, le
fait pour un gouvernement dtre soumis au
droit aide garantir la non-domination ,
c’est–dire la primaut du droit et non ltat
arbitraire de lhomme, et rend plus difficile
lutilisation du droit pour viser un individu afin
de le dominer. Peu importe son emprise sur le
pouvoir, une personne pourrait dcouvrir que
ses objectifs ne constituent pas des objectifs
publics dans le cadre dun systme de droit
public puisquun tel systme est fond sur le
respect des personnes qui sont sujettes son
autorit.
* Professor of Law and Philosophy, University of Toronto. I thank all of the participants
at the symposium for two days of exhausting but very illuminating discussion, and
Robert Leckey for written comments. Most of all, I thank our organizer, Genevive Car-
tier, not only for her work in regard to the symposium, but also because the central idea
in my contribution (the problem of the legalization of illegality) is something to which
she drew my attention some years ago.
Citation: (2010) 55 McGill L.J. 491 ~ Rfrence : (2010) 55 R.D. McGill 491
David Dyzenhaus 2010
492 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
Introduction
I.
II.
Liberalism and Its Critics
Legal Republicanism
Conclusion
493
495
497
509
RANDS LEGAL REPUBLICANISM 493
Introduction
The decisions of apex courts are immortal in the way that very famous
people are: they live on in the public record, even if much of that record
fades from active memory. And like a small group among the very famous,
a decision such as Roncarelli v. Duplessis1 lives on not only through its
place in the public record, but also because it is part of active memory. It
is remembered as having a significance beyond the fact that the apex
court of the land considered the legal issue in the case fit for its scrutiny
and resolution.
That the Supreme Court of Canada considered an issue fit for its con-
sideration is in itself supposed to indicate that the issue and its resolution
are matters of national legal importance. But often the Courts view is not
borne out by the attention the decision getsat least not beyond the rit-
ual of recital by lawyers and in judgments in related matters of its style of
cause for some years afterwards. Even when the Supreme Court of Can-
ada specifically articulates its view that a decision is a particularly impor-
tant contribution to the law, intended to provide a definitive resolution of
some difficult and important issue, that decision might earn more ritual-
istic respect than others for a time, but be doomed to fade from active
memory if neither its reasoning nor its rhetorical power suffices to cap-
ture the legal imagination of lawyers, whether in the academy or in prac-
tice.2 Rhetorical power, however, is not a necessary element for such cap-
ture. Powerful reasoning may not be memorably expressed, though many
decisions are remembered for a passage that exhibits the rhetorical power
with which a judge articulated some general insight rather than for any
bit of substantive reasoning.
Justice Rand, of course, had a rare gift for both rhetoric and reason-
ing, which is why, as Thomas Berger put it in Fragile Freedoms, his
judgements … cannot be read swiftly, and certainly they require thought,
but the truths they yield make a careful reading of them deeply worth-
while. [They] are the Canadian judiciarys greatest monument;3 that is,
Justice Rands judgments are remembered as a part of active memory,
both for what they conveyed and how they conveyed it, though as Berger
suggests, the how sometimes gets in the way of the what.
1 [1959] S.C.R. 121, 16 D.L.R. (2d) 689 [Roncarelli cited to S.C.R.].
2 I would predict that the recent decision by the Supreme Court of Canada in Dunsmuir
v. New Brunswick is an example of a case where the billing given by the Court to its
own decision will not be vindicated, precisely because there is nothing in either the rea-
soning or the rhetoric of the majority that is capable of capturing the legal imagination
(2008 SCC 9, [2008] 1 S.C.R. 190, 291 D.R.L. (4th) 577).
3 Thomas R. Berger, Fragile Freedoms: Human Rights and Dissent in Canada (Toronto:
Clarke, Irwin & Company, 1981) at 182-83.
494 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
Such obstruction happens with Justice Rand for two reasons. First, he
on occasion allowed rhetorical power to substitute for reasoning so that
even the sympathetic reader might conclude that Justice Rand was en-
gaging in rhetoric for effect rather than for insight. This is, in my view, a
markedly and unfortunately persistent feature of his extracurial writings.
Second, on occasion, his rhetoric alerts us to the fact that he needs to con-
vey an insight that is part of or encapsulates a powerful chain of reason-
ing but one that he cannot render wholly explicit. Here we might say that
rhetoric outstrips reasoning but does not substitute for it.
In my view, this second feature characterizes the judgments that
make up his monumental contribution to Canadian public law. And for a
set of judgments to have this claim, they must do more than mark some
discrete events; they must also stand as an enduring contribution to our
legal life. They must remind us of principles that would be foolisheven
dangerousto forget, since their articulation in the judgment is particu-
larly apt and, more importantly, are capable of providing us with new in-
sights into our contemporary situation.
As such, these judgments differ from even the small group among the
very famous people who live on in our active memory, for the judgments
remain live interlocutors in our debates. While those who examine the life
of another in a search for its meaning might imagine themselves as en-
gaged in a conversation, there is something different about a judgment
since, if we are engaged with it, it is as alive on its fiftieth anniversary as
it was on the day it was issued.
So one of the reasons we gathered to celebrate the fiftieth anniversary
of Roncarelli is that we started with the assumption that there is much in
Justice Rands judgment that is still capable of providing us with such
new insights. Perhaps there is even a kind of necessary aspect to the fact
that some, perhaps many, great judgments are great precisely because
their reasoning can never be rendered wholly explicit. For it is only with
time that we understand them, orso we suppose at any particular point
of timeunderstand them better.
The ambition of this symposium was correspondingly to come to some
new understandings, even if we differed about what they are. And even if,
as the symposium proved, we should now think that the dissenting judg-
ments in Roncarelli are valuable in ways that have not been hitherto ap-
preciated, we might want to be more critical of Justice Rands jurispru-
dence. For even if that critical stance is the correct one, we would not
have been led to adopt it, as the symposium also proved, if Justice Rand
had simply concurred in one of the other majority judgments in Ron-
carelli. Justice Rands judgment remains the reason we read and reread
Roncarelli, however differently we may read it at different times.
RANDS LEGAL REPUBLICANISM 495
I have already made my own attempt at a detailed reading of Ron-
carelli in a Rand Lecture at the University of New Brunswick,4 and the
five-year gap since that attempt has not been sufficient for me to develop
any new ideas. Instead, I choose here to refine and elaborate my under-
standing of the passage I used as an epigraph to that lecture. I had not
given the paragraph any sustained attention at that time, however, be-
cause while clearly I had found its message attractive, I could not then, as
it were, decipher its meaning.
The passage is taken from an extracurial writing, an essay penned
soon after Roncarelli was handed down. As we will now see, it is to a large
extent an example of rhetoric substituting for reasoning in Rand. There is
no clear connection between the individual sentences, and none of the
sentences by itself expresses a clear thought. I will argue, however, that
the paragraph, when read together with Roncarelli, expresses an attrac-
tive legal theory that is aptly termed legal republicanism.5
I. Liberalism and Its Critics
The absence of express constitutional limitations to legislative ac-
tion has not remitted the individual to the sometimes precarious and
sluggish security of public opinion and legislation. The matrix of leg-
islation in a common law parliamentary sovereignty is instinct with
the paramount purpose of sustaining democratic institutions toward
which the judicial process of interpretation should be both respon-
sive and resourceful. The urgency for their effective assertion comes
into play in times of stress and danger; it is then, in the confusion of
fear, distrust and fanaticism, that voices uttering the deep postu-
lates of free men should be heard and felt.6
As I have just indicated, this is a somewhat dark and incoherent pas-
sage. Yet there is something intriguing in it. It starts with the idea that
the legal subject is not abandoned to the play of public power in the ab-
sence of express constitutional limitations. It then indicates that legisla-
tion itself provides protection in a system of parliamentary sovereignty,
that the protection is inherently democratic, and that judges have a spe-
cial responsibility to maintain that democratic quality. Finally, Rand sug-
gests that the assertion of these democratic institutions is particularly
important in times of stress and that it is in such times that the funda-
4 See David Dyzenhaus, The Deep Structure of Roncarelli v. Duplessis (2004) 53
U.N.B.L.J. 111.
5 In this regard, my argument resonates with several of the other contributions to this
Special Issue. See especially Evan Fox-Decent, Democratizing Common Law Constitu-
tionalism (2010) 55 McGill L.J. 511; Matthew Lewans, Roncarellis Green Card: The
Role of Citizenship in Randian Constitutionalism (2010) 55 McGill L.J. 537.
6 Ivan C. Rand, Some Aspects of Canadian Constitutionalism (1960) 38 Can. Bar Rev.
135 at 160 (the original text reads the deep postulates of feee men).
496 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
mental postulates of legal orderthe deep postulates of free menmust
be forced into view.7
This set of ideas resonates with the recent revival in legal and politi-
cal theory of republicanism. Republicans present themselves as critics of
liberalism, but, as they often recognize, they have considerable difficulty
distinguishing themselves from the liberal tradition broadly conceived.8 It
is probably safe to predict that, as with the debate between liberalism and
communitarianism, liberalism will win the debate against republican-
ism since, like communitarians, republicans are not utopian thinkers
dreamers of an absolute in which all the institutions of liberal democracy
are to be smashed and replaced with some radically different, though
vaguely conceived new order. But it is also safe to predict that, while this
will be a victory for liberalism as it continues to be the dominant mode of
discourse in political and legal theory and republicanism virtually disap-
pears as a banner for critics, the discourse of political and legal theory will
nevertheless be deeply affected. We can appreciate this through a quick
look back at the debate between liberals and communitarians.
Liberalism won that debate because liberals were able to show not
only that strands within the liberal tradition had as a matter of history
been attentive to the concern of communitarianism with the place of tra-
dition and community in shaping and nurturing individual autonomy, but
also that as a matter of normative argument, it was important for liberal-
ism to revive those resources in order to focus on that kind of concern.
Moreover, the issue was, of course, not driven by considerations com-
pletely internal to academic debate. The political, social, and economic
standing of minorities, whether immigrant, aboriginal, or national within
nation states, was and is at stake in that debate; and it was no accident
that the academic debate happened at a time when such issues were com-
ing to the fore.
I do not mean to suggest that liberalism solved the problem of minor-
ity standing either at a theoretical or at a practical level. Rather, it more
or less successfully internalized the problem, making it one for liberal po-
litical and legal theory, with the concern of the communitarians taken
over by the multiculturalists, a group who argue for both theoretical and
practical reforms, but who make their arguments as liberals, not as critics
of liberalism. So my prediction for republicanism is that its critique of lib-
eralism will be internalized in the same way as the communitarian cri-
7 Ibid.
8 See e.g. Samantha Besson & Jos Luis Mart, Law and Republicanism: Mapping the
Issues in Samantha Besson & Jos Luis Mart, eds., Legal Republicanism: National
and International Perspectives (Oxford: Oxford University Press, 2009) 3 (making
strenuous and not very convincing attempts to distinguish themselves from liberals).
RANDS LEGAL REPUBLICANISM 497
tique was internalized, but also that the importance of the normative
claim that is central to the critique will force a shift in liberal inquiry.
As I have suggested, for such a shift to happen, two conditions must be
in place. First, the liberal tradition must be composed of strands that in
fact provide the normative resources for it to make the shift. Second, there
must be practical considerations that make it normatively imperative that
such a shift occur. But before I embark on discussion of these conditions,
it is important first to establish both an understanding of republicanism,
especially legal republicanism, and an understanding of why Justice
Rands decision in Roncarelli is quintessentially and illuminatingly repub-
lican.
II. Legal Republicanism
There is no uncontroversial way of describing republicanisms central
concern. However, one republican theme, most powerfully articulated by
the philosopher Philip Pettit, is that of non-domination.9 It seems to be
becoming a dominant theme, in part because of its inherent importance,
and in part because it is the best candidate for establishing republicanism
as a distinct position within political theory. In this second regard, we
should note that republicanism has to distinguish itself not only from lib-
eral theory, but also from those democratic positions that argue for an en-
hanced role for parliaments in political life and regard liberalism as com-
mitted to taking political power away from the people by making judges
the ultimate deciders on fundamental questions of political morality. I will
call these positions parliamentarist, an ugly label but one that avoids the
question begged in supposing that there is something quintessentially
democratic about them; that is, it avoids the definitional fiat that implies
that liberal positions are undemocratic.
Pettit and other contemporary republicans identify liberalism with the
idea of liberty as non-interferencean idea coined by Thomas Hobbes,10
entrenched by Jeremy Bentham,11 and rearticulated in the twentieth cen-
tury by Isaiah Berlin in his contrast between negative and positive lib-
erty.12 Republican liberty, by contrast, is what one has when one is not
subject to the domination or mastery of other individuals or the state. As
9 Philip Pettit, Republicanism: A Theory of Freedom and Government (Oxford: Clarendon
Press, 1997) at 21.
10 Thomas Hobbes, Leviathan, rev. student ed. by Richard Tuck (Cambridge: Cambridge
University Press, 1996).
11 See generally Jeremy Bentham, The Works of Jeremy Bentham (New York: Russell &
Russell, 1962).
12 Isaiah Berlin, Two Concepts of Liberty in Four Essays on Liberty (Oxford: Oxford
University Press, 1969) 118.
498 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
Pettit points out in the pioneering philosophical analysis of republican-
ism, mastery and interference do not amount to the same thing.13 One is
subject to mastery when one is in practice or potentially subject to the de-
cisions of another who is permitted to decide arbitrarily, that is, as he
likes, without regard to what will serve the interests of those subject to
his decisions.14 In contrast, if someone is allowed to interfere with me, but
only on condition that the interference promises to further my interests
and in accordance with opinions I share, there is no domination since the
interference is not on an arbitrary basis: The person envisaged relates to
me, not as a master, but more in the fashion of an agent who enjoys a
power of attorney in my affairs.15
The idea of freedom as non-domination is politically significant for at
least the following reasons. It allows us, first, to pinpoint what is wrong
with the situation where a slave has a generous master who hardly inter-
feres at all with the slaves decisions. While the master does not in fact in-
terfere in the slaves life, he is entitled to do so at whim, so the slave re-
mains in a condition of unfreedomhe is still subject to the dominion of
the master. However, on the negative liberty view of freedom, where only
actual interference counts, the slave not only is free but also has more
freedom than the citizen of a social welfare state, where the state inter-
feres constantly in the lives of its subjects.
Second, the idea does not commit its adherents to supposing that any
interference in ones life is dominating. On the negative liberty view of
freedom, any interference limits liberty and so has to be counted as a
moral loss. While this view contemplates that some interferences are nec-
essary, to keep at bay the state of nature in Hobbes, or to promote overall
utility in Bentham, the question with each interference is whether there
is some benefit that outweighs the moral loss incurred. There is thus
something morally problematic about any interference, which is why
those who hold the negative liberty view of freedom distrust state inter-
vention and are inclined to entrust the judiciary with guarding the space
of negative liberty against the interferences mandated by legislatures.
In other words, there is supposed to be a distrust of parliaments and
an emphasis on judicial guardianship of individual negative liberty inher-
ent in the liberal position. In contrast, republicans are not opposed to in-
terference as such and are open to arguments about what form of institu-
tional arrangements will make state interferences non-dominating so that
state interference in general is not regarded with suspicion, but rightly
regarded as legitimate as long as it serves the cause of non-domination.
13 Pettit, supra note 9.
14 Ibid. at 22-23.
15 Ibid. at 23.
RANDS LEGAL REPUBLICANISM 499
Third, the view of freedom as non-domination does not fall prey to a
supposed liberal obsession with state interference in the lives of individu-
als; that is, republicans oppose not interference but dominating interfer-
ence. But since what they oppose is dominating interference, they oppose
it whatever its source. Thus they will be alert in a way that liberals are
not to non-state or private sources of domination, whether the domination
is exercised by economic elites, or by men over women, or by one ethnic
group over others. Correspondingly, they will be more concerned with fur-
thering the cause of substantive equality than liberals. Since, by the sec-
ond reason, republicans have no moral aversion to non-dominating inter-
ference, and, by the third reason, will welcome interference when it not
only is non-dominating but also actively removes sources of domination,
republicans can unproblematically suppose that a state that actively pro-
motes social welfare is promoting freedom, as long as what it does is done
in the right way.
That of course leaves the question of how to tell when interference is
non-dominatingthat is, when it is arbitrary and when it is notwhich
brings us back to the first reason. Republicans tend to embrace some form
of the parliamentarist position because they associate monarchical rule
with arbitrary despotism. Thus they regard the political struggles to
wrest political power from the monarch in order to locate it in Parliament
as furthering the cause of republicanism. But, it is important to see that
there are two quite independent arguments about why this relocation of
political power might be considered republican. One we can call the repre-
sentation argument, since it puts forward the familiar thesis that with
universal suffrage in place and Parliament as the sole legislative body,
public political decisions will be decisions in which formally all can be said
to participate. Such decisions are more likely than those produced by any
other procedure to track the interests of those subject to them, and thus
will be non-dominating. The other we can call the rule-of-law argument,
since it focuses on the fact that if political power is located in Parliament,
the exercise of power will be in the form of general laws to which public of-
ficials can be held accountable; and this accountability to law is what ren-
ders the exercise of power non-dominating.
As I will argue below, there is no immediate inconsistency in putting
forward both arguments. But they do not necessarily reinforce each other.
Parliament may, for example, delegate power by law to the executive in
such a way that the executive is by law entitled to act in an arbitrary
fashion. Consider, for example, Justice Cartwrights dissent in Roncarelli.
While Justice Cartwright held the view that Duplessis had not influenced
the decision to cancel Roncarellis licence to any great degree, he was pre-
pared to assume that Duplessiss instructions to Archambault constituted
500 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
a determining factor for the purposes of the legal discussion.16 But, after
reviewing the statutory framework, he observed:
On a consideration of these sections and of the remainder of the
Act I am unable to find that the Legislature has, either expressly or
by necessary implication, laid down any rules to guide the commis-
sion as to the circumstances under which it may refuse to grant a
permit or may cancel a permit already granted. In my opinion the
intention of the legislature, to be gathered from the whole Act, was
to enumerate (i) certain cases in which the granting of a permit is
forbidden, and (ii) certain cases in which the cancellation of a permit
is mandatory, and, in all other cases to commit the decision as to
whether a permit should be granted, refused or cancelled to the un-
fettered discretion of the commission. I conclude that the function of
the commission in making that decision is administrative and not
judicial or quasi-judicial.17
And he invoked Justice Masten in Re Ashby, stating that the legislature
intended the bearer of such administrative discretion to be a law unto it-
self.18
Since an official who is a law unto himself by definition wields arbi-
trary, dominating power, republicans cannot rest with the representation
argument alone; they need to bring in the rule of law argument in order to
forestall the prospect of an executive that is empowered by statute to
wield arbitrary power. They want, in other words, not merely rule by Par-
liaments statute law, but also rule in accordance with the rule of law.
Now there is an account of the rule of law that tries to restrict its op-
eration to principles that help to determine a content of particular laws so
that the law to which officials are held accountable is the law with the
content that one can plausibly claim that Parliament, as a matter of fact,
intended that law to have. This legal positivist account of the rule of law
might well seem to be attractive to republicans, since if they accept the
representation argument, they will want the law made by Parliament to
be interpreted in accordance with principles that seek to determine a con-
tent of this sort. So, for example, Samantha Besson and Jos Mart sug-
gest that legal republicanism ought to encompass a positivist theory of
law, because it cannot rely on the existence of a natural, pre-political va-
lidity.19
Their point, though rather obscurely put and unelaborated in their es-
say on legal republicanism, is, I take it, as follows. The alternatives to
16 Roncarelli, supra note 1 at 164.
17 Ibid. at 166-67.
18 [1934] O.R. 421 at 428, [1934] 3 D.L.R. 565 (C.A.), cited in Roncarelli, supra note 1 at
167.
19 Besson & Mart, supra note 8 at 32.
RANDS LEGAL REPUBLICANISM 501
positivist accounts of the rule of law are natural law positions, which hold
that the principles of the rule of law are moral principles whose content is
established by arguments that are prior to, or transcend, politics. Since
Besson and Mart reject the claim that there are such prior moral princi-
ples, they must suppose that moral principles are simply the principles
that a republican political order decides on. It thus seems to follow that
republicanism must content itself with an account of the rule of law that
is instrumental to the production of principles. The principles of the rule
of law are then not moral. They are simply principles of efficacy, which
ensure that the rule of law is the rule of the lawof the law with a repub-
lican content.
Any such account of the rule of law must, however, cope with the prob-
lem of indeterminacy in law, those occasions when there is no uncontro-
versial candidate for the content of the law in this positivist sense. And
that problem is writ large in the administrative state because of the ex-
plicit delegations to officials of discretion as to both how to interpret the
law and how to implement it. Moreover, it is easy to see how the positivist
account can lead to conclusions such as the one Justice Cartwright drew
in Roncarelli: since the legislature did not intend as a matter of fact the
discretion to be controlled, it is uncontrolled. This problem alone thus at-
tracts republicans to richer accounts of the rule of law than the positivist
one. On such accounts, judges are entitled on the basis of principles of
fairness and justice to interpret statutes so as to find that officials are
controlled by law, where law means not only the content of the positive
law, but also whatever principles of legality are plausibly claimed to be
part and parcel of the rule of law.
On such accounts, judges are entitled to attribute to Parliament the
intention to abide by these principles so that, in a situation such as that
presented by the statute in Roncarelli, the officials discretion is controlled
by such principles. Moreover, Parliament is assumed to have that inten-
tion as a matter of its place in a legal order that is committed to constitu-
tional governmentgovernment according to law in both the senses
sketched above. And that assumption has the following consequence:
judges should always attempt to find that the content the law has is one
that furthers rather than frustrates those principles. So if one interpreta-
tion of a law seems to frustrate these principles, and another plausible in-
terpretation furthers them, judges are under a duty to choose the latter.
In this situation, if the first interpretation is the one that is suggested by
a positivistic account of the rule of law, it should still be rejected.
Consider, for example, that article 88 of the Quebec Code of Civil Pro-
cedure (C.C.P.) required that those who wished to bring suit against a
public official give that official thirty days notice, and Roncarelli failed to
502 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
give such notice. As Robert Leckey points out,20 Justice Fauteuxs dissent
in Roncarelli pivoted on an interpretation of article 88 C.C.P. as making
sense only if it was understood to confer procedural protection on acts that
might ultimately in substance be characterized as illegalan interpreta-
tion that he arrived at by reading its text literally and by looking to actual
indications of legislative intent. But this interpretation should be anath-
ema to both republicans and liberals, not only because they would be
averse to the legalization of illegality, but also because the six-month limi-
tation period during which Roncarelli would had to have given his notice
had expired.
Republicans, we can now see, find themselves in a dilemma. On the
one hand, their position threatens to collapse into the parlimentarist one,
which is that as long as the content of the law is determined exclusively
by a democratically elected parliament, the law is legitimate. It follows
from that position that one should adopt a positivist theory of the rule of
law, one that includes only principles that serve a factual inquiry into the
content of legislative intention. On the other hand, their position threat-
ens to collapse into a liberal one, which requires that legitimate law be
law interpreted in accordance with the principles of a rich account of the
rule of law, with judges given ultimate authority on matters of principle.
But we should note that there is something odd about regarding this
dilemma as particularly problematic. It exists only if one supposes that it
is crucially important for republicans to show more than that they have
an important normative claim to make about non-domination; that is,
they must show that their claim can be made only by taking a position
distinct from both liberalism and parliamentarism. The claim to such dis-
tinctiveness, while hardly limited to republicans, involves those who make
it in an endless and generally fruitless exercise of classificationan exer-
cise I have seen aptly titled the narcissism of small differences. It thus
distracts them from the task of elaborating the content of their main
normative claim.21
This exercise is a pity, especially when the claim is an important one,
as I believe the claim about freedom as non-domination to be. It is impor-
tant in political theory, especially because, as suggested above, it alerts us
to the fact that the exercise of social or private power can be dominating,
and even more oppressively so than the exercise of state power. It is not
that liberals are blind to this fact. John Stuart Mills On Liberty and The
Subjection of Women are prominent examples of a liberalism that regards
20 Robert Leckey, Complexifying Roncarellis Rule of Law (2010) 55 McGill L.J. 721 at
726ff.
21 Even a quick glance at the essays collected in Besson & Mart (supra note 8), as well as
at their own essay, will reveal that a startlingly small proportion of the book is devoted
to unpacking what words like legal and law mean for republicans.
RANDS LEGAL REPUBLICANISM 503
social power as potentially more dangerous than public power because it
tends to work more insidiously.22 Once this fact comes into view, we are
also in a better position to inquire into how legal and social mechanisms
of domination reinforce one another. It might well be the case, for exam-
ple, that inquiry would reveal that the Duplessis governments legal bat-
tle against Jehovahs Witnesses was prompted by a sense of insecurity in
Quebecs political elites and among the francophone majority about the
ability of tradition to maintain unaided the kind of cultural identity that
they considered to underpin the core of their society.
I wish, however, to focus on legal republicanismthe legal manifesta-
tion of the principle of non-domination. And in this regard I want to claim
that the principle of non-domination is the best and perhaps the only way
to understand the virtue of the rule of law. Recall that from the dark pas-
sage I quoted in Rands essay above,23 I disinterred the following ideas: (1)
the legal subject is not abandoned to the play of public power in the ab-
sence of express constitutional limitations; (2) legislation itself provides
protection in a system of parliamentary sovereignty that is inherently
democratic, and judges have a special responsibility to maintain that de-
mocratic quality; and (3) the assertion of these democratic institutions is
particularly important in times of stress and it is in such times that the
fundamental postulates of legal orderthe deep postulates of f[r]ee
men24must be forced into view.
These ideas are all exemplified and given a more coherent basis in
Justice Rands judgment in Roncarelli. It was clear to the Court that Du-
plessiss government was in the grip of a kind of moral panic about Jeho-
vahs Witnesses, as it regarded their proselytizing as a threat to the very
foundations of Quebec society. It was thus, Justice Rand thought, impor-
tant to emphasize that Roncarelli had standing as a free and equal citi-
zen:
To deny or revoke a permit because a citizen exercises an unchal-
lengeable right totally irrelevant to the sale of liquor in a restaurant
is equally beyond the scope of the discretion conferred. There was
here not only revocation of the existing permit but a declaration of a
future, definitive disqualification of the appellant to obtain one: it
was to be forever. This purports to divest his citizenship status of
its incident of membership in the class of those of the public to whom
such a privilege could be extended. Under the statutory language
22 For a collection of both works, see John Stuart Mill, Three Essays (London, U.K.: Oxford
University Press, 1975).
23 See supra note 6 and accompanying text.
24 Ibid.
504 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
here, that is not competent to the Commission and a fortiori to the
government or the respondent.25
The Jehovahs Witnesses for whom Roncarelli had posted surety bail were
being arrested under a municipal by-law that had a democratic prove-
nance. But in his case, there was no legal issue of the sort that might
make us suppose that the judges had to choose, as it were, between indi-
vidual liberty and democracy. As we have seen, Justice Cartwright did
suggest that the liquor licensing law had to be interpreted as imposing no
controls on the exercise of discretion under it other than the controls it
explicitly announced. But, as I have indicated, such an interpretation is
compelled only if one is inclined to the parliamentarist view of democracy.
Justice Rand, in contrast, supposes that a democratic state is also a con-
stitutional state (one committed to government under law) which entails
that public officials be accountable to law. For a judge to hold otherwise
would be to allow
an administration according to law … to be superseded by action dic-
tated by and according to the arbitrary likes, dislikes and irrelevant
purposes of public officers acting beyond their duty, [which] would
signalize the beginning of disintegration of the rule of law as a fun-
damental postulate of our constitutional structure.26
Indeed, the only candidate in the case for creating a dilemma between
democracy and liberty was article 88 C.C.P., which, as Leckey points out,
presented a more formidable obstacle to Justice Rands reasoning than he
and commentators, including myself, have noted. Leckey argues that the
effect of article 88 C.C.P. was not like that of a privative clause to exclude
jurisdiction, but only to limit jurisdiction.27 Since Justice Fauteux found
both that Duplessis had acted illegally and that the illegality was encom-
passed by article 88 C.C.P., he could be said to have upheld the rule of law
in a double sense: he signalled that a high official was guilty of an illegal
act and that he as a judge was constrained by law to protect that illegal-
ity. In this second regard, Justice Fauteux displayed, more than Justice
Rand, a sensitivity to the demands of the rule of law, since he took seri-
ously the idea that judges, like all public office holders, are subject to
laws discipline. Leckey wonders whether the neglect of both article 88
C.C.P. and of decades of local judicial interpretation of it by commentators
might be the effect of an effort to conscript a judicial text into the service
of a transnationalCommonwealth or globalcommon law constitution
or rule of law project, one which he believes will tend to efface the juris-
diction-specific, the local law.28
25 Roncarelli, supra note 1 at 141 [reference omitted].
26 Ibid. at 142.
27 Leckey, supra note 20 at 724.
28 Ibid. at 739.
RANDS LEGAL REPUBLICANISM 505
While I must confess my utter complicity in the effort at such con-
scription, I want to point out the complexity of determining what the local
law was. First, we should notice that Justice Rand began his judgment at
the most local level possible, in the thick of a legal dispute between an in-
dividual described with due respect for his particularity, and an account of
a political campaign against him and other adherents of his religion by
those who wielded political power. In other words, in issue in that dispute
was whether those with political power also had the legal authority to in-
flict a harm on Roncarelli.
Second, one should conclude that the effect of Justice Rands judgment
was to reduce the autonomy of Quebecs public law within the Canadian
federation if and only if the best interpretation of Quebecs public law was
that it sought to protect official illegality in cases such as that of Ron-
carelli.
Third, Justice Rands judgment has a distinct rule of law edge on that
of Justice Fauteux: it forces the Quebec legislature to decide whether it
wishes to overrule the Supreme Court of Canada by declaring explicitly
within the text of article 88 C.C.P. that the article protects particular ille-
gal acts, including acts done in bad faith.29 That is, the legislature would
have to announce publicly that this is a society that is prepared to sacri-
fice the interest of its citizens in non-domination. If the epithet public in
the phrase Quebec public law is to be given any content, one could then
argue that Justice Rands judgment also has the merit of upholding Que-
becs public law system by refusing to give it an interpretation that makes
citizens more vulnerable to dominating interferences by powerful political
actors.
This point can be elaborated by considering Diceys discussion not of a
prospective procedural protection of officials against actions on the basis
of alleged illegalities, but of the retrospective procedural protection pro-
vided by an Act of Indemnity.30 This discussion occurs in the context of
Diceys argument that even in times of great stress and danger, martial
law in the proper sense of that term, in which it means the suspension of
ordinary law and the temporary government of a country or parts of it by
29 This would still leave open the question of whether such protection would be extended
when an official had actively and in bad faith sought to bring himself within the scope
of the protection. It is important in this regard to recall that Duplessis had actively put
obstacles in the way of the various courses of legal action attempted by Roncarelli so
that, when he and his lawyers decided as a last resort to bring a private law action, it
was too late for them to give the requisite notice.
30 A.V. Dicey, Introduction to the Study of the Law of the Constitution, 8th ed. (London,
U.K.: Macmillan, 1915) at 233. I use this edition because it is the last to contain Diceys
note on Martial Law in England during Time of War or Insurrection (ibid. at 538ff.).
For an illuminating comparison of Rand and Dicey, see Mark D. Walters, Legality as
Reason: Dicey, Rand, and the Rule of Law (2010) 55 McGill L.J. 563.
506 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
military tribunals, is unknown to the law of England.31 This, in his view,
was unmistakable proof of the permanent supremacy of the law under
our constitution.32
Dicey recognized that there might be legitimate recourse by officials to
illegality in such timesthat is, to actions that cannot be justified by the
defence of necessity. It is this category of morally justified but illegal acts
that an Act of Indemnity, properly so called, is meant to cover. But the
fact that such a statute, one that retrospectively grants criminal and civil
immunity to officials for their acts, amounts (in Diceys words) to the le-
galisation of illegality,33 and vindicates, in his view, the claim that the
English Constitution does not know martial law.
As I have argued in more detail elsewhere, in order to understand
Diceys analysis of Acts of Indemnity, we have to notice the distinctions
between three classes of acts that are candidates for retrospective au-
thorization.34 First, there are those acts that would be justified by the de-
fence of necessity if an official were called to account before a court of law.
Such acts are legally justified on a test amenable to judicial evaluation
and thus, strictly speaking, there is no need for retrospective legalization,
though an Act of Indemnity may serve the purpose of economy in pre-
empting actions against officials that would otherwise have to be decided
one by one. Second, there are those acts that, while not legally justified,
are justifiable in that they were done in good faith and reasonable in the
circumstances. In respect of this class of acts, an Act of Indemnity is re-
quired, both because the officials might otherwise be vulnerable to legal
sanctions and for the sake of legality, with the latter, as I will show below,
being the more important reason. Finally, there are those acts that are
never legally justifiable because they cannot comply with legality,
whether or not these requirements are formally entrenched.
Now as a matter of practice, it might be that the third class of acts is
more likely to attract criminal prosecutions than the second, as the former
is more likely to occasion individual and public outrage. That is, in prac-
tice, the second class might not require an Act of Indemnity, at least not
to the same extent as the third. But for Dicey, on my interpretation, even
if the officials will not in practice be called to account for illegalities in the
second class, the legal order needs to repair itself for the sake of legality,
as is revealed in the following passage:
31 Dicey, supra note 30 at 283-84.
32 Ibid.
33 Ibid. at 233.
34 David Dyzenhaus, The Organic Law of Ex Parte Milligan in Austin Sarat, ed., Sover-
eignty, Emergency, Legality (New York: Cambridge University Press, 2010) 16.
RANDS LEGAL REPUBLICANISM 507
An Act of Indemnity … though it is the legalisation of illegality, is
also … itself a law. It is something in its essential character, there-
fore, very different from the proclamation of martial law, the estab-
lishment of a state of siege, or any other proceeding by which the ex-
ecutive government at its own will suspends the law of the land. It is
no doubt an exercise of arbitrary sovereign power; but where the le-
gal sovereign is a Parliamentary assembly, even acts of state assume
the form of regular legislation, and this fact of itself maintains in no
small degree the real no less than the apparent supremacy of law.35
This is the more important reason for an Act of Indemnity. In its absence,
the legal order undermines its commitment to legalityto the principle
that all official acts must be able to display a legal warrant, especially acts
that infringe the most fundamental of individual interests. Most promi-
nent is the emergency context, where the interest in liberty is protected
by habeas corpus.
In the passage above, Dicey acknowledges that a retrospectively be-
stowed legal warrant is arbitrary, since it is itself not controlled by law;
yet, he wants to claim that it is qualitatively significant since it asserts
the supremacy of law. From Diceys analysis, we can extract the thought
that all that is wrong with these executive acts is that they are arbitrary:
they are both unjustifiable by the defence of necessity and not legally au-
thorized in advance. But they are, to use my own term, legalizable, that
is, they are acts fit to be governed by a prospective regime of legality.36
They thus contrast with acts in the third class that are unlegalizable in
that they are not only legally unjustified, but also legally unjustifiable.
Dicey was, of course, aware that an Act of Indemnity could go far be-
yond what he considered to be its proper, moderate scope of covering acts
that were done in good faith and reasonably.37 But it is clear that for him,
official acts that fall into the third class, the class of legally unjustifiable
or unlegalizable acts, are not the proper subjects of an Act of Indemnity.
Thus a statute that indemnified retrospectively all acts done by the
executive in the course of dealing with the emergency would not cover
acts done in bad faith or unreasonably. To cover such an act, the statute
would have to refer to all acts, including acts done in bad faith or unrea-
sonably and even such a wide-ranging authorization would not be taken
to include, for example, the authority to torture. It is not entailed in this
position that judges are able to resist the legislature if it is explicit in this
way. But even if they must defer, the reason of the common law still con-
trols the intention of Parliament by forcing Parliament to be altogether
35 Dicey, supra note 30 at 233.
36 For Diceys discussion of a statute that gave wide powers to the Irish executive in 1881,
see ibid. at 227-28.
37 Ibid. at 230-33.
508 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
explicit should it want to legalize the unlegalizable, which is an altogether
different affair from legalizing illegality.
By analogy, if Parliament wishes prospectively to legalize the un-
legalizable, it must do so altogether explicitly. And one should be careful
not to diminish the normative significance of what judges have achieved,
even when they have to defer to a clearly expressed intention to legalize
the unlegalizable. For the judicially enforced requirement of explicit ar-
ticulation of such an intention, what is called a clear statement rule in the
United States, serves both legality and democracy by requiring a public
statement by the legislature that it wishes to govern outside of the scope
of the rule of law and thus to govern in an arbitrary way. And such a law,
in Justice Rands words, signals the beginning of disintegration of the
rule of law as a fundamental postulate of our constitutional structure.38
Note in this regard that Justice Rand displayed no hostility at all to
either statute law or more generally to the administrative state. He
clearly accepted the legitimacy of statute-based regulatory regimes but
wished to emphasize that in an era when our lives have become increas-
ingly subject to public regulation, such regulation should not be arbi-
trary.39
A better way of putting this point would be to say that Justice Rand
might suppose that there is a false opposition in thinking of the problem
the Court faced as one in which there is a choice between freedom and
democracy. Rather, on his view, it is the case that the postulates of free
men are what undergirds both democracy and the rule of law, and that
38 Roncarelli, supra note 1 at 142. I should mention that I do not in this context accept the
distinction Robert Leckey makes between legalizing an illegality and precluding an
award of damages under the applicable general law of civil liability. Leckey points out
that there are numerous legal constraints on claims for civil damages that might be en-
gaged with a government official defendanta notice requirement, a limitation period,
even restrictions on the kind of harm that is compensablethat are not appropriately
regarded, when they operate, as legalizing an illegality (Leckey, supra note 20). Thus,
in Leckeys view, my clinging to the language of legality/illegality shows some resis-
tance to recognizing that Roncarelli is a torts case and not a public law proceeding for
quashing an administrative decision. However, Diceys discussion of Acts of Indemnity
is about statutes that give either criminal or civil immunity against actions for harm
suffered as a result of official wrongdoing. It is, of course, true that Roncarellis action in
tort was aimed in part at recovering damages, as the actions would be precluded by the
civil immunity that Dicey contemplates. But, as Lorne Sossin argues, Roncarellis ac-
tion should also be seen as an action to remedy a wrong to the society as a whole; one
could say it was about restoring the integrity of the rule of law: Lorne Sossin, The Un-
finished Project of Roncarelli v. Duplessis: Justiciability, Discretion, and the Limits of
the Rule of Law (2010) 55 McGill L.J. 661. In this regard, compare Diceys discussion of
Habeas Corpus Suspension Acts, which, when properly used, only temporarily suspend
recourse to the courts for the determination of the legality of a detention (Dicey, supra
note 30 at 224-28).
39 Roncarelli, supra note 1 at 141-42.
RANDS LEGAL REPUBLICANISM 509
those postulates are unrealizable outside of a system that is both democ-
ratic and legal.
The liberty we have in a modern society is liberty made possible by
law, but it will be liberty if and only if particular laws are by provenance
and form laws that have a claim to serve the postulates of free men. And
for such law to have a hold on plausibility, any legal subject whose inter-
ests are affected by the decision of a public official must have access to an
independent forum where she can ask not only by what legal warrant the
official has acted, but also for a determination of the content of the war-
rant in light of the postulates that undergird the system.
Conclusion
I have claimed above that in order for the liberal tradition to make an
important shift, it must be the case, first, that the tradition is composed of
strands that in fact provide the normative resources for it to make the
shift, and second, that there must be practical considerations that make it
normatively imperative that such a shift take place. There are indeed
practical considerations that make it urgent for us to focus on the idea of
freedom as non-domination. After all, we are living with the consequences
today of thirty or so years of dominance of the negative liberty strand of
liberalism in public lifea dominance that is characterized by suspicion of
regulation, a taboo on progressive taxation, rising inequality, and political
apathy except on the part of those who wish the state to retreat even fur-
ther from a role in public life. At the same time, there has been a trend,
exacerbated by the reaction to the events of September 11, 2001, for the
executive to arrogate ever more power to itself, as well as troubling signs
within Western democracies of a resurgent social conservatism that seeks
an other to demonize in order to shore up a sense of collective identity.
The idea of freedom as non-domination, with its closely associated
ideal of equality, its alertness to the potentially dominating interferences
of private and social power as well as of public power, its welcoming of
state interference as long as it is non-dominating, and its aversion to the
arbitrariness of public officials who are not properly accountable to law, is
thus of particular political salience at this time. But what makes it so sa-
lient is precisely the hegemony of the negative liberty strand within pub-
lic life, which, with the other trends sketched above, manifests a disre-
spect for moral principles associated with the liberal commitment to re-
spect for the individual.
The importance of non-domination, the ideal of equality, and so on, are
all firmly part of the liberal tradition of which the adherence to negative
liberty is but one strand. Moreover, it is likely that one can make sense of
the claim that these ideas are normatively important only within that
tradition, with its insistence on respect for persons. As Charles Larmore,
510 (2010) 55 MCGILL LAW JOURNAL ~ REVUE DE DROIT DE MCGILL
at the end of an illuminating essay on republican political thought, puts
it,
[t]he freedom we prize … is always a freedom shaped by other moral
principles, principles whose authority must therefore be understood
as binding on us independently of our own will, individual or collec-
tive. Political freedom, if it is to have a shape that we today would
welcome, must take its bearings from the obligation to respect one
another as persons.40
Rands thoughts within his 1960 essay, when understood in light of his
judgment in Roncarelli, thus set out an account of the rule of law that
gives clear expression to the idea of non-domination, founded and shaped
by an ideal of respect for persons. Roncarelli was a member of a disliked
minority, who was singled out for persecution when he had done nothing
more than exercise his rights as a free and equal subject of the law. Those
who singled him out for persecution sought to achieve their ends through
law. But since government under law is valuable because it helps to se-
cure non-dominationthe rule of law rather than the arbitrary rule of
mento use law to single out an individual for domination is, as Du-
plessis discovered, rather a complex business. No matter ones grip on
power, one might find that ones ends simply do not count as public ends
within a system of public law because such a system is predicated on re-
spect for the persons who are subject to its authority.
In sum, the reason we gathered to celebrate the fiftieth anniversary of
Roncarelli is that Justice Rand took the trouble to set out in his judgment
the main elements of a republican legal theory.
40 Charles Larmore, The Autonomy of Morality (Cambridge: Cambridge University Press,
2008) c. 7 (The Meanings of Political Freedom) at 195.
