This paper was first delivered as a public lecture on May 7, 2026, at Cardus Exchange in Ottawa. It draws on an essay first published in The Hub in 2021. The text has been lightly adapted for publication.

I will begin with a fundamental fact about law and politics, which is that a Constitution is a practice, not a thing.

We have been accustomed in Canada (and the US, and frankly most of the world outside of the UK) to think of a Constitution as a codified document, the provisions of which are legally entrenched beyond the reach of ordinary legal change. But even for countries that have such codified Constitutions, including Canada, what is codified is only the most obvious part of the Constitution. The larger part of the Constitution goes unnoticed, because it doesn’t stand out from ordinary life and politics, and that is because it is made up of the ordinary life and politics of the state.

The contrary case—that only a codified constitutional document is a real Constitution—was taken up early and ferociously by Thomas Paine. After Edmund Burke had defended the constitutional rule of parliamentary government against the French Declaration of the Rights of Man and of the Citizen, Paine, with typical myopic vehemence, claimed that the only Constitutions worthy of the name were those that entrenched the rights of man and the powers of government in a written charter. “A constitution is not a thing in name only,” he wrote, “but in fact. It has not an ideal, but a real existence; and wherever it cannot be produced in a visible form, there is none.” He ended his case with a challenge: “Can then Mr. Burke produce the English Constitution?”

Paine’s zeal for prescriptive constitutionalism was so blinding that he could not see the English Constitution even after he had all but described it. After correctly insisting that a Constitution “is not a thing in name only” and has “not an ideal, but a real existence,” instead of looking for its “real existence” in the real world he demanded to be shown “an idea” scrawled on a piece of paper—the Constitution “in name only” that he had just decried.

One wants to take Paine by the hand and guide him through the streets of his own eighteenth-century London, pointing out the Constitution “in fact.” Bidding him observe the easy, regulated commerce of its meat and fish markets; the orderly commotion of the great maritime basin of the Thames dockyards; the wet ink presses of the booksellers in Paternoster Row, where his own pamphlet was printed; the clamour of the Commons and the languor of the Lords; the official and dissenting religions preached from a hundred church pulpits, and the fervour of the new Wesleyan preachers outside. To show him all this, and then admonish, in an echo of the words inscribed on the floor of the city’s great Cathedral: “If you seek the English constitution, look around you.”

Two centuries after Burke and Paine, the legal philosopher Joseph Raz insisted that constitutional and legal systems are socially embedded. He referred to this as “the assumption of the primacy of the social.” By this he meant that a legal system is always part of a larger social system that includes many subsystems, of which the legal system is just one, and all of which interact with each other according to their own rules. The most important of these rules, I would suggest, are social rules.

Social rules—the habits, beliefs, expectations, and folkways of a people—operate above, below, and around a legal system, and in the absence of formally entrenched legal rules they may act largely independent of them. To illustrate this primacy of the social, I will offer three examples of how social rules can work both independently of, and actively on, formal legal rules.

The first is a hypothetical. Imagine, if you will, a religiously homogeneous society in which there is a strong stigma against divorce, which is strictly enforced by shunning. In such a society, it would not be necessary to pass any divorce law. The social rule would perform the function that a divorce law performs in a society like ours, where more diverse beliefs and practices mean that a formal rule is necessary to coordinate social behaviour in cases of divorce.

The second example shows the way in which social expectations can, in practice, entrench some ordinary laws more deeply than other laws. The Canada Health Act is considered a third rail of Canadians politics. It is entrenched not just in law—after all, it is just an ordinary Act of Parliament, and we repeal or amend Acts of Parliament all the time—there is also a social expectation, a public demand, that governments not repeal or amend it. It is this social expectation rather than its formal legal status that shields the Canada Health Act from change.

The third example is the normal amending provision of the Canadian Constitution. In most cases, amending the Constitution requires the agreement of seven provinces whose populations add up to at least 50 percent of the national population. This is the so-called 7/50 rule. But in 1996, Parliament decided it was necessary to pass a statute requiring an even higher bar. That statute, which is still in effect, requires that before Parliament can agree to a constitutional amendment under the 7/50 rule, the proposed amendment must have the support of a group of provinces that specifically includes British Columbia, Ontario, and Quebec and at least two of the three Prairie provinces having between them at least 50 percent of the total population of the Prairie provinces, and at least two of the four Atlantic provinces having between them at least 50 percent of the total population of the Atlantic provinces. For those of you who didn’t do the math in your heads, what that means is that, in practice, Alberta needs to support it, as does at least one of New Brunswick and Nova Scotia. This means that what was intended to be a high but not impossible threshold for amending the Constitution has been superseded by a law requiring, in effect, near provincial unanimity.

Why did Parliament add this additional hurdle? Presumably because the legislators thought it was required by the social conditions of Confederation. This means that the constitutional amending formula that we have in practice is a social requirement, entrenched in an ordinary statute, and which effectively supersedes the amending formula agreed upon in 1982 and set out in the text of the Constitution.

This interplay between legal rules and non-legal forms of entrenchment is not unique to Canada. In every state, the practical effect of the constitutional text will be determined by and ultimately depend upon the social rules—or to put it more broadly, the culture—in which it is embedded.

It follows that we cannot fully understand a state’s Constitution until we understand its culture. Alexis de Tocqueville had something like this in mind when he wrote that “if we would become acquainted with the legislation and the manners of a nation . . . we must begin by the study of its social condition.” Implicit in this advice is the assumption that societies differ as much in their cultures as in their Constitutions, and that most of the differences in the latter are due to differences in the former.

Charles de Gaulle, who knew something about Constitution-making, believed that “a Constitution is a spirit, a set of institutions, and day-to-day practice.” If it is curious that he does not mention the text, it may be because, as he said elsewhere, “[the French] know what Constitutions are worth; we have had seventeen in 150 years, and the nature of things is stronger than texts drawn up by politicians.”

In de Gaulle’s distinction between the codified text of a Constitution and the deeper and more powerful “nature of things,” we can hear the appeal of Montesquieu, for whom a Constitution was a social organism in which the laws reflected the spirit of a particular society by responding to the unique combination of social and political forces at work in it. For Montesquieu, these forces included everything from religious morality and custom to the quality of a state’s soil and its climate. (I think he was on to something there about soil and climate, and I will come back to this point later in my discussion of federalism.)

These differing and contingent conditions that obtain in different states are why, as the eminent English constitutional scholar Martin Loughlin has admonished, “pace the liberal conviction, there can be no such thing as an ideal best state: a constitution can only be assessed within a particular socio-political context.”

If we want to assess the success of a Constitution or of constitutional government, we must do so always bearing in mind the internal and external conditions that a particular government must accommodate. Some of these conditions will be beyond the control of government, but many are within its control. Good governments are those that, given existing conditions, including those it cannot control, are able to foster a civil state that conduces to peace and allows the people to flourish, individually and together. That state of affairs, when achieved, is a healthy constitutional order.

Even when it is achieved, a healthy constitutional order can never be final or static. Change, sometimes small, sometimes dramatic, is inevitable—“the great law of Change,” as Burke called it—and for a state to maintain order over time, it must be able to change itself in ways that minimize discontinuity and social disruption and maintain social harmony. Ideally, this means that change, where it must happen, is organic.

Organic change is change that is intelligible to the people of a state. Seen in the light of all relevant circumstances and conditions, such change accords with the internal logic of the social Constitution. Organic change is experienced by the people as following if not an inexorable logic, then an orderly and evolving process that is consistent with the spirit of the Constitution. As Michael Oakeshott put it, legal or constitutional change should follow a “principle of continuity,” by which “authority is diffused between past, present, and future; between the old, the new, and what is to come.”

For change to be felt as diachronically continuous in this way, it must be change by and among a specific people with a broadly common understanding of having a shared past, and living in a shared present, with a shared commitment to continuing to live together in the future.

The more the processes of change within a state can rely on informal social and political rules and norms, rather than on fixed and entrenched legal rules or codified constitutional provisions, the more organic and continuous it is likely to be.

In this regard, the constructive unamendability of Canada’s Constitution, which its framers did not intend (but should have anticipated), is a standing impediment to organic and continuous change. That is a problem Canada will have to deal with eventually, but it is not a problem I propose to solve today.

Another way to facilitate organic and continuous change in a state is to arrange it so that, to the greatest extent possible, its members share common or similar pre-political interests, beliefs, and prejudices. The more homogenous and stable this social substructure is, the less the state will need to rely on rigid and formal legal and constitutional mechanisms that inhibit organic change.

Encouraging this condition of social stability has obvious and important implications for a state’s education policy, its cultural policies, and its immigration policy. Most importantly for my lecture today, it also explains the importance of federalism for realizing the common good in a large, regionally diverse, and culturally diffuse state.

This brings me to part 2.

The common good is a universal goal that cannot be realized universally. It can only ever be the best possible local instantiation of universal principles of the good among a specific people in a specific time and place. In other words, like a healthy constitutional order, it is not an end that can be achieved but a project that must be carried out in common and for as long as possible.

Again, like a healthy constitutional order, the common good must be promoted by the social habits of a people as well as by laws and political institutions that must be adjusted through contingent circumstances marked by the thousand natural shocks that flesh—and states—are heir to.

This will be easier if the people of the polity in question are linked together not just legally but socially. Aristotle called this kind of social connection, which he believed was a precondition for social order, a “civil friendship” (politike philia). “Civil friendship” exists on a spectrum of social ties somewhere between the bonds of a family and the bonds of a commercial enterprise, but closer to the family than to the corporation. A man might willingly die for his family or his country, but only a fool would die for a corporation.

The knowledge necessary to hold together a social order is immanent in the understanding—overt and implicit, individual and collective—of the people living within that order. As it has been put, “Man lives in the polis, but the polis also lives in man.” This intimate and vital relationship between and among the people of a polity is necessary to realize the common good in practice as well as to support the social Constitution through the processes of organic change required to keep it healthy and ordered.

I said that the civil friendship is closer to a family bond than to a corporate bond. And, like a family, a society has natural limits that allow it to remain ordered. These limits are set by the practical limits of civil friendship. We can and should love beyond our kin and beyond the kin of our kin, but at some point the universe of potential love exceeds the practical reach of politics. The size of a self-governing polity properly conceived should be the largest social group to which we can effectively extend the condition of civil friendship.

This is first a practical limit. That is, we can engage in a common civil project only with those with whom we can commune socially and politically. But this practical limit has moral implications. Except under unusual circumstances, it is morally insulting to believe that we can impose our political will on those with whom we do not share a social bond. To attempt to govern beyond the civil friendship of the polis is an insult to the stranger.

This is why the European Union’s commitment to “ever closer union” is so dangerous. Its goal is to break the bonds of civil friendship within nations by stretching them so far that they snap. It also presumes a power to govern strangers who do not have a say in, and do not see themselves reflected back in the vision of, a transnational government or a transnational court.

So, is the Canadian federation like the EU’s “ever closer union”? Is it also a dangerous and presumptive exercise of anti-civil authority? I would say no—so long as we respect the differences of the natural and evolved polities within its boundaries. Where federalism works, it does not simply divide a state into smaller polities for ease of governance. It recognizes that in a large and diffuse state, even people who share a common citizenship and a common legal Constitution may be more or less strangers to each other.

This is true, of course, even within a province or a large city. We no longer live in the sort of city state that Aristotle considered a natural polis, where approximately 20,000 Athenian citizens had the right to vote and participate in government. Personally, I subscribe to the idea that a community is what you can walk across in a day, which in our constitutional order roughly corresponds to a municipality.

But given all that we expect of government today, that is far too small a unit for effective self-government. So, we need something between the city and the state, and that, for better and worse, is the province. Provinces make possible in a large and sprawling country like ours what Roger Scruton has called “the politics of the first person plural.” The “we” in “we the people.” They allow for the creation of stable units within the larger and inevitably less stable unit of the country. Orders within order, producing greater order.

Canadian federalism is a policy much older than official bilingualism, official multiculturalism, or the Charter—the three other ways we’ve tried to accommodate social difference within our country. It was born out of the need to accommodate regional differences that were acknowledged as real—and probably intractable—long before Confederation.

From the beginning, federalism has been a way to reassure divers parts of the country that they will not be dominated by a single cultural or economic vision driven by the most populous and electorally powerful provinces. Remember that Prince Edward Island, despite being the site of the 1864 Charlottetown Convention, refused to join Confederation just three years later in 1867. The reason, as its premier Edward Palmer said at the time, was, “We would submit our rights and our prosperity, in a measure, into the hands of the general government and our voice in the united Parliament would be very insignificant.” Whether it was the little islanders of PEI or the linguistically, religiously, and legally distinct French Canadians, Canadian federalism has, from the beginning, been a response to the fear of domination. Federalism is not a solution to our regional differences and divisions—it was never meant to be—it is a recognition of them.

You can quibble with where the borders were originally drawn, but over more than a century and a half those borders have created natural regional affinity groups based on shared geography, shared industry, shared language and dialect, and even shared soil and climate. Just as Montesquieu would have predicted—particularly if you include within “soil” the natural resources grown in and found below it. The constitutional function of the province today is thus not solely legal and political; it is also social. It helps cohere distinct regional identities that are partly real and partly mythopoetic, identities that are imposed or self-imposed until they become social facts.

You can say that there is more that we have in common as Canadians than there is that divides us, and you would be right. But that does not deny that the divisions and differences that do exist are real. A functioning federalism would not just acknowledge these differences, it would respect them. Unfortunately, the opposite has happened. Pace Mark Carney, for too long we have been too European in this respect. Over time, Ottawa has arrogated to itself more and more power over provincial matters. From health care policy to the federal carbon tax to recent childcare, dental, and pharmaceutical mandates, Ottawa has taken on, and taken over, the proper role of provincial governments. This is not healthy, and it is not how federalism is supposed to work.

One of the most eloquent recent statements against this centralizing tendency came from the Alberta Court of Appeals in its decision striking down the federal carbon tax. Writing for the majority, Chief Justice Cathy Fraser admonished that

Time has not eroded the provinces’ rights to have the powers assigned to them under our Constitution sedulously respected. While some may view the division of powers as anachronistic or a barrier to uniform action . . . the division of powers remains key to our federal state. It is part of the fabric of Canada itself. . . . The federal government is not the parent; and the provincial governments are not its children.

An earlier defence of federalism came from, of all people, Pierre Trudeau. In a 1957 essay, Trudeau surprised his readers by siding with Maurice Duplessis against Prime Minister Louis St. Laurent in the latter’s decision to provide funds to provincial universities. In words that would make a libertarian economist’s heart flutter, Trudeau wrote that “if a government has at its disposal such a surplus of funds that it can undertake to support a part of the common good that does not lie within its jurisdiction, one may suspect that this government controls more than its share of taxation.”

Trudeau returned to the theme in a later essay titled “The Practice and Theory of Federalism.” “Since regionalisms do exist in Canada,” he posited, “such feelings should be exploited to further the cause of democracy: each community might enter into a state of healthy competition with the others in order to have better ‘self-government.’” More recently, Carl Vallée described Stephen Harper’s successful accommodation of Quebec’s desire for provincial autonomy as “dictated by conservative logic: we believed it was the best way to make such a large and diverse country with competing regional interests function.”

Note, this constitutional defence of federalism is fundamentally different from the related idea that federal systems create “laboratories of democracy” in which provinces can test different economic and social policies, with the others eventually adopting those that prove most successful. The goal of provincial difference should not be national convergence. True federalism assumes that different provinces may—and, in fact, must—choose different ways of achieving the common good that are suited to their culture and their circumstances and not to others’. The virtue of federalism is not that it allows for policy experimentation, although it does, but that it does not impose a single prescription for good government on a diverse country.

Federalism is our best hope for reconciling our national differences because it is the best way of respecting those differences. Of respecting the mutuality of civil affection within each provincial polity and of allowing those polities to respect each other in turn. By contrast, a vision of Canada imposed on Quebec by Ontario and the West will only exacerbate national tensions. Just as a vision of Canada imposed on the economically productive Western provinces by the economically moribund, have-not provinces of Ontario and Quebec cannot be felt as anything but an alien and hostile provocation.

True federalism is not incompatible with a strong federal government, but it reminds us that a strong federal government need not be a large or intrusive federal government. Ottawa can be strong in the areas needed to defend our country’s national interests, such as defence and foreign policy. And although I would transfer the criminal law power to the provinces, Ottawa would still be free to use its legitimate constitutional powers to address truly national matters and emergencies. Otherwise, it should leave the provinces to govern themselves.

If this all sounds like a rather stiff and formal arrangement, and possibly a sterile one, I would like to make the case that it is in fact the makings of a love story. Some of you may know that I occasionally dabble in poetry, so forgive me if I move into a poetic register to end my talk.

The title of Hugh MacLennan’s classic Canadian novel Two Solitudes is taken from the Austrian poet Rainer Maria Rilke. In Letters to a Young Poet, Rilke wrote that “Love consists in this, that two solitudes protect and touch and greet each other.” That is one translation of Rilke’s notoriously elusive style, but there are others. Mary Herter Norton’s 1934 translation rendered the line as “The love that consists in this: that two solitudes protect and border and salute each other.” And in 1945, the year MacLennan published his novel, Reginald Snell translated it as “The love which consists in the mutual guarding, bordering and saluting of two solitudes.”

There are subtle differences here: “Protecting” is not quite the same as “mutual guarding”; “touching” sounds rather different than “bordering”; and “greeting” sounds more friendly and mutual than “saluting.” But all the variations have this in common: Love both depends upon, and must accommodate, natural differences.

This makes sense. You don’t love your husband or wife because they are like you, or because they are different from you, but because they are themselves. Love is the recognition of another person as someone who deserves to be loved as themselves, just as you would want to be loved as yourself.

This idea is obviously related to Kant’s insistence that we treat others as ends in themselves and not as means to our own ends. Love is not instrumental, it is selfless—it does not demand that your beloved change to fit your idea of them, even if you think that you have their best interests at heart and are sure you know what would be best for them.

The same idea is expressed in the idea of love, or true friendship, that Aquinas describes in his discussion of caritas in book II of the Summa Theologica. The moral and legal philosopher John Finnis summarizes it this way:

In friendship one is not thinking and choosing “from one’s own point of view,” nor from one’s friend’s point of view. Rather, one is acting from a third point of view, the unique perspective from which one’s own good and one’s friend’s good are equally “in view” and “in play.”

Moving back from friendship to federalism, love—caritas—means not imposing one vision of Canada on all regions. It means not asking why other parts of the country can’t be more like yours, like Henry Higgins wondering, stupidly, “Why can’t a woman be more like a man?” Nor does it mean wishing that other parts of the country would become more like your idea of what would be best for them.

Federalism embraces, or at least facilitates, the idea of respectful difference, through something like the principle of subsidiarity as it was originally described in Pius XI’s 1931 encyclical Quadragesimo Anno. Like federalism, subsidiarity is not primarily an efficient solution but a principle of natural justice. Finnis again:

The principle [of subsidiarity] is one of justice. It affirms that the proper function of association is to help the participants in the association to help themselves or, more precisely, to constitute themselves through the individual initiatives of choosing commitments . . . and of realizing these commitments through personal inventiveness and effort in projects.

Self-government is a never-ending process of decision, debate, reflection, revision, rejection, and new decisions. It is the process that constitutes the act of making and being a society. To pre-empt this process, to take away a province’s process of self-government, is a denial of its dignity and the right of its people to self-determination within their subsidiary sphere. Such presumption is the opposite of love.

With a much more complex Canada today than the “two solitudes” we struggled to reconcile for the first century after Confederation, I submit that a federalism that embraces the original vision of subsidiarity (rather than the purely practical understanding that is often applied to it today) is our best hope for realizing something close to the idea of love, caritas, within a sprawling country of regions, and the best way for the people of each province to achieve the civil friendship, the politike philia, that is a precondition for the realization of the common good.

This view has sometimes been disparaged as a vision of a “loose confederation.” Pierre Trudeau often used that phrase pejoratively. To which I would respond, What is wrong with a loose confederation? Where critics see looseness as national instability or fragility, I see suppleness and flexibility—a play in the constitutional joints that allows for ease of movement, adaptation, and adjustment. It is a fixed and brittle constitutional rigidity that we should fear. And goodness knows we have more than enough of that in Canada already. The last thing we need is to compound the problem by imposing uniform social strictures in the free spaces still left between the textual walls of our sclerotic legal Constitution.

It has always been the case in our country that to be well ordered federally, Canada must be well ordered provincially. That is true now more than ever. If we can’t maximize the benefits of our social Constitution at the federal level—and given our current federal constitutional stalemates, it appears we can’t—then we should be especially enthusiastic to promote the conditions of social solidarity and civil friendship where the scale of government is most amenable and the natural commonality of interests is strongest: in our provinces. And at all costs, we must resist the cultural and social straitjacket that tightens every time provincial policy is dictated nationally by those whose view of Canada extends no further than the horizon visible from windows of the Rideau Club.