As I have suggested before, I consid6er Chandran Kukathas (who is a good friend of this blog) as the standard-bearer of what we may call ‘right-liberalism’ today. (Perhaps he identifies as a ‘classical liberal,’ but because I am no friend of that term I avoid using it.) Recently, he sent me the transcript of his 2023 MAS Keynote Lecture, “On The Very Idea of An Open Society.” I have been unable to find the version of this paper online.
Now before I get to Chandran’s analysis, it is worth saying something by way of background. It seems Bergson coined the term, and then Popper made it famous in his (1945) Open Society and Its Enemies. In my view Popper means something different with ‘open society’ than Bergson did, but this matters very little for what follows because Bergson’s version had little direct impact on subsequent liberal thought. Unfortunately, it is by no means easy to extract a definition or definite description of an ‘open society’ from Popper’s work. (I don’t view that as a criticism.)
In the Open Society, Popper does write that “In what follows, the magical or tribal or collectivist society will also be called the closed society , and the society in which individuals are confronted with personal decisions, the open society.” (Chapter 10, p. 165 of the single volume edition; this is the ‘definition’ cited by Wikipedia.)
In fact, what it means to be ‘confronted with personal decisions’' is by no means obvious. Popper goes on to write (after a paragraph on the closed society) on the same page:
The aspects I have in mind are connected with the fact that, in an open society, many members strive to rise socially, and to take the places of other members. This may lead, for example, to such an important social phenomenon as class struggle. We cannot find anything like class struggle in an organism.
So, a reader may be excused if they infer that for Popper an open society is characterized by a class struggle in virtue of the fact (to use a Smithian expression) the desire to better one’s condition including one’s relative positions. This is not, I submit, what is emphasized by those who use ‘open society’ self-consciously in its Popperian sense.
In the Stanford Encyclopedia of Philosophy article on Karl Popper, Stephen Thornton avoids using Sir Karl’s own words to define an open society, and suggests (not implausibly) that “in an open society the rights of the individual to criticise administrative policies will be safeguarded and upheld, undesirable policies will be eliminated in a manner analogous to the elimination of falsified scientific theories, and political differences will be resolved by critical discussion and argument rather than by coercion.” He then writes, “The open society as thus conceived of by Popper may be defined as”
Levinson’s work is, by the way, a defense of Plato against Popper’s charges.
None of this would matter all that much if Hayek hadn’t also claimed that “the expression 'the Great Society', which we shall frequently use in the same sense in which we shall use Sir Karl Popper's term 'the Open Society,'” (Law, Legislation and Liberty, Volume 1: Rules and Order, p. 148, note 11,) and identified both these terms (great and open societies) with the “spontaneous order which Adam Smith called 'the Great Society', and Sir Karl Popper called 'the Open Society'.”" (Law, Legislation and Liberty, Volume 1: Rules and Order, p. 2.)
Back in 1976 already, Richard Vernon had lampooned this identification by Hayek in a classic article (see here). Unfortunately, Vernon is not interested in providing a better definition of an open society. Much to my horror, the late Jerry Gaus also did not define the open society in his (magisterial) The Open Society & Its Complexities (OUP, 2021). For Gaus, the open society is a “new form of civilization,” (p. 1) or a “complex social system” (p. 12); “a self-organized order” and so “constantly evolving” (p. 245). In an “Open Society, individuals come together at various levels to solve collective problems, but always in the context of a system in which one group’s democratic decisions require reflexive adjustments by others.” (P. 245) There is a Millian strain here, that becomes clear in other places, “Given that the Open Society seeks to free human intellect to search for new ways of living and thinking.” (p. 10) But Gaus himself treats Mill (not implausibly) as imposing a normalized or technocratic diversity, and rejects this for the open society.
The closest Gaus comes to a definition is his claim that an “Open Society involves autocatalytic diversity, producing ever greater connected diversity.” (p. 239) Or put differently:
In this way our moral nature, forming diverse moral networks, is the foundation of the Open Society. Rawls called a just society a “social union of social unions”—we might call the Open Society a network of moral networks. (p. 240)
To be sure, Gaus closes his book while subtly distancing himself from Hayek: “Hayek was deeply mistaken: without justification, the Open Society could not self-organize.” (p. 246) But a network of moral networks under-describes what an open society must be. (Angels could have a network of moral networks.) This description ignores the role of politics and institutions (something Gaus attends to in the book).
In an earlier (2013) paper, Gaus had noted that pluralism is intrinsic to an open society: “The very nature of the open society is that one must deal with strangers, who cherish very different ideals, and whose character traits are largely unknown.” (p. 4) And he had defended the idea that this also requires a ‘'rule-based morality for the open society.’' Yet what the open society is, is left unclear.
All of this by way of set up to praise Kukathas attempt to be clear what he is talking about. Kukathas’ views on the open society are not all that far removed from Gaus’. For Kukathas an open society is a ‘regime of regimes.’ Here’s how he puts it:
“Is an open society itself a kind of regime?” that the answer is yes. Is it a regime of regimes? Yes. And we can ask, is it an order of open regimes? Then the answer is no.
How am I going to defend this particular conclusion? I'm going to start by suggesting that an open regime is governed by norms of toleration. This is the key underlying conceptualization that I want to defend in thinking about an open society. The question then becomes, what is toleration? I'm going to give you a definition which is not particularly controversial.
To tolerate is to forbear or refrain from exercising power, to compel others to act or not act in ways that do or do not meet with our approval.
After making this more precise, Kukathas concludes with the following illustration of his conceptualization:
What I want to suggest is that an open society is what exists at the intersection…where immigration is free…[and] where powerlessness is noted. And the intersection…gives you the heart of the open society. That's the thesis. And the further away you go from that core, the further away you get from the open society as I've conceptualized it here.
So, Kukathas re-affirms that the open society “is a regime of toleration.” In particular, “it's a regime of toleration that must depart [sometimes] from toleration.” This focus on a certain kind of toleration does not come as a surprise to readers of Kukathas’ magnum opus (2013) The Liberal Archipelago: A Theory of Diversity and Freedom. In reflecting on his work I have come to feel that liberalism should not rest its self-understanding historically or conceptually on toleration (recall this essay; and this essay, both with extensive remarks on toleration). Let me explain.
In particular, toleration is the wrong sort of virtue or characteristic to be taken as central for liberalism. When taken as a kind of mental state or attitude, true toleration is difficult to achieve,[1] and, as Kukathas argues, requires forbearance.[2] It is, however, a feature and not a bug of liberal political thought, one of its true insights, that it is supposed to be much less demanding of ordinary citizens than, say, its Republican, deliberative, or Marxist counterparts which demand from citizens much more intense political participation and time-investment as well as more informed knowledge of political circumstances than liberalism prescribes.
By contrast, since Adam Smith, the liberal takes the advanced (cognitive) division of labor of mass society for granted; political life then requires division of powers, electoral polycentrism, effective parties, a robust civil society, press freedoms, legal accountability, and the involvement of other intermediary bodies as effective checks on power and as mediating institutions to translate the needs, interests, and desires of individuals and public associations into informed policy. (I return repeatedly to the significance of the advanced (cognitive) division of labor of mass society.) Living in a liberal polity should be relatively undemanding and without anxiety, so that we have cognitive, emotional, and financial resources to pursue spiritual, artistic, and other collaborative ends that provide meaning to our lives.
Furthermore, usually, and originally, toleration involves an attitude taken by a majority toward a minority. It is then very much treated as a privilege extended by the former, understanding itself as a physically and morally superior majority, toward the later. In fact, and now speaking of my native Holland, often the tolerated minority is never allowed to forget this. That is, the very idea of toleration also presupposes that such a privilege can be revoked at the majority’s discretion. This is a recipe for permanent mutual suspicion, even hostility (as it often is when the minority is growing). Latent power or demographic imbalances can, then, never be ignored because they might re-open the status of toleration, or even reverse its arrow. If toleration is subject to the prudence and judgment of the governors, their prudence and judgment becomes a source of political anxiety and promotes the politics of suspicion.
In fact, avoiding the use of power is no way to ease the concerns of the relative powerless. This is the core insight that liberalism takes from eighteenth century republicanism. So, what’s evidently needed is a different conception of the open society, one that centers on prevention of concentrated power alongside receptivity toward newcomers.
To be continued….
[1] Susan Stebbing (1939) [2022] Thinking to Some Purpose, London: Routledge, p. 256.
[2] Kukathas (2003) op. cit. p. 22.
I had a go at this in Crooked Timber a few years ago https://crookedtimber.org/2019/12/27/tolerance-acceptance-deference-dominance/ Here are my somewhat scattered thoughts.
Tolerance, acceptance, deference, dominance
I’ve been thinking about the various versions of and critiques of identity politics that are around at the moment. In its most general form, identity politics involves (i) a claim that a particular group is not being treated fairly and (ii) a claim that members of that group should place political priority on the demand for fairer treatment. But “fairer” can mean lots of different things. I’m trying to think about this using contrasts between the set of terms in the post title. A lot of this is unoriginal, but I’m hoping I can say something new.
Starting from the left (in more senses than one), tolerance involves the removal of legal barriers to being recognised as a participating member of the community, with legal freedom from persecution, voting rights, property rights and so on. Women, gays, religious minorities and people of colour have all had to struggle to obtain this recognition. But, as has been pointed out many times, mere legal tolerance is demeaning and discriminatory. Identity politics involves a demand not merely for tolerance but for acceptance.
Jumping to the right, the idea of tolerance implies the existence of a dominant group that does the tolerating, either as a result of moral suasion or as a response to political pressure. Moving from tolerance to acceptance implies an erosion of that dominance. It becomes unacceptable for members of the formerly dominant group to express or act on the view that the other group is inferior: such views, once expressed openly without fear of adverse consequences, are now criticised as racist, misogynistic, homophobic.
The most difficult term in the series is deference. In sociology/anthropology, it’s typically used in counterpoint with “dominance”, as the attitude displayed by one submitting to dominance. But in the context of identity politics, I think there’s something more subtle going on.
Members of the formerly dominant group may be willing to extend acceptance to others, but they still expect a kind of deference in return. Most obviously, they expect to be treated as the default identity for the community as a whole, as “typical”, “real”, “true”, Americans, Australians, Finns or whatever.
When that expectation of deference is not fulfilled, the choices are to accept the new situation, or to support what might be called default identity politics. More or less inevitably, that implies an alliance with those who want to reassert or restore the group’s dominant position: racists, theocrats, and so on, depending on which aspect of the dominant identity is being challenged.
That makes default identity politics a “double or nothing” bet. If it’s political successful, it’s dragged further and further towards entrenched minority rule by members of the dominant racial or religous group, and typically towards some form of personal dictatorship. If it’s unsuccessful, the divisions it creates risks a reversal of the previous order. Instead of being accepted as one element of a diverse community, the formerly dominant group becomes the object of hostility and derision. The signs of that are certainly evident, particularly in relation to the culture wars around religion.