Chesterton's fence and Rousseau's
Liberalism, conservatism, and rationalism on institutions and tradition
I have either noticed or remembered (at the limit these two feel phenomenologically identical) that the very interesting Further or Alternatively wrote a post before Christmas that I was namedropped as having prompted. It is a little rude for me not to have even acknowledged that—apologies!—but consider this post an acknowledgment, a ‘thank you’, a recommendation, and a response.
One of the theses of FoA’s post is this: a certain way of thinking about the value of institutions—exemplified by the ‘Chesterton’s fence’ principle—is essentially conservative rather than liberal. For the uninitiated, here’s the fence passage:
In the matter of reforming things, as distinct from deforming them, there is one plain and simple principle; a principle which will probably be called a paradox. There exists in such a case a certain institution or law; let us say, for the sake of simplicity, a fence or gate erected across a road. The more modern type of reformer goes gaily up to it and says, ‘I don’t see the use of this; let us clear it away.’ To which the more intelligent type of reformer will do well to answer: ‘If you don’t see the use of it, I certainly won’t let you clear it away. Go away and think. Then, when you can come back and tell me that you do see the use of it, I may allow you to destroy it.’ [G. K. Chesterton, The Thing, chapter 4.]
I will have more to say about what this principle means below, but for now let’s stick with FoA. He makes no attempt to define ‘conservatism’, self-consciously and (I think) correctly. But he does give us an account of what liberalism is:
Liberalism is famously a broad church. But if there is a core sense to liberalism—a sense in which people on the Left can ultimately be pinned down as being liberals rather than, say, socialists, or people on the Right can be pinned down as being liberals rather than, say, conservatives—it must surely be something to do with individual liberty. (Or freedom, or political liberty, or autonomy, or...) At root, it’s all about people (in some sense of the word ‘people’) getting to exercise personal choice (some more or less informed and more or less restrained kind of choice) about the things that matter to them and the polities to which they belong.
I think FoA intended this as a basically uncontroversial truism about liberalism, essentially a purely etymological or definitional point; so I’m sure he’ll be happy that I’m going to controversialise it.
As lots of people have noted—I like Helena Rosenblatt and Marilynne Robinson—the term ‘liberal’ (whence ‘liberalism’) is not connected to ‘liberty’ in the nice straightforward way that FoA and the Stanford Encyclopaedia of Philosophy and everyone else assumes. Originally, ‘liberal’ was the adjective form of ‘liberality’, Latin ‘liberalitas’, which in Roman thought was the name for the virtues that were particularly fitting for a freeman as opposed to a slave. It had connotations of generosity, giving freely (Robinson’s essay on liberalism is titled ‘Open Thy Hands Wide’) with a kind of noble condescension, back when ‘condescension’ was a good thing. Over time, social attitudes shifted, and ‘liberality’ was reclaimed from below in a couple of different ways, with changing ideas of who and what a ‘free man’ was. At some point in the eighteenth or nineteenth centuries,1 ‘liberal’ came to be solidly associated with a political programme rooted in these developing ideas. And eventually someone stuck an -ism on the end, and now ‘liberal’ is just the adjective form of ‘liberalism’ and nobody says ‘liberality’ any more unless they’re putting on an old play.
This is all to say that, while there is a genuine connection between the idea of freedom and the history of liberalism, it’s a weird one. Etymologically, liberalism is less the political theory of freedom, and more the political theory that has certain virtues such that the analogous virtues in humans are well-suited to free men. And of course, words are not bound by their etymology, and in the two centuries or so since we all forgot about ‘liberality’ all sorts of new ideas have taken their place in the liberal canon. To be sure, if you don’t have any place for freedom in your political philosophy, you can’t be a liberal; but this is a much wider ‘core’ of liberalism than the one FoA proposes.
I bring this up not as a ‘gotcha’, but because it changes our perspective on how liberals can think about institutions. FoA writes: ‘if asked “what is this (good) institution good for?” or even “what is it for?”, the liberal will have to give the answer “freedom”’. Why? Because liberalism is ‘all about’ freedom. So if the liberal looks at an institution like (FoA’s example) the National Trust, they will tend to ask questions like ‘is this helping cultivate a strong civil society that can counterbalance state power and offer people choice?’ or ‘does this help self-interested parties combine for the purposes of rent-seeking and exclusion?’ (plausibly both should be answered with ‘yes’). By contrast, the conservative can value institutions on their own terms: they can answer ‘what is the National Trust for?’ with ‘preserving nice old buildings’ and just stop there; that can be valuable in itself, not just as a means to the end of liberty.2
Another way to put it is that FoA thinks liberals have to be reductionist: part of liberalism is that you can only care about institutions insofar as you can care about them from the point of view of freedom. This is at the heart of FoA’s argument for why liberals cannot accept Chesterton’s fence (click through to make sure I’m not misrepresenting it): that principle is all about thinking about a thing in terms of its ‘real purpose [FoA’s paraphrase]’, but liberals have in effect walled themselves off from this way of thinking about the world. Liberals, so the accusation goes, cannot simply value institutions on their own terms; we have to stand back and see them ‘from the outside’, as it were. You can clearly see how this reasoning could blend into accusations that liberalism is atomising and destructive of community, although FoA doesn’t make this explicit.
But if we free ourselves from the notion that liberalism is ‘all about’ freedom in a relatively straightforward and uncomplicated way, we can open ourselves up to versions of liberal political theory that are not reductionist. To be sure, some versions of liberalism really are reductionist, I don’t want to deny that. But I do want to at least say that (a) it’s not necessary for liberals to be reductionist and (b) some of the most important historical liberals were not. If (b) is true then (a) follows, so I’ll focus on (b).
In what is almost certainly the most famous chapter of The Spirit of the Laws, Montesquieu feigns to walk through (step-by-step from first principles) what a constitution designed to protect liberty would look like. And what do you know, it looks exactly like the English constitution! Generations of students have thought of him as an Anglophile as a result. But in fact, Montesquieu thought English liberty was fragile and indeed destined to fail; he much preferred the French constitution (at least as it was before Louis XIV).
Why? Well, thinking in terms of liberty and from first principles, Montesquieu thought that each individual institution of the French monarchy could be improved by moving closer to the English model. But it is a fallacy of composition to imagine that this means that France as a whole would become freer by reforming its institutions in an English direction. At most, Montesquieu was committed to viewing the entire system of society through the lens of freedom; but this is compatible with, and indeed (he argued) necessitates, evaluating each individual institution of society from a more local, less reductionist perspective. Montesquieu would have been annoyed, would have objected, to the nobles turning away from their smug vainglory and organising themselves into a sort of Maison des Seigneurs in Paris; and while he might have been motivated by liberty as an ultimate value, he nonetheless had a test for the success of the nobility (as an institution) that was common with other nobles who didn’t feel the same way about freedom.
Now, I don’t actually think we should call Montesquieu a liberal, although that’s contestable.3 But what’s uncontestable is that his thought deeply influenced subsequent liberals—Smith, Constant, and Tocqueville to name just three with obvious debts to him—and that this influence included his constitutional thought and his particular way of thinking about the contributions of political actors and institutions to overall freedom. Many subsequent liberals were, like Montesquieu, committed at most to evaluating society as a whole in terms of freedom, not necessarily to so evaluating each institution of society.
Now, I think even that’s a little too strong, and liberals can go even weaker. For an example, take Isaiah Berlin. For all that he is read as saying the opposite (and for all that it’s his own fault, for being insufficiently clear in crucial parts of his argument), Berlin believed that other values could come into conflict with liberty at any level of thinking, and that sometimes those other values should come out victorious. These other values could well include those formed ‘internal’ to existing institutions. None of this stopped Berlin from saying that (negative) liberty had a special place in a society like ours, which is what makes him a liberal; but this special place was not ‘at the top at all times, with all other values necessarily subservient’.
So liberalism does not have to take a reductionist view of institutions. But nonetheless, I do think FoA is right in saying that liberals cannot accept the reasoning behind Chesterton’s fence, which serves the purpose of a conservative break on reform. FoA helpfully sums up Chesterton’s fence as the idea that ‘until you understand the real purpose of a feature of life, you have no business getting rid of it.’ I think it’s this idea of a ‘real purpose’ that is problematic for liberals; but I think they’re right in thinking it’s problematic, and non-liberals should join them in rejecting it.
The above-quoted passage from The Thing continues with Chesterton arguing that his principle
rests on the most elementary common sense. The gate or fence did not grow there. It was not set up by somnambulists who built it in their sleep. It is highly improbable that it was put there by escaped lunatics who were for some reason loose in the street. Some person had some reason for thinking it would be a good thing for somebody. And until we know what the reason was, we really cannot judge whether the reason was reasonable. It is extremely probable that we have overlooked some whole aspect of the question, if something set up by human beings like ourselves seems to be entirely meaningless and mysterious. There are reformers who get over this difficulty by assuming that all their fathers were fools; but if that be so, we can only say that folly appears to be a hereditary disease. [Emphasis mine.]
What FoA calls the ‘real purpose’ of an institution, Chesterton makes clear, is just the intention behind its construction, the ‘reason’ why it was thought to be a good idea. While a fence is a one-person construction, institutions usually take a group, but the idea generalises in a straightforward way: the real purpose is the set of intentions of those who have shaped the institution. This is why FoA places such emphasis on the ‘institution’s own goals’. The purpose of the National Trust is to preserve nice old buildings; we know this because that was the intention of the founders of the institution and is the reason its members support it.
Now, this way of thinking is often reasonable. In some circumstances, there really will be a particular intention (whether individual or collective) that has directly shaped an institution, that explains why that institution takes the form it does; and in these cases, it is crucial to understand that intention before embarking on reform. My particular brand of liberalism places a lot of emphasis on ideas from the public choice tradition, and to some degree this is because it helps you figure out who built various fences, tells you that their motivation was self-interested rent-seeking, and shows you how building the fences fucked over everyone else. But as a general principle, Chesterton’s fence is wrong.
The assumption of both Chesterton and FoA is that the ‘true purpose’ of an institution can be identified entirely from within the point of view of the institution itself. As FoA argues, conservatism can exist inside, while liberalism must look in from the outside. Now, this assumption seems to require that the inner life of the institution is essentially uncomplicated, that there is a single coherent set of motivations that guide those within it. (In chapter 4 of The Thing Chesterton was concerned with the institution of family, and his dismissal of Ibsen’s A Doll’s House as ‘feeble’ suggests quite strongly to me that he was engaged in a project of flattening the complexities of family life.) But I will set this point aside for now.
What the assumption certainly requires, however, is that the institution is uncomplicated in a weaker sense: that it has been formed according to the motivations of those who shaped and continue to shape it in a fairly straightforward way. This doesn’t necessarily mean that those motivations are coherent, but it does mean that the motivations have to in some sense be about the institution. The institution must be the way it is because of how people were trying to shape it.
But sometimes in social life, to turn an old Buddhist example on its head, one of us tries to build a wall, another a spear, another a fan, another a bit of thick tubing, and we end up building an elephant. Very often, an institution is not a particular way because anyone thought it would be a good idea; people thought various other things would be good ideas, and the institution emerged chaotically out of the interactions of those intentions. A classic example of this in liberalism is the story of English liberty told by Smith: contra Montesquieu and his own friend Hume, Smith said that English freedom emerged not from constitutional design, but from the legacy of Roman decline on economic development and the desire of nobles for trinkets and luxury goods.
Smith of course thought English liberty was a good thing. But if we decide a particular institution is bad thing, this reasoning shows that we need not understand its ‘true purpose’ before reforming it if, in fact, it has no true purpose. Indeed—and I don’t know if Chesterton was thinking of this when he wrote about his fence, but I’d not be surprised—one of the most important and compelling passages in the history of political thought is about a literal fence, and the importance of tearing it down:
The first man who, having enclosed a piece of ground, to whom it occurred to say this is mine and found people sufficiently simple to believe him, was the true founder of civil society. How many crimes, wars, murders, how many miseries and horrors Mankind would have been spared by him who, pulling up the stakes or filling in the ditch, had cried out to his kind: Beware of listening to this impostor; you are lost if you forget that the fruits are everyone’s and the Earth no one’s… [Jean-Jacques Rousseau, second Discourse, trans. Victor Gourevitch, II.1, italics in original, bold mine.]
The entirety of the second Discourse is, essentially, one long explanation of the motivations that led to this particular fence, and it is very clear that these motivations were not about private property, neither for or against. According to Rousseau, private property emerged just because ‘things had by then reached a point where they could not continue as they were’. Of course, this is not the claim that there was no individual intention behind the putting up of the fence; the ‘founder of civil society’ probably wanted to stop cattle from grazing on the land or whatever. But Rousseau is clear that this intention is not the ‘true purpose’ of the fence, that indeed there is no ‘true purpose’, no society-wide explanation in terms of function or telos.4 To explain the fence, he asks us to ‘take up the thread earlier, and try to fit this slow succession of events and of knowledge together from a single point of view, and in their most natural order.’ Private property has a historical cause, but not a ‘true purpose’.5
Rousseau was (importantly) not a liberal in any useful or coherent sense, not even in the Montesquieu way where it’s debatable. He advocated for small, homogeneous republics founded on a paradoxical kind of coercion. Another non-liberal with liberal debtors was Nietzsche, who gave even shorter shrift to the idea of thinking about institutions and ideas in terms of their ‘true’ function or purpose. Here he is writing about our institutions of punishment (which of course many liberals wish to reform):
… things are not as our naive genealogists of morality and law have thus far assumed, all of whom thought of the procedure as invented for the purpose of punishing, as one once thought of the hand as invented for the purpose of grasping… the concept ‘punishment’ in fact no longer represents a single meaning at all but rather an entire synthesis of ‘meanings’: the previous history of punishment in general, the history of its exploitation for the most diverse purposes, finally crystallizes into a kind of unity that is difficult to dissolve, difficult to analyze and—one must emphasize—is completely and utterly undefinable. (Today it is impossible to say for sure why we actually punish…) [On the Genealogy of Morality, trans. Maudemarie Clark and Alan J. Swensen, II.13, italics in original, bold mine.]
Of course, there is a long tradition of thinking about society in terms of purposes, uses, and functions. Anthropologists and sociologists call it ‘functionalism’, but historically you’d have used Aristotle’s terminology and talked about final causes: the idea, essentially, of what an institution is ‘good for’. Many liberals, including the greatest (Smith), talked and still talk in these terms. But it is crucial that this does not get us to Chesterton’s picture: that specifically requires that the purpose of an institution is an intention or set of intentions.
As such, sure, there are liberal reasons to reject Chesterton’s fence. But there are also a great many non-liberal reasons. Overall, there are just good reasons to reject Chesterton’s ostensibly ‘common sense’ principle. But FoA is, I think, right to hint (though he does not outright say) that conservatives tend to want to accept it. This is not to say that conservatives must accept Chesterton’s fence and think about institutions only from the ‘inside’; like the idea that liberals must only think about them from the ‘outside’, I think this is too essentialist by half, belied by the history. But there is certainly a pressure in conservative thought to simplify the complexities of existing institutions, to think of them in relatively-simple terms so as to have a relatively-simple defence of their value, so as not to have to grapple with calls for reform.
To sum up. The liberal does not have to judge institutions solely instrumentally, in terms of their contribution to freedom—what I above called ‘reductionism’. But neither should they adopt Chesterton’s fence. And the space between these two views is, I would say, attractive independent of liberalism; there are strong arguments pushing us into that space, and it is home to some of the great non-liberal minds in the history of ideas. So the fact that liberalism is able to occupy it, whereas conservatism struggles, is a point for the former.
Ok, then, so much for liberalism. What of rationalism? FoA writes not just about conservatism and liberalism, but also of LessWrong-style rationalism; his argument is that rationalists can join conservatives ‘inside’ institutions, and that rationalist ideas can therefore be helpfully used by political conservatives (by which FoA does not necessarily mean Conservatives).
Now, I think the rationalists tend to use the term ‘Chesterton’s fence’ more broadly than FoA does, indeed more broadly than Chesterton himself. But regardless, we can ask: do rationalists genuinely sit alongside conservatives on this topic?
Well, as FoA himself observed, even where rationalists do not go all the way to utilitarianism, they tend to tip-toe right up to the line. The movement’s founding documents tell us that ‘[w]hoever knowingly chooses to save one life, when they could have saved two… they have damned themselves as thoroughly as any murderer.’ And that money is ‘the measure of how much society cares about something.‘ And that aiming for satisfaction by doing good in any but the most effective manner is just the selfish pursuit of ‘warm fuzzies’. And it might seem that utilitarianism, including the 98% utilitarianism of the rationalists, would be reductionist in the same way FoA accuses liberalism of being, thinking of value only in terms of utility.
But FoA points to the increasing awareness among rationalists that utilitarian outcomes might require non-utilitarian reasoning and motivations, an idea that (in some form or other) has been in the air since at very least Sidgwick. Some utilitarians insist stridently that it cannot be right, but I think they are increasingly in the minority. The rest have thought through its consequences to various levels of subtlety. Most historically interesting (though relatively uninfluential compared to his other work) was Derek Parfit’s treatment in part one of Reasons and Persons, which I wrote about at length in my master’s thesis and which I may eventually talk about on this blog. But a basic and—I think—basically representative response comes from Peter Singer and his co-author Katarzyna de Lazari-Radek, who argue that utilitarians permissibly can (and indeed should) adopt an ‘esoteric’ approach to their own morality: saying in public what they do not believe in private, to encourage non-utilitarian motivations in others while making their own decisions for purely utilitarian reasons. Bernard Williams famously compared this view to the haughty and secretive attitude of colonial administrators. The utilitarians are perhaps willing to grant that other people should care about institutions from the inside; but they themselves must stand outside, caring only about utility.
After all, while ‘liberalism’ may not derive in any direct manner from ‘liberty’, ‘utilitarianism’ comes straight from ‘utility’. And while the liberal tradition has spread outwards and diversified since the early nineteenth century, the utilitarian tradition that started life as a squarely public movement for reform has narrowed in to professional philosophy, where it has become more reductionist and more totalising.6 Indeed, rationalism (along with its sister movement, effective altruism) is essentially the utilitarian tradition’s first sustained period outside the academy since Sidgwick.
Now, given their move into the public sphere, the rationalists have room to throw off some of the accumulated baggage of utilitarian philosophy. I myself think there is rich potential in attempts to make utilitarian and consequentialist ideas less reductionist. But (for reasons both social and intellectual) I don’t have great hope that the rationalists will go all that far towards developing that potential.
As Popper observed in the context of utopianism, there is something almost inherently self-stabilising about reductionist views of value: they can easily begin subsuming other values, twisting them to fit into the reductionist scheme, discarding nuance and subtlety and richness. I think this why FoA, and so many others, understand liberalism as essentially reductionist. Reductionist liberalism is certainly not the only version of liberalism, but it has a great gravitational pull on liberals, who have a tendency to desperately try to argue that other political goods can be reduced to freedom or at least will never conflict with it.7
Some argue that contact with the realities of power and politics can push one out of reductionism, by making one grapple with the complexities of the social world. But it can just as easily harden reductionism into ideological belief. This is a problem for rationalists, liberals, and (indeed) conservatives. Liberals, though, are lucky in at least one regard: they have a rich tradition of non-reductionist intellectual ancestors they can draw on. By contrast, the rationalists—having started from reductionist utilitarianism—will have to do their own thinking for themselves if they want to escape it. The rest of us will just have to wait and see how that goes.
Eric Schliesser says the first use of the word in this sense was in Smith’s Wealth of Nations, but Rosenblatt’s pick—Benjamin Constant—is more popular.
You might try to escape this argument by saying that ‘preserving nice old buildings’ is not a political point, that liberals are entirely free to have place non-political value on preserving nice old buildings, and that it is only in the realm of politics where they have to look at everything through the lens of freedom. FoA rightly rubbishes this attempt to make the political / non-political distinction sharper than it is. If you have the time, an interesting take on the role of this distinction in the history of liberalism that I really enjoyed is my friend Keir Bradwell’s master’s thesis.
For my part, I think it’s clear that the Marxists have the measure of him: he was clearly an aristocratic class reactionary. How else are we supposed to interpret the eminent judge Baron de Montesquieu telling us ‘the most important thing for France right now is to strengthen the power of judges and barons’? Where the Marxists go wrong is that this pretty much doesn’t matter; ideas have a life of their own, and Montesquieu’s conceptual frameworks were so rich and his arguments so interesting that they could fuel centuries of thinking across the political spectrum.
In slightly more technical language, Rousseau is not a (certain kind of) methodological individualist, but he is what Bernard Williams would call a formal individualist. This seems to me to be the objectively correct stance to take towards these matters.
Going back to public choice, insofar as my version of public choice is Georgist (insofar, I might suggest, as all public choice should be Georgist), then it will take many leaves out of Rousseau’s book, and will not necessarily take the form I attributed to it above. Rent-seekers may not actually be seeking rent, in terms of their defined and steady intention, so long as there is some kind of historical or causal story as to why their behaviour is enabled by the presence of rents and as to why their behaviour leads those rents to accrue to them.
I am not trying to pick on Toby Ord by citing him twice for this—he’s just a very clear representative of the trend.
Sometimes, as with Rawls, going so far as to just stipulate this—Rawls ensures that liberty will take complete, ‘lexical’ priority over other values by simply declaring it to be so.