Hayek’s 1960 book The Constitution of Liberty was criticized when it came out for being unsystematic in its normative commitments. Its structure is more of a series of considerations on a theme – certainly less tidy than a deduction from Rothbardian non-aggression or Randian egoism. And on the question of coercion, presumably central to any “restatement of liberal political economy”, he sounds downright evasive.
Nevertheless, the various chapters do circle around and suggest a moral first principle, even though – as far as I recall – it’s never stated explicitly. Specifically, it suggests the principle that the point of morality is the preservation and advance of the extended order – both in a functionalist and, in some respects, in a practical normative sense.
Hayek approached the question of norms and morals from the standpoint of the emergent macro-properties of a society where people follow particular behavioral rules – most explicitly in The Fatal Conceit, but the approach is in the background of The Constitution of Liberty too.
Since then, his approach has caught on in a big way in cognitive anthropology, though apparently without any direct influence. The big question in cognitive anthropology is, what kinds of behavioral rules are necessary to make cooperation work, and how far outside of our narrow interests do they have to be? These rules, to the extent they’re not biologically obligatory, constitute morality. Indeed, the human capacity to internalize moral rules, and thus to be able to commit to cooperative strategies, is a key distinctive from other forms of animal cooperation.
In addition to the basic question of cooperation or not, there’s also the question of cooperation at what scale – and as an economist, this is what Hayek was primarily interested in. Morality in general will (usually) get you some form of cooperation. This is an important and non-trivial hurdle from a zoological perspective. But there’s also a huge difference between hunter-gatherer-scale cooperation and Western-civilization-scale cooperation that depends on the specific content of the moral rules. Identifying that specific content – what norms enable large-scale anonymous cooperation? – was a big part of Hayek’s overarching project, and the foundation for his commitment to liberalism as a political philosophy. The success of Western civilization, after all, was precipitated by the spread of the bourgeois virtues, though the specific connection took some centuries to uncover.1
In a functionalist sense, the conscious experience of ought in humans is directed toward cooperation as its end, and specifically cooperation on a supra-animal scale. This being the case, it’s a very small leap across the is-ought gap2 to the conclusion that the scale of human social cooperation – the extended order – has overriding moral weight.
The extended order is a rather odd moral first principle compared to others like non-aggression or egoism. Most importantly, it’s far too telescopic to be of much use in deciding the questions that moral rules are usually called upon to decide. How should I treat my wife? How much should I participate in my neighborhood association? How far out of my way should I go to include the quiet guy at work in my lunchtime conversations? It’s not clear what effect any of these things would have on the extended order. And in any case, we got the extended order without orienting our moral rules toward it. If we need other principles in order to get to concrete action anyway, why not just save a step and use those other principles as first principles instead?
In “The Theory of Complex Phenomena“, Hayek distinguishes between pattern predictions and point predictions – basically the level of detail with which we can describe a system and predict its outcomes. The extended order is a huge macroscopic system whose contours and preconditions we can only meaningfully talk about in terms of broad patterns. Our moral life, however, is mostly lived on the point level of particular decisions. The patterns of the extended order consist of nothing other than the interactions among particular decisions (methodological individualism!) – however, there remains a practical epistemic gulf between them: it’s not possible analytically to reduce one to the other. As he argues in The Sensory Order, a system (the brain) cannot, in principle, comprehend the particulars of a system more complex than it (the extended order) in anything except more or less broad patterns. A system is necessarily less complex than a system it’s a part of, so the human mind cannot control, direct, or predict the wider outcomes of particular decisions except in broad outlines. This is the core of Hayek’s most famous contribution, the knowledge problem.
The usual takeaway of this point is the impossibility of central planning and socialist calculation – indeed that was Hayek’s first foray into the problem. But it also establishes the gulf between – let us call them – practical ethics and metaethics. The extended order as an ultimate moral end doesn’t tell you how to act.
That being said, Hayek’s rough theory of group selection points in the direction of practical ethics being functionally oriented toward larger-scale social cooperation. “Men did not foresee the benefits of rules before adopting them”, but those that favored larger-scale social cooperation were able to spread. Because ethical systems mainly compete in terms of fitness and spread rather than on a rational basis, systems of practical ethics usually purport to have something else other than the extended order as their first principle, but nevertheless contain a great deal of tacit knowledge: rules that foster extended social cooperation without anyone having intended for them to do so. It’s not obvious how marital norms bear on the extended order, but – given the selective process that he posits – odds are that they do in some way or another.
So even given that morality is functionally oriented toward large-scale cooperation, it might be the case that that fact has no practical moral implications. So, having at least established that Hayek took this problem seriously, we ask again: given that the epistemic gulf exists between practical ethics and metaethics, why bother with the metaethical question at all?
Hayek wrote a great deal against scientism and the hubris of applying the methods of the natural sciences to the social sciences. So it surprises people sometimes to see Hayek’s more sanguine reflections on the potential of the social sciences. He was, after all, a social scientist, and never backed away from calling economics a true science.
In particular, and contrary to James Buchanan’s remark on The Constitution of Liberty, Hayek did not advocate the uncritical acceptance of evolved institutions and moral rules, including ethical systems. Indeed, the earlier quote continues: “Men did not foresee the benefits of rules before adopting them, though some people gradually have become aware of what they owe to the whole system” – namely, through (social) science.
The Achilles’ heel of any system of practical ethics is the legitimacy of its first principle. And for any first principle, there are conditions that render it irrelevant in the face of other considerations. This is why moral philosophers get such a kick out of edge cases and reductiones ad absurdum. Is it permissible to kill an innocent person if that person is patient zero for a globe-threatening plague? Not on an ethical system based on non-aggression or egoism – and yet, I suspect most people would regretfully pull the trigger in that situation.
Any system of practical ethics will have exceptional edge cases like this, that fall outside its organizing principle. This fact might seem to delegitimize ethics as such. If there is and can be no general system of practical ethics, private interest is the only remaining imperative – social cooperation is founded on quicksand and the abyss of the bellum omnium contra omnes yawns before us.
But this is exactly where the extended order steps in as a moral first principle. It is, as we’ve seen, incapable of serving as a guide to practical moral action. But what it can do is organize the exceptions to the systems of practical ethics. Practical ethical systems cannot organize their own exceptions, but that doesn’t mean those exceptions are unorganizeable.
This is the role of social science. It cannot make point predictions about the consequences of specific ethical imperatives, but it can comprehend increasingly more detailed patterns. In other words, the advance of social science lets us see with increasing (if never total) clarity just how norms of one sort or another bear on the possibility of an extended order. The epistemic gulf between practical ethics and metaethics may be unbridgeable in its entirety, but it can be seen across at certain points, at least in broad outlines. The better telescopes we build, the more refined our analytical tools, the more adeptly we can navigate exceptional situations; the more confidently we can take the leap into the dark where our tacit knowledge fails us. In such a situation, the extended order is a metaethical lodestar – an alternative to private interest in the wake of the breakdown of a system of practical ethics, and a foundation for the eventual relegitimation of another (or the resumption of the same) system.
Take liberalism as a system of practical ethics, one that Hayek himself had no small fondness for. Given that there are conceivable situations where liberal institutions are incompatible with their own preservation, the advance of social science lets us more prudently ascertain when such a situation in fact exists, and the least disruptive way forward.
Of course, such a decision must always be made with fear and trembling. The advance of social science makes it more likely that we can successfully navigate the exceptions to practical ethics. But the nature of tacit knowledge is such that we don’t know how much we don’t know. If navigating the exception has a certain probability of success, it’s impossible to know whether a doubling of the analytical power of the social sciences (so to speak) moves us from 30% to 60%, or from 1% to 2%.
Still, if the exceptional circumstances are sufficiently dire, it’s worthwhile to have the tools at hand to navigate it as deftly as possible.
Leave a Reply