The priority of social order

Bodo ethics

To inflict a bit of realism on our story, let us go back to the 11th century to see the social evolution of ethics and of intellectual debate on ethics (Hardin, 1999). Axel Leijonhufvud characterizes the village society of 11th century France in which the villager Bodo lived. We have detailed knowledge of Bodo’s society from the parish records of the church of St. Germaine. Today one would say that that church is in the center of Paris, but in Bodo’s time it was a rural parish distant enough from the center of Paris that many of its inhabitants may never have seen Paris. Virtually everything Bodo consumed was produced by about 80 people, all of whom he knew well. Indeed, most of what Bodo consumed was likely produced by his own family. Perhaps no one other than these 80 people touched anything he consumed.

In life as organized in Bodo’s village, everyone could assess everyone else and could judge their contributions to their own families’ lives, as well as to the lives of all other villagers. The villagers had the epistemology to do for their fellow villagers what modern states can still do only relatively poorly today. Everyone in that village was intimately known to all others. The situation of Bodo’s village, although not of France in general, was very nearly that of Gerrard Winstanley’s vision for the organization of economic life, each family producing primarily for itself.

Ethics in Bodo’s society must have been a compound of religious values and social reciprocity. We must rely on each other to some extent and our failures of reliability will be known to everyone. Any of us who are utterly unreliable are likely to become pariahs to everyone else. Much of the simplistic morality of common sense fits such a society fairly well. That morality is a set of minatory constraints: do not fail to keep your promises, do not lie, do not cheat or steal, and so forth. Bodo ethics is essentially a set of rules, norms, or expectations for regulating daily life in small communities with generally close-knit relations. Bodo ethics may still be much of the quotidian ethics of people today, but it cannot ground, say, a principle of quasi egalitarian distributive justice in a large and complex society. It cannot even be very good for regulating broader relations with those – virtually everyone – outside a very close-knit community. More complex and encompassing orders build on simple systems such as Bodo’s. Primitive ethics can help us achieve enough order for us to build far beyond such systems.

A striking feature of Bodo ethics is that it is relatively easily enforceable by the community. An individual need not rely on self-regulation to be moral. The knowledge that the whole community has of each individual’s adherence to the local moral code allows community members to sanction miscreants. An enormous part of the debate about morality in the modern secular world is about how individuals can be motivated to act morally. That question is answered easily for Bodo’s world. The community spontaneously enforces its morality as a set of compulsory norms. This does not guarantee compliance, but it exacts a toll for non-compliance. The exaction would typically be quick and aimed at the right person.

Human societies have moved from regulation of behavior almost exclusively at the communal face-to-face level to substantial regulation by broader institutions. Communal regulation is still common within small groups such as close-knit families that have a high degree of transparency. This fact means that our moral lives are varied and not rigorously defined by any standard moral theory or principles. Our actual moral lives are more interesting and sensible than any moral theory. Those lives are compartmentalized, perhaps so thoroughly that many of us must occasionally seem to be different persons in different contexts. We face incentives from the outside that may parallel or cross our own incentives from within. There may be very few generalizable moral rules that have had sway over the centuries.

Societal-level explanations of social order

In Bodo’s world we do not need morality to keep us all in line because the transparency of all our actions is virtually total. Regulation of behavior through rich face-to-face sanctioning is likely to work in such a world. It is only with increasing scale that we need more than village knowledge, because complete common knowledge breaks down in a larger society. Our first response in a world of compelling religious belief is historically to invoke an all-knowing god to supersede all-knowing neighbors. Eventually this, too, hails.

European political philosophers have proposed at least three ways to motivate obedience to the state. The oldest is a theological vision that we must obey god’s commands (see Filmer, 1680/1689).Perhaps the most important command in this view is implied by the divine right of kings. In actual practice this is one of the most heinous and stupid of all views of how citizens must see their duties. Its only merit – arguably a great merit – is that it must stabilize societal expectations. The power of this theory is that it is both normative and also rational or self-interested for those who have the relevant beliefs and accept their god’s command. Combining the two grounds for behavior makes this an especially powerful theory IF we can believe it. The actual behavior of European monarchs, including their occasional accessions to power by assassinating their predecessors, is very hard to reconcile with the theory. That is what a god wanted? The list of dishonorable actions by monarchs must have been deeply distressing to supporters of the divine right of kings.

With the rejection of the theological view, three courses were left open in the era of the decline of belief in the divine right of kings. One is rational; one is normative; and a third claims pure reason as its source. The rational theory is best exemplified by Thomas Hobbes who holds that political obedience can be motivated by the powerful threat of coercion. This requires the construction of institutions to deliver the relevant coercion. Hobbes does not give us compelling institutions. To this day we have collectively failed to construct the relevant institutions, although Hume makes some progress with this with this project. Gaus and many others in our time grapple with this apparently very difficult task.

We are finally left with the task of designing institutions that will motivate cooperation and otherwise orderly interactions. Rather than a holistic theory, we can try to develop piecemeal bits of theory in the hope of eventually putting together enough pieces to show how to go further in the endeavor. The principal difficulty in this enterprise will be our limitations in thinking through and holding in mind the whole of the complicated story of social order. I propose a good place to begin is with Hume’s conventions, which, so long as they are adequately understood, are rationally binding on us, but they may often become moralized. For example, driving according to the local convention is rational but we might come to think that it is morally binding. The moralizing of the convention is not wrong in this case because it would be immoral knowingly to violate the standard convention; hence, following the convention is moral. Self-interest and morality coincide in this happy instance.

After the failure of theology, the normative theory that is supposed to elicit the requisite cooperative behavior of citizens is politically appealing but unsuccessful contractarianism as founded on consent so that our obedience to a regime to which we have consented makes it morally required for us to obey it. Despite more than three centuries of work on it, contractarianism remains to this day a failed enterprise. It can generally give us justifications for varied actions, but it cannot give us the motivations to act according to the theory. The theological theory is deeply rational because it holds out the threat of punishment in an afterlife to motivate people in this life. Contractarianism does not offer any such built-in punishment for misbehavior. It is plausible that contractarianism has seemed compelling to many political philosophers just because it gives a moral ground for obedience to the state: one has agreed to it and is now therefore morally compelled to keep one’s agreement. That this would be a worthlessly weak argument in real life relations should eviscerate it for political theory.

Promise-keeping can be fully, or at least substantially, self-regulated by the interests of the promisors to establish and maintain reputations for reliability that will ease their way to further promises to each other or to others. For this perspicuous reason, we have a strong interest in keeping our promises. A social contract in which we each promise to all others to cooperate in maintaining order for organizing collective actions for the benefit of all is the large number strategic analog of two-person cooperation. Unfortunately, contractarian agreement on a government cannot be regulated by mere self-interest in the way promise-keeping or simple two-person exchange can be. In the traditional logic of collective action, if each member of some large group or even of a whole society bears a share of the costs of providing a collective benefit to the group, they all can benefit by more than their share of the costs. Some members, however, might freeride on the efforts of the others and collective provision might therefore commonly fail (Olson, 1965). A mere norm is unlikely to override self-interest in many such contexts. Some members might be sufficiently motivated by moral commitments, but we cannot generally expect everyone to be, especially when the stakes are high.

Our limited capacities

Given the motivational failure of strictly normative theories – theological and consent – we want a self-interest account of social cooperation. To spell out your interests, mine, and others’, and how all these fit together is beyond what we can do unless there are very systematic patterns that generalize across many individuals. The apparent variations among individuals recommend against assuming this. Let us assume that there are systematic features, such as self-interest and lots of biological needs. But suppose your tastes and mine are different. We might gain a step on our task if we assume, as Hume does, that there is a virtually necessary connection between self-interested human nature and the form of such institutions as those for securing justice in Hume’s sense of justice as order (again, not distributive justice). Human nature might restrict the range of institutional structures that set general laws in a particular era. But then more particular laws might be required for a given jurisdiction (Hume, 1739–1740/2000: T3.2.3.11fn). ¹ If we follow our natural proclivities, we will take few actions for the advantage of others from disinterested views (Hume, 1739–1740/2000: T3.2.5). In the 18th century and earlier, many philosophers believed that their god instilled moral principles and motivations in us. No major philosopher today considers this a credible position, although one might concede that this was a clever move in its time.

In one of his most famous dicta, Hume (1739–1740/2000: T3.2.5.12-13) says that interest is the first motive for keeping promises. There is a vast literature that moralizes promise-keeping, some of it falsely attributing these moralistic views to Hume. Against such views, it is typically in our interest to keep our promises because doing so enables us to enter other promises and contributes to our beneficial reputation. Generally, there need be no further consideration to motivate us to keep our promises. Some people do feel morally bound to keep their promises. Some of those who think it merely a matter of their own interest might sometimes assess their interests wrongly and might therefore fail to keep promises some of the time. As a rule, if the matter at stake in our promise is very large, we might rightly quail at staking very much on another’s promise-keeping. In those cases we might therefore recur to the legal regulation of our interaction. In sum, we can expect that most people, at least some of the time, will follow the moral rule on keeping our promises; most of us would insist, however, on contractual backing for major issues. Between these two categories we might have mixed – moral and self-interested – motivations.

We are inherently hard-wired to be moral at least in the limited sense that we are wired to attend to our own interests. (If we were not, we would be overrun by people who look more assiduously after their own interests.) However, this is a confused motivational structure that might run self-interest and beneficence together much of the time, especially when the stakes are not very high. Certain of our institutions will have to work in the face of extensive, contrary self-interest.

In a line that might have come later from Adam Smith, Hume (1739–1740/2000: T3.2.6.6.) says, “The general system comprehending the interest of each individual is also advantageous to the public though it be not intended for that purpose by the inventors.” This can rightly be seen as Hume’s central formula in the use of self-interest to accomplish general good. Let us spell out his general view by focusing on his account of justice. Justice, he says, handles defects of man that can be remedied by society (Hume, 1739–1740/2000: T3.2.2.3 and T3.2.4.1). It is “only from the selfishness and confin’d generosity of men, along with the scanty provision nature has made for his wants, that justice derives its origin.” Therefore justice cannot be either a prior or a universal principle. It is clearly not abstractly or otherwise deducible but is entirely contingent, not a priori rational (Hume, 1739–1740/2000: T3.2.2.18; see also T3.2.6.6). Public “utility is the sole origin of justice, and … reflections on the beneficial consequences of this virtue are the sole foundation of its merit” and, as he says in a later work (Hume, 1998: 3.1.1.1), “self-interest is the original motivation to the establishment of justice” (Hume, 1739–1740/2000: T3.2.2.24). In “order to establish laws for the regulation of property, we must be acquainted with the nature and situation of man” (Hume, 1998: 3.2.27). One could go on listing such insights, which, taken together must stymie any a priori or universalistic ethics. Incidentally, all these invocations of “justice” concern justice as order, as in a criminal code or a legal regime of property ownership and not distributive justice.

Hume makes this final point: that justice as order makes no sense in those states in which conditions are misfit in many ways. For example, he says that the fictional state of nature or a golden age can be useful in showing that justice is contingent because it would be irrelevant in some kinds of society or among some kinds of people, whose character must differ from ours. His great achievement is to make collective outcomes very different from the human material going into that result. This is his answer to the 18th-century puzzle of how to explain seemingly moral behavior.

Hume largely shares Thomas Hobbes’s basic assumption of what we now call homo economicus, of individuals as largely self-interested, although Hume’s position here is considerably milder than Hobbes’s. One must qualify the homo economicus claim with “largely,” because both Hobbes and Hume allow concern for others. For example, Hume enunciates a principle for altruism or beneficence. He says, “A man naturally loves his children better than his nephews, his nephews better than his cousins, his cousins better than strangers, where everything else is equal.” (Hume, 1739–1740/2000: T3.2.1.18). In response to claims that altruism must be genetically programmed into us through natural selection, the great population geneticist, JBS Haldane, perhaps under the influence of the pint he was drinking, is supposed to have quipped that he would lay down his life for two of his brothers or eight of his cousins (Richards, 1987: 541). Hume is the original Haldane of beneficence. He elaborates the point to show that it is an important part of human nature and morality, attributing this partiality to “the original frame of our mind” in which “our strongest attention is confin’d to ourselves; our next is extended to relations and acquaintance; and ‘tis only the weakest which reaches to strangers and indifferent persons” (Hume, 1739–1740/2000: T3.2.2.8).

Hume has two extraordinary insights that have been slighted, even ignored, in the vast body of commentary on his work. One of these is the psychological phenomenon of mirroring (Hardin, 2012) and the second is the sociological account of convention in the spontaneous, iterated coordination of large numbers of people in mutually beneficial ways. I will address this latter issue, that of convention, because it is central to Hume’s political program (Hume, 1739–1740/2000: T3.2.2.9-12).

The capacity for thinking strategically is part of human nature. As is true of other parts as well, this one is differentially rich in people’s capacities. Hume has astonishing capacity for strategic thinking, so much so that he can do strategic analyses without the modern vocabulary of game theory that makes the task radically easier, so that ordinarily bright undergraduates can do such analyses today. Others for whom strategic thinking comes naturally include Thomas Schelling and many economists. Many people, including many philosophers, are put off by it. Hume did his conventions before modern traffic patterns presented us with what is now seen as the definitive example of a convention: the rule of the road. Yet he grasped the phenomenon clearly enough to find it in many forms of social interaction. Hume identifies at least 22 conventions in a remarkable variety of contexts (see the list in Hardin, 2007: 85). He obviously did not need the matrixes of game theory to recognize iterated coordination interactions and to grasp their strategic significance.

Another sense of normative theory is as in a theory of norms. Hume’s strategic account of conventions is de facto an explanation of the formation and working of many norms, both positive and negative. Cooperativeness is a norm in many contexts that involve coordination, but so is racist discrimination. These have normative force only in the sense that they get us to act in ways that are mutually advantageous, albeit sometimes at great cost to those somehow excluded from the norm. The force of a convention that establishes a norm comes from its securing our compliance with it by making it our interest to comply, not from the moral rightness of it. Hence, we can count following such a norm as individually rational because it eliminates any individual costs even while it produces a collective benefit. To some extent, moral and self-interested norms over-determine our actions.

That a particular norm can benefit some while harming others means, of course, that the spontaneous creation of a norm through voluntary actions of members of a group does not make it normatively good. A convention can be represented game theoretically, but this is not at all necessary. Indeed, Hume, the inventor of the category and its strategic character, was two centuries before game theory and he analyzed conventions entirely verbally but with great clarity.

Within his own assumptions, Hume sometimes fails to consider that, if conventions can be unintended, then they might be contrary to the interests of those under their sway. He says, wrongly of many cases, that conventions are “intended” (Hume, 1739–1740/2000: T3.2.2.16). This can be true; typically, for example, when a government imposes a convention already extant elsewhere that may have come into being there through spontaneous, unplanned development. In a famous example, in 1967 Sweden switched its convention from driving left, as in the British Isles, to driving right as in most of Europe. Plausibly, only a government could have organized such a change so efficiently at this scale.

To break a convention often requires putting a new one in its place. This can be done with law as it was in Sweden and is done regularly in common law decisions and in legislation. It can be virtually impossible to change a convention spontaneously, as it surely would have been for the Swedish driving convention even though that convention probably arose at least partly through spontaneous coordination. A joke in Sweden at the time of the switch was that the conservative unhurried Swedes would change the rule gradually, with trucks switching to the right one week, busses the next, and cars last of all. A massive police presence, heavy advertising, and instant changing of traffic signs and lights in the middle of the night helped make the move remarkably safe.

Hume implicitly raises the suggestion of functional explanation for conventions. We need a rule for some context but there might be many possible, different rules that are equally good, as in the driving convention and in the details of the civil law. These can vary greatly even though they might address very similar problems. As Hume (Hume, 1998: 3.2.31) says, sometimes “society may require a rule of justice in a particular case, but may not determine any particular rule, among several, which are all equally beneficial.” For example, the many societies in the European Union have widely varied legal systems to govern contractual relations. This diversity of laws poses an obstacle to easier economic relations and development. Any of these systems would most likely be a good one for all of Europe today. Writing a new European body of contract law or switching to any of the extant systems would, however, be very difficult.

Convention arguments may be the clearest applications of the assumption of instrumentalist self-interest that we virtually all share in relevant circumstances, and that, if we are able to analyze our actions, we would acknowledge to be incumbent on all of us. Yet, I adhere to a convention not because it is best for all but because it is best for me. It is, of course, best for all that I adhere to the local driving convention, but that is no part of my motivation, which is to prosper by driving safely. Nevertheless, achieving order by having all individually coordinate on following some rule is what happens. It would be odd to say that coordination is part of selfish human nature, but the motivation de facto to go along with a coordination convention is part of human nature because in this case that selfish nature produces a collective benefit. This conversion of action from self-interest into action for collective benefit is a compelling case of the standard insight that self-interest can be mobilized for public good. In this case it happens at essentially no cost but only benefit. Note that this is essentially a two-stage or two-level argument. At a higher level – of theory – we collectively or institutionally adopt a rule and at the lower level of action we individually follow the rule.

Levels of analysis

The 20th-century effort to read earlier utilitarians, such as John Stuart Mill and Hume, as really rule-utilitarians (e.g. Urmson, 1953) misses the central point of their concern. That point is about institutional arrangements that would secure good consequences and not about individual-level rule following. Mill and others do not have the confused deontological leanings that others assert. They are “institutional utilitarians” (Hardin, 1988). It is true that we cannot imagine living without rules in our daily lives but the main body of such rules is those that are institutionally promulgated and enforced (Hardin, 1995: chs 4 and 5). Following many of these rules is in our mutual interest (again, as in driving conventions). In any modern nation the main body of such rules is the body of law, but there are countless others that are specific to particular professions or organizations. Institutional utilitarianism is about the structure of legal and other institutions, such as systems of rights and of practices. It is a second-order theory of how first-order choices are to be made.

To make clear sense of this position, consider a standard criticism of Hume, that he violates his own vision to see moral and political philosophy as a single, coherent enterprise. Barry Stroud (1977) and others argue that Hume’s move from moral to political theory violates his own vision in just this way. Supposedly, it seems not to be coherent in that the devices that secure moral action do not work to secure compliance with prescriptions of institutions. John Mackie (1980: 86) frames the complaint by asking, “does Hume show that self-interest together with [primarily strategic] understanding is sufficient to make each man accept the system of justice?” (Mackie goes on to answer this challenge with an all too obvious and compelling self-interested reason for our compliance with the law – it can punish and its threat of punishment will usually motivate us. Mackie sees Stroud’s complaint but rightly does not think it is a problem. He says that Hume’s “explanation of the artificial virtues is essentially sociological, while his explanation of the natural virtues is essentially psychological.” Stroud is right. Does this fact matter? No. These are the central natures of the two categories of the issues we must address: sociological and psychological, or institutional and individual.

Concluding remarks

The combination of moral and pragmatic motivations for action in some contexts is powerful in that some people will be driven by one of these motivations while others will be driven by the alternative and seemingly contrary motivation. I may be a Kantian moralist about promise-keeping while you may be a Humean pragmatist. We can both rely on the other’s promise. I the Kantian might think you morally depraved, but that does not matter in determining the outcome of our promising. We will succeed. Your instrumentalism can be beneficial for me so long as you find Hume’s claim that promise-keeping is generally self-interested compelling.