The main idea: judging from the historical performance of non-territorial poly-legalism, tomorrow’s rights enforcement agencies will probably be driven to seek a minimal yet universal agreement delimitating the legitimate territorial boundaries of their clients, a point in favor of Rothbard’s conception of a Free Society.
1. Rothbard vs. Friedman
1.1 Two rather different visions of a future Free Society have been put forth by two well-known libertarian thinkers. Murray Rothbard’s The Ethics of Liberty goes on to produce a rather detailed vision of a universal libertarian legal code to be enforced by private enforcement agencies, the services of which are to be freely purchased by the public.
1.2 David Friedman’s The Machinery of Freedom envisions a set of legal codes, driven by market demand, again enforced by private agencies. Yet Friedman not only doesn’t discuss what the resulting legal code may be, but indeed insists that there will be no single such code, and that instead the laws to be applied to a specific dispute will depend on the specific disputants, or rather on what their respective enforcers have agreed to beforehand. Such agreements between enforcers A-to- B and A-to- C need not be the same.
1.3 Leaving aside Rothbard’s critique of his peer, Friedman has said of “The Ethics of Liberty” that both the implicit assumption that a single code will apply to all disputes and that the actual content of the code can be known in advance are highly unlikely.
1.4 If we disarm ourselves of natural rights theories (as Friedman does), there may well be no way to judge such claims until the actual emergence of a Free Society. Or again, one could look into the history of poly-legalism - arrangements by which no single set of law applies to every dispute – to see whether some empirical understanding may help tip the scales.
2. Poly-legalism is practice
2.1 Finding a rather rich history of poly-legalism is as simple as looking at a list of multi-cultural (by their own time’s standard) societies. In all cases where the cultural minorities in such societies have been willing and/or able to oppose assimilation or deportation – and provided they were not small enough to be safely ignored – some sort of accommodation had to be provided for them. If we disregard city-states – which can often avoid tensions arising from heavily multi-cultural societies due to the ease of exit – than we can see that historically such an accommodation has taken either of two forms.
2.2. The simplest form of poly-legalism meant applying different sets of laws to the members of different communities, a most notable example being the millet system of the Ottoman Empire. Under this system, religious minorities where allowed a degree of self-rule based on their own laws, managed by their respective religious authorities. Muslims being the dominant group, every case affecting a Muslim citizen was adjudicated under Muslim law, but otherwise two orthodox Christians could call on their own church to adjudicate their quarrel.
2.3 The second form implies allowing different laws to apply to different locations, instead of different persons. An historical example can be seen in the old European Jewish ghetto. Unlike the Ottomans, no European states I know of allowed Jews qua Jews to settle their disputes according to their own laws, religious or otherwise. Instead, it was generally understood that specific areas – the infamous ghettos – would be left under the power of local elites. We may not assume that such an arrangement was always lamented by the minorities themselves and, as Kakugo points out, in at least one instance (Venetian Jews) ghetoisation was explicitly requested.
2.4 The difference between the two approaches to poly-legalism should be clear: you either let different laws apply to different persons or to specific places, letting the persons who fall under those laws sort themselves out.
2.5 “Segregation” of any kind being frowned upon nowadays, the modern approach to tackling issues of multi-culturalism is distinctively of the first kind, although far from the extremes to which the Ottomans where willing to take the principle. Nowadays it is “rights” that follow a person no matter where he goes, and no longer laws. Thus, a Walloon or Quebecois should be able to exercise their ‘linguistic rights’ no matter where they live in Belgium or Canada. This implies widespread bi- or multilingualism throughout the country, and especially within the civil administration (sadly, often the prime target of minority right activists).
2.6 By contrast, the second approach lives on in Switzerland, which subdivides its territory in 28 cantons, and (by and large) leaves each canton to sort out its linguistic preferences, counting that the rather minute subdivisions will leave a negligible minority in any given canton.
2.7 The historical record of these two approaches seems biased: I know of no instance of the first approach surviving intact to our days as many societies which counted on it to avoid internal turmoil have experienced just such turmoil. The Ottoman Empire was torn apart by its disaffected minorities, and the modern-day separatist movements in Canada and Belgium are well known.
2.8 By contrast, the last episode of communitarian conflict in Switzerland (the Jura crisis of the ‘70) emerged precisely because the Swiss model was not being applied in full force (the conflict was resolved with the creation of the canton of Jura). City-states aside, Switzerland remains, to my knowledge, the only fully successful model of a multi-cultural society to date.
2.9 Likewise, when the Hapsburg Empire quit its efforts to accommodate the Hungarian elites and allowed them their own fief to (mis)govern, it turned a militant minority into a loyal and supportive group for the Dual Monarchy.
2.10 Of course, many instances of territorial poly-legalism failing to prevent the eruption of civil unrest can be found (Spain in the making?), while the case that the failures of ‘peer-to-peer’ poly-legalism are to ascribed to exogenous causes (taxation, conscription, assimilation drives, external wars, etc.) can be made. But still, one will have to admit that the chances that there is something unworkable in ‘peer-to-peer’ poly-legalism in itself are higher that parity. A meager conclusion, but such are the fruits allowed to those who stray from the path of praxeology.
3. The reasonable Rothbard
3.1 The empirical record of the two models of poly-legalism can help us further the Friedman-Rothbard discussion. On first sight, Friedman’s comments seem to apply, as it would seem unlikely (at least for those of us who do not take natural law theories seriously) for a society to just agree on a single code of law.
3.2 Further, Freidman’s model would only require individual enforcement agencies to strike a number of dual deals, clearly easier and much more practicable than somehow hectoring in all enforcers into a single, mutually agreed code. An enforcement agency sponsoring a number of such deals could offer better protection to its clients right now, instead of having to wait for all agencies to agree in the future to one code.
3.3 Still, the probable unviability of the first type of poly-legalism (a generalization of which Freidman’s structure is) makes one think again. The “laws following persons” model, in and of itself doesn’t seem to do enough to contain the displeasure of living with people of widely divergent world-views (otherwise yours truly would still be writing from an Ottoman border province).
3.4 One can expect this poor record to translate in monetary terms for Friedman’s enforcers. Perhaps customers would be willing to pay a lower price for a company that couldn’t provide a certain and fixed set of laws in advance, or else the frequency and severity of quarrels between clients of different companies (hence, of companies themselves) would be higher. These costs should tend to increase the less culturally homogenous society is.
3.5 On the other hand, an enforcer which operated on the base of the territorial conception of poly-legalism, i.e. which applied its protection to a specific area instead of a client, could be expected to receive a monetary boon in the form of higher premiums and/or lower costs. If this monetary incentive will turn out to be strong enough (I suspect it will) one can see how a territorial enforcer could, all else being equal, outcompete a Fridmanite one.
3.6 But in order for territorial enforcers to dominate the market some sort of general agreement on territorial boundaries will have to be agreed. Allowing Jura to set its own laws would do no good if the parties disagreed on the borders of Jura. The maximal value to the enforcers could be gotten by agreeing on a general, yet minimal set of term for delimitating proper territorial boundaries between their clients.
3.7 Such a system, I say, already exists in the form of the libertarian theory of just property, discussed by Rothbard. Again invoking no natural right theories, one could say that the Rothbardian program imposes the least amount of rules necessary to properly delimitate the areas of ‘insurable interest’ of different people. Such a program is, to borrow Nozick’s term, a framework for utiopias instead of a utopia itself, and a group of enforcement companies eager to make money by applying territorial poly-legalism could reasonably be expected to quickly come to a general agreement of this kind among themselves.
3.8 None of this is to mean that no services as those envisioned by Freidman could ever survive (plenty would, I say), but that some sort of minimal and universally agreed set of rules should still emerge from a Free Society. Also, not every last detail of the Rothbardian code can be afforded such a defense, and only the general outlines should count.
3.9 Still, Friedman may have been too hasty in denouncing the larger points of Rothbard’s vision as unlikely to be agreed by all. History seems to suggest that the only way to let people of different worldviews live in peace for extended periods of time, and even let them have some sense of community, is to provide competitive sets of rules tied to specific territories, which in turn requires a general agreement on the definition of a just territorial claim. Or then again, the Swiss may be deluded and the Walloons and Flemish just in a playful mood.