According to Augustine, Aquinas, and Martin Luther King, Jr., an unjust law is no law at all. I’ll call this thesis ULNL, relying more on Aquinas’s version of it than Augustine’s or MLK’s. ULNL is, famously or notoriously, a staple of natural law theorizing. Though sympathetic to what he calls “the minimal content of natural law,” H.L.A. Hart takes issue with ULNL in The Concept of Law on both theoretical and deliberative grounds.

In this post, I want to sketch a line of thought that gives ULNL a plausibility that Hart fails to discuss.  The missing link in Hart’s discussion is the law’s connection to the common good. According to Aquinas, it’s essential to a system of law that it somehow aim at the common good of the governed. A system that fails to do this fails to satisfy a necessary condition of law. Properly understood, I think Aquinas is right about this, and that Hart’s dismissal of ULNL fails due to his failure to discuss this “missing link.” 

Following Hart, let’s suppose that law is a union of primary and secondary rules involving a rule of recognition. Such a system can bear at least four different connections to the common good. It can…

1. Generally aim at the common good, and generally achieve it.
2. Generally aim at the common good, and not generally achieve it.
3. Not generally aim at the common good, yet generally achieve it.
4. Neither generally aim at the common good, nor generally achieve it.

I’ll assume that a union of primary and secondary rules involving a rule of recognition and exemplifying (1) is, uncontroversially, a system of law. But what about cases (2)-(4)?

Cases (2)-(4) in some sense fail with respect to the common good–either with respect to outcome (2), aim (3), or both (4). Obviously, failures of this kind will tend to differ by degree. Cavalier as it may sound, we might describe some of these failures as merely “garden variety,”  at least going by the broad sweep of legal history, and others as “egregious.” The more egregious the failing, I want to suggest, the more plausible ULNL becomes.

Take case (2). Case (2) is one in which the officials behind a legal system (framers or sustainers or both) aim at the common good, but fail to achieve it in a systematic way–or rather, systematically fail to achieve it. Put differently, despite their intentions, there is some systematic flaw (or set of flaws) in their legal system that vitiates the promotion of the common good. In particularly egregious cases, we can imagine that the flaws are so deep that they literally subvert the common good. In the most egregious cases, we can imagine that they subvert the common good more thoroughly than would be the case had there been no system of law at all.

The plausibility of this supposition turns on the kinds of systemic errors we imagine. So imagine a system in which many of the most important features or provisos of the law are not publicly promulgated; where the scope given to official discretion is so wide as to make the law completely unpredictable; where the laws are themselves unclear and internally inconsistent; where the laws presuppose false factual assumptions; and/or where the laws demand the impossible of the governed. A system of this sort might be called a system of law, but wouldn’t in fact be one.

We don’t, in general, describe an artifact as a genuine, functioning instance of its kind if it positively subverts the aim it was designed to bring about. A knife that was worse at cutting things than the use of one’s fingers would not really be a knife. A fork (even a plastic fork) that fell apart at its first contact with any food requiring a fork would not be a fork, even if it looked like one.  A computer that never turned on or booted up (and couldn’t be fixed so as to do so) would not be a computer. A computer program that never (or almost never) did what its designers intended it to do would not be a program. A blog post consisting of pure gibberish would not be a blog post. A system of medicine that achieved worse outcomes than no treatment at all–where the incidence of iatrogenic injury exceeded what would be the case without treatment–could not really be regarded as a system of medicine, no matter what anyone called it. It seems to me that a system of law with the same design flaws follows the same logic. A legal system that can’t govern is like a knife that won’t cut and never could.

It’s a worthwhile interpretive question whether Hart would himself agree with the preceding line of reasoning. (It’s an open question whether Hart is a legal positivist, and if so, how much of one he is.) Though I’ve put things in a different way than Hart does, case (2) seems like one that runs afoul of what Hart himself calls “the minimum content of natural law”–minimum content that he seems to regard as a necessary condition of something’s being a legal system (The Concept of Law, pp. 193-200). Case (2) may also may run afoul of the mean he intends to adopt between formalism and rule-skepticism in Chapter VII of the book. But I’m not certain of any of this, so I leave the interpretive question open. My point is that Hart does not clearly address the case I’ve just described with respect to ULNL.

Now take case (3). Case (3) is one in which the officials behind a system (again, framers and/or sustainers) have no intention whatsoever of promoting the common good. They just happen, fortuitously, to succeed at doing so. In other words, imagine that these officials have aims that are so egregiously nefarious that they amount to using an enforceable union of primary and secondary rules to facilitate what would normally be described, at least morally speaking, as a criminal scheme. What I have in mind is not simply an unjust or defective system of law, but a system of rules, mimicking law, that aims to subvert even the most minimal purpose of having the rule of law.*

I grant that it’s implausible to imagine such a system’s fortuitously promoting the common good, but it’s certainly possible. Imagine a system that, by intention, aimed to dominate, exploit, victimize, or even destroy the governed in the long term, but happened to promote their good in the short term. Or imagine a system whose adverse effects on the population were so incremental (yet ultimately lethal) as to be indiscernible until well into the effectuation of the scheme. The promotion of the population’s good might, within such a scheme, come about either through some failure of calculation or execution, or by deliberate design. Either way, a system that by its officials’ explicit intentions aims at the subversion of the common good seems less like a legal system than an elaborate blueprint for crime, even if it happens to achieve the common good. Though not literally self-contradictory, it seems problematic to think of a legal system as the essential instrument of a criminal scheme.

Again, it’s an interesting (non-self-evident) interpretive question whether Hart would agree with the preceding reasoning, but I see no clear textual warrant for thinking he would. If anything, what Hart says about “legal validity and moral value” (Concept of Law, pp. 200-212), and in particular, thesis (vi) under that heading (Concept of Law, pp. 207-212), suggests that he wouldn’t. But I’m not sure I fully understand Hart’s position, so I leave the issue open.

Case (4) is just the conjunction of the worst parts of cases (2) and (3). Intuitively, if both (2) and (3) give ULNL some plausibility, then (4) combines what is plausible about both cases to give ULNL more plausibility than either case has on its own.** For this reason, as I see it, ULNL has plausible application in all three cases, but clearest application in case (4).

So far, I’ve just listed cases where one could adopt ULNL. But why adopt it? Hart adduces theoretical and deliberative or practical reasons against doing so. Here are some theoretical and practical reasons in favor.

As a theoretical matter, ULNL fixes a baseline such that putative legal systems that fall below it fail, even in a minimal sense, to aim at and/or achieve the goods distinctive to law. A system of rules that fails at realizing the threshold functions of a legal system is not (as I see it) a legal system. There is a useful theoretical distinction to be drawn between a system of rules that has the outward forms of law but subverts the goods distinctive to law and a system that has those forms but promotes them. The value to historiography, political science, and political philosophy of such a distinction seems obvious. Just as we need a way to demarcate science from pseudo-science, or medicine from quack medicine, or a business scheme from a fraud, we need a way to demarcate bona fide legal systems from pseudo-legal systems, even while granting the (garden-variety) defects that a bona fide legal system may have while still qualifying as one.

As a practical or deliberative matter, the distinction between a legal system and pseudo-legal system codifies the thought that there is such a distinction to be drawn in practical life, so that when  a “legal system” fails to meet certain necessary conditions, it lacks even the minimal claim to respect that is due to (the officials of) a garden-variety unjust or defective legal system. Where the failures are clearly culpable, as in case (3), it becomes permissible to wage war on the system, to destroy it and its sustainers, thinking of them as having initiated a war on the “governed” via the (so-called) legal system. Where the system’s failures are not (necessarily) culpable, as in case (2), the question arises of ending the system, nullifying its rule of recognition, and starting over to create a completely new system. Ordinarily, as in the case of a defective but remediable legal system, such drastic measures would not be called for. But if the system in question was never a legal system at all, it merely stands as an impediment to an actual system. What is needed is an actual system.***

I’ve discussed ULNL at the level of systems rather than individual laws, and for egregious rather than garden-variety failures. It may be that ULNL fails in the cases I haven’t discussed. But in cases where we have wholesale, egregious, system-wide failure, the thesis strikes me as plausible enough, and as given somewhat short shrift by Hart and his positivist successors.

*Though Hart is very insistent on the principle of nulla poena sine lege (no punishment without law), I would insist that we can define a set of moral crimes independently of any law (Hart, Concept of Law, p. 211). A proper subset of the so-called common law felonies (murder, rape, robbery, arson, kidnap, burglary/larceny, and assault/battery) would (I would argue), morally speaking be crimes in a Lockean State of Nature lacking any system of law. Contrary to Hart, it’s a legitimate question whether punishment is justifiable with or without law, in which case nulla poena sine lege does less normative work than he seems to realize.

**As an example of case (4), consider the thought-experiment I used to justify “terrorism” in my critique of Vicente Medina’s Terrorism Unjustified. The Machiavellian-Lockean scheme I described there would not, on my view (and on the generic Augustinian-Thomistic-King view I’m defending), count as a bona fide legal system, but I think it would on Hart’s. (To be clear, I what I was justifying was “terrorism” in Medina’s sense, not what I myself regard as terrorism. So don’t go calling the cops on me.)

***I grant that historically, there are probably no clear cases of this kind (or at least very few), and that most attempts to conceive of legal systems as complete pseudo-systems requiring wholesale destruction have probably been mistaken. But the conceptual possibility is worth bearing in mind even in those cases.