Our reading today is a great “parable” of legal philosophy. Fuller's case of the grudge informers is as famous as Hart’s ‘Mob Island’. It raises similar questions in clearer fashion.  The case is particularly good for showing the normative consequences of adopting a definition, concept or descriptive theory of law.

We consider a situation (fairly obviously modeled on Nazi Germany) in which a recognized, continuous legal system falls into the hands of The Purple Shirts (Hitler's violent youth gangs were known as the Brown Shirts). They pass secret laws, convict people ex post facto, have secret trials without the right to know who or what accusation is being made and so forth. The Purple Shirts would sometime simply execute those it considered enemies without trial but in open, legal-like executions. Other times they would do it secretly. Some people took advantage of the situation to “report” on their personal enemies. They would secretly accuse them of opposing the Purple Shirt regime and the Purple Shirts would carry out a quick judicial execution on their marital rivals, for example.

Later, after the Purple Shirts were overthrown, the new government is inundated with people who wanted justice not only against the purple shirts, but against the “grudge informers” who used the legal framework for killing others.

This question is posed to five “deputies” who argue for different policies based on their legal philosophies. The point is that which legal philosophy we accept has consequences for what we should do about this case.  Hence the "correct" concept of law is a normative issue. The question is not simply which deputy has the right legal philosophy, but which legal philosophy ought we to have considering this case?  (Notice the same structure actually underlies Hart's mob island argument.) What conception justifies the rule of law and which conception (language game) would it justify? It reminds us that the issue is not abstract definition but concrete actions that depend on what we think of as law.

First deputy seems to be a classical positivist. He argues that the Purple Shirts had effective control of the legal apparatus.  (This looks like Kelsen and Austin – "the principle of legitimacy is limited by the principle of effectiveness"). What they did may have been morally repug­nant, but it was legal. Hence, we should do nothing about it. This is the clear separation of law and morals.

Second deputy is a little harder to figure. The formulation makes him seem a little like natural law or a modified version of it. Law has certain moral require­ments and the Purple Shirts behaved with such arbitrariness all we can say is there was no rule of law. Their laws were immoral, hence invalid.  There were no laws at the time. Since there was no rule of law, the grudge informers did nothing illegal and could not (should not) be punished.  This was a state of nature – a war of every man against every other man.

Notice that the conditions the second deputy takes to be signs of the rule of law are broadly procedural.  This Kelsen should allow.  However, Fuller extends "procedure" and shows that we require a procedure with a specific moral content. This begins the blurring of procedural and moral content of the concept of law. He is interested is things like the uniformity of application of the law, how clearly it is promul­gated, how much intervention there is of non‑judicial political power in judicial process etc. The question of the openness of the law and its fundamental fairness seem to be relevant.

Third Deputy has some of Gray's realism though again he takes a more subtle position. There were, he argues, legal processes governing the lives of citizens, courts were making rules for rights and duties and people following them.  Marriages, births, inheritance and the like went on as normal and people recognized and relied on the system in place before and after the Purple Shirts. Also, some incidents were extra‑lawful. What we should do is carefully sift through these and deal with clear cases both ways.  We do it as the courts do: decide matters case by case.

The Fourth objects to the picking and choosing among the cases. We don’t want to be like the Purple Shirts. We want uniformity in our legal process. He recommends that we pass a law, a uniform rule, for dealing with grudge informers. It would be an ex post facto law, but a realist (Gray) might say that courts do this all the time. It is more important to have uniformity (to what? Fairness?) so we don’t have the different attitudes of all the judges affecting who gets punished and who doesn’t.

The fifth, then, reacts to the fourth. He places the value of avoiding ex post facto laws above that of uniformity.  He counsels ignoring the problem and letting private revenge and vigilante action take over (again abandoning the rule of law).

Notice a feature of human behavior that has come up a couple of times already–the human taste for vengeance.  Notice also how the formal definition issues blend into normative questions. We think of the question not as a sociologist would but as an advocate or participant would. In effect, we substitute "What deserves the name 'rule of law.'?" For "What concept captures 'law'?"

This illustrates how the issue of the concept of law becomes a practical, normative problem. The “analytic” theories turn into normative theory about the range of practices that (should) constitute the "rule of law." We are not talking about necessary and sufficient conditions for existence of law, but choosing to have a language game of certain description. This shows that Hart’s internal-external distinction can apply to the very definition of law as it does to the secondary rules.

Hart softens his rigid separation of law and moral in the face of these kinds of problems.  He admits that there are many connections between law and morality. Morality causally affects and reinforces law. However, he insists, something's being against the law semantically implies nothing about its morality. That weakening of the classical Natural law position is no longer at issue; however, the question raised by these reflections is new one. Must the whole system of law meet some substantive moral criterion to count as a rule of law?

The question Fuller wants to put comes from the second deputy. Is there some moral minimum for the entire system? Fuller obviously has in mind things like equal enforce­ment, general application, publicly known and available, consistent inter­pretation, etc. Would a system count as “rule of law” if there were no equality of enforcement? If no laws applied generally? If no one could find out what they were?  It seems like Kelsen's basic norm could not dispense with these implicit procedural requirements and still count as a norm of law. (Even Austin allowed they must be general commands and made known.)

The opposite of the rule of law is the rule of man. The more unformulated intuition and arbitrary, unpredictable results characterize the system the more it becomes rule of man and less rule of law.

Now both sides agree that these prescriptive consequences are relevant to the conceptual question “What is law.” Both appeal to moral consequences of conceiving of law in one way rather than the other. An amoral, positivist conception focuses on certain formal "trappings of law." It says, no matter how arbitrarily it actually functions, how much of a facade it is, we must judge it to be a rule of law if it primary rules. The realist thinks that if the facade covers something arbitrary, then it's being in the form of rules backed by hurting people does not qualify it as the rule of law. The mob island with rules is still the mob island.

Consider a game analogy.  If one is playing Go ³ò ´Ñ   and very occasionally cheats (moves a peace after it is placed) then we would say that he cheats at Go.  However, if he systematically and regularly moves pieces after placing them–then we would simply deny that he playing Go. He may be playing some other game.  Similarly, if some vicious, punitive thugs have a legislature writing laws but they consistently keep these laws secret, pass them ex post facto, apply them inconsistently, alleged them to exist without proof and so forth, then we would not describe that as the rule of law.  That is the rule of man -- lawlessness.

The features of law that Fuller regards as essential are those that make the system as a whole predictable and reliable. Implicitly, even Hart’s argument about the common man accepted that predictability was part of what justifies law as an institution. These are broad moral restraints, Fuller argues, on the principles guiding any system that would count as the rule of law. Our concept of law must account for our fidelity to law.

The value or utility of law, he argues, comes from its role in adjudicating claims. We would not be satisfied with just any system for settling claims (e.g. flipping a coin).  The system has to be seen not merely as giving an answer but giving the right answer with some level of reliability.  (Arguably, when people do accept something like trial by combat or the mandate of heaven, they typically assume that divine intervention makes the outcome of the procedure accord with justice.)

At a minimum, we would always object to any lack of formal justice--treating like cases alike and different cases differently.  Procedural fairness is a requirement for our fidelity to law.  Similarly, Hart's argument that a judge ought to use the interpretation that would "spring to the mind" of the common man appears to be a moral one.  We would regard it as unjust if the interpretation were utterly unpredictable or counter-intuitive. This kind of consideration also lies behind the principle that laws must be promulgated or made public. We should not allow ex post facto laws for a similar reason—predictability.  The courts should follow the principle of stare decisis (let the decision stand–the court should reach a result that is consistent with its former decisions).  We have a moral right that they not do this--it is the morality implied in the very concept of law, according to Fuller.