Epistemic status: Marginalia on Scott Shapiro's "The 'Hart-Dworkin' Debate: A short guide for the perplexed". This is my first serious reading on the Anglo-American version of legal positivism and, thus, my understanding of the framings used by Hart and Dworkin are those provided by Prof. Shapiro. Unrevised.
I have never been a fan of legal positivism, at least in its civil law version. Not so much because I think it's not workable or because it fails to provide a sound basis for law. After all, modern civil law - at least in the infra-constitutional level - is grounded in positivism: the idea that an unfair law is still the law is a creature that can only make sense due to the existence of a Code. It's an empirical matter: positivism can be the grounds for a workable and sound legal system.[^1]
My issue with positivism comes from its lack of resiliency when faced with authoritarian attacks. I am not sure that you can blame Kelsen for the legal crazyness in Nazi Germany but I think it's not unfair to say that the judiciary, along with all other institutions, failed to parry the attacks. My view was that positivism's obsession with the Code makes it so that any successful attack on the Code becomes self-implementing, since the courts or the government bureaucrats can't fix it. This critique of positivism might be a mere critique of centralized (or at least hierarchical) sources of law. Thus, it is possible that I was equivocating civil law with positivism and actually criticizing the former instead of the latter. This description of common law version of positivism given by Shapiro made me rethink my groundings. However, I still believe that positivism is not resilient to authoritarian attacks, but for different reason.
Shapiro nicely describes an evolving debate between the dworkian and the positivist respective on how we identify the legitimate sources of law, i.e., what are the proper grounds for the rule of recognition. Hart's view is that the rule of recognition relies fundamentally on social facts whereas dworking appears to believe that everything is, fundamentally, a moral judgment. In the civil law tradition, the rule of recognition is given by the code[^2]. By contrast, the common law tradition has a multi-institutional approaches: both legislatures and the judiciary have independent law making roles.
The debate, as narrated by Shapiro, goes like this. Dworkin initially challenged Hart's positivism due to its incapacity to explain why judges sometimes depart from settled law and resort to moral reasoning. The solution to that problem isn't difficult by resorting to a "meta" approach that says that the use of the moral principle is either (1) itself a rule recognized by the rule of recognition and that the importation into jurisprudence is a recognized rulemaking process ("Exclusive Positivism") or (2) not distinct from the use of any rule because a judge would never use a moral principle that does not obey the rule of recognition ("Inclusive Positivism").
The response by Dworkin to theses replies is basically that these responses do not solve the problem made explicit by the fact that judges and scholars have theoretical disagreements. That is, two judges sometimes agree on what the law says and what is the gap needed addressing but they disagree on what to do next. And it's here, in my view, that Dworkin touches on the trouble with positivism. Positivism's approach says that the solution must be found in society, in some manner. Shapiro's approach that addresses this question is successful. The shared social fact is the "legal designers purpose" and the theoretical disagreement is on the issue of what that purpose is.
I think that the idea that this is the norm to be followed is a perfectly fine proposition, in abstract. The problem is that that is just not what judges do. They, instead ask themselves "what is the right thing to do here" and "will the law allow me to do this"[^3]? In fact, dare I say, that is what every rulemaker does. Their purpose, whether a judge or a congressmen, is tailored to a question at hand, "how far can I go to serve a purpose I care about." As such, the "legal designer's purpose" is a fiction. The original US Constitution had a slavery compromise. The civil war amendments appeared to support "separate but equal" in the initial years. And more practically, since the Supreme Court says what the constitution is, they are an important component of the "legal designer" and their jurisprudence is not quite an example of judicial stability.
At heart, the issue that I have with positivism is that it is an attempt at removing the impact of the sociological elements of lawmaking and jurisprudence by pretending that it they are not important, either by claiming it's irrelevant or claiming it's subsumed within the principled stance. In some sense, positivism believes that there are parties in the system that are involved in law finding and others that are involved in law making. In truth, all parties are doing both. In different degrees and constrained by their institutional roles, to be sure. But it is dangerous to think that a judge that is overturning solid jurisprudence did so because the overturning itself, the new rule, or the principle that drove the judge to overturn the rule, that either of these obey the rule of recognition. It is must more likely that the judge did it because they think it's the right thing to do, and the fact that we can rationalize a positive view of what they did just adds a "social fact legitimacy" to something that was a moral judgment call.
[^1]: A fact that many civil law systems have issues is no objection, since most legal system have issues. A system that is unworkable in this sense is one where the grounds create issues that prevent a system from "looking like" there is a rule of law.
[^2]: Or the will of the people, which the code attempts to express, if you prefer a purposivist approach instead of textualist one.