By Terence Yung
For today’s post, I will discuss Hart’s conventional foundations of the concept of law and the rule of recognition. In focusing on consensus among officials, does it ignore disagreement as to what counts as law in the legal system? I will argue that while Hart has been a popular target, his theory is defensible.
- Hart’s foundations
I want to first (briefly) set out Hart’s conventional foundations of the concept of law. Hart says that law is a system of rules that form a normative regulatory structure. We can have rules and habits. Rules are different from habits because rules require an internal point of view (whereas habits don’t). This manifests via a critical reflexive standard that involves criticism, demands for conformity, and an acknowledgement that criticism/demands are justified. Rules can be moral or legal. Legal rules are different from moral rules because we have to use the rule(s) of recognition to identify legal rules (but not moral ones). Legal rules can be primary and secondary. Primary rules impose obligations. One can be obliged (coercion) or be under obligation (duty). Hart, therefore, allows the law to make both a positive (facilitative) and negative (coercion, sanction) contribution. But primary rules suffer from three defects from the state of nature: (1) the need to have certainty as to the content of primary rules (2) stasis – the need to change rules, and (3) inefficiency – the need to resolve disputes arising from particular cases. Secondary rules are meant to deal with these defects. Thus, (1) the rule of recognition deals with uncertainty, (2) the rule of change deals with stasis, and (3) the rule of adjudication deals with inefficiency. There are two important features I would like to underscore. First, Hart notices the internal aspect of law. He says the participant’s attitude is necessary for us to understand the nature of law. Second, he notices secondary rules – these are rules that other theorists missed or misdescribed.
- Rule of recognition
I want to now set my sights, more specifically, on a particular secondary rule – the rule of recognition (which has been subject of much criticism). I talked about how the critical reflexive standard distinguishes between rules and habits. The problem is that this is not distinctly legal. We need a way to tell the difference between legal rules and other rules (e.g., social rules, rules of a game, rules of etiquette, etc.). This is the defect of ‘uncertainty.’ The rule of recognition deals with this. It lets officials tell the difference between what is a valid legal rule (from what isn’t). We can state the rule generally thusly – a rule can be legally valid only if it satisfies the criteria laid down by the rule of recognition. More specifically, Hart says in the UK, this means rules are legally valid only if the King in Parliament enacts it. There are a number of takeaway points here. First, the rule of recognition is different from Bentham in that it doesn’t rely on realpolitik (as Bentham/Austin so). Second, it relies on a right to rule without needing to involve sovereign. So, unlike Bentham/Austin, Hart can (in more natural terms) appreciate the difference between a bank robber and a tax collector. Third, in keeping with Hart’s positivism, the right to rule is formal/legal (not moral). Fourth, the rule of recognition is an ultimate rule. While it tells us what is valid, it can’t be validated itself. We have to assume it is valid (which is an assumption outside of the system). In this sense, it is like Kelsen’s grundnorm. The critique here would be that accepting the rule of recognition would involve considerations external to law.
- Ignoring disagreements
First, some difficulties flow from the rule of recognition itself: (1) what it means and (2) how many rule(s) it consists of. First, there is disagreement over what the rule of recognition means. Part of this is Hart’s own fault, because he talks about the rule of recognition in three (slightly different) different ways. First, he says they are linguistic entities that designate what primary rules are by designating criteria for legal validity. Second, he says they consist in certain linguistic entities. In other words, they are criteria telling us what primary rules are. If a rule fails these criteria, it is excluded (i.e., it isn’t a legal rules). Third, he says they are a social rule and a specific type of social practice. The extent to which it is a social rule/social practice has been subject of some debate. Waluchow agrees with Hart. Coleman disagrees. He says the rule of recognition is a rule, whereas a practice is constituted by behaviour. The former has conditions of satisfaction, the latter doesn’t. So, the rule of recognition can’t be equated as a social practice. Overall, Prof. Zipursky concludes that this ambiguity as to what the rule of recognition means undercuts the authority of Hart’s endeavour. A number of elements of his theory turn on the rule of recognition, and the fact that it means 3 things at the same time injects a lot of confusion. Second, there is a debate over how many rules of recognitions are out there. Is there one? Or, is it several? Well,Hart injects some confusion here as well. Hart says we can state the rule generally thusly – a rule can be legally valid only if it satisfies the criteria laid down by the rule of recognition. But he also says more specifically, in the UK, the rule of recognition means rules are legally valid only if the King in Parliament enacts it. MacCormick has tried to set out a rule of recognition for a fictional state. He comes up with six rules dealing with the constitution, laws of the legislature, delegated legislation, court rulings, custom/usage, and rules in force. If we look at it this way, then the question becomes whether the rule of recognition boils down to a list of sources of law. Prof. Wacks says we can respond to this question in at least two ways. On the one hand, there is an essential connection between what is law and sources of law. On the other hand, this could be interpreted as more than a list. The rule of recognition is one rule, but it has six criteria ranked in order of importance. In sum, Hart does not provide a clear enough answer to the question of how many rule of recognitions there are. Prof. Wacks (I think rightly) counters that these considerations are more tangential to Hart’s main idea. We can have uncertainty in these areas, yet this would not be fatal (and Hart’s main idea could still be sound).
Second, Dworkin further says Hart’s rule of recognition ignores disagreements over what counts as law in two ways – (1) it can’t identify a rule and (2) it requires interpretation. One, it ultimately can’t identify a rule to be applied. On Hart’s rule of recognition, in order to be a rule of law, you have to be able to identify it with certainty (if not, it’s not a rule of law). But hard cases are, by definition, cases where there is some controversy over what counts as law. But Dworkin says if something is controversial, that can’t be law. Hart says judges have to exercise strong discretion: (1) they can make new laws (2) they can apply this law to the defendant. Hart says the judge is ‘finding’ the law, but for Dworkin, this is a distortion of usage. What Hart really means is that the judge is just making law or stretching old precedents to fit new situations. But Dworkin says a judge doesn’t have that kind of strong discretion – making law contravenes separation of powers and stretching old laws to fit new situations is impermissible retroactive. So, Dworkin concludes that ultimately Hart’s positivism is unsuccessful –it can’t identify which law to apply and so it can’t solve the defect of uncertainty. Two, Dworkin thinks Hart is wrong to suppose there’s a descriptive sense of what a rule means before interpreting it. Hart looks a rule in terms of a core meaning (an uncontroversial meaning) and a penumbra (uncertain meaning). But Dworkin says whether the meaning is core or a penumbra, either requires interpretation. A clear meaning is clear because you interpreted it (and this is, a fortiori, the case with an unclear meaning). In response to both critiques, I think Hart’s best defender is Raz. And basically, what Raz says is, yes, before the legal official makes the decision there is some uncertainty (and thus no law). But things don’t have to be uncertain forever. In fact, after the legal official makes a decision, what the legal official decides becomes an exclusionary reason and this is what solves the problem of uncertainty (and thus we have law).
I have tried to show that while Hart has been a popular target, his theory is defensible. I looked at two difficulties flowing from the rule of recognition – it’s not clear what it means and it’s not clear how many rule(s) there are. I argued these types of uncertainty could be regarded as somewhat tangential to Hart’s overall point, so as not to detract from the explanatory power. Dworkin’s critiques, though, seemed to strike at something more fundamental. He says Hart’s theory can’t yield law and that Hart is wrong to assume there’s a descriptive sense of what a rule means. I argued that probably Hart’s best defender is Raz. The implication of this is perhaps Dworkin is wrong to say Hart’s theory can’t yield law. It can, after a judge interprets it. While Dworkin might be right to say there is no descriptive sense of what a rule means before interpreting it, I think Raz (and Hart) would not find that problematic. As long as we can come up with a descriptive sense of what the rule means after a judge decides, they would consider that the defect of uncertainty has been dealt with.
About the author
Terence Yung is an international arbitrator with a commitment to resolving complex disputes in a fair and impartial manner. His interests span across various industries, including financial markets, intellectual property, energy, and infrastructure projects. He has been an Associate Member (ACIArb) of the Chartered Institute of Arbitrators since 2022. In addition, he is affiliated with the German Arbitration Institute, the London Court of International Arbitration, the International Bar Association, the International Centre for Settlement of Investment Disputes, and the Swiss Arbitration Organization. He was awarded an LL.B. from the University of London with First Class Honours. Prior to that, he was awarded a B.A. and a B.M. from the University of Houston, both with summa cum laude honors. He also holds course certificates from Yale University, the University of Pennsylvania, the University of Edinburgh, and the SDA Bocconi School of Management.
Email: adrwithyung@gmail.com