09 Jul Concluding Post on Making Sense of Customary International Law
[Monica Hakimi is the James V. Campbell Professor of Law at Michigan Law School.]
Thanks again to all the contributors to this symposium. It’s hugely rewarding to have such an extraordinary group of international lawyers and scholars engage with my work. It’s all the more rewarding that some of them find my argument at least somewhat persuasive. Candidly, I expected much more pushback than I received. Even the respondents who resist my argument do so on terms that leave my core claims intact. They do not deny my descriptive claim that, in the everyday operation of CIL, it works nothing like a rulebook. They do not touch my analytic claim that the rulebook’s methods are inadequate (or worse) for those who need to understand or navigate the real-life practice of CIL. And they do not take issue with my normative claim that the rulebook conception has led us systematically to devalue CIL. So, what do they disagree with? In this post, I want to home in on what I take to be the sharpest criticisms of my argument and explain why they are either not responsive or not compelling.
I’ll start with Kevin, who is the most critical of the bunch. Although Kevin claims to “disagree with nearly every word” in my article, his two “most important objections” are orthogonal to it. First, he contends that states themselves buy into the rulebook conception; they think of CIL as a set of rules, and they invoke the two-element test to advance their normative positions on CIL. Kevin is absolutely right on this point. But in pressing it, he simply exemplifies the problem that I am trying to address. Yes, people (including people who advise states) imagine that CIL is like a rulebook. My claim is that they are wrong to imagine it in this way. Their theory of CIL is completely dissociated from their practice on CIL. When states and other actors actually engage with CIL, they do not follow anything resembling a rulebook.
Kevin focuses on the rulebook’s supposed secondary rules. He concedes that they do not really operate like rules. They are not stable or consistently applied but contested and highly malleable. In Kevin’s words, they are “admittedly fuzzy” because “states disagree, often intensely about . . . how international courts, NGOs, and even other states apply CIL’s secondary rules.” That’s part of my descriptive claim. The supposed secondary rules of CIL are not at all like rules. The idea that they are is misinformed. And to be clear, there’s nothing unusual about people—or states—having mistaken theories of their own behavior. It’s routine. (If you need a concrete example, here’s an easy one: The U.S. government’s official name for the 2003 Iraq War was “Operation Iraqi Freedom.”)
Kevin’s second objection is that, instead of discarding any secondary rules, I am simply replacing the rulebook’s “admittedly fuzzy secondary rules with secondary rules that are even more fuzzy.” This objection also misses the mark. My argument is not that we should replace one set of secondary rules with another. My argument is that CIL does not have any secondary rules. One of the article’s section headings is even entitled, “No Secondary Rules.” By this I mean that the criteria for establishing or identifying CIL do not function as rules. They are not stable, intelligible, or generally applicable. They are much more fluid, contestable, and variable than any rule would be. They are, to quote Kevin, “fuzzy.” The rulebook conception is wrong to suggest otherwise.
Other respondents raise objections that are more jurisprudential in orientation—about how I define law in relation to the social practice. Jean suggests that I mistakenly “dissociate” the law (which he sees as inherently rule-like) from the real-world practice. As he describes it, I engage in “a form of ontological dualism that is very common in international legal scholarship whereby international law and the world to which international law is applied are strictly distinguished.” By contrast, Martin suggests that I do the opposite—that I collapse the law’s normativity (i.e., the ought) into its materiality (i.e., the is). I guess this means that the real Monica Hakimi must stand up. In my view, CIL cannot meaningfully be dissociated from the social practice that gives life to it. I therefore reject the “ontological dualism” that Jean accuses me of. Perhaps the reason Jean misses this point is that he assumes that law must be rule-like and thus that the part of the social practice that is not rule-like must live outside of or be antithetical to law. Put differently, when he says that “there is just nothing outside the rule,” he is in my view conflating “the rule” with “the law.” Or perhaps Jean thinks that rules are inescapable at the level of epistemology or semiotics. Even if they are, it wouldn’t follow that they’re inescapably part of CIL. As I argue in the piece, CIL is a kind of law that can be but is often not rule-like. So, while Jean might be correct that there is nothing outside the law, I disagree that “there is just nothing outside the rule.” There’s a lot of law outside the rules.
At the same time, the social practice that I describe does not belong entirely in the world of the “is.” I am describing a legal—and therefore an inherently normative—practice. It is about what states and other actors may do, not just about what they can do. For example, when global actors invoke the ICRC’s Interpretive Guidance to conduct, justify, condemn, analyze, or argue about particular targeted killing operations, they are engaging simultaneously with the is and with the ought. The two cannot be disentangled. To be clear, this does not mean that the ought collapses into the is. If anything, the fact that global actors constantly use CIL to argue about the ought suggests that that dimension of law matters much more than old-school realists recognize.
Jutta and David raise a related objection. They both suggest that we need criteria for distinguishing the law from the parts of the social practice that are not law. This point goes to a much larger, metaphysical question about how we define law. I address it to some extent in the article (see Section III.B), but it requires more discussion than I could give it there or, for that matter, in this post. I plan to return to it in my current book project. For now, let me say that not all definitions of law require it to satisfy certain abstract and external criteria (e.g., generalizability) that distinguish it from not-law. People in this world conceive of law in many different ways. In the article, I’m less concerned with this metaphysical question than I am with how the material that people refer to and treat as CIL actually works.
This point warrants further emphasis: the conduct norms that are treated as CIL in the real world do not always operate as rules. Many of them have CIL effect in some settings, by some actors, and for some purposes but not in, by, or for others. For example, when a national court invokes the ICRC’s positions in the Interpretive Guidance to absolve someone of responsibility, it treats these positions as CIL. The positions are for all intents and purposes CIL in that setting, even though they are not CIL and are contested in many other settings. When the ICC Appeals Chamber asserts that CIL entitles it to exercise jurisdiction over heads of state, that position is, for the ICC’s purposes, CIL. It is CIL in the ICC, even though other actors in other settings reject it. When the Dutch and UK governments use the unable-or-unwilling position to explain to their national parliaments that the anti-ISIS operation in Syria is internationally lawful, that position has CIL effect. And so on down the line. The point is that this contingent, variable, and still contestable normative material is a kind of CIL. Or at least: in the settings in which it is treated as CIL, it is not distinguishable from the stuff that is “really” CIL. Recognizing this point is critical if we are to understand and advise people on the real-life operation of CIL.
Finally, Alonso takes issue with my suggestion that CIL’s non-rule-like qualities help to limit bias in international law. Alonso uses the example of the development of the law of the sea to show that, in CIL’s competitive “marketplace,” powerful states prevail on their preferred positions (in this example, on the continental shelf) much more easily than countries of the Global South prevail on their preferred positions (on what became the exclusive economic zone). Alonso thus concludes that “[a]bandoning the rulebook conception is not enough” to protect the Global South from oppression “because CIL is not a leveled playing field.”
I wholeheartedly agree. Alonso is also right that more work should be done to assess when and how the Global South can use CIL’s malleability to their advantage. But in my view, discarding the rulebook conception is a necessary first step. Precisely because the Global South operates at a structural disadvantage, the constant push for CIL to be more rule-like—for it be more or less the same for everyone, everywhere—will systematically work to the detriment of the Global South. These states less often expend the resources thinking through how a nascent norm might affect them, pushing hard for their preferences, or resisting the positions of the more powerful. Those positions get entrenched as CIL when analysts, convinced that CIL must be rule-like, “find” the rule on the basis of the only evidence that is available to them. Recognizing that CIL is more fluid and contestable than the rulebook acknowledges would preserve space for the forms of active and passive resistance that Alonso describes. Again, this is not enough; there might be circumstances in which having a norm that operates like a rule works to the benefit, rather than to the detriment, of the Global South. The answer would be to find ways to entrench that CIL norm, not to assume that all CIL norms must be rules.
So, where does this leave us? In the end, I think everyone can agree to discard the rulebook conception. Just kidding. I know that many people still cling to it. But the case for it has not yet been made—and, in my view, cannot be made. The sooner we come to grips with CIL’s contingent and fluid nature, the better equipped we’ll all be to engage with it and evaluate when, why, and how it might be worthwhile.