There is a fundamental, eternal and unresolvable conundrum at the heart of customary international law (CIL): we do not know on what we should base our arguments. The debates on the theory of CIL continue unabated, because there is a strong belief that we need it in order to keep international law working. But is that legal reason enough to consciously or subconsciously change the mechanics of customary law to suit these needs? The ILC’s CIL project is suffused with the spirit of pragmatism, but it could not avoid taking a stance on theory. On the other side are foundational critiques of CIL, like Jean d’Aspremont’s: custom is merely a story international lawyers tell themselves. Both methods have virtues and very dangerous vices and contain the seeds for their own destruction. One aim of this chapter will therefore be an effort to show the relative merits of these techniques from a theoretical point of view. My portrayal usually stops at the recognition that we cannot find the law which tells us the rules of CIL-making. In this chapter, I will attempt to go a step further to a new method for conceptualising this: the Approximatively Plausible Empowerment Norm.

Can a doctrinal scholarship which is based on the normativist-positivist framework espoused in this book do any better than mainstream legal scholarship? Starting from a critique of certain hyper-systematising approaches, the chapter introduces Kelsen’s idea of legal scholarship properly so called, one devoid of external influences because these make it impossible to correctly cognise the law. Reconstructed in this manner, doctrinal scholarship can provide a structural analysis of the law, both on the macro-level of system-coherence and on the micro-level as frame-determination. It can also give practical information on how awards have exercised their freedom within the frame of possible meanings.

Is arbitral investment case-law on expropriation precedential in a legally relevant sense? Orthodox approaches are marked by agreement on a narrow set of arguments, namely that international law is not a common law and arbitral awards do not have stare decisis power, that jurisprudence is hugely important and tribunals rely on it, and that there must therefore be a sort of de facto system of precedents in operation. In effect, ‘factual importance’ is fashioned into a source of legal authority. However, few arguments are given as to why this transfer from fact to law would occur and they do not provide a foundation for a general legal value for precedents. Yet the weight of arbitral jurisprudence is both too great to ignore and too helpful in discovering what ‘’ means in a pragmatic sense. Precedents are statements about general norms; outside the common law, judge-made law is merely an interpretation of a general norm in a judgment. Not even a constant tradition of decisions can turn such a statement into a norm.