IR scholarship fails to entertain the possibility that recourse to IL in the process of actualising interests and normative beliefs could have a distinguishable effect on behaviour, or so the previous chapter showed. However, I submit that from IL’s causal dependence on prior interests or non-legal norms it does not follow that it makes no difference for behaviour whether or not actors’ instrumental or principled considerations are subject to law. By the same token, epistemic dependence does not mean that drawing on IL to make a strategic or aspirational argument has no distinguishable effect. Since the effects of compliance with IL on behaviour have so far not been an object of investigation in IR theory, this chapter undertakes the task of providing a theoretical account of how recourse to law can afford a counterfactual added value. It introduces a theory of IL’s behavioural relevance and discusses its implications for the distinctiveness of legal norms. The chapter concludes that IL is indeed ontologically dependent. It is dependent, yet separate.

How international law works: intellectual and motivational effects

Chapter 1 outlined Martti Koskenniemi’s influential explication of the structure of international legal argument and his contention that IL does not provide an independent standard for evaluating action. Koskenniemi’s view is exemplary for the position, implicit in much of IR scholarship on IL: IL’s dependence means that recourse to law makes no difference for behaviour. An actor might as well have followed her interests or normative beliefs directly. From his observation that IL has ‘no meaning that can be determined independently from power and or morality’, Koskenniemi infers that IL is merely ‘a means to articulate particular preferences or positions in a formal fashion’. This, in turn, grounds his conclusion that ‘legal arguments do not produce substantive outcomes but seek to justify them [italics in original]’. Over the following paragraphs I uncover and dispute the assumptions that lead to this equation of IL’s dependence with a lack of behavioural relevance. The aim is to show that, in spite of its threefold dependence, IL can have a distinguishable effect on behaviour – an effect that makes norms of IL distinctive.