Les crimes contre l’humanité, les crimes contre la paix, les crimes de guerre, le génocide constituant le noyau dur des crimes internationaux représentent les actes qui causent les dommages les plus graves à la sécurité humaine et internationale. Ils sont codifiés dans les statuts des tribunaux pénaux internationaux. Les crimes transnationaux ou les crimes de traités ne sont pas codifiés dans ces statuts ou sont codifiés d’une manière assez limitée. L’apartheid, la torture, l’esclavage, la piraterie maritime, le terrorisme sont les exemples de ces crimes. Selon les règles de transformation du système international de Morton Kaplan, une variable du système international cause des transformations ou joue un rôle dans l’essor des systèmes internationaux. Le noyau dur des crimes internationaux est abordé dans notre travail comme l’une de ces variables du système. Dans ce but, le système bipolaire (1945- 1991), le nouvel ordre mondial (1991-2001) et le nouveau système depuis les attentats de 11 septembre 2001 ont été analysés. Les Tribunaux Internationaux de Nuremberg, de Tokyo, pour l’Ex-Yougoslavie, pour le Rwanda et la CPI sont devenus les études de cas. / Crimes against humanity, crimes against peace, war crimes and genocide constitute the hard core of most threatful acts to human and international security. They have been codified through the statutes of international criminal tribunals. Transnational crimes or treaty crimes are not codified in these statutes or they are narrowly codified. Examples of such crimes include the apartheid, torture, slavery, maritime piracy, and terrorism. According to the rules of international system transformation as defined by Morton Kaplan, one variable of the international system entails transformations and plays a role in the emergence of international systems. In this study, the hard core of international crimes is considered as one of the variables of the system. Thus, the bipolar system (1945-1991), the new world order (1991- 2001) and the new system after September 11 attacks are analysed. International Tribunals such as Nuremberg, Tokyo, for the former Yugoslavia, Rwanda and the International Penal Court constitute case studies.

International crimes prosecutions have become more common since 1993, both domestically and at international courts and tribunals. The advance of this norm confronts realist state interests causing debate about the norm's status. Kathryn Sikkink views a norm as cascading when enough states adopt it to cause international influence, without domestic pressure, to procure levels of conformity. This thesis considers the degree of conformity by observing the level of case selection independence to determine whether this norm is cascading. By identifying the jurisdictional and functional elements of case selection independence, I develop a framework for observing the interface between politics and law. While Sikkink errs towards the quantity of international crimes prosecutions, I focus on the quality. This project examines case selection independence at the International Criminal Tribunal for Rwanda, the Special Court for Sierra Leone and the International Criminal Court, in Uganda. The project considers whether case selection has become more or less independent at these courts - whether the norm of international crimes prosecution has cascaded or contracted. In observing the various case selection independence elements I attempt to explain the observed cascades and contractions at each court. I then consider whether a cascade or contraction occurred during the period of the courts' collective design and function. The research qualitatively observes a cumulative justice contraction. The research observes a combination of factors affecting case selection independence, including shifts in power dynamics between and among weak and powerful states, increasing state sophistication in international court engagement, a shift in jurisdiction triggering actors and forums, and realist state co-option of norm entrepreneurs via endearing explanation of independence-diminishing policies.