The chapter maintains that different forms of democratic constitution making are susceptible to authoritarian dangers and deformation to quite different extents. Nevertheless, I argue, that because of the significantly path-determined nature of the specific methods adopted, normative preference cannot always lead to the utilization of what is deemed best or even better. I try to make the case by analysing and comparing the major assembly forms of constitution making: extra-ordinary constituent assemblies, ordinary legislatures, American-type conventions, and round tables linked either to ordinary legislatures or to limited constitutional assemblies. I next try to show that even when the political path of transformation does not favour the adoption of the best paradigms of constitution making, it is possible to learn important things from both legitimating principles and specific mechanisms of the latter. I conclude with some reflections on the role, both negative and positive, of international actors in the process of constitution making. I distinguish here between types of intervention and the levels of influence on domestic actors.
The chapter considers the strong yet problematic relationship of revolution and constitution in the modern world. Relying on Hannah Arendt, it affirms both a historical relationship as well as a logical “elective affinity” between revolution and constitution. Nevertheless, I argue that if interpret constitution in the normative sense, i.e. as some version of constitutionalism, then revolutions more often than not culminate in the contrary: in forms of institutionalized (rather than just temporary) dictatorships. According to the argument, this too is a logical and not merely a historical relationship, although not absolutely necessary or unavoidable. Using the lessons of post-revolutionary democratic transitions in South Africa and Central Europe, and considering contemporary examples such as Colombia and Tunisia, I argue that it depends on the form of constitution making whether or not the dictatorial logic and outcome can be avoided in revolutions or quasi-revolutions. In particular, the replacement of a single stage model relying on a sovereign constituent assembly, should be replaced to whatever extent possible by multi stage approaches in which no agent can be considered sovereign.