We generally expect the modern state to take the fortunes of society in hand and to lead it, by judicious steering, to a better future. Political science has devoted a substantial branch of the discipline to this question with its research into policy and its implementation. This research constitutes the third great pillar upon which political science supports itself, the other two being the investigation of the political community as a form of human group-life, and research into politics as a struggle for power between the parties and interest groups (Scharpf, 1989, 11-12; von Beyme, 1991: 90-144).
Over the years, this concept of the state as the center of society's political regulation has been exposed to considerable doubt; the question being whether the state is remotely capable of acting out this role, and whether a social theory which takes the complexity of modern societies seriously can reasonably expect the state to take on such a role. These doubts arise mainly from the perspective of Niklas Luhmann's theory on autopoietic systems, which has, in the meantime, become the prevailing paradigm (1981, 1984, 1986, 1988, 1989). In its wake, a theory of decentralized contextual control has been developed, which places its hopes in the ability of society's so-called differentiated functional systems to regulate themselves (Willke, 1983, 1984, 1987a, 1987b; Teubner and Willke, 1984; cf. also Glagow, 1984; Glagow and Willke, 1987; Mai, 1990a, 1990b; Berger, 1991).
Here, I intend to show that neither the traditional nor the modern understanding of political regulation does justice to the complexity of the phenomenon and that both operate using strong simplifications. Indeed, the traditional understanding fails to appreciate that political control is not simply the result of a one-way relationship from the state to society, but that it occurs when the society's various functional systems act together. More recent systems theory, conversely, fails to take into account the fact that this collaboration does not occur within a system-environment relationship of autopoietically operating systems, but in the negotiating processes between representative actors in the interpenetration zones of the functional systems, in which reciprocal transformations of political power, influence, truth and money are carried out (Parsons, 1969; Munch, 1987, 1991: 135-283). The theory of decentralized contextual regulation is also denied access to these processes as a result of its heavy dependence on Luhmann's autopoiesis theory. Inevitably this theory imprisons itself and the processes it describes in the straitjacket of autopoietically observing and self-managing sub-systems: processes which in actual fact occur between these sub-systems.
Models of Political Regulation
A quick glance at the discussion on neo-corporatism and pluralism, and also the empirical research into policy and its implementation offers us numerous indications as to the accuracy of this thesis (eg. Schmitter and Lehmbruch, 1979; Lehmbruch and Schmitter, 1982; von Alemann, 1981; Berger, 1981; Cawson, 1985; Schmidt, 1988; Mayntz, 1980, 1983, 1987, 1988; Brickman, Jasanoffand Ilgen, 1985; Vogel, 1986; Jasanoff, 1986, 1987, 1990). In order to limit the scope of the question, I will concentrate on the control of technological risks currently being discussed under the heading of Risiko-gesellschaft (risk society) (Beck, 1986, 1988, 1991). Let us look at the example of the laws on the use of dangerous chemicals; laws, which were tackled in earnest in the 1970s in several countries, put into force and then executed within the same decade or in the early eighties (Brickman, Jasanoff and Ilgen, 1985; Schneider, 1985, 1988). In analyzing these processes of legislation and implementation, one should always bear in mind that they take place in the "cross-border" interaction between the society's political system and a series of its other functional systems and this also involves both an input from the bordering functional systems into the political system and an output from the political system to the other functional systems in society. …