It should be obvious that the autonomy of science is a mirage if it is supposed to state a requirement that the agents of science, individuals and organizations are to lay the constitutional rules in order to regulate their activities by themselves and for themselves. Science will remain ipso facto fundamentally heteronomous, as long as it is conducted in a territory where the fundamental rules are laid by the agents of a state that controls violence – unless the agents and the scientists are identical, something that is largely only a theoretical possibility. The scientific conventions, the moral rules of science and the scientific techniques, can help maintain a domain where informal rules will provide decisive normative guidance to the participants to the game of science, but the right to pursue scientific activities can ultimately only be granted by the state. Science can never be entirely self-governed. The constitutional question is to determine the extent and specification of this right and the philosophical task consists in the provision of arguments to answer this question.