Abstract
Many philosophers have used the concept of belief system or some related notion as a basic tool of epistemological discussion and analysis. A belief system is a set of propositions or statements which represents a person’s doxastic state or credal state in a certain situation; it consists of the propositions which the person either explicitly or implicitly accepts in the situation. One of the many concerns of epistemologists is to attempt to formulate general “conditions of rationality” for belief systems. I want to suggest in the present paper that in this endeavour philosophers usually treat belief systems as if they were artifacts or tools made for various epistemic purposes: for providing satisfactory answers to interesting questions, for helping to find answers to new questions, and for providing resources for argumentation and research. Insofar as epistemology and philosophy of science can legitimately be viewed in this way, they may be regarded as “sciences of the artificial” in Herbert Simon’s sense, and not as sciences of the “natural world” or parts of natural science, as some proponents of “naturalised epistemology” have recently suggested.