On the Validation of Computer Science Theories,

We address normatively the demarcation problem for Computer Science How can Computer Science research be conducted scientifically First we attempt to delimit the subject matter of Computer Science, and conclude that it is not computers but programs. Since programs are not physical objects, it is difficult to see how they can be studied empircally. The rest of the paper is devoted to an explanation of how this can be done. This method is first illustrated by a hypothesis of narrow scope, analogous to a physical law. Next it is illustrated by a theory of wide scope - the Turing Machine model of computers. The approach is summarized in the conclusions.