As a recent discipline, computer science, and program-ming language research in particular, have so far eluded the eyes of philosophers of science. However, we can gain interesting insights by looking at classical works in philosophy of science and reconsidering their meaning from the perspective of programming language research.

This is exactly what I attempt to do in this essay – I will go through the established theories of science and look what they can say about programming language research. Then I suggest how we can improve our scientific practice in the light of these obser-vations.

First, I discuss how understanding the research programme is important for evaluating scientific contributions. Second, I argue that overemphasis on precise, mathematical models in early stage of research may limit the creativity. Thirdly, I propose how to design stand-alone (theory-independent) experiments in program-ming language research and how this can help to integrate the vast amount of knowledge gathered by software practitioners.