Abstract
The aim of the paper is to analyze how language affects scientific research, from planning experiments and interpreting their results, through constructing models and the testing their predictions, to building theories and justifying their principles. I try to give an overview of the potentialities of language of science. I propose to distinguish six potentialities: analytic, expressive, methodical, integrative, explanatory, and constitutive power of language. I will shortly characterize each of these potentialities and illustrate their contribution to scientific research. Although I believe that the theory is valid for a wide range of scientific disciplines, all illustrations are taken from physics.