This project examines images (as opposed to texts) as a medium of technical/scientific communication. Our research will focus in particular on the logic and argumentative power of visual images – by this we mean graphical, schematic images but also images produced by sensors and computed images – in the natural and engineering sciences. Based on historical graphical and visual representational techniques (primarily the images in Diderot’s Encyclopedia), we will analyse technical images, ranging from graphics to analogue (autographic) imaging processes to computerised, digital imaging processes. We consider these to be essential argumentative tools in modern engineering disciplines and the natural sciences, and will compare them to text-based logics. The use of programmed processes in the analysis of complex sensor data is giving rise to new forms of argumentation (in medicine, the natural sciences and engineering). This is challenging both the written form of logic in the sciences and its definition as a formal logic in mathematics and the natural and engineering sciences; indeed, it even goes as far as to challenge Sir Karl Popper’s »Logic of the Social Sciences« and the theory of science. Based on new developments in image generation in computer science, we will present and analyse these transformations from a transdisciplinary perspective.