The idea of the empirical cycle of research is to conduct controlled experiments – not only to explore unchartered terrain but also to test concrete predictions inferred from theories or models (de Groot, 1961/1969). This method has roots in physics, is characteristic of experimental psychology, and is the dominant method in cognitive neuroscience. Although experimental data are often multi-interpretable, they can be used as evidence for or against ideas and assumptions in theories or models.

In Part I, I sketched how SIT's coding model developed in interaction with empirical research, and here, I go into more detail on the regularities that may be exploited to arrive at simplest perceptual organizations. In Chapter 3, I discuss a formalization which establishes the unique mathematical status of the hierarchically transparent and holographic nature of the regularities that are proposed to be visual regularities, that is, regularities to which the visual system is sensitive.

This formalization belongs to the theoretical cycle of research, but I prefer to present it here because, as I discuss in Chapter 4, it leads directly to a quantitative model of the detectability of single and combined visual regularities, whether or not perturbed by noise.