According to modern concepts, the idea of a paradigm is associated with a certain common approach within the natural sciences in description of a large class of processes, objects, systems. If we’re talking about a global paradigm, this means such (global) paradigm should cover the huge classes of objects in nature and society. Today we highlight three global paradigms throughout modern science, which cover three global clusters of all processes and objects of living nature and inanimate nature. W. Weaver was first who proposed such statements in 1948 in his famous publication “Science and complexity”. However, during nearly 70 years almost no one paid any attention to these statements (though he spoke very simply about the most important things). W. Weaver has divided all objects and systems in nature into three giant clusters: the simplest systems (simplicity), which are described now in the framework of deterministic theories and models, unorganized complexity (stochastic system) and system of third type — organized complexity (organized complexity). Under the system of the third type he understood all living systems, but no special features in their organization has been identified and learned by W. Weaver. It is now clear that he could not do the study within the framework of modern science because this requires a different (third) paradigm and different science.

The discussion of crucial problems of psychophysiology and natural sciences is presented, in general: are there any recurrences of measured values of psycho­physiological parameters (functions) of a person. It is shown that it is impossible to obtain arbitrarily a repetition of samples for tremor and electromyogram (when trying to maintain static muscular tension), statistical distribution function of samples f(x) is continuously changing. This dynamics is typical for tremor (conditionally involuntary movement) and electromyogram when the higher nervous activity of individual oversees the muscular effort. The involuntary movement (tremor) shows the number of coincidence of electromyogram samples (at unrestricted muscular tension) k, which is similar to such in the matrices of pairwise comparisons of tremor samples. Shannon’s entropy in this case does not change significantly.