The Generation of Semantics in Natural Language and the Formation of Brain Intelligence

The foundation of life phenomenon is the abilities of representation, memory and behavior of a life form. Human natural language can describe and interpret this matter world and naturally support the subjective initiative of humans. Not as a computer system, the brain neurons can realize complex interconnections with each other in 3D physical space, such as in which no word is stored repeatedly, and form the matter base of consciousness activity. It is known that the radial basis function representation principle is widely used to represent color, sound, smell and gustation, this paper tries to use a quantitative analysis method to reasonably estimate the number of single wavelength color attributes being represented by our brain. It is also known that some simple neurons in the primary cortex are sensitive to the direction of short line, this paper introduces the representation principle of single basis inclination which can represent the inclination attribute of a shorter line, and our brain can further extract more topology attributes of a shape through introducing the concept of vision event that is the evolution foundation of cognition ability. By the mechanisms of semantics lumpability and semantics abstract our brain can construct and represent more complex semantics. This paper emphasizes that both the semantics in natural language and the original semantics only perceived by our brain exist in the neural fiber plexus nodes of brain, they are twins. We humans always tries to consider this world as a whole, just as at a certain moment the neurons activated by the brain attention form an expanded semantics, while the shifting of attention forms a consciousness stream, so the semantics and syntax can be rationally expanded, which can be represented through the semantics abstract mechanism, specially it is our behavior with the subjective initiative making us distinguishing from animals to have shaped the consciousness stream structure. This paper also underpins the principle of matter primacy. We also explain why the brain consumes less power. In current technique stage, humans endowing a machine or a system with intelligence is feasible. That developing a kind of computer language which can analyze the attributes of semantics and semantics abstract in physical world can implement a Digital-Twin system, and through a great deal of practical training and simulation, we can help a machine to construct a semantics vector-base which contain much many semantics and semantics abstract with a hierarchy structure, which concern with all the behaviors that occurred. Therefore, based on the semantics vector-base and applying the real-time planning technique, we can construct an intelligence automation machine or system such as a humanoid robot and an AI agent.