A new approach to solving the problems of creating automated methods to build a semantic web for life science is considered: these key ideas about a semantic basis for a scientific knowledge system language would allow the gap between ideas represented by natural and formal languages of different mathematical disciplines and computer science to be bridged.