The paper analyzes the shortcomings of existing models of text documents, it proposes an uniform content model of text that formed on the basis of synthesis of logic and linguistic models of its sentences. The work shows basic steps of the proposed algorithm for constructing content models of text