This article continues the study of the problem of systemic reconstruction of legal norms (See: Voronin, M. V., Avramenko, A. V., Belous, A. A. (2022). Systematic reconstruction of legal norms: needs, methodology, prospects. Legal world, 9, 30-38).
 The aim of the present article is to develop and describe a method of systemic reconstruction of legal norms in its linguistic aspect to solve the problem of converting legal texts in natural language into a database. The advantage of the method is a comprehensive interdisciplinary approach that combines jurisprudence, linguistics and subject matter expertise. This approach ensures high quality of the converted data.
 Drawing on the case of The Good Manufacturing Practice (GMP), the substantial factors that make it difficult to interpret unambiguously the regulatory material are analyzed; with these factors in view, a model of a formalized wording is proposed. The model makes it possible to present normative material in a standardized form and to combine it into a database. Based on the reconstruction experience of the Rules of GMP, lexical and syntactic transformations of the source text, as well as typical semantic tasks that the expert solves at the synthesis stage are described. The article shows the effectiveness of transformational grammar rules at that stage to achieve semantic equivalence of the original and transformed textual information within the framework of the described method.
 The method of systemic reconstruction in its linguistic aspect is significant for further work related to improving the language of technical regulations.