Mainly focused on solving abstraction problems, programming paradigms limit language expressiveness, thus leaving unexplored natural language descriptions that are implicitly expressive. Several authors have developed tools for programming with a natural language subset limited to specific domains to deal with the ambiguity occurring with artificial intelligence technique use. This article presents a review of tools and languages with naturalistic features and highlights the problems that authors have resolved and those they have not addressed, going on to discuss the fact that a “naturalistic” language based on a well-defined model is not reported.

While the use of natural language for software development has been proposed since the 1960s, it was limited by the inherent ambiguity of natural languages, which people resolve using reasoning in a text or conversation. Programming languages are formal general-purpose or domain-specific alternatives based on mathematical formalism and which are at a remove from natural language. Over the years, various authors have presented studies in which they attempted to use a subset of the English language for solving particular problems. Each author approached the problem by covering particular domains, rather than focusing on describing general elements that would help other authors develop general-purpose languages, instead focusing even more on domain-specific languages. The identification of common elements in these studies reveals characteristics that enable the design and implementation of general-purpose naturalistic languages, which requires the establishment of a programming model. This article presents a conceptual model which describes the elements required for designing general-purpose programming languages and which integrates abstraction, temporal elements and indirect references into its grammar. Moreover, as its grammar resembles natural language, thus reducing the gap between problem and solution domains, a naturalistic language prototype is presented, as are three test scenarios which demonstrate its characteristics.