Rekonstrukcja i rewitalizacja zagrożonych wymarciem języków z wykorzystaniem narzędzi lingwistyki komputerowej

Abstract

RECONSTRUCTION AND REVITALISATION OF ENDANGERED LANGUAGES USING COMPUTATIONAL LINGUISTICS TOOLS The monograph presented here consists of an introduction, six chapters and a conclusion. The first chapter briefly discusses the current linguistic situation, paying particular attention to the languages with a small number of speakers, which are threatened with extinction, and provides a thorough analysis of reasons for the disappearance of such languages. In the following chapters, namely chapters two, three and four, the authors reviewed the current linguistic situation in the western part of Europe. They briefly discuss selected languages spoken in this area which are threatened with extinction and belong respectively to the Germanic (second chapter) and Romance (third chapter) language groups. The fourth chapter discusses the languages in danger of extinction in Europe which belong to other groups of the Indo-European language family, with particular emphasis on the Celtic language group. The fifth chapter contains a description of the syntactic structure generator of Norwegian, developed by the authors. This kind of software may serve as a model for building analogous systems for other languages in danger of extinction. At this point, it is worth mentioning that the Norwegian language chosen by the authors exists in two official variants. One of them – New Norwegian (Nynorsk) – is currently perceived as a language potentially threatened with extinction. On the other hand, the sixth chapter of the monograph presents further prospects for the development of the system elaborated by the authors. In particular, the possibilities of its evolution into Machine-Aided Human Translation software have been considered. The task of this software would be to support the process of translating texts into the endangered languages. It is important to mention here that, thanks to the use of machine learning techniques, especially deep learning artificial neural networks, the issue of computer translation has already been solved in a largely satisfactory manner in the general sense, which not so long ago seemed like a scenario from a science fiction novel.