This paper discusses applying corpus linguistics principles to compile text corpora reflecting any special knowledge domain. Hence, the special feature of semiotically correctly assembled text corpus as reflection of extra-corpus reality is under discussion. The two reciprocally balanced approaches of corpora generation are described. The first is deductive approach. It means a methodology based on the movement from general description of corpus to the program based algorithm of proper text selection. The properly balanced methodology of these two logical principles could be taken from so called object oriented programming. So, the OOP methodology of reflecting outer world reality in a class construction procedure can be applied to the corpus of texts reflecting any special knowledge or linguistic domain.
展开▼