Language variation aspects as object of corpus research

The article reviews the status of researches of different forms of language variation (temporal, genre-based or sociolinguistic variation) within the corpus approach, and of those homogenous methods which are based on the common principle of quantitative and comparative analytical study based on existing corpus material. In particular, the role of synchronous and monitor corpus in diachronic and synchronic research is explained. The choice of text-centered or human-centered approach to unit of variation is proposed to be viewed as a manifestation of statistic or organic corpus ideology adopted by such a research.