This book is an essay on relevant problems of epistemology (the theory of knowledge) related to computer science. It draws a continuous line between the earliest scientific approaches of epistemology, starting with the Greek Classics and the recent practical and theoretical problems of computer modelling, and by that the appropriate application of computers to our present problems. Uncertainty, logic and language are the key issues of this road leading to some new aspects of cognitive psychology and unification of the different results for a modelling procedure. The book is not a textbook but a critical survey of usual and advertised methods with an evaluation of them from the point of view of their applicability, reliability and limits. Probability, Bayesian, Dempster-Shafer, fuzzy and other approaches are treated in this way in uncertainty, different worlds' concepts, non-monotonic logic and other methods and views in logic. The emphasis in linguistics is put on the meta concept, and in cognitive applications of the pattern concept.Written mostly in an entertaining style, this book provides a more palatable reading of a profound subject.