Towards a classification of knowledge and a «maintenance» of the training systems for psychologists: notes on current critical issues and possible alternatives