The traditional classification of adverbials into local, temporal, modal, and causal, such as those used in our syntaxes, and their applications for schooling are suitable for these purposes. For empirical processing under the Functional Generational Description, which also serves, among other things, the tasks of automatic language processing (especially for machine translation, but essentially for the translatological problems associated with the teaching of Czech as a foreign language), a finer classification seems necessary.
On the analysis of the material of selected adverbials and in terms of the problems that accompany their solution, we propose a way of capturing them.