Issues of POS Tagging of the (Diachronic) Corpus of Czech : Preparing a Morphological Dictionary

Many important decisions concerning the part-of-speech categorization remain unexplained in the current practice, only reported in corpus manuals. The aim of this paper is to offer a different perspective on the problems of morphological annotation of corpora – the perspective of mapping and analyzing conceptual problems in the annotation. Focused mainly on function words in Czech, we discuss the possibilities of the POS tagging of the inherently ambiguous category of particles and we introduce criteria for distinguishing particles from interjections.