The article deals with a number of challenging points in the annotation of different categories (parts-of-speech, classes and subclasses) in the texts attesting the history of the Bulgarian language. The focus is on category changes which affect mainly the so-called function words (auxiliary verbs, coordinating conjunctions, subordinating conjunctions, pronouns, among others) in historical texts, with a goal of building a unified annotation scheme for the stages of the Bulgarian language. Annotation approaches followed by other projects will be discussed to arrive at practical solutions with a view of the theoretical assumptions and purposes of the data. Keywords: