This chapter is about semantic annotation, discussed from formal and computational perspectives. Annotation is viewed as a scholarly technique or methodology. This chapter explains what annotation was in general and is now, how annotation has gradually developed and become applicable to the automatic building of larger data in language, and what applications semantic annotation aims at and what principles govern the modeling of semantic annotation schemes. There are two governing principles discussed: the partiality and situatedness of information to be annotated. I also mention the use of machine learning techniques for the automatic annotation of language data, which is represented either in a textual form or in a graphic form.