Typology, as a universal method of interpretation, is not used in art history to the same degree as in some other branches of science (biology, archaeology, psychology, linguistics et al.). It is hard to imagine the theory of art history encompassing a kind of taxonomy - general principles of classification which could also include typology. Nevertheless, typological inferences and issues are widely used, albeit not always theoretically defined. Typology is unavoidable when it is necessary to classify archaeological material that could be interpreted as artefacts. It is used to investigate and describe groups of artworks or other objects relevant to art history. The concept of a theoretical model with fixed or substantial traits (attributes) seems the most appropriate in this case. Attributes are important only from the aspect of the aim of a particular work of research. This does not mean that typological models are arbitrary mental constructs or an ideal generalisation of metaphysics. On the contrary, 'types' should be tied to the concrete phenomena of research. World art history, as well as Latvian art history, is full of generalisations that could be regarded as the products of typological research. Their verbal exposition can be supplemented by graphic schemes (more common in texts on architecture). We can recollect the typology of classical orders that have been described and shown in drawings in countless reference books, the typology of the kuroi and korai statues in the Ancient Greece, or the medieval type of the 'beautiful Madonna' and many others. Two different examples from Latvian art history can also be mentioned: Paul Kampe's 'central type of church building in Vidzeme' and Tatjana Kacalova's examination of the types of compositional structure in the landscapes of Vilhelms Purvitis.