Cartographic generalization yesterday and today

Abstract The author presents evolution of views on cartographic generalization since it was defined by Emil von Sydow (1866) until today. It is divided into three chapters which present the evolution of views on cartographic generalization, models of generalization and digital generalization, respectively. Views on the topic of generalization evolved in the direction of broadening the term itself and towards a different perception of its nature. Originally generalization was understood as a process which can be performed on maps only. Now the prevailing understanding is that it begins earlier, at the conceptual stage of map making. Determination of the method of contents’ presentation is an indication of such generalization. The character of generalization is another important aspect of the discussion on its nature. The notion of a subjective nature of generalization, expressed, among others, by Max Eckert (1921), was originally predominant. Later there also appeared different opinions, allowing its objectivization (K.A. Saliszczew 1998). This direction helped to result in automation of the process of generalization of map contents. Currently a dualism in perceiving generalization can be observed, with a strong bias towards its objective aspect. In a separate chapter the author discusses conceptual models of generalization proposed by: L. Ratajski (1967, 1973), J. Morrison (1974), B.G. Nickerson (1988), K.E. Brassel and R. Weibel (1988), as well as R.B. McMaster and K.S. Shea (1992). They are divided into the universal models of theoretical character and those constructed for the purpose of computer automation of the process. Attempts at digital generalization which currently develop in the context of generalization of general, and especially topographic maps, are discussed separately. Most important algorithms concerning generalization of linear objects are presented chronologically, concluding with a description of comprehensive generalization systems. The summary presents two main conclusions. Firstly – work on generalization will continue to consider the geographical context during the process. Secondly – generalization of thematic, and especially statistical maps is the prospective direction.