This paper briefly summarizes the progress of scientific thinking in the course of 20th century focusing on the emergency of system theory and its implications for urban design research. This is done in the quest for the verified method and conceptual framework applicable to use when dealing with a complex and elusive problem such as a city is. It was observed, that when dealing with questions concerning a city we rarely start from a point in which we define what kind of a scientific problem a city poses. Not doing so we mostly put forward definitions that don’t indicate scientific method appropriate to use when exploring this phenomenon. It exposes us to a danger that we might unconsciously use methods which can’t lead us to correct answers. The nature of a problem, not our preference for certain way of handling it, should guide our decision.
There are three basic types of scientific problems: problems of simplicity, problems of disorganized complexity and problems of organized complexity. They can’t be approached in the same way. Problems of simplicity are problems of two variables for which the methods that rely on linear causality can be used. Problems of disorganized complexity are problems of many variables behaving in “haphazard” way, and the questions they pose, can be answered by the techniques of the probability theory and statistics. Problems of organized complexity involve several variables whose behavior is interconnected, all are varying simultaneously but nevertheless they create an organized whole. Before the emergence of system theory in the first half of the 20th century, science was not able to successfully deal with these kinds of problems. City, too, happens to be a problem of organized complexity and therefore it is wise to approach it as such.