In reply to an article in this journal by A.D. Armand "Landscape as construction' (volume 120(2), 1988), the authors stress the need for a clear distinction between the scientific and commonplace concepts of landscape. Numerous examples are used to show the objective nature of natural territorial complexes identified by landscape research. It is emphasised that the isolation of landscapes is an extremely complex procedure, demanding, first a thorough knowledge of natural interdependences, second proficiency in the methods of field and laboratory research and, third scientific intuition and a special perception of nature. Further objectivisation of landscape research is possible with the study of natural territorial complexes as spatial-temporal systems. -translated by P.Cooke