The past two decades have witnessed a large number of proposals for the classification of periodontitis. These proposals are all founded in an essentialistic disease concept, according to which periodontitis is a link between the causes and the signs and symptoms of periodontitis. Essentialistic definitions are necessarily rather imprecise and thereby subject to multiple interpretations. Consequently, it remains unknown to what extent current knowledge regarding 'different' forms of periodontitis is based on the 'same' type of patients. However, periodontitis is a syndrome, the clinical manifestations of which may come in all sizes. Thereby, periodontitis has no diagnostic truth, just as there is no natural basis for a sharp distinction between health and disease or between 'different' forms of periodontitis. Recognition of these facts and adoption of a nominalistic approach to the definition of periodontitis is needed to provide a rational framework for the development of a classification system that meets the needs of both clinicians and scientists.