The answer to the question which phenomena should be seen as a disease changes over the course of history. This article discusses, on the basis of the historical development of the concepts of arthritis and arthrosis, that the painful joint conditions that we now call 'arthrosis' have been viewed, and - importantly - treated, over the past 150 years, as both aging versus disease, and as systemic versus local. The specialists who were most involved - rheumatologists and orthopedic surgeons - followed different explanatory models and scenarios for treatment. Over the course of the 20th century, arthrosis emerged as an independent clinical entity, which some concluded may be the result of a normal aging process. This discussion leads to the idea that diseases do not have an essence separate from diagnostic and therapeutic practices.