A host of methodological problems attends to the study of the history of clothing, particularly its various descriptions of internal (qualitative) developments. Conventional historical treatments are incapable of answering such basic questions as “precisely when does an article of clothing change in function or form?”. Periodization is proved to be too narrow, and many researchers have wrongly approached dress as a purely historical event, in the hope of adumbrating the spirit of the times. Whatever its primary origin and function -such as protection or decoration — , clothing constitutes a system. That is, every article of clothing is subsumed under a formal, normative system instituted by society, In Saussurian terms, the clothes of an individual -or even his/her hair style— are the equivalent of parole, while the wider system to which they belong -the rules governing sartorial combinations and uses, the constraints and uses, degrees of permissiveness and toleration- constitutes the underlying iangue. Fashion (mode) is thus always a “fact of dress as a system” (costume). But parole can turn into Iangue , e.g. when a social group collectively adopts an individualized type of dress – which often happens in men’s fashion; conversely, Iangue can turn into concrete parole, as often occurs in female haute couture. In post-Darwinian terms, individual habillement corresponds to a biological organism, whereas costume to a species. As a system, dress (costume) can and should be analysed both syn chronically and diachronically. To return to my opening reservation: dress, though a product of history and especially susceptible of change in periods of turbulence, should be studied above all in the light of values that are oppositional and structured as social models.