68 The Language of Fashion
fashion clothing we have chosen, from the outset, ‘an homogenous
level of description’, to which we have tried to hold as rigorously as
possible; the pertinence chosen is that of semantics;^7 we have decided
to look at contemporary Fashion clothing from the point of view of the
meanings that society attributes to it, to the exclusion of all other points
of view.^8
Pertinence is at first glance a costly operation since the analysis
which accompanies it has to exclude from the phenomenon studied
a large number of factors which seem quite reasonably to be crucial
parts of it: for example, we are taking Fashion here as if this institution,
whose signifying element alone we are considering, lived free of charge
amongst humans and without any economic, sociological, or historical
basis. Everyone knows however that the Fashion phenomenon is linked
to a certain economic gap within societies, characterized generally by the
need to sell an object (clothing) at a rate which is faster than its wearing
out; and that in the renewal of Fashion models, in their organization
and dissemination, there intervene elements for which psychosociology
alone can account;^9 and, if the contents of Fashion cannot be directly
attached to the contents of history at the level of event, as Kroeber
and richardson have shown, then the phenomenon itself could not be
explained except by recourse to a specifically historical, mental category:
there is a definite link between Fashion and history at the structural
level. By overlooking these categorically fundamental determinations
(and considered fundamental especially today), semantic pertinence
seems to be undermining the most arrogant of explanatory principles in
contemporary social sciences: the principle of totality.
however, it appears that in applied research the totality principle
has a fairly low level of payback; either it remains theoretical (as
in the sociology of G. Gurvitch) or else, whilst claiming a totality,
research in fact is beholden to a particular pertinence (socioeconomic,
phenomenological or psychoanalytical); and so the desire to exorcise
the myth of determinism in recent years has been a very pious one.
So it is better then not to set totality directly in opposition to the
pertinence principle and to let this principle develop freely with all these
consequences: perhaps then we will see that it is less costly than one
thought and that it has a better chance of returning, in its own way, to
the totality of a phenomenon than if it performed its analysis, in however
voluntary a way, from a single point of view.