130 The Language of Fashion
from sociology. Indeed, in the early period of the 1960s Barthes was
moving swiftly away from sociology, despite references to Durkheimian
method.^15 Carter (2003: 152–53) rightly underlines the sociological
impulse to Barthes’s writings on clothing history in the late 1950s,
especially in the rejection of an evolutionary view of clothing forms. But
Barthes now, in the early 1960s, wanted to leave behind the Gurvitchian
sociological method he had defended in the early writings on clothing.
rather than seeing social phenomena in relation to the sum of
human individuals, Barthes now insisted that clothing showed ‘the
privileged example of a completely pure dialectic between the individual
and society’, as he put it in his 1962 piece on dandyism (Chapter 6
here). This seemed to be at odds with Gurvitch’s totalizing sociology,
and is perhaps a central distinction between a sociology of clothing and
a semiology of fashion.
So nairn’s strong reservations on Barthesian formalism in The Fashion
System point then to a slippage operated by a tactical separation of
semiology and sociology; it was a gap which allowed a structuralist
optic, with the formalist and functionalist analyses that this entails, to
dominate in The Fashion System. This growing formalism in Barthes’s
work on fashion explains perhaps why the ‘rhetorical’ analysis, the
ideological critique of fashion, is relegated to the end of the study, and
has little of the social and political engagement evident in Mythologies.
however, as always in Barthes’s work, discarded or relinquished
positions always return at another, higher point in the spiral. Perhaps
aware of the formalist, even empty, nature of his magnum opus on
fashion—and influenced by the seismic radicalization of may 1968,
which also exposed structuralism’s technocratic tendencies—Barthes
then moved in the post-1968 period towards much less formalist
analyses of clothing.
But the main point that Barthes seemed to be making at the very end
of the early preface to The Fashion System—which had distinguished
semiology very clearly from sociology—was that semiology can (should)
be used by sociology, by political critique and by ideological analysis.
here is the prelude to Barthes’s (briefly held) view that semiology was a
‘meta-language’, the discipline that trumped all other disciplines because
it recognized its and every other discipline’s status as language. This
was an idea to be heavily criticized by henri Lefebvre (see Sheringham
2005: 305–6), and on which more in a moment.