The Writings of Roland Barthes 131Despite this (temporary) scientistic belief in the power of semiology
it is important to stress the ‘provisional’ status of many of Barthes’s
theories and methods, especially in his use of semiology in relation to
structuralism and to sociology. as with much of Barthes’s work, his
work on fashion emerged from teaching at the Ecole Pratique des
hautes Etudes in Paris, a fact reflected in the tentativeness with which
he presents his scientific results. For example, though a forerunner of
The Fashion System,^16 ‘“Blue is in Fashion This year”’, written in 1960,
shows Barthes making surprisingly regular self-reference—‘me’, ‘I’, and
not ‘we’ (something evident also in the early preface to The Fashion
System)—suggesting a rather cautious, pre-scientific approach to a
subject which was after all not what he had looked at so far in his
semiology, or in his analysis of clothing across history.
The article’s list of eighteen points suggests a rather modest and
clumsy approach, gesturing to the ‘note’ in the article’s subtitle. Barthes
was working in highly uncharted waters here (not even Kroeber managed
this kind of detail); and though ‘“Blue is in Fashion This year”’ clearly
picked up on the concluding remarks of his 1959 article ‘Language and
Clothing’, Barthes now appears slowly methodical in 1960, rather than
glibly essayistic. This may be the stirrings of his ‘little scientific delirium’,
a rather fastidious and calm search for a method and an object of study,
culminating in the (triumphant) thoroughness of The Fashion System;
but it was also the origins of what Jonathan Culler (1975: 35) calls
Barthes’s methodological ‘neglect’ in The Fashion System.
Culler’s critique of the method in The Fashion System seems to
revolve around Barthes’s implicit decision to abandon sociology in favour
of a structuralist use of semiology. In its linguistic analysis of what is in
fashion, Barthes chose or neglected, argues Culler, to suggest what the
‘functional distinctions’ were within any one fashion utterance. as Culler
puts it, ‘It does not follow that each descriptive term [in any fashion
utterance] designates a feature without which the garment would be
unfashionable’ (35). Similarly, Culler regrets Barthes’s restriction of his
corpus to just one year. If Barthes is interested in the fashion system in
general, surely more than one year should be analysed, argues Culler.
Furthermore, the oppositions in fashion that one finds between years—
say, large one year, or small another—cannot be analysed simply on
distributional grounds: they have a reality within fashion diachrony,
suggests Culler. Barthes’s retort no doubt would be that to include any