Page 125 - Culture Media Language Working Papers in Cultural Studies

114 THE IDEOLOGICAL DIMENSION OF MEDIA MESSAGES

colour discrimination, faithfully serve under her flag’. Barthes does not make it
clear why this second-order meaning, myth, is different from, rather than a
special case of, connotation. We would like to suggest that the difference
between myth and connotation depends on the amplitude of the lexicons from
which the concepts are drawn. The connoted meaning in ‘pig=policeman’ and in
‘pig=male chauvinist’ are clearly linked to the lexicons of identifiable sub-
groups. By contrast, myth seems identifiable with the lexicons of very large
groups, if not of the society as a whole. Myth therefore differs from connotation
at the moment at which it attempts to universalize for the whole society
meanings which are special to particular lexicons. In the process of
universalization, these meanings, which in the last instance are particular to certain
lexicons, assume the amplitude of reality itself and are therefore ‘naturalized’.
Thus, we might say, myths are connotations which have become dominant-
hegemonic.
12
In WPCS 3 we emphasized that the ideological level always refers to the
connotative aspect of the message. This was one of the strong criticisms
advanced by Terry Lovell in her review of WPCS 3 in Screen. 13
Some misunderstanding here can be attributed to our failure to explain clearly
enough how we were using the concept of ‘denotation’ (see pages 133–4 below).
The concept of ‘denotation’ was not sufficiently clarified. By ‘formal-
denotative’ we were following the argument by Barthes in Elements in
Semiology, where ‘denotation’ is not given a special status as ‘natural meaning’,
but simply refers to the first system of signification which generates a second
system ‘wider than the first’ (which is the plane of connotation).
In part, the problem is to understand precisely what is meant by ‘level
of signification’. By referring to a ‘formal-denotative’ level, we were employing
the term as an analytic concept, useful for distinguishing between different levels
of the organization of meanings. Veron, for example, has observed that ‘ideology
is a level of signification which operates by connotation’. Because of our lack of
clarity on this point, Lovell assumed that we therefore subscribed to the idea that
‘denotation’ represented a pre-ideological or ‘neutral’ state of the message. But,
in our view, the denotative level cannot be identified with a ‘neutral state of
language’: there can be no ‘neutral state’ because denotations also must be
produced by the operation of a code. To distinguish between different levels of
the operation of codes is not, therefore, to imply that messages can be produced
without a code (see pages 133–4 below).
This point has been subject to further confusion because in the texts which
followed Elements of Semiology (and to some extent already in Writing Degree
Zero) Barthes appeared to subscribe to the notion of a ‘zero degree of writing’
14
and to the idea of an ‘empty text’. But whatever the metaphorical status of
these concepts, we cannot subscribe to the idea that there is a level of ‘denoted’
meaning which is free of any ideological operation. In this sense, ideology is
beyond and involves the whole universe of the sign as such—denotative and
connotative. It is inside the coded sign that an analytic distinction can be usefully