But I am running ahead of my argument. The point I want to make about the liberal pluralist account of variation and difference is that it implicitly assumes a closed universe of readings, making up a contained diversity of audience groupings with definite identities, equivalent to the liberal pluralist conception of electoral politics where voters are distributed over a fixed repertoire of parties. It is this sense that liberal pluralist discourse conjoins the marketing discourse of market segmentation (where consumers are neatly divided up and categorized in a grid of self-contained demographic or psychographic ‘segments’), which is not so surprising given that both discourses are two sides of the coin of ‘democratic capitalism’. This concept of diversity presupposes that ‘society’ is a finite totality, a ‘unity in diversity’, or more precisely, a unity of a diversity of meanings and identities. This concept of social totality is conceived as ‘the structure upon which its partial elements and processes are founded’, that is to say, as ‘an underlying principle of intelligibility of the social order’ (Laclau, 1991: 90-1). In this sense, difference and diversity refers to the structured partition of that unitary order – say, the imaginary global village – into fixed parts, such as identifiable readings and audience groupings (to be uncovered by ‘audience research’).
The idea of indeterminacy of meaning, however, enables us to put forward a much more radical theorization of difference and diversity, one that does away with any notion of an essence of social order, a bounded ‘society’ which grounds the empirical variations expressed at the surface of social life. Not order, but chaos is the starring point. Variation does not come about as a result of the division of a given social entity into a fixed range of meaningful identities, but represents the infinite play of differences which makes all identities and all meanings is not the expression of a structural predetermination within a social order. On the contrary, it is the (temporary and provisional) outcome of, in Laclau’s (1991) terms, the attempt to limit the infinite play of differences in the site of the social, to domesticate the potential infinitude of semiosis corroborated by the principle of indeterminacy of meaning, to embrace it within the finitude of an order, a social totality which can be called a ‘society’. From this perspective, this ordered social totality is no longer a pregiven structure which establishes the limits within which diverse meanings and identities are constituted. Rather, since the social is the site of potentially infinite semiosis, it always exceeds the limits of any attempt to constitute ‘society’, to demarcate its boundaries. This is way, as we all know, a ‘society’ can accomplish only a partial closure, a partial fixing of meanings and identities, a partial imposition of order in the face of chaos. That is, any containment of variation and difference within a limited universe of diversity is always-already the product of a determinate ordering by a structuring, hegemonizing power, not , as the functionalist discourse of liberal pluralism would have it, evidence of a media and audiences is not why there isn’t more homogeneity, but why there isn’t more heterogeneity!
To illuminate how this altered notion of difference effectively subverts the closure of liberal pluralist discourse, let me briefly return to the argument I have put forward in Desperately Seeking the Audience (Ang,1991), where I have discussed the history of the corporate practice of ‘audjence measurement’, or, more popularly, ‘ratings’. Over the years, there has been a progressive sophistication of measurement methods and aimed at the ever more detailed and accurate determination of size and demographic composition of the audience at any particular moment, for any particular programme or channel. (…) (T) he latest device currently being tested in the respect is the so-called ‘passive people meter’, a kind of computerized eye roaming people’s living rooms in order to catch their gaze whenever it is directed to the TV screen. The industry’s hope is that this technology will deliver ratings statistics that can tell the television companies exactly who is watching what at any split second of the day. However, this very search for the perfect measurement method, which I have characterized as desperate, is based on the implicit assumption that there is such a thing as an ‘audience’ as a finite totality, made up of subdivisions or segments whose identities can be synchroniclly and diachronically ‘fixed’. I have suggested that this assumption is a fiction, but a necessary fiction for a television industry which increasingly experiences the audience as volatile and fickle. A hegemonic, empowering fiction which is positively constructed as true by the creation of simulations of order in the ranks of the audience in the form of ratings statistics and other market research profiles.