First,there issomething to be said for an approach which explores the structure of public and private space,and which is sensitive to the proximity of individuals to real conditions in assessing media usage and influence (see Gumpert and Drucker 1991).
The structure of public and private space is different for the two groups investigated. The White community, through the benefits it has reaped from segregation and Apartheid, does have access to more public amenities - for leisure,sports and entertainment,and may not be wedded to either the mass media as leisure pursuit or information source to quite the same degree as the Indian community,which had been systematically dispossessed of its access to public space. We are on tricky ground here,for while this has been the case,itis beginning to change quite rapidly with the integration of residential areas closer to the city, as well as
changes of education and income levels among younger people. The point is that there has never been the pressure on the White community to tum domestic space into public space, even though its affluence allows it the luxury of choosing to do so. But this really only tells us of the potential for mass media to playa larger role in the Indian community.
Furthermore,there is a vibrant associationallife in the Indian community which may in fact nullify this loose hypothesis.
Secondly, in terms of proximity to real conditions,there does not seem to be much in it - both White and Indian communities are literally miles away from the townships, which are the predominant site ofviolence. In fact, it may be that the Indian community is closer to violence,particularly in the PMB situation where many African people will shop in a section of the city with shops owned and patronised by members of the Indian community.
Thirdly, from the point of view of what Cohen, et al (1989) call the dimension of
psychosocial context,there may be interesting differences between these communities.It should be noted at the outset that the construction of personal space for these two
communities differs markedly,with the Indian community characterised by close kinship networks as opposed to a more open or fluid social networking characteristic of the White community.
Furthermore,the Indian community,or significant parts of it, may fmd themselves in a situation of political and social insecurity, which may not apply to any significant part of the White community. Historically it is a community that has been squeezed between Black and
White (Carrim 1993;Chetty 1992). This kind of insecurity,painted in very rough strokes here, is to some extent demonstratedby the fact that almost half of the Indian respondents who identified themselves with the political policies of one or more of the African National Congress or the South African Communist Party also identified themselves as supporters of National Party policies.
If there is a grain of truth in this formulation, could it be that Indian respondents identify the mass media as the main source of information as a result of attempting to keep 'in the know' of national developments,and high reliability scores for the mass media do not necessarily reflect a belief in the content but in their own need to be informed?
On the other hand,the White community has benefited from decades of White minority rule and have a sense of security in their own future - perhaps even an arrogance which is no better illustrated than by the militancy of the White Right. This is a community that may not need the news,good or bad.This is probably overly cynical,and is belied by the fact that the group identifying contacts as a main source of information do tend to provide answers to some of the attitudinal questions which are in line with academic research fmdings.
The notion of cognitive engagement is useful here,and when translated by Zaller (1992: 43) into 'political awareness', could be a fruitful tool for exploring the contact/mass media differential that exists within the two communities studied. His argument is that people with a higher political awareness are more likely to receive political messages, which falls into line with the fmding that White respondents of a LeftlLiberal persuasion are more generally
sceptical ofthe mass media (preferring contacts for information),and of television in particular. Ifwe also continue to explore the idea that the Indian community is insecure, we end up with a pool of respondents who are politically aware,but whose main sources of information are quite different and lead to different attitudes about when violence started, its degree of severity and the toll.
On the face of it,this research falls into line with research that has shown,
the greater ability of the more involved viewer to counter argue with news stories (which is) expected to produce a degree of immunity to framing effects (Iyengar 1991, 118).
This is a real possibility within a section of the White community and reinforces a
Left/Liberal understanding of political violence as being a result of social forces rather than simply a victim/perpetrator narrative. However for the Indian community such a conclusion is not immediately apparent. The social structure of this community may predispose its members towards 'involvement',but this does not produce the counterargument that Iyengar identifies as its result.
This research also fmds resonance with Susan Booysen's (1991) work on the differential between mass media and social networks as sources of political information among English and Afrikaans speaking White university students. She has established that there does exist such a differential and that cultural categories can be successfully used in the study of political communication in South Africa.