Greg Philo
Much of the work of the Glasgow University Media Group, of which I am a longstanding member, has highlighted the role of media in relation to the development of social attitudes and beliefs. In the UK, this has gone against a trend in media and cultural studies which emphasised the active nature of audiences and their capacity to resist messages as well as to create their own. Our work did in fact show that audiences can indeed sometimes be active and critical but we also found strong evidence pointing to the power and influence of media. This dimension of media power is often neglected in current scholarship. In this chapter, I will examine some of the major arguments which have led to this ‘reduced’ view of media effects and then re-evaluate these in relation to contemporary evidence on the nature of reception processes.1
One of the major stimuli for the development of active audience theory was Stuart Hall’s (1980) well-known encoding/decoding model. The discussion around it has often been very confused so I will return to it first and then explain what actually happens when messages are produced and received. The main elements of Hall’s thesis are well known. In essence he is concerned with ‘the class struggle in language’. Following Barthes, he uses the concept of the ‘code’. This is a system of meaning which relates visual signs and spoken and written language (linguistic signs) to the different ideological positions by which a cultural order is either legitimised or contested. As he puts it, codes ‘contract relations for the sign with the wider universe of ideologies in a society’. The codes:
refer signs to the ‘maps of meaning’ into which any culture is classified; and those ‘maps of social reality’ have the whole range of social meanings, practices and usages, power and interest ‘written in’ to them.
The point is that what is being written constitutes a ‘dominant cultural order’, which imposes a ‘taken for granted’ knowledge of social structures (Hall, 1980, p. 134).
This is what he terms a ‘hegemonic’ viewpoint. This ‘carries with it the stamp of legitimacy – it appears coterminous with what is ‘natural’, ‘inevitable’, taken for granted about the social order’ (Hall, 1980, p. 137). Hall argues that television news or a current affairs programme will be ‘encoded’ within this viewpoint. The language and visual images that it uses will be organised within this taken for granted knowledge.2 At the time of his writing there were major conflicts between British governments and trade unions. Inflation was high and, to reduce it, the official policy was to hold wages down. The state presented this policy as being in the national interest, but the trade unions’ perspective was that they were being made to pay for a crisis they had not produced. Some believed that the policy simply preserved an economic and class structure from which they did not benefit.
In our own work, we analysed these conflicts and the manner in which they were presented on television. We showed how words such as ‘reasonable’ and ‘rational’ were used in news coverage to describe what was presented as acceptable behaviour by trade unionists. They could be seen as ‘code’ words within Hall’s formulation and as linked to an overall system of meaning (an ideology) which legitimised a dominant order. For example, in an exchange between a journalist and trade unionists, the latter say that they are only trying to hold wages level (i.e. they were not asking for anything ‘excessive’ or ‘unacceptable ‘). To this, the journalist replies:
But as reasonable men and responsible citizens can you say that is all you are trying to do and all you are interested in when you hear warnings from the Chancellor to the effect that increases of this sort are going to wreck the national economy? (ITN 13.00 24.2.95 in Philo et al., 1982, p. 60).
Hall’s main concern was to show how the ‘hegemonic viewpoint’ would then be ‘de-coded’ by viewers.3 He sets out three positions. The first is the ‘dominant hegemonic position’ where the viewer takes the meaning ‘full and straight’ – i.e. de-codes the message in the same terms within which it was encoded (p. 136). Here the message relates directly to what the viewer sees as normal, natural and as ‘taken for granted’. The second is the ‘negotiated’ position which contains a mixture of ‘adaptive and oppositional’ elements. Here viewers might accept the hegemonic viewpoint at a general level, but seek particular exceptions in terms of their own beliefs or behaviour. So a trade unionist might believe that there is a ‘national interest’ and that it is wrong to have ‘excessive inflationary wage demands’ but this belief might co-exist with the view that his/her own wages are very low and therefore the person is willing to go on strike to get a better settlement.
The third of the positions is what Hall terms the ‘oppositional code’. Here the viewer decodes the message in a ‘globally contrary way’. The message is ‘re-totalised’ within an alternative frame of reference. As Hall writes:
This is the case of the viewer who listens to a debate on the need to limit wages but ‘reads’ every mention of the ‘national’s interests’ as ‘class interest’ (Hall, 1980, p. 138).
This is an important paragraph since it had the effect of sending a large number of people in media and cultural studies up a very long and ultimately pointless path. In particular it encouraged the belief that the language of news texts (visual and verbal) was polysemic – that it could have a variety of meanings to different groups. So viewers of the same news could be seeing and hearing it differently. What they saw and heard would be defined by their own class, gender or ethnicity. The assumption was that people would thus be ‘closed off’ from the intended (encoded) message and could in fact create their own meanings. A crude example of what was being suggested was the supposed behaviour of football supporters who do not ‘see’ the fouls of their own side but only those of the other team. The assumption was also justified by pointing to cross cultural studies, which suggested that people from various cultures would receive TV messages differently. In a soap opera, for example, behaviour seen as villainous in one culture might be interpreted as heroic in another. The conclusion of some theorists was that meanings were created in the encounter between the reader and the text and that each new encounter could potentially create a new meaning.4
The theory that texts were polysemic and subject to divergent meanings swept through cultural studies. In this example Oliver Boyd-Barrett (2002) describes the work of Hall and its later development by David Morley:
(Hall) developed a theory of ideology which allows that texts are ‘polysemic’, which is to say that they offer the possibility of a diversity of readings, even if a ‘preferred reading’ is inscribed within the text by its producers. Through the work of Morley, in particular, this crucial insight has been further explored to reveal the divergent meanings that different groups, whether defined in terms of social class, gender or ethnicity, could draw from texts (2002: 45).
There is a fundamental error here, it seems to me, in what is being suggested about how audiences can reject messages. It might be that people from very different cultures do not understand each other’s cultural symbols, as for example when Western explorers appropriated artefacts, which had deep religious significance and thought they would make nice wall decorations. It is also the case that some media, literary or artistic products are deliberately made to be open to a variety of interpretations, as in a poem or work of art. But our work on TV news showed that audiences within a culture do not typically create a new meaning with each ‘reading’ or encounter with an encoded message. Rather, they are likely to criticise the content of the message in relation to another perspective, which they hold to be correct. They are therefore aware of the encoded meaning and the manner in which it has been constructed – they just don’t agree with it. We studied audience perceptions of the miners’ strike in 1984/5 in Britain. At this time, TV news repeatedly showed images of miners and pickets in violent clashes with police. People from different class and political backgrounds saw these images and understood in the same way what was being shown (i.e. that the miners and pickets were responsible for the trouble). Thus, a politically conservative person in our sample commented on what she saw as the bad behaviour of ‘this rabble’. The miners who we interviewed did not ‘see’ the images differently in the sense that they thought they were looking at news pictures and language about pickets being kind to police. There was no new meaning created in this sense. The miners understood the encoded message but they thought that the news had left out images which showed their perspective (which was that the police had started the trouble).
We can now look back to Hall’s hypothetical example of the oppositional viewer. Here again the person may understand the hegemonic view that a wages policy is supposed to be in the national interest, but simply not believe it and instead think that it will mainly benefit the rich.5 Crucially, it is also possible for someone who accepts the hegemonic view to be aware of the alternative ‘radical’ view and even to see how such a perspective is being excluded by the encoded message. My argument is then that (1) viewers do not typically construct a new meaning with each encounter with a news text but rather, they can share an understanding of what is being presented and differ in their responses to it (acceptance or rejection), and (2) viewers do not necessarily occupy their own ‘sealed off’ cultural space, unaware of the values and definitions offered by others. The second of these points was actually confirmed by Morley’s work (against his own expectations) when he tried to test Hall’s model. Hall had suggested that people who accepted the encoded message would be ‘living within’ the taken-for-granted hegemonic ideology. They would therefore be unaware of the processes by which the ideology was being ‘preferred’. Following this position, Morley (1980) writes that:
For some sections of the audience the codes and meanings of the programme will correspond more or less closely to those which they already inhabit in their various institutional political, cultural, and educational engagements, and for these sections of the audience their dominant readings encoded in the programme may well ‘fit’ and be accepted (1980, p. 159).
Morley believed that such people would see the television account as being simply commonsense – and also that the alternatives which television has excluded will be ‘invisible’ to them. But his own evidence in fact showed examples of both conservative and radical groups who were clearly aware of alternative positions on the economy. In one case, a group of bank managers who were politically conservative actually criticised a programme they were shown, for promoting the conservative position without giving what they saw as a proper ‘balance’ of views. In Morley’s terms they were able to ‘deconstruct’ the programme. In this case the programme had included an interview with an accountant who was presented as a neutral commentator. In the event this ‘expert’ gave a very conservative view of economic policy, focusing on the need for tax cuts to provide incentives and the need for cuts in public expenditure. The group of bank managers who watched this commented that:
Particularly that accountant from Birmingham … was … very much taking a view very strongly, that normally would only be expressed with someone else on the other side of the table … (1980, p. 106).
Morley, in fact, found many examples of audiences being able to de-construct programmes by, for example, ‘seeing through’ loaded questions. But this ability to de-construct did not necessarily mean that the audience members rejected the view that was being promoted. As he wrote:
The recognition of ‘preferring’ mechanisms is widespread in the groups and combines with either acceptance or rejection of the encoded preferred meaning; the awareness of the construction by no means entails the rejection of what is constructed (1980, p. 140).
But not everyone is aware of the ‘preferring’ mechanisms, so Hall’s original depiction of some people who live within a taken for granted ideology is still useful.
This leaves us with three central questions on the relationship between media, audiences and ideologies (which we take to mean perspectives which are linked to and legitimise social interests). The first question is what are the conditions under which people accept or reject a perspective when they are aware of the range of alternatives? This is a very fundamental issue and relates in part to how humans choose between value systems. Such choices are affected by many factors including our class and cultural history, by notions of self interest or more altruistic beliefs, by our accumulated experience of attempts at social change and by the immediate economic and political conditions which we face. All these can influence preferences for social democracy or free market capitalism or socialism or whether people decide that no change is possible or simply take to drink or drugs or consult astrologers or in other ways try to forget all about it. So it does not follow that just because people know the range of alternatives they will necessarily choose the radical, oppositional one. Still, one factor in the choices that are made and in attitudes which are formed is whether people do actually know of alternative ways of understanding and of a range of possible actions.
The second question is then, what are the conditions under which information about these alternatives is either made available or is limited in public discourse, and what happens to Hall’s group of people who are living within the hegemonic ideology if they are given different information? We have argued in our work that the media are a key element in the construction of public understanding. When we studied attitudes to tax and redistribution of income, we found that most people in our focus groups did not know that in Britain 10 percent of the population owned well over half of the private wealth or that it was the poor who paid the highest proportion of tax relative to their income. Some participants were angry when they found out that this was the case. This does not mean that such responses necessarily transform into political action but none the less it was felt by people in the groups that TV news was not properly informing the public when it discussed government policy on tax spending and budgets (ESRC 2004). Another example from our work was of how understandings of the Israeli–Palestinian conflict were affected by news accounts of the motives and actions of those involved. We showed how the news gave very little explanation of the history of the conflict. This history was contested by both sides, but the Palestinian perspective rested crucially on their claim that they had lost their homes and land when Israel was established and also that they were now living under Israeli military occupation. Without knowledge of this, there was no apparent rationale for Palestinian actions when they attacked Israelis. The Palestinians could thus be seen as ‘starting’ the violence. The Israelis were portrayed in the news as ‘responding’ to these attacks that had been made upon them and some in the audience clearly took this message from what they saw. As one participant expressed it:
You always think of the Palestinians being really aggressive because of the stories you hear on the news. I always put the blame on them in my own head … I always think the Israelis are fighting back against the bombings that have been done to them. (cited in Philo and Berry, 2004, p. 222).
But when people did encounter new information on the history of the conflict, this could then affect their understanding:
I just thought it was disputed land. I wasn’t under the impression that the Israeli borders had changed or that they had taken land from other people.
And another:
The impression I got was that the Palestinians had lived around that area and now they were trying to come back and get some more land for themselves – I didn’t realise they had been actually driven out (Philo and Berry, 2004, pp. 216–217).
It is clear that how actions are perceived and the legitimacy of different positions can be dramatically affected by the context in which they are understood and the information which is given. Inasmuch as the media influence this information flow, then
they do have a key role in sustaining and developing some ideologies.6
The third question is under what conditions can such ideologies (in the encoded message) be critiqued by audiences. For Hall and Morley the key issue is the class and cultural location of the viewer (at least in the periods of their work which I am discussing here). As Morley put it, the hegemonic meanings encoded in a television programme will jar ‘with those produced by other institutions and discourses in which they are involved – trade union or deviant subcultures for example’ (1980, p. 159). In our own work we also found evidence of this. For example, one of our interviewees rejected television images of violence in the miner’s strike, because she had direct experience of being involved in another strike at Chrysler, the car company. She believed that TV had misrepresented this strike and then generalised her experience to what she saw later of the miners’ dispute (Philo, 1990, p. 63). But it was not always the case that a cultural and class affiliation was sufficient to produce such a rejection. We also interviewed a group of retired people, many of whom had trade union experience and were sympathetic to the miners. Yet these also had a high regard for television as an information source and some expressed their sadness at how the ‘violence’ reflected on the miners’ cause (Philo, 1990, p. 93). None of these people had actually attended a picket line. This turned out to be a crucial variable and we found that everyone who had actually seen a picket line (whatever their politics or class) rejected the television depiction (p. 150).7
Another key variable in terms of audience criticism was the use of conceptual processes such as logic. Thus when people were asked if they believed that picketing was mostly violent, some reasoned that it could not have been so, given the huge numbers of people involved. As one noted, ‘if they had been really violent, the police couldn’t have coped, it would have been the army’. Their conclusion was that television images of violence must have been very selective. Such a use of logic could occur in different groups and was not necessarily related to support for the miners. Another participant who said she ‘would have shot’ striking miners also argued that ‘because of the amount who were actually on strike … it can’t all have been violent’ (Philo, 1990, p. 151). We found the same phenomenon of the use of logic in our study of beliefs about the Israeli–Palestinian conflict. We asked large samples of people about the casualties suffered by each side. In fact, many more Palestinians had been killed than Israelis (about 2–3 times as many). But TV news coverage had focused disproportionately on Israeli casualties with extensive coverage of suicide bombing. Many people did in fact take their beliefs from this presentation and accepted the TV account rather uncritically. They believed either that the Israelis had had the highest casualties or that the numbers were about equal for the two sides. In this example, a participant describes the source of her belief as being TV news:
Well basically on the news coverage they do always seem to make the Palestinians out to be the ones who are the suicide bombers, so it’s like, I would imagine, it’s going to be more casualties on the Israeli side, but it is purely from television, that’s where I’m getting my info from, that’s how it is being portrayed on television (cited in Philo and Berry, 2004, p. 234).
Only a minority of those interviewed actually believed at this time that Palestinians had significantly more casualties than the Israelis. For these participants, one of the key factors in making this judgement was the use of logic.8 These participants commented on the contrast between Palestinians with ‘sticks and stones’ and Israelis with ‘guns and tanks’. The conclusion was:
If that’s all they have got to fight with, presumably they are killed more often than someone with a gun (cited in Philo and Berry, 2004, p. 235).
This is a dimension of audience activity which is completely missing from the encoding/ decoding approach. The model as offered by Hall and Morley was in some ways useful in that it focused attention on how class and cultural factors could produce different responses to encoded messages. In his later work Morley (1986) writes of how a range of ‘experiences of life’ can incline people to different readings (1986, pp. 42–43). But the model is weak both in that it misses important dimensions of audience activity but also underestimates the power of the media in shaping ‘taken for granted beliefs’. The conceptual arrangements appropriate to a class position are seen in the model as the key variable in evaluating a new message from the media. However, this neglects the issue that the conceptual structures include ‘knowledges’ about what typically occurs and assumptions about the rationality and legitimacy of action which may already have been subject to prior exposure to media messages. There is little room in the encoding/decoding model to investigate such a possibility. In one noteworthy case, Morley in his own work showed the response of audience groups to TV images of car workers. The workers had actually won £600,000 on the football pools. But soon as the image of the factory was shown, several of his audience groups assumed the story was about a strike. One person commented that ‘first thing you think, whenever you hear of British Leyland (the car manufacturer) is “who’s on strike this time?” ’ Morley (1980) comments on the groups responses that:
Their decoding of this item is informed by, and leads into a generalised exposition of, a stereo-type of the ‘greedy car worker/mindless union militants’ presumably derived, at least in part, from the media (1980, p. 127).
But Morley does not then pursue this with his interviewees to trace the constituent elements of the belief systems or their origins, as the research design does not really permit this.
However, the main problem which I have with the encoding/decoding model is the impact which it had on the subsequent development of media and cultural studies. The view which many took from it was that audiences could resist messages, safe in the conceptual boxes of their class and culture, and renegotiating an endlessly pliable language. This led eventually to the serious neglect of issues of media power. In contrast, the Glasgow University Media Group has resisted the tendency to underestimate the capacity of audiences to actively engage with texts. Today there is a powerful body of evidence which shows the influence of media messages on the construction of public knowledge as well as the manner in which evaluations are made about social action and what is seen as necessary, possible and desirable in our world. For us, media power is still very much on the research agenda.
1 An earlier version of this chapter appeared as: Philo, G. (2008) ‘Active audiences and the construction of public knowledge’, Journalism Studies, 9(4): 535–544.
2 Theorists such as Fairclough have shown how this can work in relation to news and other texts. Fairclough analysed the speeches of Tony Blair and showed how they gave preference to right wing, neo-liberal views. Fairclough used concepts such as hyponymy (meaning a presupposed semantic relationship) to show how ‘globalisation’ and ‘economic progress’ might be treated as hyponyms (2003: 213). Van Dijk in other work has looked at the importance of headlines, story structures, and issues such as ‘agency, responsibility and blame for actions’ (1998: 44). For a further account of these studies, see Philo (2007).
3 The term ‘hegemonic’ is used here in discussing Hall’s original model. It implies the legitimising of a dominant perspective but we should not assume that media output in contested areas of social and political life consists only of the promotion and reproduction of such perspectives. There is some space for alternative accounts. When we studied TV news coverage of the Israeli–Palestinian conflict, we found for example that the Israeli perspective was heavily featured but some news such as that on Channel 4 featured more critical commentary of Israeli spokespeople than did the BBC news. Journalists certainly felt very pressured when covering this area but there were instances such as in the deaths of the Palestinian children where strong criticisms were made of Israeli policy across various news outputs including the BBC. We describe media in our work as a ‘contested space’ though the contest is certainly not equal and structures of access and power have a strong influence on media content. For a fuller account of this see Philo (2007).
4 See, for example, Fiske (1987). For a critique of this position see Philo (1990) pp. 190–199.
5 This understanding of Hall’s model is tenable within his original formulation as he says of the oppositional position that it is possible for the viewer ‘to understand both the literal and connotative inflection’ (plus their own oppositional understanding). In his later work Hall moved towards a discursive approach in which language was constitutive of the real and speakers occupied different positions in relation to ‘truth’. The exchange of meaning becomes a ‘process of translation’ between different speakers. For a critique of this perspective see Philo and Miller (2001, pp. 33–35).
6 This is not their only function since the media also endorse value systems as well as influencing how we understand action in relation to such systems. So media references to being ‘reasonable’ and ‘responsible’ are actually the advocacy of a value system, while the reporting of what unions are asking for, shaped the information which viewers can use to decide whether the unions are acting in accordance with these values. (For a further discussion of this see Philo, 1990, pp. 4–6).
7 This was such a strong research result that we took it to be almost axiomatic that direct experience would overrule a contrary media image or presentation. This was until we undertook work on mental distress and media. In this study we asked people whether they associated mental illness with violence. We found that such was the level of fear generated by media images such as in horror films, that in some cases these associations overruled direct experience. For example, a young woman working in a special hospital with elderly people who were not in any way violent, none the less made this association and was scared of them because of what she had seen in films and plays (Philo, 1996, p. 104).
8 Other groups of people also judged correctly the ratio of casualties for each side. This was because they had studied the subject and/or had an intense support for one or other side in the conflict. So both those who were strongly pro-Palestinian and as those who were pro-Israeli stated that Palestinians had the highest numbers of deaths. They did not of course agree on everything and many events were highly contested, but it is interesting that both accepted the ‘facts’ in this case.
Boyd-Barrett, O. (2002) ‘Theory in Media Research’ in Newhold, C., Boyd-Barrett, O. and Van Den Bulck, H., The Media Book, London: Arnold.
ESRC study (2004) Audience Receptions of Television News, Current Affairs and Documentary Programmes End of Award Report, ESRC R/000/23/9669.
Fairclough, Norman (2003) Analyzing Discourse, London: Routledge.
Fiske, J. (1987) Television Culture, London: Methuen.
Hall, Stuart (1980) ‘Encoding/decoding’, in Hall, Stuart, Hobson, Dorothy, Lowe, Andrew and Willis, Paul (eds) Culture, Media, Language. London: Hutchinson/CCCS.
Morley, D. (1980) The Nationwide Audience, London: BFI Television Monograph.
Morley, D. (1986) Family Television; Cultural Power and Domestic Leisure, London: Comedia.
Ofcom (2007) New News, Future News. Office of Communications, London.
Philo, G. (1990) Seeing and Believing, London: Routledge.
Philo, G. (1996) Media and Mental Distress, London: Pearson/Longman.
Philo, G. (2007) ‘Can discourse analysis fully explain journalistic practice’, in Journalism Studies, April, London: Sage.
Philo, G. and Berry, N. (2004) Bad News from Israel, London: Pluto.
Philo, G., Hewitt, J., Beharrell, P. and Davis, H. (1982) Really Bad News, London: Writers and Readers.
Philo, G. and Miller, D. (2001) Market Killing, London: Pearson/Longman.
Van Dijk, Teun (1998) ‘Opinions and Ideologies in the Press’, in Bell, A. and Garrett, P. (eds) Approaches to Media Discourse, Oxford: Blackwell.