Cultivation theory (sometimes referred to as the cultivation hypothesis or cultivation analysis) was an approach developed by Professor George Gerbner, dean of the Annenberg School of Communications at the University of Pennsylvania. He began the 'Cultural Indicators' research project in the mid-1960s, to study whether and how watching television may influence viewers' ideas of what the everyday world is like. Cultivation research is in the 'effects' tradition. Cultivation theorists argue that television has long-term effects which are small, gradual, indirect but cumulative and significant.
They emphasize the effects of television viewing on the attitudes rather than the behaviour of viewers. Heavy watching of television is seen as ‘cultivating’ attitudes which are more consistent with the world of television programmes than with the everyday world. Watching television may tend to induce a general mindset about violence in the world, quite apart from any effects it might have in inducing violent behaviour. Cultivation theorists distinguish between ‘first order’ effects (general beliefs about the everyday world, such as about the prevalence of violence) and ‘second order’ effects (specific attitudes, such as to law and order or to personal safety).
Gerbner argues that the mass media cultivate attitudes and values which are already present in a culture: the media maintain and propagate these values amongst members of a culture, thus binding it together. He has argued that television tends to cultivate middle-of-the- road political perspectives. And Gross considered that 'television is a cultural arm of the established industrial order and as such serves primarily to maintain, stabilize and reinforce rather than to alter, threaten or weaken conventional beliefs and behaviours' (1977, in Boyd- Barrett & Braham 1987, p. 100). Such a function is conservative, but heavy viewers tend to regard themselves as 'moderate'.
Cultivation research looks at the mass media as a socializing agent and investigates whether television viewers come to believe the television version of reality the more they watch it. Gerbner and his colleagues contend that television drama has a small but significant influence on the attitudes, beliefs and judgements of viewers concerning the social world. The focus is on ‘heavy viewers’. People who watch a lot of television are likely to be more influenced by the ways in which the world is framed by television programmes than are individuals who watch less, especially regarding topics of which the viewer has little first-hand experience. Light viewers may have more sources of information than heavy viewers. Judith van Evra argues that by virtue of inexperience, young viewers may depend on television for information more than other viewers do (van Evra 1990, p. 167), although Hawkins and Pingree argue that some children may not experience a cultivation effect at all where they do not understand motives or consequences (cited by van Evra, ibid.). It may be that lone viewers are more open to a cultivation effect than those who view with others (van Evra 1990, p. 171).
Television is seen by Gerbner as dominating our 'symbolic environment'. As McQuail and Windahl note, cultivation theory presents television as 'not a window on or reflection of the world, but a world in itself' (1993, p. 100). Gerbner argued that the over-representation of violence on television constitutes a symbolic message about law and order rather than a simple cause of more aggressive behaviour by viewers (as Bandura argued). For instance, the action- adventure genre acts to reinforce a faith in law and order, the status quo and social justice (baddies usually get their just dessert).
Since 1967, Gerbner and his colleagues have been analysing sample weeks of prime-time and daytime television programming. Cultivation analysis usually involves the correlation of data from content analysis (identifying prevailing images on television) with survey data from audience research (to assess any influence of such images on the attitudes of viewers). Content analysis by cultivation theorists seeks to characterize ‘the TV world’. Such analysis shows not only that the TV world is far more violent than the everyday world, but also, for instance, that television is dominated by males and over-represents the professions and those involved in law enforcement.
Audience research by cultivation theorists involves asking large-scale public opinion poll organizations to include in their national surveys questions regarding such issues as the amount of violence in everyday life. Answers are interpreted as reflecting either the world of television or that of everyday life. Respondents are asked such questions as: ‘What percentage of all males who have jobs work in law enforcement or crime detection? Is it 1 percent or 10 percent?’. On American TV, about 12 percent of all male characters hold such jobs, and about 1 percent of males are employed in the USA in these jobs, so 10 percent would be the ‘TV answer’ and 1 percent would be the ‘real-world answer’ (Dominick 1990, p. 512).
Answers are then related to the amount of television watched, other media habits and demographic data such as sex, age, income and education. The cultivation hypothesis involves predicting or expecting heavy television viewers to give more TV answers than light viewers. The responses of a large number of heavy viewers are compared with those of light viewers. A tendency of heavy viewers to choose TV answers is interpreted as evidence of a cultivation effect.
In a survey of about 450 New Jersey schoolchildren, 73 percent of heavy viewers compared to 62 percent of light viewers gave the TV answer to a question asking them to estimate the number of people involved in violence in a typical week. The same survey showed that children who were heavy viewers were more fearful about walking alone in a city at night. They also overestimated the number of people who commit serious crimes (Dominick 1990, p. 512). One controlled experiment addressed the issue of cause and effect, manipulating the viewing of American college students to create heavy- and light-viewing groups. After 6 weeks of controlled viewing, heavy viewers of action-adventure programmes were indeed found to be more fearful of life in the everyday world than were light viewers (ibid., p. 513).
Cultivation theorists are best known for their study of television and viewers, and in particular for a focus on the topic of violence. However, some studies have also considered other mass media from this perspective, and have dealt with topics such as gender roles, age groups, ethnic groups and political attitudes. A study of American college students found that heavy soap opera viewers were more likely than light viewers to over-estimate the number of real-life married people who had affairs or who had been divorced and the number of women who had abortions (Dominick 1990, p. 512).
The difference in the pattern of responses between light and heavy viewers (when other variables are controlled), is referred to as the 'cultivation differential', reflecting the extent to which an attitude seems to be shaped by watching television. Older people tend to be portrayed negatively on television and heavy viewers (especially younger ones) tend to hold more negative views about older people than lighter viewers. Most heavy viewers are unaware of any influence of television viewing on their attitudes and values.
Cultivation theorists argue that heavy viewing leads viewers (even among high educational/high income groups) to have more homogeneous or convergent opinions than light viewers (who tend to have more heterogeneous or divergent opinions). The cultivation effect of television viewing is one of 'levelling' or 'homogenizing' opinion. Gerbner and his associates argue that heavy viewers of violence on television come to believe that the incidence of violence in the everyday world is higher than do light viewers of similar backgrounds. They refer to this as a mainstreaming effect.
Misjudging the amount of violence in society is sometimes called the 'mean world syndrome'. Heavy viewers tend to believe that the world is a nastier place than do light viewers. Pingree and Hawkins (1981, cited in Condry 1989, p. 127) studied 1,280 primary schoolchildren (2nd-11th grade) in Australia using viewing diaries and questionnaires. They found that heavy viewing led to a 'television-biased' view of Australia as a 'mean and violent' place. The children with the bleakest picture of Australia were those who most watched American crime adventure programmes. Oddly, they did not judge the USA to the same extent by these programmes.
Gerbner reported evidence for 'resonance' - a 'double dose' effect which may boost cultivation. This is held to occur when the viewer’s everyday life experiences are congruent with those depicted in the television world. For instance, since on television women are most likely to be victims of crime, women heavy viewers are influenced by the usual heavy viewer mainstreaming effect but are also led to feel especially fearful for themselves as women. The cultivation effect is also argued to be strongest when the viewer's neighbourhood is similar to that shown on television. Crime on television is largely urban, so urban heavy viewers are subject to a double dose, and cultivation theorists argue that violent content 'resonates' more for them. The strongest effects of heavy viewing on attitudes to violence are likely to be amongst those in the high crime areas of cities.
A correlation between television exposure and the beliefs of viewers do not, of course, prove that there is a causal relationship, although it may suggest the possibility of one. There could be a another common factor influencing the apparently associated ones. Hawkins and Pingree could not find conclusive proof of the direction of the relationship between television viewing and viewers' ideas about social reality. Rather than heavy TV viewing leading people to be more fearful, it may be that more fearful people are drawn to watching more television than other people. There might be a reciprocal relationship: 'television viewing causes a social reality to be constructed in a particular way, but this construction of social reality may also direct viewing behaviour' (Hawkins & Pingree 1983, cited in McQuail & Windahl 1993, p. 101). In any case, surveys cannot establish causation.
Cultivation research does avoid the artificiality of laboratory experiments - it is based on normal viewing over a long period - but it is subject to the usual criticisms of both content analysis and surveys.
Some studies have shown that careful controls of various variables tend to reduce or eliminate cultivation effects. Doob and MacDonald (1979, cited in Condry 1989, p. 130) report that in the study of the topic of violence, controls for neighbourhood were more reliable than the controls for income used by Gerbner. Hirsch (1980, cited in Livingstone 1990, p. 16), argued that an apparent relationship between exposure to violence on television and fear of crime can be explained by the neighbourhood viewers live in. Those who live in high-crime areas are more likely to stay at home and watch television and also to believe that they have a greater chance of being attacked than are those in low-crime areas. Cultivation theorists do tend to underplay the point that heavy and light viewers do vary in other ways in addition to their TV viewing habits, such as in age, sex and education.
Pingree & Hawkins have argued that breakdowns by content type are more useful than measures of total viewing, because viewers are selective. More specifically content-based measures would show stronger correlations in cultivation analysis (Condry 1989, p. 128). Over- reliance on content analysis misses subtleties and assumes that meaning resides 'in' television programmes (although Gerbner does emphasize connotative rather than denotative meaning unlike many in 'effects' tradition). Also, different genres - and even different programmes - contribute to the shaping of different realities, but cultivation analysis assumes too much homogeneity in television programmes (though some commentators would argue that there is increasing homogeneity in television programmes which may make the cultivation case stronger).
Asking viewers for their estimations of crime statistics is a crude measure of their beliefs about crime. Doob & MacDonald note that there is evidence of a cultivation effect with social questions (e.g. 'How many muggings were there in your neighbourhood last year?') but less so with personal questions (e.g. 'Are you afraid of being mugged?'). Even in the context of a symbolic function, some critical theorists go further than cultivation theorists, arguing for instance that the relative absence of female characters on television is a symbolic statement about their lack of importance in current social reality: women are 'symbolically annihilated'.
Condry (1989, p. 139) makes the point that viewers don't usually use people on television for 'social comparison'. We are not worried by contrasts between how people on television look and live and the way we do. If we were, then the heaviest viewers would be most concerned about their appearance, health and weight because television actors and actresses tend to be young, thin and attractive. But the heaviest viewers are in fact least concerned about their health and weight.
There is relatively little evidence of cultivation effects outside the USA. Wober (1978, cited in Condry 1989, p. 130) found no British evidence of a link between heavy viewing and insecurity. But this may be because there is less violence on British television than in the USA, and Condry suggests that there may be a critical level of the televisual distortion of social reality before it is reflected in the attitudes of viewers. Or it may be that Britain has a more diverse media culture.
More recent theories stressing the active viewer downplay the power of television to influence viewers which is assumed by cultivation theory. Cultivation theory focuses on the amount of television viewing or 'exposure', and does not allow for differences in the ways in which viewers interpret television realities. Viewers do not necessarily passively accept as 'real' what they see on television. Television programmes are open to varying interpretations. The degree of identification with characters by viewers may play a part. Motivations to view also vary greatly. Joseph Dominick comments that ‘individuals who watch TV simply to pass time or because it becomes a habit appear to be more affected than people whose viewing is planned and motivated’ (Dominick 1990, p. 514).
Cultivation theorists tend to ignore the importance of the social dynamics of television use. Interacting factors such as developmental stages, viewing experience, general knowledge, gender, ethnicity, viewing contexts, family attitudes and socio-economic background all contribute to shaping the ways in which television is interpreted by viewers. When the viewer has some direct lived experience of the subject matter this may tend to reduce any cultivation effect.
There is some evidence that lower socio-economic groups tend to watch television as a source of information more than other groups, but the viewer's framing of television 'reality' also needs to be considered here. It is often argued that cultivation may be enhanced when the viewer interprets the content of programmes to be realistic; sceptical viewers may be less likely to be affected. There is some evidence that ethnic minorities exhibit more sophistication in 'perceived reality' than others do (van Evra 1990, p. 169). There is also evidence that working class mothers are more likely to confirm the realism of programmes offering negative depictions of undesirable behaviour to discourage such behaviour, whereas middle-class mothers may tend to make less directive comments.
Daniel Chandler, UWA
This page has been accessed times since 18th September 1995.