Full Text Version

Sectarianism, Children and Community Relations in Northern Ireland.

  Chapter 2:
Learning From Research on Children and Prejudice



The previous chapter offered a comprehensive review of what we know from the research literature concerning the impact of sectarianism on children’s lives in Northern Ireland. Unfortunately, given the limited individualistic focus of much of the research and its particular methods used, our understanding of this area remains severely limited. In order to gain some appreciation of what we can reasonably assume in terms of the nature and forms that sectarianism takes among children, it is worth assessing what we can learn from research that has been conducted more generally on the topic of children and prejudice. It is accepted that there are obvious limitations to adopting such a comparative approach given the unique nature of sectarianism in Northern Ireland. These will be outlined and addressed later in this chapter. Nevertheless, such an analysis will at least help to guide researchers and practitioners in terms of what age we may reasonably expect children attitudes to begin to be affected by sectarianism and how those affects develop with age. This can therefore stand as a hypothesis to be tested through empirical research and practice.

The chapter will begin by outlining what is known about the acquisition and development of racial prejudice among children. It will then proceed to assess the application of these findings to the situation in Northern Ireland.

Research on children and ethnic prejudice

In attempting to examine the broader literature on young children and prejudice, it is useful to begin with that which has focused on ‘race’ and racism. While it is accepted that the findings of such work cannot simply be taken and applied to the situation here in Northern Ireland, not least because the social cues used to identify racial groups are obviously much more overt and visible than those cues used to identify and distinguish between Protestants and Catholics (Cairns 1980; Brewer 1992), they do offer an important overview of the changing nature and influence of prejudice in children’s lives. This will provide the overall context within which the developing influence of sectarianism among children in Northern Ireland will then be discussed.

p. 29 p. 30

Jump to Top

   Contents Page

The effects of cognitive development on children’s racial prejudice

In relation to studies of young children and ‘race’, there is now a colossal body of research that has attempted to understand at what age children first become prejudiced and how that prejudice changes with age (for summaries see Milner 1983; Aboud 1988). In essence such work has shown that from about the age of three onwards, White children are already capable of expressing racially prejudiced attitudes towards others (Comer 1989). As Aboud (1988: 29) summarises, ‘most White children between the ages three and five choose a Black person as looking "bad", as having negative qualities or as being least preferred as a playmate’. At this age, however, while White children may be able to categorise others as Black and to associate such a grouping with negative connotations, such categorisations are understandably likely to be quite limited and rudimentary. The general process appears to be that children from around the age of three onwards begin to categorise others by skin colour and then, as they develop, attach more sophisticated and complex meanings to those categories. As Bar-Tal (1996: 353) recently summarised, ‘the categorisation basis changes with age. While young children form their social categories on the basis of behaviours and traits, older children are able to use defining features to define a social category. Also, the social categories of the older children are more elaborated and differentiated’. In other words, while very young children may be able to develop the concept of a Black person based upon their skin colour and their presumed violent behaviour, as those children get older they tend to use more complex defining features such as religion, nationality and language/dialect to categorise and evaluate Black people.

Interestingly, this developing ability to categorise in more sophisticated ways does not relate in any simple way to children’s levels of prejudice. On the contrary, as Aboud (1988) explains, children’s levels of prejudice tend to increase over the early years reaching a peak at around the time they start school and then, from about the age of seven onwards, their levels of prejudice tend to decrease somewhat (see, for instance, Bigler and Liben 1993; Corenblum and Annis 1987; Ramsey 1991). This mismatch between the increasing ability to categorise and the decreasing levels of prejudice found among children as they grow older has been examined recently in a study by Doyle and Aboud (1995) who conducted one of the first longitudinal studies of children’s prejudice. They charted the changing nature of 47 White Canadian children’s attitudes towards Black people and Native Indians 

p. 30 p. 31

Jump to Top

   Contents Page

from the age of six to nine. As with previous research, they concluded that ‘prejudice is high in kindergarten and decreases over the early years of middle childhood’ (p. 222). However, what is extremely interesting from their study is that, over this time, the children’s general ethnocentric bias (that is their positive evaluations of their own group and negative evaluations of Black people and Native Indians) did not change.

In order to try and explain the fact that the children’s ethnocentric biases remained intact while their overall levels of prejudice decreased, Doyle and Aboud looked to other factors. In particular, they found that a correlation existed between the levels of prejudice and the number of ‘counter-biases’ that the children developed. In other words, the more that children developed attitudes that ran counter to their general evaluations of themselves and others (i.e. some White people are mean; some Black people are intelligent and so on), the more that their general levels of prejudice reduced. It was thus not so much that children progressively came to forget or reject the negative evaluations they had about particular minority groups that led to prejudice reduction but, as Devine (1989) has also found, that they also developed contradictory attitudes alongside these that they could use to re-assess their overall beliefs.

For Doyle and Aboud (1995) they postulated that this increasing development of counter-biases related to developments in the children’s cognitive abilities, particularly those associated with reconciliation. In brief, it has been found that the ability of children to appreciate other people’s perspectives (which Piaget suggested does not happen until about the age of seven or eight in most children), they first develop a recognition that others prefer members of their own social group (termed reciprocity) and then that such preferences are legitimate and acceptable (i.e. reconciliation). For Doyle and Aboud, they found that this cognitive development in the ability to recognise and cope with differing beliefs also correlated with the children’s development of counter-biases.

The importance of social context in influencing children’s racial prejudice

As suggested earlier however in relation to research on children’s attitudes in Northern Ireland, it is important not to develop too deterministic an understanding of children’s prejudice that simply relates their attitudes at any given time directly to their perceived level of cognitive development. While such developments are obviously 

p. 31 p. 32

Jump to Top

   Contents Page

useful in terms of the basic context that they can set for understanding how children’s attitudes and behaviour change, they should be viewed as guidelines only. The particular way and rate at which children move through these stages of cognitive development will rely heavily on the particular social contexts within which they are located. As Aboud and Doyle (1993) have recognised in a separate paper, the development of prejudice is also influenced by non-cognitive factors most notably social factors including children’s particular experiences and the complex ways in which they are socialised at home, school and within their peer-groups. The importance of social factors was something highlighted by Aboud and Doyle (1996) in a more recent paper which reported upon a study of 88 White Canadian children aged eight to 11 years. They tested the children’s general levels of prejudice and identified those whose scores indicated that they were either high- or low-prejudiced. Children were then paired (one high and one low) and given two index cards one with a positive attribute written on it, the other with a negative attribute. The children were then asked to discuss between themselves who they should assign each of the attributes to (i.e. to a White child, a Black child, a Chinese child or more than one child). Following that task, the children’s levels of prejudice were measured once again. Interestingly, Aboud and Doyle found that talking to low-prejudiced children had a positive effect on reducing the prejudice of those identified as high-prejudiced. In other words, their study offered supporting evidence for the importance of children’s peer-groups in mediating their prejudicial attitudes and behaviour.

Moreover, recently published research by Ocampo et al. (1997) offers further support to the importance of social factors in influencing children’s ethnic identities. Their study involved a sample of 103 Mexican American children aged four to 10 years. Two tests were conducted to ascertain the children’s levels of ethnic self-identification, ethnic constancy and ethnic knowledge. A further test was administered to measure the children’s levels of cognitive ability in relation to their understanding of conservation and classification. Interestingly they found no relationship between the children’s level of cognitive ability and two of the three measures of their ethnic awareness. As Ocampo et al. (1997: 493) summarised:

The age differences in cognitive ability did not account for the age differences in ethnic self-identification and ethnic constancy as predicted by cognitive developmental theory. 

p. 32 p. 33

Jump to Top

   Contents Page

The correlations of age with ethnic identification and ethnic constancy remained virtually unchanged when the effects of level of cognitive ability were statistically removed.

The only measure that did positively correlate with cognitive ability was ethnic knowledge which can, logically and unsurprisingly, be assumed to grow and develop in any case with age and exposure to the child’s own ethnic culture. In order to explain what can account for the development of ethnic self-identification and ethnic constancy, three separate studies carried out by these researchers (see Knight et al. 1993a, b, c) found a clear correlation between the Mexican mothers’ own ethnic identity, how Mexican culture was taught to their children and then their children’s own ethnic identity. In other words, they drew attention to the significant role that socialisation plays in the development of a child’s ethnic attitudes and identity.

Finally, more qualitative, ethnographic research on young children aged five and six in an English multi-ethnic primary school by Connolly (1998a) has further shown just how important children’s social experiences are in relation to the development of racially prejudiced attitudes and behaviour. Connolly found that levels of racism among the children were not constant but varied between the children and from one context to the next. In a similar way to the findings of research on older primary school-aged children (see Carrington and Short 1989; Troyna and Hatcher 1992), Connolly showed that children as young as five were highly active and competent in the way in which they drew upon, re-worked and reproduced racist discourses in order to make sense of their own particular experiences and/or justify their own behaviour.

In summary, it is clear that children can become prejudiced at a very early age. Around the age of three, children already demonstrate the ability to categorise others and, crucially, develop negative attitudes towards those categorised as the out-group. This then has what Devine (1989) has termed an ‘imprinting effect’ where the initial categories identified by children become embedded within their ways of thinking and the children, over time, come to develop and identify more sophisticated and elaborate defining features of the different social categories. While children’s ethnic understanding may begin simply with ethnic self-identification and progress through to an increasing awareness of ethnic constancy and ethnic knowledge, these stages should not be ‘cast in stone’ nor limited to particular ages. The particular 

p. 33 p. 34

Jump to Top

   Contents Page

nature and way in which children develop ethnic knowledge and prejudice must be seen as equally influenced by non-cognitive factors including, most prominently, those social factors relating to their experiences, socialisation within the home and school and the influences of peer-groups.

Finally, it is clear that this growing awareness of ethnicity is not, in itself, the cause of prejudice. It appears to be the culmination of broader social factors and the increasing knowledge and understanding that this brings to children. In particular, it is the specific social experiences that children gain that tends to influence whether they ascribe positive or negative attitudes towards particular ethnic groups. Within this, the more that children are able to acquire attitudes that run counter to the general ethnocentric biases that they have learnt and continue to maintain, the more that their overall levels of prejudice reduce.

Applicability of research on children and prejudice to the study of sectarianism and children in Northern Ireland

At the beginning of this chapter it was recognised that such findings in relation to children and ‘race’ cannot simply be applied, without adaptation, to the situation in Northern Ireland in respect of children’s ethno-religious attitudes. As a consequence it has been argued by Cairns (1980) among others that the age at which children begin to categorise others in relation to religion in Northern Ireland is likely to be older than that in relation to ‘race’. This is because racial categorisation involves obvious, physical cues (i.e. skin colour) while ethno-religious categorisation involves more complex cues such as names, residential areas and faces. It is argued that these latter cues rely on stereotypes which, in turn, can only be understood by more mature and cognitively developed children. It is such reasoning that therefore enables researchers in Northern Ireland to maintain the belief that children, on the whole, only develop the ability to categorise on a denominational basis at about the ages of 10 or 11 even against the background of a vast array of research that has shown that children, more generally, are able to exhibit prejudice at a much earlier age. Before we can begin to examine the implications of the foregoing discussion for what we can reasonably assume about the developing influence of sectarianism in children’s lives in Northern Ireland, it is important in this section that we first ascertain the relevance of the broader literature outlined in the previous situation to the Northern Ireland context.

p. 34 p. 35

Jump to Top

   Contents Page

Young children’s categorisation of others in the absence of physical cues

To do this it is useful to discuss the findings of research recently conducted by Bar-Tal and his colleagues concerning Israeli children’s concepts of the ‘Arab’ (for a summary see Bar-Tal 1996). They found that many of the Israeli children first developed the concept of the Arab from around the age of 24 - 30 months. However, in broad agreement with the literature on ‘race’ and children outlined above, it was only from about the age of two-and-a-half to three years that the children began to ascribe negative evaluations of the category. As before, those initial evaluations are relatively simple and crude and a more detailed and elaborated set of negative qualities concerning Arabs are learnt over time. Moreover, the process of categorisation itself begins simply with reference to little more than either appearance and/or particular traits (i.e. Arabs are ‘nasty’) and develops over time to include more defining features such as their religion, nationality and language.

In one sense, therefore, the findings that Israeli children categorise and develop prejudiced attitudes towards Arabs at about the age of three and that this then provides the basis from which they develop more sophisticated means to categorise and more elaborated negative evaluations of Arabs as they grow older are not particularly new nor surprising. However, what is of specific interest is the fact that such a process occurs in a very similar way and to a relatively equal extent among Israeli children who have had no contact with Arabs at all. Indeed, and of particular relevance to the Northern Ireland context, while children as young as three had developed the concept of Arab and had begun to negatively evaluate it, when asked to draw an Arab and a Jewish person they could not distinguish between the two. In other words, the ability to develop an awareness of an outgroup (in this case Arabs) and express prejudice towards it, does not necessarily need to rely upon physical cues and/or the ability to recognise that out-group. This is also a finding supported by Hirshfeld (1993, 1994) in relation to children and ‘race’. For these specific Israeli children then, Bar-Tal (1996: 350) reports that they did not develop knowledge about Arabs through contact but, when asked, the majority gained their knowledge either from television (86.7 per cent) or their parents (80.6 per cent); the latter appearing to offer support to the findings of Knight et al. (1993a, b, c) in relation to socialisation within the family.

p. 35 p. 36

Jump to Top

   Contents Page

Reassessing young children’s ability to categorise on a denominational basis in Northern Ireland

If we now return to the situation in Northern Ireland, there is no reason to do other than assume that children can also develop a rudimentary understanding and negative evaluation of the concepts Catholic and/or Protestant. Moreover, the work of Bar-Tal and his colleagues demonstrates that this does not have to rely upon children having the cognitive abilities to deal with stereotypes in the form of Protestant and Catholic names and faces or particular residential areas in order to develop an awareness of the two main ethno-religious groups. Indeed, of the limited research data available from Northern Ireland, as discussed in the previous chapter, it is clear that children as young as five are already demonstrating an awareness of the concepts Protestant and Catholic (Jahoda and Harrison 1975; Cairns 1978, 1980; McWhirter and Gamble 1982). Moreover, with the central influence of socialisation through family and television as highlighted by Knight et al. (1993a, b, c), Bar-Tal (1996) and Ocampo et al. (1997), it is only logical to assume that such an awareness also begins to develop in Northern Irish children from about the age of three onwards. As with the research reported on by Bar-Tal (1996), there are equal and widely-held stereotypes about Catholics and Protestants and the dangers associated with coming into contact with them that are also presumably taught to children from a very early age, just as there are those concerning Arabs taught to young Israeli children. The anecdotal evidence available from those working in nurseries and P1 classes in primary schools in Northern Ireland certainly supports this belief that children aged three onwards are capable of categorising and negatively evaluating the other main ethno-religious group; in other words capable of being sectarian (Playboard 1997).

In summary, and with no evidence to the contrary, it therefore seems reasonable to assume that children’s levels of sectarian beliefs in Northern Ireland follow the same trajectory as that concerning children’s prejudicial beliefs more generally. In particular, we can assume that they have the capacity to develop an awareness of one or both of the concepts Catholic and Protestant from about the age of three onwards and also, at this time, be able to evaluate them negatively and/or positively in a rudimentary manner. It can also be assumed that the main sources of information concerning this awareness will not be visual cues (i.e. the ability of a child to recognise those of the other main ethno-religious grouping by some physical trait) but will be verbal 

p. 36 p. 37

Jump to Top

   Contents Page

i.e. information mainly passed on through the family and by television and other children. Over time, and as the children develop cognitively and also expose themselves to wider social experiences, these basic ethno-religious distinctions that they make can be assumed to develop into more elaborate and sophisticated forms to include awareness of the meanings associated with, for example, certain names, colours, faces and residential areas. Viewing the development of children’s sectarian beliefs in this way helps to explain exactly why the particular tests discussed in the previous chapter used to identify young children’s abilities to categorise on a denominational basis have been largely inappropriate and have been unable to identify the extent to which young children think in ethno-religious terms. Children may well only begin to categorise others using the more sophisticated stereotypical cues such as names and faces at an older age. However, as argued above, this does not mean that they do not categorise others at an earlier age, just that they do so in a different, more rudimentary form. However, throughout all of this discussion it needs to be stressed that such conclusions need to be approached with a degree of caution and can only ever be regarded as provisional given that there is no direct data to support them.


The purpose of this chapter has been to draw upon the broader literature on children and prejudice to assess what we can reasonably assume about the nature and forms that sectarianism takes among children. The two core conclusions reached are: first, we can expect that some children will begin to develop a knowledge and understanding of the categories ‘Protestant’ and ‘Catholic’ and be able to negatively evaluate them from about the age of three; and second, that such knowledge and understanding will develop with age as they expose themselves to an increasing array of social influences and contexts.

Clearly, at this present time, we know very little about exactly how young children come to first appropriate and use these two categories. Indeed, it would be quite premature to even speculate what terms they use for these categories. More than anything, the foregoing discussion highlights the importance of approaching the study of sectarianism and young children with an open agenda. As argued in the last chapter, rather than attempting to test their levels of understanding in terms of pre-existing categories and frameworks, it is important that young children are allowed to express what attitudes 

p. 37 p. 38

Jump to Top

   Contents Page

and knowledge they have in their own words and without guidance or prompting. As will be discussed in Chapter Four, the importance of more qualitative, largely unstructured methods is quite clear for children of this age.

However, such a methodology is as important for older children if we are to construct an understanding of how these ideas develop with age. The exploratory study reported in Chapter Five and Six offer a brief glimpse of the type of attitudes and behaviour that 10 and 11 year old children are capable of in terms of their responses to the conflict in Northern Ireland. From some of the explicit views and relatively sophisticated understandings demonstrated at these older ages, it is clear that there is much research to be done to fill in the gaps in terms of how children get to that level of awareness from the ages of three or four.