Full Text Version

Sectarianism, Children and Community Relations in Northern Ireland.

  Chapter 1:
Research on Sectarianism and Children



This chapter focuses on the research that has taken place on the impact and effects of the conflict on children’s lives in Northern Ireland. The aim will be to assess the contribution that such work has made to our understanding of the nature and forms that sectarianism takes among children. To do this, we must begin with a brief discussion of the term ‘sectarianism’ and how it is defined and to be used in the present Report. This, in turn, will provide the basis from which a comprehensive review and assessment of what we currently know about the particular ways in which sectarianism manifests itself in children’s attitudes and behaviour can then take place.

The term ‘children’ has been, and continues to be, used to refer to a variety of different age-groups. For the purposes of this Report, it will be restricted to those of nursery and primary age (i.e. aged 3 to 11 years). Those 12 to 17 years of age will be referred to as ‘young people’. When limiting the definition of children to this age group it is interesting to note how little we know about the nature and influence of sectarianism in their lives. As we will see below, the consistent findings of the research that has been conducted in this area appears to show that while children may begin to develop an ability to categorise others on a denominational basis by the ages of five or six, it is not until about the age of 10 or 11 that the majority of children develop the particular skills required to do so on a consistent basis. If we take such conclusions at face value, therefore, it would seem that concerns about the influence of sectarianism among children are generally unwarranted.

However, it will be argued in this chapter that such conclusions are rather premature. Following the review of the literature, the chapter will look again at the research that has been conducted and will offer a detailed critique of it. It will be shown that the research has been of little help in developing an understanding of the nature and extent of sectarian beliefs among children generally and especially those in their early years. This will provide the core rationale for the argument that an alternative methodological approach is needed. The nature of this alternative will be develop in the following chapters.

p. 5 p. 6

Jump to Top

   Contents Page

Defining sectarianism

‘Sectarianism’ is a term that is frequently used in social, political and academic circles but is rarely defined. For some people, it conjures up images of violent attacks and indiscriminate shootings and is therefore restricted to such actions. Perhaps many more people tend to widen the definition slightly to refer to all actions of discrimination arising from political or religious prejudice. The refusal to employ someone because they are a Catholic, or the belief that all Protestants are bigots, would be two examples that would clearly be labelled as sectarian within this understanding.

The problem with such definitions, however, is that they can often provide a convenient framework of denial. Sectarianism becomes a problem ‘out there’, something that other people do. For many of us, once we have convinced ourselves that we do not hate or harbour any grudges against those of the other main tradition, we can therefore absolve ourselves of any responsibility for sectarianism. Indeed, we may be happy to employ, work alongside and/or mix socially with Catholics/Protestants and thus we believe that sectarianism is clearly not an issue for us.

However, there is an increasing recognition that sectarianism is not just about individual attitudes and behaviour but is embedded within broader socio-cultural, political and economic structures (Brewer 1992; Logue 1993; McVeigh 1995). Thus, while we may, personally, hold no discriminatory or negative attitudes towards others, we may be located in specific social contexts, political organisations and/or economic structures that do tend to discriminate and reproduce religious divisions. Simply by being a part of such contexts or organisations, therefore, our actions may well contribute to these broader discriminatory and divisive processes, however unintentionally.

Of course, because the term ‘sectarianism’ has become so politically loaded and has become largely associated with such overt and political acts of discrimination and violence, it is often very difficult for us to accept that any aspect of our own lives may be tainted by it. While this is totally understandable, it often acts to restrict debate and activities aimed at addressing sectarianism within society. It is therefore important to begin to think about sectarianism more broadly, to take on board its complex and multi-facted nature. As will be seen in the chapters to follow, it is only when this is done will we be in a position to develop programmes of action that are able to more adequately address the influence of sectarianism in our lives.

p. 6 p. 7

Jump to Top

   Contents Page

What follows, therefore, is a working definition of sectarianism. It is by no means meant as a definitive and conclusive statement on the nature of sectarianism. Rather, it is simply offered to set out how it is being used in this present Report and thus to acknowledge the parameters of the problem. It is clear that more work is needed to develop and refine a definition that is not only practical and workable but one which more accurately reflects the complexities of sectarianism.

In the definition to follow, I have drawn upon and slightly revised the definition of sectarianism proposed by McVeigh (1995: 643). At this level, the definition is purposely left open and inclusive to reflect the complex and multi-faceted nature of sectarianism:

Sectarianism in Ireland refers to all of those changing sets of ideas and practices which, whether intentionally or unintentionally, tend to construct and reproduce divisions and inequalities between ethnically-defined Catholics and Protestants.

Such a definition draws attention to the main determinant features of sectarianism. First, as McVeigh (1995) points out, it is not simply a theological dispute but is rooted in the specific nature of Irish society. The divisions and inequalities that it reproduces in terms of the distinctions between Catholic and Protestant are not just religious but are more ethnically based. In other words, being a Protestant or Catholic in Northern Ireland is about a shared sense of community and belonging (i.e. ‘ethnicity’) that is reflected in various customs and traditions that are often only loosely related to the doctrinal beliefs of Catholicism or Protestantism. It is this that accounts for the apparently contradictory fact that you can be a ‘Protestant’ or ‘Catholic’ in Northern Ireland while also being an agnostic or atheist.

Second, sectarianism is therefore defined, ultimately, in terms of all those ideas and practices that contribute towards the construction and reproduction of the ethnic identities of Catholic and Protestant. Most importantly, it shifts the definitional focus towards the outcomes of particular ideas and practices. It is thus not just the specific content of an idea or the intention that lies behind a particular practice that is important, it is rather the effects those ideas and/or practices have. Thus, an individual or organisation may have no intention of being sectarian and/or may be acting in a certain way for the most laudible of 

p. 7 p. 8

Jump to Top

   Contents Page

reasons. However, if their actions nevertheless tend to reproduce divisions and/or inequalities between Catholics and Protestants, in whatever way, then they can be defined as sectarian.

Third, in stressing the changing sets of ideas and practices associated with sectarianism, the above definition also emphasises the context-specific nature of sectarianism. As McVeigh (1995) argues, sectarianism cannot be understood simply as a fixed and static set of ideas and practices but as a combination of social processes that develop and change over time. The particular ways in which sectarianism manifests itself today through popular ideas and practices is very different to the ways it was expressed in the early 1970s. In addition to this stress on the historically changing nature of sectarianism, it is also important to recognise that, within any given period, the particular nature and forms that sectarianism will adopt will change from one context to the next. For example, it will be expressed differently in working class and middle class areas, in urban and rural localities, among young men and women and so on. The key point, therefore, is that sectarianism can only be understood in its specific historical and social contexts. As will be seen in relation to the present Report, this has important implications for how it should be studied among children.

Finally, sectarianism is also, inevitably, about inequalities. For McVeigh (1995: 644) it involves an ‘asymmetrical power relationship’ where, within the broader context of Northern Irish society and how it has been constructed, Protestants have power, as a group, over Catholics. Given the context-specific nature of sectarianism, however, it is important that such inequalities in power are not understood in absolute terms. While the majority-status of Protestants affords them power at the more structural levels of society, it is important to note that in certain contexts and situations, Protestants can also be disadvantaged and discriminated against. The point then is that each expression of sectarianism, and its consequent effects, needs to be studied in its specific context.

There are clear implications from the foregoing discussion in relation to how research on sectarianism needs to be approached. The move away from intentionality, means that particular sectarian ideas and practices may well manifest themselves in very subtle and unintentional ways. Research therefore needs to adopt a focus that begins with the divisions and inequalities that exist and is then able to identify the many different ideas, processes and practices that contribute towards these. As such, it is useful to identify at least four different 

p. 8 p. 9

Jump to Top

   Contents Page

levels within the social formation at which sectarianism may well be reproduced: Subjective, Sub-Cultural, Structural and Systemic levels. It is worth briefly examining each of these in turn


Subjective level


This relates to an individual’s own attitudes and behaviour. The focus here is on how they come to think about themselves and to perceive others. It is also how their own resultant sense of identity comes to influence and shape their behaviour and, most importantly, the many different and often subtle and unconscious pre-dispositions they may have to relating to those from the other main tradition.

The most obvious examples of this would be where an individual holds overtly negative and derogatory views of ‘the other side’ and who, thus, actively avoids contact with them, discriminates against them and/or is intentionally antagonistic towards them. Such practices would clearly be seen as contributing towards the broader divisions and inequalities that exist and would thus be defined as sectarian. However, other individuals may bear no negative attitudes nor malice towards the opposite tradition but may, nevertheless, simply feel apprehensive about them and/or be unsure how to behave towards them. This may well manifest itself in behaviour which can be regarded as ‘cool’ or ‘awkward’ when interacting with those from the other tradition. Again, while there may well be no intentionality in this case, such behaviour can possibly have the effect of reinforcing divisions, however subtly, and thus can also be regarded as a particular process which is sectarian in its outcome.

Given the politically-loaded nature of the term, there may well be a degree of resistance to the use of sectarian in describing this last example. Again, this illustrates the need to reassess and move beyond popular understandings if we are to fully understand and tackle sectarianism. There is clearly no comparison between the above example and a violent sectarian attack. However, both do contribute towards, in very different ways and to very different levels of intensity, the sum of processes and practices that tend to reproduce and reinforce divisions and inequalities.

Sub-cultural level

This level refers to the broader attitudes, behaviours and practices associated with groups of individuals. It refers to all those identifiable sub-cultures that exist in local areas and communities and in specific

p. 9 p. 10

Jump to Top

   Contents Page

organisations and institutions. Most fundamentally, it refers to those situations where particular modes of action and behaviour have developed and are now expected from those in the group. An example could include a group of young people who associate together in a certain area. Certain forms of behaviour, values and knowledge will be distinctly valued within the group and afford status for those involved while other forms may well result in an individual’s exclusion from the group. Another example could be the sub-cultures that exist on the ‘shop-floor’ of a particular factory, or the offices of a specific business. Again, particular forms of behaviour, values and knowledge will develop which will effectively become the markers of those who are included and excluded from the group.

As before, the task is to assess the particular ways in which these sub-cultures, and the interactions that take place within them, may possibly tend to produce and/or reinforce divisions and inequalities. Again, some of these may well be overt and intentional, such as a clear, shared value among the particular group which is anti-Catholic, for instance. However, there may well be other, much more subtle, unintentional and unconscious forms which also tend to contribute to existing divisions. If we take an office environment as an example, this could involve a situation where the majority of workers are Catholic and where the general culture and banter that exists often draws upon knowledge and values from that community. While there may not be any anti-Protestant sentiment expressed, it may still contribute to any Protestant workers feeling uncomfortable or excluded from these interactions. Obviously, the same processes could also be equally true for Protestant-dominated environments and the potential for Catholic workers to feel excluded.

Inasmuch as this latter example may tend to discourage individuals from applying for jobs in organisations which are seen as being largely composed of a Protestant or Catholic workforce, then the sub-cultural processes can clearly be seen as reinforcing existing divisions and thus as being sectarian. The clear point from this example, as before, is the lack of intentionality involved and thus the need for research to be sensitive to the more subtle and unconscious ways in which sectarianism may operate at this level.

p. 10 p. 11

Jump to Top

   Contents Page

Structural level

This level refers more to the overall organisation of particular institutions and how they are managed. Institutions in this regard refer to a range of organisations including: political organisations, social services, the police, schools, churches and the wide range of statutory and voluntary agencies that exist. Examples of overt sectarian practices include those where recruitment and promotion procedures clearly and intentionally favour one group above another or where the services that are provided do the same. However, as before, there may well be more subtle and unintentional forms of sectarianism manifest at this level. For example, guidelines and formulae for funding specific services in certain areas may have been developed with all of the best intentions and with the need to ensure equal treatment. However because of demographic and social changes, it may emerge over time that the particular formulae now operate in such a way as to inadvertently favour one community over another. Thus while there may be no intention to discriminate against a certain group in this instance, failure to monitor funding patterns and accommodate changes as necessary can lead to an outcome that creates difference and inequalities in service delivery and is thus sectarian in outcome.

Systemic level

This is the broadest of the levels identified here and refers to general political and economic processes at play. As Ginsberg (1992) has highlighted in his specific focus on racism in housing, many economic and political processes take place that have no intention of discriminating against certain groups but which, by default, tend to disadvantage some more than others. He uses an example of the UK Government’s ‘Right to Buy’ housing policy of the early 1980s to illustrate this in relation to ‘race’. It was a policy that, for the first time, gave council house tenants the right to buy their own property and proved to be very popular. As Ginsberg points out, it was not a policy led by any form of racist intent, indeed it formed part of the core element of the Thatcherite project of increasing privatisation and enhancing the notion of a ‘property-owning democracy’. Nevertheless, the effects have been to significantly reduce the public housing stock generally and, with the more attractive properties being more likely to be bought, increasingly leaving a residual stock of less desirable properties. The fact that Black people are far more likely to be dependent upon council housing because of their generally disadvantaged 

p. 11 p. 12

Jump to Top

   Contents Page

economic status means that such a housing policy has tended to disproportionately affect them as the number and quality of public housing stock is diminished. For Ginsberg, therefore, while quite unintentional, this is an example of racism at the broader system level.

Similar examples may well be evident in relation to sectarianism in Northern Ireland. For instance, the fact that Catholics are more likely to be employed in the least secure and most vulnerable jobs means that, in times of economic recession, the Catholic population is likely to be hit relatively more severely. This, in turn, will contribute to inequalities between Catholic and Protestants. The point arising from this, as before, is the often subtle and complex way in which sectarian processes may operate and for the need to be sensitive to the different levels in which they can take place.

It should be clear from the above discussion, therefore, that any understanding of sectarianism needs to adopt an approach that can accommodate the many different processes, from the more overt to the more covert and unintentional, which combine to reproduce division and inequality between Protestant and Catholics. It is equally clear that such an analysis needs to recognise the way in which sectarianism operates at all of the different levels of society – from the individual subjective level to the broader economic and political systems that are in place – and how these particular processes tend to influence and shape one another. It is this central task that research on sectarianism in Northern Ireland needs to set itself.

Having set out a working definition of sectarianism and the broader parameters that demarcate its nature and form, it is now possible to assess the contribution that research has made to our understanding of sectarianism among children. Following a detailed review of the literature, the chapter will offer an assessment of it in light of the above definition and discussion.

Research on children and sectarianism in Northern Ireland

When reviewing the research literature on children and sectarianism in Northern Ireland it is interesting to note that, as Trew (1996) has recently pointed out, very little work has been undertaken or published since Cairns’ (1987) seminal review of the research that appeared just over ten years ago. Of the work that had been done up until that time, it appears to have been carried out in response to the outbreak of violence during the early 1970s and the concerns raised by a number of what now appear to have been rather sensationalist and anecdotal research 

p. 12 p. 13

Jump to Top

   Contents Page

studies conducted at the time concerning the long-term effects of such violence on the socialisation and moral and emotional development of the children witness to it (Fields 1973; Fraser 1974). The work reported here, then, is disproportionately concentrated within the period between 1975 and 1986 with only a handful of research studies having been published since that time (see Houston et al. 1990; McAuley and Kremer 1990; McClenahan et al. 1991; Toner 1995) and the remaining literature being composed largely of summaries and reviews of the research carried out over this earlier period (Cairns 1987; Gough et al. 1992; Trew 1992; Cairns and Cairns 1995; Cairns et al. 1995).

The effects of the conflict on children’s attitudes

Overall, such work has been dominated by a plethora of psychological tests which have, because of the sensitivity of the subject matter, attempted to ascertain and measure children’s attitudes towards the conflict and one another using indirect means (Cairns and Cairns 1995: 109). One of the first such studies was published by Jahoda and Harrison (1975) and focused on 60 boys aged six and 10 drawn equally from two schools in ‘troubled’ areas of Belfast - one Catholic, the other Protestant. A further control sample of 60 boys from Edinburgh was also used. All the boys were asked to complete four game-like tasks. The first involved the boys being presented with 12 linear figures dressed in different clothes to represent a range of social roles, four of which were directly related to the conflict (i.e. police officer, soldier, Roman Catholic priest and Protestant minister). The boys were then asked to rank who they most and least ‘liked’. The results indicated that even the attitudes of the younger boys, aged six, appeared to have been influenced by the conflict in that both Catholic and Protestant boys of this age demonstrated a more negative evaluation of the Police Officer than the control group from Edinburgh. Moreover, both religious groups from Belfast gave a low ranking to the ‘outgroup figures’, represented by the Protestant Minister and Catholic Priest respectively. Jahoda and Harrison found that these attitudes developed with age so that the attitudes of the 10 year old boys towards the two religious figures had become stronger and more polarised. Moreover, while the six year old boys from both religious backgrounds indicated a rather neutral attitude towards the soldier, the 10 year old Catholics in Belfast demonstrated a ‘very much more unfavourable’ attitude compared to their Protestant peers in Belfast and both Protestant and Catholic peers in Edinburgh.

p. 13 p. 14

Jump to Top

   Contents Page

One of the other tasks that Jahoda and Harrison gave their sample of boys also supported these findings regarding the influence of the conflict on their attitudes and how these attitudes develop with age. The task involved the boys being given 16 geometric shapes of the same size in four forms (circle, semi-circle, square and trapezium) and each in four different colours (red, blue, orange and green). The boys were then asked to sort the shapes. They found that, at the age of six, boys from both Belfast and Edinburgh showed no overall preference in terms of whether they sorted the objects by colour or shape. However, at the age of 10, all of the Belfast boys sorted the objects according to colour and all of the Edinburgh boys sorted them according to shape. Moreover, while Jahoda and Harrison explained that some of the six year olds who sorted by colour did justify this in terms of religious connotations, they found that over half of the 10 year olds spontaneously mentioned the religious symbolism of their choice, with one child, for example, reported as having justified their actions by arguing that ‘cause red and blue are Protestant colours and orange and green are Catholic colours’ (p. 14).

A number of consequent studies have further shown the influence that the conflict has had on the attitudes of Northern Irish children. As regards younger children, for instance, Cairns et al. (1978, 1980) conducted a comparative study of five and six year old children from a fairly trouble-free area of Northern Ireland with a control group from a South London suburb. Each were shown a photograph of a derelict row of houses and asked to explain what they felt had happened to them. Cairns et al. (1978, 1980) found that the children from Northern Ireland were much more likely to make reference to ‘terrorist bombs’ and ‘explosions’ in their explanations than the group of children from South London. In another study, McWhirter and Gamble (1982) conducted a standard word definition test with a total of 192 children aged six and nine from three different areas of Northern Ireland; one from an area that had a history of sectarian conflict and the other two from different relatively ‘peaceful’ towns. The children were asked to define a series of words including those of ‘Protestant’ and ‘Catholic’ which were embedded within the list so as to disguise the focus of the research. Their conclusions, to quote, were that ‘about half the children in Northern Ireland, at six years of age, have some understanding of at least one of the category labels, Protestant and Catholic, and that by nine years of age the majority of Northern Ireland children are aware, to some degree, of what both terms denote’ (p. 122).

p. 14 p. 15

Jump to Top

   Contents Page

Age-related differences in children’s ability to categorise others on a denominational basis

It is clear from the few studies that have been conducted with young children that, at the very least, they are beginning to develop a rudimentary understanding of the violence in Northern Ireland, witnessed by their reference to ‘bombs’ and ‘explosions’ and that a significant number of them are developing an awareness of the basic categories of ‘Protestant’ and ‘Catholic’. However, the findings reported above in relation to Jahoda and Harrison (1975) and McWhirter and Gamble (1982) do point towards the way in which children develop their knowledge and understanding of the conflict with age. This has been supported by a number of other studies which seem to suggest that, on the whole, children appear not to develop significant and/or consistent attitudes towards religion and the conflict until the ages of 10 and 11. Cairns (1980) for instance reported on a number of unpublished studies which tested children’s ability to recognise stereotypical Catholic and Protestant names (see Cairns 1976, 1977, 1978; Barr 1977). In essence children of various ages were presented with a random list of names including names that were generally recognised as stereotypically Protestant and Catholic and ‘foreign’ names. The children were then asked to recall as many of these names as possible. The tests aimed to measure the degree to which the children clustered the Catholic and Protestant names as an aid to recalling them. The more that they did this, the more it could be concluded that they were aware of this ethno-religious distinction. Overall the findings were relatively consistent and indicated that, as Cairns (1980: 122) summarised, ‘while some children in Northern Ireland are capable of making ethnic discriminations based on first names, at least by age seven years, most children do not achieve this skill until age 11 years or older’. Furthermore, he goes onto state that ‘the accumulated results of these studies ... provide some evidence to suggest that children as young as five years do not discriminate among the names employed here on an ethnic basis’ (p. 123). More recently, these findings have gained additional support from a similar name-recall study published by Houston et al. (1990) involving 80 children aged seven and 11 years from two primary schools, one Catholic and the other Protestant, located in a provincial town in Northern Ireland with little experience of violence. In line with the previous research, they also found that it was only the 11 year olds who could consistently discriminate between the Protestant and Catholic names.

p. 15 p. 16

Jump to Top

   Contents Page

In a similar vein, a series a studies have also attempted to measure the age and extent to which children are able to recognise stereotypical Catholic and Protestant faces. Stringer (1984) identified 20 photographs of male faces half of which were consistently rated as ‘very Protestant’ by his adult respondents and half as ‘very Catholic’. Consequently Cairns (1985) used four of these photographs from each of the two categories and presented them in a random manner to a sample of adults, 10 and seven year old children from Northern Ireland. They were all asked to sort the eight stereotypical faces into two categories - either Irish/Not-Irish for the Catholic respondents or British/Not-British for the Protestant respondents. He found that although the seven year olds were not able to distinguish between the faces, some of the 10 year olds were beginning to develop the ability to recognise stereotypical Catholic and Protestant faces. As regards the adults, Cairns’ findings tended to support those of Stringer (1984) that the adults were in almost total agreement as to what constituted a Protestant and Catholic face. A further study, published by Stringer and Cairns (1983) measured adolescent (14- 15 years) Catholic and Protestant children’s rating of these stereotypical faces on scales such as ‘trustworthy - untrustworthy’ and ‘clever - stupid’. They found that ‘as far as out-group attitudes are concerned Protestants are relatively more ethnocentric than Catholics’ (p. 245) in that while the Protestant adolescents rated the stereotypic Catholic faces in a rather negative manner, the Catholic adolescents did not demonstrate any signs of negatively evaluating the Protestant faces and tended to evaluate their own group as equal to Protestants.

This evidence in relation to age-related differences in Northern Irish children’s attitudes to ethno-religious stimuli is further supported by research conducted by Trew (1981) on a sample of 609 Catholic children aged nine and 11 year old children taken from an area that was, at the time, experiencing a high level of violence. She asked them to give twenty answers to the question ‘who am I?’. While she found that 46 per cent of the children overall referred to themselves as Catholic, this was more likely to occur for the 11 year old children in comparison with the 9 year olds. However, in reviewing this work, Trew (1983) cautions against drawing any premature conclusions from this and refers to a similar study by Lawless (1981) who found that the vast majority (94 per cent) of nine year old Catholic children who were asked directly ‘Are you a Catholic?’ answered ‘yes’. It seems more likely to conclude from this that, as Trew (1983) contends, while the majority of children 

p. 16 p. 17

Jump to Top

   Contents Page

are aware of their religious identity they do not always consider it to be a salient feature of their identity when describing themselves. Thus while Trew’s (1981) own research has shown that the salience of this aspect of the children’s identity in her study did increase with age, this does not mean that younger children are not at least aware of it. Indeed such a conclusion accords with the study by McWhirter and Gamble (1982) outlined earlier where they found that just under a half of six year old children could at least identify at least one of the two labels of Catholic and Protestant. This is an important point to bear in mind and will be returned to in the following sections of the Report.

Children’s political awareness

Alongside these studies which have consistently shown that it is only by the age of 10 and 11 that children in Northern Ireland develop a significant and consistent knowledge and understanding of the ethno-religious categories of Catholic and Protestant, a number of other studies have attempted to explore the nature of that understanding among children aged 10 and 11 and how it has affected their political awareness. One such study, by Cairns and Duriez (1976), focused on 10 and 11 year old children from two schools - one Catholic, the other Protestant - in an area that was relatively unaffected by the conflict and found that both sets of children had already developed a strong sense of national identity. The children were all asked to listen to a tape recording of a man reading a passage in a middle class Belfast accent. They were then asked a series of questions to measure their ability to recall the information included in the passage. A few weeks later the same children were read the same passage this time by a male speaker with either a middle-class Belfast accent, a middle-class Dublin accent or a standard English (RP) accent. What Cairns and Duriez found was that the Catholic children who listened to the RP accent recalled much less of the information compared to the others as did those Protestants who listened to the middle-class Dublin accent. Cairns and Duriez argued that these results indicate that the children have already developed a significant sense of national identity by the age of 10 and 11. As they argued: ‘it is possible that Catholic children possess, as a compliment to their Irish national identity, a negative attitude towards things English including an RP accent. This may have led them to attend less closely and thus retain less’ (p. 442).

More generally, against the background of the findings of McGrath and Wilson (1985) who had surveyed 522 10 and 11 year olds and found that, among other things, 20 per cent reported that they 

p. 17 p. 18

Jump to Top

   Contents Page

had heard a bomb explode, 20 per cent had lost a friend or a relative to the conflict and a further 12 per cent felt that they lived in an area that was ‘unsafe’, it could be argued that this experience would translate into a higher level of political awareness among the children. However, the findings of research conducted on this topic are inconclusive. Interestingly, Hosin and Cairns (1984) reported that Northern Irish children in their study demonstrated a depressed political interest compared to those in Southern Ireland, Jordan and Iraq that they also studied. A sample of children aged nine, 12 and 15 from each of these countries (totalling just under 3,000) were asked to write an essay entitled ‘My Country’. Through a contents analysis of the essays written, Hosin and Cairns found that, at all ages, children from Northern Ireland mentioned politics on fewer occasions than their peers from the other countries. A further study by Whyte (1983) of 91 Protestant and Catholic children aged 11 and 12 from West Belfast found that while they did appear to have a heightened interest in politics generally, evidenced by the fact that more of them (59 per cent) regularly watched the television news and read a newspaper compared to a control group in England, this interest in politics was largely confined to events outside of Northern Ireland. However, while they did not indicate an interest in Northern Irish politics, a significant number of the children clearly expressed an awareness of the conflict in their response to a question asking them what their ‘wishes’ were for the future. Again, alongside the studies of children’s self-identity, we can tentatively conclude that while the children do tend to have an awareness of the conflict in Northern Ireland, this does not always translate into being a salient aspect of their day-to-day lives.

Having said that, however, other research has indicated that some children do exhibit a keen interest in political affairs in Northern Ireland. Cairns and Cairns (1995: 103), for example, draw attention to two large scale surveys of young people conducted in 1982 and 1985 respectively that indicated that between 17 and 19 per cent of respondents claimed to be either ‘interested’ or ‘very interested’ in politics. More reasonably, therefore, it can be claimed that, in line with Whyte (1983), a significant number of children aged 10 and 11 are aware of the broader political context of the conflict to varying degrees, but that this awareness will become more or less salient for particular children depending upon the specific contexts within which they are located.

p. 18 p. 19

Jump to Top

   Contents Page

In summary, while there have been a large number of research studies reported above, adopting very different methods, a number of consistent themes do emerge from the findings. It would appear that by the age of five and six, a significant proportion of children are already developing an awareness of the categories Catholic and Protestant. Moreover, some are expressing an understanding of ‘terrorist bombs’ and ‘explosions’ (Cairns et al. 1978) and the political significance of certain colours while also demonstrating more negative attitudes towards the police and ‘outgroup’ figures such as Catholic Priests and Protestant Ministers respectively (Jahoda and Harrison 1975). However, it would appear from the research reported above that a significant and consistent attitude towards religious identity and the conflict does not develop until a later age. As Cairns et al. (1995: 138) summarise in one of the more recent reviews of the literature: ‘these studies have shown that although children have some understanding of the denominational group labels at an earlier age, it is not until about the age 10 or 11 that the majority of children acquire the specific skills required to discriminate between the two groups’.

Critique of research on children and sectarianism

Clearly, the body of work highlighted above has played an important role in focusing concern on the issue of sectarianism among children. It has also offered some important clues as to how the ways in which sectarianism comes to influence and shape children’s attitudes changes over time. The research reported above tends to suggest that sectarianism begins to impact upon children’s lives at about the age of five or six where they begin to recognise and use one or both of the categories of ‘Protestant’ and ‘Catholic’. With age, children then appear to be able to consistently understand and apply these categories and, presumably, also increase their understanding and knowledge of what these categories represent.

However, if we take such conclusions at face value there is a danger that they can be interpreted in such a way as to support the sentiments that community relations work with children is, at best, unnecessary. At worst, it could be argued that such work could actually be quite dangerous given that it may encourage children to think in terms of the Catholic/Protestant divide who until that point had not yet even developed an interest in doing so. Such arguments are serious and deserve a detailed response. It is for this reason that this final section will return to these broad conclusions drawn and will argue 

p. 19 p. 20

Jump to Top

   Contents Page

that they are rather premature. It will look again at the research outlined above in the light of the definition of sectarianism discussed earlier. It will be argued that, because of the indirect and rather artificial approach adopted by the research, it has tended to overlook the complex and context-specific nature of sectarianism and thus has not provided a particularly insightful guide to the nature of sectarian beliefs among children.

In offering a critical assessment of the existing research literature, it is important to begin by noting that the researchers themselves are only too aware of the potential limitations of their work. It is quite understandable that researchers, especially in the particularly violent and politically volatile context provided by the conflict in the 1970s and early 1980s, should adopt a cautious and indirect approach to the study of the effects all this was having on children in Northern Ireland (Cairns 1987; Cairns and Cairns 1995). However, one of the potential and unavoidable consequences of this indirect approach, as Cairns (1987: 101) explains, is that it ‘raises the possibility that children in Northern Ireland may be aware of the existence of the two major social categories at a younger age than [the researchers] suggest but that this knowledge has not yet generalised to such things as colours or first names’. While this is an important point, it masks the more fundamental limitations to such work. Six of the more central limitations will be outlined and developed here.

1. Inappropriateness of the particular measures used

The first relates to the problems with the particular measures of sectarian attitudes being used. More specifically there is a clear sense in which these tests are applying adult cognitive frameworks and ways of thinking to children without acknowledging the fact that children may well think about and categorise others on a denominational basis in a qualitatively and quantitatively different manner to adults. It should not come as too great a surprise therefore that if researchers choose adult ways of categorising Protestants and Catholics - in terms of the ability to discriminate in relation to names, faces and political colours for instance - as a means of measuring children’s ability to think in terms of the religious divide then the findings will inevitably reach the conclusion that children, and especially young children, are largely unaffected by sectarian attitudes and beliefs. There is thus a clear mismatch between how and what children think in relation to sectarianism and the measures that are being used. To put it another way it is similar to concluding 

p. 20 p. 21

Jump to Top

   Contents Page

that because children may not have an understanding of monetarism, the Exchange Rate Mechanism or the concept of inflation then they therefore have no awareness of economics at all. What is required therefore is more sensitivity to the different and age-related ways in which young children think about and categorise on the basis of ethnicity. In relation to the point made by Cairns above therefore, it is not enough simply to acknowledge that young children may well be aware of the two major social categories at an earlier age but we also need to develop a more appropriate methodology to be able to gain an understanding of what precisely the nature and level of that awareness is.

2. Artificial nature of the methodology employed

Second, attempts to measure children’s social attitudes towards Protestants and Catholics are potentially further inhibited by the essentially artificial nature of many of these tests. The fact that younger children seem to find it difficult to complete many of the tasks given to them may quite simply be due to the ‘alien’ nature of the tasks themselves and their meaninglessness to the children’s lived experiences. Not only are the measures themselves quite possibly unfamiliar to the children but they may also lack motivation to complete the particular tasks set for them because they find it difficult to understand their relevance or utility. There is a danger therefore that the conclusions drawn about the (in)ability of younger children to categorise and think on a denominational basis may simply derive from the particular methods used rather than reflecting reality. Thus by their very nature, such tests do not provide a particularly sound basis from which to develop broader generalisations about the influence of sectarianism on children’s lives.

3. Establishing children’s ability to categorise others as Catholic or Protestant tells us little about their levels of sectarianism

Third, even if a more appropriate and age-sensitive methodology of measuring young children’s ability to recognise and categorise others on a denominational basis could be found, there is then the question of whether establishing the ability to categorise others as Catholic and Protestant actually tells us much about the nature and influence of sectarianism in their lives. As Aboud (1988) has successfully demonstrated in relation to children and prejudice more generally, the fact that a child may be able to exhibit a clear ability to distinguish between different social groups, and even express a strong sense of 

p. 21 p. 22

Jump to Top

   Contents Page

identity in relation to belonging to one of them, cannot, in itself, tell us whether she or he tends to evaluate those groups positively or negatively or even give them much significance at all above and beyond the belief that they ‘belong’ to one of them. In other words, most of us are able to make distinctions between African/Caribbeans, Indians and Chinese people, for instance, but that is no predictor of the types of attitudes that we may have towards these different ethnic groups. This is also the conclusion reached by Stringer and Lavery (1987) who researched ethnic categorisation among university students in Northern Ireland. As they concluded, ‘the ability to distinguish individuals on the basis of stereotyped cues, and the value and use that individuals make of this ability, can be viewed as separate processes’ (p. 163). It is therefore of limited use simply to demonstrate children’s ability to make distinctions on a denominational basis without including some focus on and consideration of the particular values, if any, that the children then ascribe to the different ethno-religious groups they distinguish between. In other words, there is a need to compliment the type of experimental, quantitative studies reported above with more qualitative research offering children the space to express and explore their own values and attitudes. This is a point that will be returned to shortly.

4. Over-rigid and decontextualised understanding of children’s identities

Fourth, even if research can successfully measure the particular values and attitudes that children apply to the social categorisations that they make, there is also a need to develop an appreciation of how significant those attitudes are to the child’s own sense of identity. As Cairns (1987: 106) argues,

while the question of children’s awareness of their social identity is a relatively simple one to answer the question of the importance or salience of this identity poses a greater problem. The problem is that of finding a way to decide if a child or young person’s identity as a Catholic or Protestant is something which dominates the way that child thinks about him or herself than age or gender?

However, even this implies a fairly rigid and static conception of identity where, presumably, one can develop methodological tools to test, measure and compare the differing influences of factors such as religion, 

p. 22 p. 23

Jump to Top

   Contents Page

gender or age in order to ascertain which is the most dominant as Cairns (1987) implies (see also Weinreich 1982). Rather, as Trew (1992) proposes in her own review of the existing research in this area, there is a need to reject the idea that we can measure and identify a person’s identity as if it were a singular, fixed entity which remains constant regardless of social context and, instead, accept that an individual’s identity will change from one social setting to the next (see Connolly 1996, 1998a). Whether the gender, age or religious/nationalist aspects of a child’s identity become the most salient will always depend upon which particular social context they find themselves in. In the classroom when interacting with the teacher, it may well be the child’s age that is of primary importance to how they see themselves (and, in turn, how they are seen) at that instance. In the playground and within the context of a group of male or female friends it may well be their gender which becomes most salient while, when walking home through an interface area between the two main communities, it may well be their religion that becomes the most immediate and influential aspect of their identity.

To be fair, the importance of context in mediating children’s ethno-religious identities is something that has been touched upon by a number of the research studies. Fee (1981), for instance, showed pen portraits of individuals varying in denomination and political beliefs to 100 male students and then measured the students’ perceptions of that figure and their acceptability as a friend. He found that the figure’s religion was not enough, on its own, to stimulate prejudice among the students. Rather, they also needed additional, contextual information concerning the figures’ political beliefs to then make a decision. In relation to children, the study by Cairns et al. (1978, 1980) referred to earlier where they showed children photographs of derelict houses and asked them to say what they thought had happened found that the frequency with which the children mentioned ‘bombs’ and ‘explosions’ reduced markedly between the first time the task was carried out (in 1976) and when it was repeated the following year. As Cairns (1983) explains, on checking the records relating to the actual number of bombs and explosions that took place during these two years, the researchers found that ‘the frequency with which the children mentioned bombs and explosions mirrored an actual decrease in the frequency of such events in Northern Ireland’. More recently, Ferguson and Cairns (1996) conducted a test to measure the moral development of 421 Northern Irish children and adolescents from two areas of high political violence and two areas that have experienced relatively low levels of conflict. 

p. 23 p. 24

Jump to Top

   Contents Page

They found that levels of moral development differed depending upon the area that the children lived with those coming from areas with high levels of political violence scoring at a lower level than their peers.

However, even in recognising the importance of social context in mediating children’s attitudes about themselves and the conflict, the implications of such findings have rarely been explored and researched in any detail. Moreover, even the limited data that does exist fails to capture the essentially fluid, contradictory and context-specific nature of children’s identities as touched upon above and thus recognise the importance of taking account of what Trew (1992: 344) refers to as ‘the multiplicity of identities which can be subsumed under category labels’. As with the previous point, such a recognition raises a number of important questions about the type of methodology that is most appropriate in order to gain such an understanding. While this will be a point that will be addressed shortly, it is clear that the types of quantitative experimental tests employed to date tend to lack the flexibility and sensitivity required to pick up the complex processes and nuances associated with children’s multiple identities.

5. Largely descriptive nature of the studies and their lack of explanatory power

The fifth point concerning the research reported above relates to the fact that even if we accept the findings at face value they remain at a rather descriptive level and offer very little understanding of why children in Northern Ireland develop their ethno-religious awareness in the way that they do. The inevitably individualistic focus adopted by these attitudinal tests, coupled with their indirect and ‘artificial’ nature, prevents any focus on the broader social processes that come to influence and effect children’s sectarian beliefs found at the sub-cultural, structural and systemic levels as outlined earlier. Thus the influence of peer-group relations, the painting of curb stones and political murals, the erection of peace lines, the very real fear of attack and thus the inability to live in certain areas, discrimination in the workplace, the lack of political representation - these are all aspects of the broader social structures experienced by the children to one degree or another and which inevitably come to influence and shape their sectarian beliefs. Moreover, within the context of Northern Ireland, there is also an additional exogenous layer of influence - the political involvement of the British and Irish governments - which provides yet another crucial element in understanding the origins, nature and extent of sectarianism 

p. 24 p. 25

Jump to Top

   Contents Page

(Fitzduff 1995). Thus, in order to develop a more grounded and comprehensive understanding of sectarianism and to explain its particular influences upon children, we need to understand the social interactions and contexts within which they are located. Rather than removing them from their ‘natural surroundings’ we therefore need to study precisely what their experiences and perceptions of these surroundings are and how, exactly, they appropriate and re-work sectarian beliefs to make sense of these.

It is clear from the above that it is only when an analysis of social context is included can we begin to unravel the complex causal relations and processes that can help to understand and explain the salience of sectarianism in children’s lives (see also Connolly 1996). The type of ‘laboratory-style’ attitudinal tests referred to above can therefore tell us very little about precisely when and where children draw upon sectarian attitudes and behaviour in their lives. Moreover, the fact that children do frequently move between various social contexts means that the particular way that they come to express sectarian sentiments will also be effected by the particular setting they find themselves located in at any one time.

6. Underestimation of children’s social and cognitive abilities

The sixth and final point, relates to the way in which researchers have tended to significantly underestimate children’s abilities to think about and reflect upon their own attitudes and behaviour. For Jahoda and Harrison (1975: 3) for instance, their decision not to question children directly about their attitudes was due to their belief that ‘such an approach is ... not suitable for young children’. Presumably, such a belief reflects the cognitive developmental orthodoxy that children as young as five and six are located within what Piaget termed the ‘pre-operational’ stage of development. As such they are assumed, as Connolly (1997: 162) explains, ‘to be egocentric, unable to attend to more than one idea at a time and lacking the ability to think beyond their immediate experience to form more abstract concepts’. From this point of view, as Connolly goes onto argue, it makes little sense to attempt to elicit children’s perspectives directly as they are simply socially and cognitively unable to reflect upon their attitudes and behaviour and express themselves in an appropriate manner. It seems to be fair to conclude that it is partly because of the popularity of such assumptions that research on children and sectarianism in Northern Ireland has been almost totally dominated by the types of structured, attitudinal tests outlined above.

p. 25 p. 26

Jump to Top

   Contents Page

The problem with this type of approach is that it inevitably becomes self-fulfilling. The types of tests developed rarely allow the children to explore their own perspectives and express themselves in ways that are meaningful to themselves. In effect they act to prevent such data emerging and therefore leave the perceived wisdom concerning children’s social and cognitive (in)abilities unchallenged. And yet, relatively recent research that has sought to elicit young children’s perspectives on a variety of subjects including the environment (Cohen and Horm-Wingerd 1993), war and peace (Hall 1993) and racism (Connolly 1998a) has shown that young children’s cognitive understanding and social abilities often far exceeds the self-imposed limits set by developmental theories. Indeed a recent study by Marcos and Verba (1997) has shown that even children as young as 18 months are able to attach meaning to an adult’s attitudes and adapt their own behaviour accordingly.

It is clear, therefore, that there is a need to put aside any preconceptions about children’s social and cognitive abilities and engage with them more directly. It is only when children are allowed to talk more openly and candidly about their own experiences and perspectives that we can begin to develop an appreciation of the complex and context-specific nature of their ethno-religious attitudes and identities. It is interesting that while a small number of commentators have been advocating such a shift in methodology since 1980 (see Beloff 1980, 1989; McWhirter 1983), there is still very little published work that has actually done this. While there will be a number of methodological considerations to address in adopting such an approach, as will be discussed in more detail in Chapter Four, recent research on the impact of racism among five- and six-year-old children in England has shown that it can be done (Connolly 1998a). There is no reason therefore why such an approach cannot be applied to the study of children’s attitudes and behaviour in Northern Ireland.


This chapter began by offering a definition of the term sectarianism and examining its nature and the diverse forms that it takes. It has been argued that there is a need to move away from restricting sectarianism to an individual problem to one which accepts it broader structural nature. As such, it was proposed that sectarianism needs to be understood as operating on at least four different levels of the social formation: the Subjective, Sub-Cultural, Structural and Systemic levels. 

p. 26 p. 27

Jump to Top

   Contents Page

Research on the nature and influence of sectarianism therefore needs to engage with the ways in which sectarian processes operating at each of these levels tend to influence and shape one another.

This general definition of sectarianism provided the basis from which the chapter moved on to assess the contribution that research has made to our understanding of sectarianism among children in Northern Ireland. Through a comprehensive review of the literature in this area it was found that while children may already be developing an awareness of the categories of Protestant and Catholic by the ages of five and six, the general consensus of research opinion appeared to be that a significant and consistent attitude towards religious identity and the conflict does not develop until a later age. More specifically, it is argued that it is only by the ages of 10 or 11 that children acquire the specific skills to discriminate clearly and consistently between the two main communities.

However, in looking again at the research in this area it was argued that such conclusions were a little premature. More specifically, because of the particular ‘adult-oriented’ measures used and the rather artificial and alien nature of the tests set for children, the findings concerning children’s ability to categorise others on a denominational basis cannot be taken to be conclusive. Moreover, the basic finding that such attitudes only concretely develop among children from about the age of 10 or 11 onwards may well be as much a result of the particular methods employed than an accurate reflection of reality. In addition, it was argued that the ability to categorise does not relate in any direct way to the level or nature of a particular child’s prejudice. It was thus suggested that such work, however accurate it may be, is of limited use in contributing to an understanding of the influence of sectarianism in children’s lives. Finally, and in light of the definition of sectarianism outlined at the beginning of the chapter, it was argued that such work is crucially limited by its inability to consider the importance of context in influencing the salience of sectarianism in children’s lives and also the lack of focus on broader social processes and practices which prevents, in turn, an understanding of why children develop and maintain such attitudes.

Overall, it is clear from this chapter that the research community has the potential to make a much greater contribution to our understanding of sectarianism in children’s lives than it has done to date. To be fair, it needs to be recognised that much of the above research never set out to study sectarianism in the way outlined at the beginning 

p. 27 p. 28

Jump to Top

   Contents Page

of the chapter. Indeed, considering that the vast majority of the research has been carried out by psychologists, then their concerns have been understandably limited more to the effects of sectarianism on children’s attitudes and understandings. In many ways, then, we should be careful not to criticise a body of research for failing to achieve something that it never originally set out to do.

However, given that this is all that has been done in terms of research on the effects of the conflict on children, it is a legitimate enterprise to assess what contribution it has made to our broader understanding of the nature and influence of sectarianism in their lives. Ultimately, it is now time for other academic disciplines to make more of a contribution to this area. In particular, it is clear that there is much that sociology can potentially contribute to our understanding given its core concern with the social and structural nature of society. Moreover, from the discussion above, it is also clear that the methodology that needs to be employed should be a more unstructured, qualitative one that is able to tease out the complex and contingent nature of children’s attitudes and accommodate the influence of particular social contexts and processes on these. These methodological concerns provide the focus of Chapter Four and one particular way in which an alternative methodology can be adopted will be modelled out in the exploratory case study reported in Chapter Five and Six.