Centre Publications

Community and Conflict in Rural Ulster

by Brendan Murtagh
Published by the University of Ulster, Coleraine 1999
ISBN 1 85923 119 5
Paperback 64pp £5.00

Out of Print

This material is copyright of the Centre for the Study of Conflict and the author(s) and is included on the CAIN web site with the permission of the publisher. Reproduction or redistribution for commercial purposes is not permitted.

COMMUNITY AND CONFLICT IN RURAL ULSTER
Brendan Murtagh

I would like to thank Dr Dennis McCoy, Central Community Relations Unit for funding the research on which this report is based. I am also indebted to the people who responded to surveys, requests for information and who supported the conduct of the research. Finally, I would like to thank Professor Seamus Dunn at the Centre for the Study of Conflict for publishing the work.

This project was supported by the European Union’s Physical and Social Environmental Programme.

Brendan Murtagh’s work on the sectarian geography and sectarian interfaces of Northern Ireland represents an important contribution to our understanding of the microsocial world of the ‘Hidden Ulster’. It adds to the growing and significant body of related geographical study pioneered by Fred Boal at Queen’s University and added to by Mike Poole and Paul Doherty of the University of Ulster. The Centre for Study of Conflict has been privileged to publish a two volume study by Poole and Doherty (in 1995 and 1996) of ethnic residential segregation in Belfast and in Northern Ireland generally, and we are very pleased to add this new study to our list.

It is perhaps not surprising that our knowledge of the urban - and, in particular, the Belfast - version of Northern Ireland is considerably more detailed and extensive than that of the small scattered ‘country’ towns and their associated hinterlands. The complex patterns of social and religious separation and integration that characterise rural life are often visible only to the insider. In addition, research that tries to map this complexity is often perceived as intrusive and threatening, and this ensures that valid and reliable data are hard to establish.

Dr Murtagh’s study is therefore all the more interesting in that he has managed to apply a whole set of sophisticated research tools to questions of importance about rural life. In particular he examines the nature of conflict in a divided rural community, along with issues of territoriality, how these manifest themselves, and the clearly fundamental role played by territoriality in the continuance of the conflict.

Seamus Dunn
May 1999

CONTENTS

SECTION 1
INTRODUCTION

1.1     Background

Traditional analysis of the geography of conflict in Northern Ireland has concentrated on urban environments and, in particular, the regional capital, Belfast. Boal’s (1969) seminal work on territoriality set the context for studies on sectarian demography (Poole and Doherty, 1996), the process of segregation (Keane, 1990) and the experiences of enclave communities (Darby, 1986). There has been a number of locality based studies following ethnographic methodologies in rural towns and villages (Donnan and McFarlane, 1986) but the balance of empirical research has been with survey and analysis of urban communities. Moreover, the literature on rural geography, development and planning has paid scant attention to the ethno-religious dimensions to life in rural Ulster. Indeed, much of the available information on rural experiences of conflict are found in local histories (Adams, 1994), travel literature (Murphy, 1978) and poetry (Heaney, 1975).

This research attempts to fill, in part, the research deficit on rural conflict and community life. It is based on a number of research questions regarding space and society in Northern Ireland and, in particular, questions whether the processes that have produced 14 peacelines in an urban context produce equivalent spatial relationships in rural areas. There is clearly no reason to suggest that the sorts of social and community processes that created brutal boundaries to sectarian space in Belfast end at the green belt, and so the relationship between spatial and cross-community relationships forms a central focus to the empirical research.

1.2     Research Objectives

The research was guided by three objectives:

• to investigate community relations attitudes and behaviour in rural Ulster;
• to specifically explore the presence of an ‘interface’ in a rural community;
• to draw implications for policy and community relations practice in Northern Ireland.

1.3     Research Design
The problems associated with researching social attitudes and community contact have been well documented in Tutt’s review of community relations research in Northern Ireland. He highlights the problems using qualitative and quantitative data to assess valid attitudes and behaviour, shifts in attitudes over time and predicted behaviour (Tutt, 1994). In particular, he questions ‘the appropriateness of research designs which consistently employ uni-factorial analysis of the community. Research which highlights divisions rather than coherence within the community may, if taken in isolation, not fully inform policy. Moreover, its publication may, in some instances, emphasise divisions in society’ (Tutt 1994, p. 19). Knox and Hughes point out that there is no single definitive method or ‘right way’ to undertake community relations research and that ‘there is merit in combining both a qualitative and quantitative approach in evaluating community development policies’ (Knox and Hughes 1994, p. 248).

Boal has also picked up on a number of related weaknesses in research on the geography of segregation and recognised that ‘using participant observation and exploration greatly enriches survey oriented or census approaches’ (Boal, 1987, p. 120) and he laments that geographers have not been notable as ‘urban explorers’ (Boal, 1987, p. 120), nor as geographer anthropologists have they been prepared to live in the areas they wish to study as participant observers. Boal consistently highlights the need for new case study analysis to understand the processes of segregation and their impact on community cohesion and fracture. Drawing on this short review of methodological advice, a combination of research methods were applied within a single case study design. The selection of the study area, weaknesses of the method and steps taken to improve data quality are discussed below.

1.4     Selection of the Case Study
The case study area was selected in three stages. The first was to confine the search focus to areas of high rural violence. Poole’s index of violence provided a useful tool to target District Council Areas with significant rural populations that experienced the highest level of violence. Poole concluded that ‘rural violence’ is governed by proximity to the border,

‘... sixty two percent of the rural fatal incidents which have occurred have been concentrated into six District Council Areas adjoining the border, especially Newry and Mourne, Dungannon and Fermanagh. In fact, the rate of fatal incidents per 1,000 people for rural areas is 1.38 in these six border District Council Areas but only 0.48 elsewhere in Northern Ireland’ (Poole, 1983, p. 175).

At the second stage a focused group discussion was then held with Community Relations Officers (CROs) from these six districts together with Armagh, Strabane and Limavady (also identified by Poole as areas of high rural violence, Poole, 1983, p. 171). The aim of this discussion was to consider the nature of rural conflict, community relations and cross-community attitudes and areas of interest within each of the districts. At this stage the Newry and Mourne and Armagh District Council Areas where identified as areas for further analysis. The initial discussions showed that these areas had experienced high levels of violence, internecine conflict, population shifts between Catholic and Protestant heartlands, localised problems relating to marches and traditional festivals and problems related to segregated schooling.

The final part of the case study selection process involved a further group discussion with local CROs, community activists and representatives from the Rural Development Council. This discussion attempted to draw a mental map of high conflict across the middle part of the study area. This established a centre line that would form the focus for empirical data gathering (see Figure 1.1 at end of section). The journey along this line was not designed to be prescriptive but rather to guide initial investigation, scoping communities and defining issues and areas for further analysis. The spatial concentration of urban communities in urban areas lends itself to targeted area based survey approaches.

Where communities are dispersed, contact relationships are less precise and alternative spatial interactions have a more complex relationship with social processes. More investigative, less precise approaches, or what Measor refers to as ‘rambling’ (Measor, 1985, p. 67) research is required. By necessity this involves the application of qualitative and ethnographic based research techniques and these have been combined within a case study design as shown below. Table 1 a shows the overall design of the research which begins at the broad spatial scale and progressively narrows to focus on the mid-Armagh case study.

Table 1a Research Methodology

Demographic analysis sets the context for more focused analysis of life in small town Ulster, a two village survey and observational research in the rural study area.

1.5 Case Study Methodology
The Armagh case study involved secondary analysis of census data, a small scale survey and observational methodology. Yin (1994) defines the methodological principles to be followed to ensure that data collected is both valid and reliable. These principles, and the response to them in this research, are shown in Table lb.

Construct validity
In case study designs the problem of construct validity relates to the development of an operational set of measures and the subjective judgements that are used to collect them. To counter this, Yin suggests the use of multiple sources of evidence and being explicit about the development or chain of analysis in the process of data collection. As Table 1a shows, an attempt has been made to triangulate in the design by combining analysis of secondary data, quantitative survey and qualitative observations.

Internal validity
By this we mean inferring that a particular event is related to a particular causal factor or factors. The process of pattern matching compares an empirically based pattern with a predicted one. If the patterns coincide, the results can help to strengthen the internal validity of the case study. The point of departure for this research was the established patterns of territorial behaviour among Belfast’s peaceline communities. Replicating the socio-spatial dynamic that created fourteen physical barriers between the two communities is the central empirical focus for the study. If similar patterns can be observed in rural Armagh as on the Falls Road, then internal validity will be supported.

External validity
Yin defines the third test as external validity which deals with the problem of knowing whether a study’s findings are generalisable beyond the immediate case study.

‘A common complaint about case studies is that it is difficult to generalise from one case to another The analyst falls into the trap of trying to select a "representative" case or set of cases. Yet no cases, no matter how large, is likely to deal with the complaint’ (Yin, 1994, p. 37).

In his approach to neighbourhood change he therefore argues that a theory or hypothesis must be tested through replications of the findings in a second or even third neighbourhood, where the theory has specified the same results should occur. Once such replication has been made, the results might be accepted for a much larger analysis of similar neighbourhoods. Given the pilot nature of this design there are inevitable weaknesses in the extent to which generalisations can be drawn from limited neighbourhood field work.

Reliability
The objective of reliability is to minimise and control error and bias in data collection, interpretation and presentation. Yin gives clear guidance on the need for transparent and detailed documentation of approaches and research output. Thus an auditor could repeat the same procedures and arrive at the same results. Table 1 a shows the approach adopted in this design, the relationship between the various elements of the research and the embedded location of case studies in the design.

1.6     Structure of the Report
This section has defined the broad direction for the research and has outlined the combination of methodological approaches employed within a case study design. Section 2 uses survey data to highlight differences in community relations attitudes between urban and rural areas. Section 3 introduces census data to provide an important demographic context for small scale survey data in subsequent sections. These are initially developed in Section 4 in an analysis of life in a small Ulster town. Section 5 explores life in two villages and the issue of interface is developed in Section 6 in which of the significance of land and ownership are discussed. The implications of the research are circumscribed by the limitations of a case study design but preliminary ideas are drawn out for discussion and further research.

Figure 1.1 Study Area

SECTION 2
LIFE IN RURAL ULSTER:
URBAN AND RURAL SOCIAL ATTITUDES

2.1 Introduction
Large scale attitudinal surveys are now commonly used in Northern Ireland to measure and monitor social distance between the two communities (Gallagher 1995). The Baseline Study for the Peace and Reconciliation Programme for Northern Ireland and the Border Counties of Ireland used province-wide attitudinal research to evaluate the impact of the fund (Stutt et. al., 1996). The use of research of this type experiences problems of validity, in that, it often fails to link recorded attitudes with actual behaviour. For example, Social Attitudes Surveys in Northern Ireland have consistently shown supportive attitudes to integrated living, despite that fact that segregation levels in urban areas are high (Poole and Doherty, 1996). However, with these limitations in mind, this section aims to compare and contrast differences in community relations attitudes in urban and rural areas. This should be viewed as a crude first attempt to illicit differences in the attitudes and values that might produce different social relationships in rural space. It also provides useful benchmark with which to compare indicators from individual locality studies conducted as part of this research.

The survey sampled 1050 people across Northern Ireland during October and November 1995 using random systematic sampling techniques. The urban and rural classification is based on DOE(NI) planning guidelines which categorise settlements in Northern Ireland by type and whilst this does not provide a pure system it is useful to create simple analytical categories for the purposes of analysis.

2.2 Social Distance in Urban and Rural Areas
The survey data suggested that rural areas were more likely to be religiously mixed than urban communities. Forty four per cent of rural dwellers defined their neighbourhood as mixed compared to 34.6% for urban areas.

Table 2b shows some degree of divergence in attitude in the main analytical categories. However, it was ‘religion’ rather than ‘location’ that explains variations in the data. Overall, 12% of the surveyed population wanted to live in a neighbourhood of their own religion compared to 76% who preferred mixed housing. While the proportion preferring own-religion neighbourhoods is higher for urban than rural areas, the Table shows that Protestants were more likely to prefer own religion localities. Indeed, the contrast is most marked in rural areas where 18.3% of Protestants preferred own religion neighbourhoods compared to 4.6% for Catholics. However, roughly even proportions in urban and rural areas had friends of their own religion. Seventy eight percent of urban dwellers are in this category compared to 79% for those in rural areas. Similarly, 91% of those in both urban and rural areas had friends mostly of their own religion.

Table 2c shows that attitudes in rural areas were more positive than in urban areas. A total of 73.6% urban dwellers defined community relations as good or very good compared to 89.7% of rural respondents.

2.3 Experience of Daily Life in Urban and Rural Areas
The more positive attitude to community relations in rural areas was supported by the daily experiences of people in the two settlement types. For example, 4.4% of urban dwellers stated that they had problems using key services and facilities because of the sectarian divide compared to 0.5% in rural areas. Moreover, 18.9% of urban and 13.6% of rural dwellers had moved house in the last three years. While no-one in the rural areas identified sectarian factors in their decision to move or choice of location, 11.7% of those moving in urban areas cited this as a factor.

The broad perception, that rural areas enjoy better community relations than urban localities, is supported by Table 2d. Most respondents (67%) felt that rural areas were less violent than urban areas. However, this is more keenly felt among the rural community (76%). Similarly, 65% of respondents thought that rural communities were more integrated and again, this was more strongly felt by rural dwellers (72%) than those in urban areas (62%).

There was significant divergence between the two groups when the health of community relations was measured. Again rural respondents felt that community relations were better in their area than in urban locations but less than a quarter (24%) of those living in urban areas felt the same way.

2.4 Land and Territory
The general picture to emerge from the attitudinal survey is of a harmonious, less divisive and more optimistic rural community particularly when compared to urban based respondents. However, settlement ceases to be a meaningful explanatory variable when attitudes to specific practices are measured.

* Rural areas only

For example, 38% of urban respondents felt that traditional parades should be allowed to march through the area of the other ethno-religious group (out-group) compared to 36% in rural areas. However, 52% of Protestants agreed with this statement compared to 20% of Catholics. Moreover, one-third (33%) of the Northern Ireland population agreed that people in rural areas do not generally sell their land to people of a different religious denomination. This was a view more widely held in rural areas (35%) than in urban area localities (32%). Most people in Northern Ireland stated that they would sell their land and property to members of the opposite religion (80%). Nine per cent said they would not do this. This was lower in urban areas than in rural localities where 16% said they would not sell to a member of the out-group. There seems to be an important contradiction in the attitudes of the rural community, in that, general measures of social distance would seem to suggest more negative practices among rural communities.

2.5 Conclusion: The Pivotal Role of Land
Central to this contradiction is the issue of land. Stockdale highlighted the ‘intense emotional dimension to land ownership’ (Stockdale, 1992, p. 419) as a key variable in property ownership and development in rural areas. This broad statistical analysis has shown that it is as much a problem in rural communities as it is in urban areas. How it is manifest is a key task for this research but the attitude to land is clearly a key variable to examine.

3.1 Introduction
Demographic trends form a crucial backcloth to any ethno-religious study of local communities. The literature pays particular attention to the relative numeric proportions as a deciding factor in the retention or exit decisions of a particular ethnic group. Darby (1986) observed that perceived threat in the social environment has significant consequences for community demographics.

‘General environmental pressures result, not from changes within the community, but from pressures on the community itself These pressures have been strong enough to produce enforced population movements, not because individuals have been attacked or threatened, but because the communities to which they belong have themselves become isolated and vulnerable. There are many examples of communities becoming detached from their broader religious heartlands through gradual demographic change’ (Darby, 1986, p. 55).

Population balance is also an important conditioning factor in the type and intensity of inter-group contact and therefore behaviour. In a fascinating study of racial attitudes to school desegregation in Texas (USA), Pettigrew and Riley (1972, p. 183) noted that ‘southern counties with a large percentage of Negroes tended to have smaller lynching rates when the size of the Negro community is held constant’. This graphic portrayal of the influence of demographic balance may have important implications for Armagh where segregation rates vary considerably.

A context is set for the review of the data by examining recent demographic shifts, and, in particular, rural re-population in Northern Ireland. This analysis is divided into three main parts. The next subsection looks at current population size and characteristics but focuses on the difference between the two communities and segregation at ward level. This is followed by an analysis of trends since 1971. Trend analysis at the micro spatial level is beset with problems as administrative boundaries have changed considerably since 1971. However, the analysis attempts to infer general conclusions from the available data. The final sub-section highlights some key socioeconomic differences between the two communities and profiles the area against trends at the Northern Ireland regional level.

3.2 Demography and Ethno -Religious Space
In a recent exhaustive account of segregation outside the Belfast urban system, Poole and Doherty (1996) describe the processes of change in towns over 5000 people in Northern Ireland. They classify the 39 towns they examined according to the levels of Dissimilarity and Dominance. The Dissimilarity Index measures unevenness in the spatial distribution of the population while the Dominance Index measures the deviation of a social composition from a situation where each social group is found in equal numbers. Finally, isolation is analysed using an ‘exposure index ... which measures the extent to which minority members are exposed only to each other, together with an interaction index which measures the extent to which the minority is exposed to the majority population’ (Poole and Doherty, 1996, p. 26). Applying each indicator gives a different analysis of division at town level. For example, they found that high dissimilarity combined with high dominance happens in only 5 out of the 39 towns but that these contained 58% of Northern Ireland’s urban population. However, there are 17 towns classified as highly segregated in terms of dominance and they contain 78% of the province’s population.

Of the 20 towns whose change was analysed between 1911 and 1981, 15 experienced an increase in the Catholic share of their total population. ‘In consequence, Protestants adjust to the threatening Catholicisation of their town by moving house just enough to ensure that they continue to live in the same kind of local environment as before’ (Poole and Doherty, 1996. p. 248). They compare the results of their research with that of Massey and Gross (1991) on American housing where a fundamental control on the amount of desegregation which could take place was the attitude of the white community to having a black minority in its neighbourhood and they suggest as little as 5% as a threshold. Poole and Doherty argue that analogous Northern Ireland thresholds are substantially higher in most places but that Protestant populations engage in a degree of residential adjustment.

3.3 Population Change in Rural Ulster
Rural population shifts have received considerable attention in academic literature. In a broad based analysis of population change between 1981 and 1991 Shuttleworth (1992) described the broad process of loss in the Belfast Urban Area and net gain in rural areas. The District Council Areas (DCAs) with the highest rates of increase were Derry, Newry and Mourne, Limavady and Magherafelt. These DCAs experienced rates of natural increase between 11% and 16% over the 1981-1991 intercensal period. Shuttleworth commented on the rate of increase:

Stockdale has detailed the process of ‘counter-urbanisation’ and has identified the DCAs of Armagh and Newry and Mourne as experiencing the highest levels of rural population growth. This process of rural repopulation dates back to the mid 1970s when ‘the distribution pattern of rapid population growth can be described as representing isolated pockets with three principal areas of concentrated growth: the border region of Newry and Mourne DCA; the Lakeland area of Fermanagh DCA and the south-west of Londonderry DCA’ (Stockdale, 1991, p. 76). By 1985-87 not only were these wards involved, but the previous ‘pockets’ had been consolidated to produce several major axes of growth in which Newry and Mourne was one.

In a positive departure from the rather descriptive nature of much of the research, Stockdale (1992) posits three explanations for rural re-population: voluntarist, non-voluntarist and intervention. Adopting a behavioural approach, Stockdale firstly emphasises the importance of human motivation and individual residential choice as a factor influencing this trend. Secondly, the importance of structural forces operating in society as a whole is stressed. Finally, and most importantly for Stockdale, intervention is seen as the factor most influencing the ‘turnaround’. In particular, she argues that the relaxation of planning controls associated with residential developments in the open countryside since 1978 is the key causal factor:

This factor will be further explored in this study but the significant point is the extent to which state intervention and the use of planning instruments can be used to control differential demographic shifts in peripheral rural areas.

3.4 Population Size and Structure
The overall population of the study area is 77,538 according to the 1991 Northern Ireland Census. Of this total 44% (38,837) is in the north sector and 17% in South Armagh. A total of 40% (30,761) is in the interface zone which is the main empirical focus for the research. Immediate differences in the population characteristics of the two religions become apparent when average household size is examined. Compton (1995, p. 69) points out that ‘the religious differential is most marked and persistent in family size’. In the interface zone the average household size was 3.7 persons and 3.0 persons for Catholics and Protestants respectively. At the last 1991 Census the overall household size for Northern Ireland was 3.2 persons.

Figure 3.1 (see end of section) shows that in the north zone, Protestants and Catholics are in relatively even proportions. However, the population pyramid does reveal some important distinctions between the two religions. Firstly, there are greater numbers and proportions of Catholics in the younger age cohorts. In the nortin the younger age cohorts are relatively fewer and the proportion in the older age cohorts relatively higher. In the interface zone, more than one-fifth of the population are young Catholics (under 16 years old) while young Protestants make up just 7% of the total population.

Indeed, there are more Protestants over 50 years old than there are under 16 years of age. The reverse is the case for the Catholic population where 12% of the population are over 50 years old.

This has significant implications for population growth in that the Catholic population has greater long term potential for renewal while the Protestant population is ageing relative to Catholic trends. In the interface zone 59% (18,212) of the population are Catholic compared with 28% (8,646) for Protestants. This numeric difference coupled with differential demographic structures adds to the vulnerability of the Protestant community. Compton and Coward have noted the wider cultural and political significance of these trends:

The link between the political aspirations, demography and territory is perhaps most vividly demonstrated in South Armagh. Only 1% of the population is Protestant. Moreover, nearly one third of the Catholic population is under 16 years old.

3.5 The Spatial Distribution of Religion
Figure 3.2 (see at end of section) maps out the distribution of Protestants and Catholics at ward level. The main feature of the maps is the monolithic block of Catholic South Armagh where the proportion of Catholics in each ward exceeds 90%. However, the maps also emphasise the relative segregation at ward level in the interface zone. Indeed, of the 32 wards that make up the wider study area only 7 are ‘mixed’, (defined by having Protestants and Catholics in a proportion of between 30% and 70% in each ward).

Finally, the analysis shows the relative concentration of Protestants in the north part of the study area, particularly in the wards to the south and the east of Armagh city. While the graphic symbols of segregation in highly concentrated urban communities may be absent here, the data show that ethno-religious segregation is a significant feature of living patterns in rural communities. Given this backcloth, it would be surprising if segregation and integration did not condition inter-community attitudes and behaviour. The primary aim of the empirical research will be to explore how the two communities act out differences in a rural context. This will be strongly influenced by the changing levels of the existing balance of the population.

3.6 Trend Analysis: Border Flight?
It has been noted previously that changes in administrative boundaries make it difficult to compare precise population change over time. However, it is possible to consider broad demographic shifts and, in particular, differential changes between the two communities in the study area. The analysis compares census data between 1971 and 1991. The overwhelming trend has been the reduction in the Protestant population from around 19% in 1971 to 1% in 1991. The corresponding increase for Catholics was from 67% in 1971 to 91% in 1991.

Precise changes in the religious balance of the population are possible when specific settlements are analysed and the case of Keady illustrates the relative ‘greening’ of this part of Armagh. Between 1971 and 1991 the proportion of Catholics rose by 10% from 79% to 89%, whereas, the Protestant population declined from 9% to 3%. The reverse is the case in Tandragee where the Protestant population increased from 46% to 64% in the same period while the proportion of Catholics declined from 16% in 1971 to 13% in 1991. The dominant demographic trend seems to be a proportionate increase in the Catholic population in South Armagh and a proportionate increase in the Protestant population in the north of the study area.

This is having a profound effect on the interface community which is experiencing these centrifugal forces first hand. This is particularly important where one community is in a minority. In most cases this will be where the minority population is Protestant, declining and displays relative ageing in the demographic structure. Population structure and change are key variables that condition inter-group attitude and behaviour.

3.7 ACORN Population Analysis
The use of ACORN data to profile socio-economic conditions in the area provides useful three dimensional summaries of the distinctions between the three broad zones. ACORN is a cluster analysis based tool, which, while heavily used in marketing, has important policy relevance. It draws together a range of census variables to produce ‘lifestyle’ profiles that go beyond the descriptive analysis. Table 3a shows the three areas as described by the ACORN typology and provides a useful descriptive summary of the study area.

The main point to note from this table is the changing social structure when moving south through the study area. The dominant group is Protestants in Group 56. This is titled Affluent Rural Communities with Good Communications Links and Low Unemployment reflecting the link between that community and urban centres in north Armagh and Belfast. A smaller proportion of Protestants are in Group 10, Affluent Working Families with Mortgages which is characterised by households in the higher socio-economic classes who ‘love buying new gadgets and appliances and are very keen to keep up to date with new technology’ (CACI, 1993, p. 22).

Catholics in Group 58 are a feature of all areas but increase in importance toward the south zone. This Group, called Local Occupations, Some Unemployment and Large Families, is characterised by employment in agriculture, high unemployment (an average of 24%) and 22% have an average household size above six persons (CACI, 1993, p. 72). Group 57 again dominated by Catholics, is similar to Group 58 and is titled Remote Properties and Small holdings, with Employment in Agriculture and Tourism is characterised by poor infrastructure, land quality and housing standards. The final Group in Catholic South Armagh is 33 and is described as Council Areas, Some New Home Owners.

• This analysis provides an important backcloth for the detailed study of attitudes and behaviour within the studyarea. A number of points need to be emphasised.
• The overall study area is highly segregated and only 7 of the 32 wards could be considered ‘mixed’.
• Protestants tend to concentrate in the wards to the north and east of the study area, particularly around the city of Armagh, whereas, all the wards in the south are in excess of 90% Catholic.
• Trend analysis shows that, over time, the north of the study area has become increasingly Protestant whilst the reverse is the case in the south zone.
• These changes have had a significant impact on the interface area where the day-to-day processes of changes are having most impact on delicately balanced ethno-religious communities.
• There are significant demographic differences in the profiles of the two communities, in that Catholics tend to be characterised by larger household size, a younger age structure and relatively fewer numbers in the older age cohorts than Protestants.
• The sustained decline in rural population has been reversed in this part of Northern Ireland. This is in common with trends elsewhere in the Province.
• The study area has a generally poor social and economic profile with a reliance on a poor agricultural base to the south, resulting in relatively high unemployment rates.

4.1 Introduction
In their respective studies of small town life, Leyton (1975) and Harris (1972) have pointed to a common regional loyalty in rural districts and to what Leyton called ‘shared values’ between Catholics and Protestants. Darby (1986) argued that recovery from conflict and intimidation in mixed towns has been eased by a common desire to preserve an integrated community. This section reviews the literature on small town life and community conflict. As the analysis shows, this is principally anthropological, which in itself has inherent dangers for understanding the significance of ethno-religious cleavages in spatially defined communities. Empirically, the section looks at life in a town, called here Oldtown, in the West of the study area. It has experienced a high degree of violence, particularly in the 1970s, and the town centre is dominated by a large police base. The town is the centre for the local agricultural economy with hardware, feed supply and the local ‘mart’ linking the town to its rural hinterland. This section links the analysis of survey data with in-depth interviews conducted in the locality. The population of the town at the 1991 Census was approximately 3600 people with a religious split of 59% Catholic and 39% Protestant. A sample of 400 households was drawn from the electoral register and a total of 202 (51%) responded to the postal survey.

4.2 Communities and Conflict in Small Town Ulster
Whilst the literature on rural segregation and conflict is not as well developed as the case material for urban environments, there are some important studies on the experiences of micro-communities. The seminal work in this area was carried out by Rosemary Harris in the early 1950s in an area she called ‘Ballybeg’. Written in a period of relative peace it not surprisingly highlighted the frequent and positive nature of contact between neighbours. However, she also highlighted the existence of quite separate worlds were religion and social institutes dictated mutually exclusive patterns of behaviour and interaction. Describing shopping patterns in the village she pointed out that ‘the advantage offered by one shop over its rivals had to be very considerable before a Protestant owner could attract Catholic customers and viceversa’ (Harris, 1972, p. 6). Leisure activities were also highly segregated in Ballybeg partly because they were organised by the respective churches and she pointed out that mixing was often the preserve of middle class social contexts in tennis and golf clubs.

Leyton’s (1975) study of the small Protestant rural community of Perrin observed that its inhabitants ‘see their village as a bastion of Protestant morality and Protestant virtue’ (Leyton, 1975, p. 11-12) but highlighted that in areas experiencing high levels of violence Protestants emphasised their political rather than their religious identity. Similarly, in their study of a small border village of Daviestown, Hamilton et. al. (1990) highlighted the damaging consequences for community relations of a prolonged paramilitary campaign in the area.

‘In such a small and close knit community, the resultant deaths had had a traumatic effect, amusing suspicion and fear and leading to an almost total polarisation and lack of understanding between Catholics and Protestants ... The violence had strengthened the constraints which had always existed, leading to increased social segregation and polarisation’ (Hamilton et. al., 1990, pp. 54-55).

In an extensive review of the anthropological literature on locality conflict, Donnan and McFarlane (1986) highlighted the significance of diverse social, kinship and ethnic cleavages in rural life. Table 4a attempts to summarise their analysis of the relationship between different cleavages in micro-societies. Kinship patterns are important in establishing new business while neighbours are more likely to be involved in sharing arrangements between farmers (Leyton, 1975). Political and religious differences are evident in the attitudes of the respective communities to their identity and self description of core values as well as the way in which local services are used on sectarian lines. In her work, Harris (1972) highlighted the common identity produced by attachment to place and the sense of ownership and pride in a local community and its long term success. Finally, Donnan and McFarlane highlighted the significance of local social structures and hierarchy based on influence and wealth which cut across kinship, religious or community bonds.

* Based on Donnan and McFarlane (1986).

Thus, Donnan and McFarlane conclude that it is difficult to say whether kinship, social class or religion is the determinant variable in explaining social relations generally and inter-group contact in particular.

Writing from a geographical perspective, Kirk (1993) offers one approach to prioritising types of social contact by distinguishing between individual interests (for example, by sharing labour and machinery between farmers) and the group interests of preservation. In his exhaustive analysis of land transfers, he points out how Protestant and Catholic farmers accepted lower values for land by selling it within the ethnic group. Thus, he concludes,

The survey presented below also suggests that anthropological studies have tended to underplay the significance of ethnic cleavages. Adams (1995) analysis of Cashel, Co. Fermanagh highlighted the danger of over-reliance on a single research design in community conflict. In this recent contribution to the anthropological literature, Adams (1995) highlighted the importance of a sense of common community in overcoming religious tension in the village,

However, her report contains some significant contradictions in that it also touched on the presence of separate schooling, different sporting and social infrastructure and the exodus of Protestant families from this border region after the murder of a number of local members of the security forces. This, in part, explains some of the weaknesses in the anthropological literature for it almost always treats the community under research as a closed system unconditioned by the socio-political and economic context around them. Donnan and McFarlane recognise the limitations of research ‘carried out in communities without an eye to events which are taking place outside the community’ (Donnan and McFarlane, 1986, p. 135). For example, an analysis of the attitudes and behaviour of Protestants in Oldville would be incomprehensible without an understanding of the wider demographic changes in the area and, in particular, the shift of the Protestant population away from the border. These condition attitudes which themselves affect population movements as feelings of insecurity are acted out by the people moving to more secure heartland territory.

4.3 Religion and Demographic Change in the Town
The household survey showed that the population of the area is relatively settled with roughly even numbers of Protestants (89.9%) and Catholics (79.9%) resident in the town for longer than 10 years. However, Protestants were much more likely to have lived in the town all their lives (63.3%) than to Catholics (41.3%). This is significant given the differential demographic shifts in the community over the last 10 years as this part of the study area has become increasingly Catholic in composition. As noted before, these demographic shifts have had a powerful impact on the behaviour and attitudes of isolated communities. Indeed, Protestants were more likely to describe the town as Protestant (75.5%) rather than Catholic (24.4%) and more likely to want to live in an area of their own religion (46.9%) compared to 25% of Catholics who would prefer to live in own religion neighbourhoods (Table 4b).

It is difficult to reliably compare community relations attitudes in Qidville with other localities or to gauge whether rural areas are more or less integrated than urban environments. However, Table 4c compares Qidville with other areas and surveys on attitudes to integrated living. This shows that attitudes among both Protestants and Catholics are more extreme in the town than among all rural dwellers and Northern Ireland as a whole. Only people living in Belfast’s peaceline zones hold more negative attitudes to the question of mixing. Again, this would at least question the hypothesis that rural communities are generally more harmonious than urban ones. Moreover, the available data suggest that the history and experience of violence and the sense of community sustainability are key determining variables explaining community attitude and behaviour.

1. Gallagher (1995, p. 19)
2. Murtagh (1994)
3. Rural Interface Research Project (see Section 2)

The reality of demographic changes was reflected in the 61.2% of Protestants who felt that the area was becoming more Catholic compared to 32.5% of Catholics who felt that way (see Table 4d below). Only 2% of Protestants felt the area was becoming more Protestant while 34.7% felt that it had stayed the same. Catholics were most likely to feel that the religious composition of the area would stay the same (56.3%).

4.4 Self Perception of Identity
In common with other studies on self-expression of identity most Protestants expressed themselves as British (55.1%) or Protestant (24.5%). Catholics were most likely to describe themselves as Irish (71.3%) or Catholic (17.5%). In peaceline areas in Belfast, Protestants were more likely to call themselves British (51.7%) or Protestant (22%) while Catholics in Peaceline areas were more likely to describe themselves as Irish (44%) or Catholic (33.3%) (Murtagh, 1994). Dunn notes the powerful interaction of identity and the desire for separate living and territory,

The consequences of these processes are lack of understanding, reinforced nationalism among respective communities and ultimately fear and suspicion. The survey revealed these processes at work in the micro-community of Oldville particularly when social distance was measured. Forty three per cent of Protestants said they would not allow a member of their family to marry a member of the out group compared to 9% of Catholics who felt the same way. As the table below shows the same trend was evident for joining clubs and societies and for integrated living.

* Figures in percentages

More than 14% of Protestants would not welcome a member of the out-group to live in their street or neighbourhood compared to 1.3% of Catholics in the town. Contacts at a personal level were also relatively limited. Three quarters of Protestants (75.5%) had most friends of their own religion which was roughly equal to the figure for Catholics (73.8%). This is broadly in line with the statistics for all rural areas in Northern Ireland (see Section 3). In their study on three communities in Northern Ireland Hamilton et. al. noted that,

Hamilton also argued that each community is unaware of its own behaviour and attitudes and this ‘cultural blindness’ explains the contradictions in the survey data which show a strong perception that community relations in the town are good while at the same time showing minimum contact between the two religions and segregated use of local services and facilities. Thus they point out that an area can, overtime, be defined as Protestant or Catholic, and there is then a reluctance by the out-group to visit that area or use facilities in the area,

These themes are reflected in the survey data from Oldville. Thus, a significant paradox in attitudes is evident in attitudes to community relations and violence. A total of 79.2% of people thought that community relations in the town were good although this was slightly lower among Catholics (77.5%) than among Protestants (87.6%). However, more than half of Protestants (53.1%) thought that the area was more violent than other areas in Northern Ireland during the Troubles compared to only 28.8% of Catholics (see Table 40.

Interviews with two Protestant community leaders suggested that this was due to the different experiences of the two groups over the 25 years of the conflict,

4.5 Contesting Territory
Poole’s analysis of violence suggested that this was indeed a period of considerable violence and during the late 1970s IRA attacks on local UDR and part-time police officers were at their height. There are four road-side monuments or plaques to commemorate members of the security forces killed during the Troubles, and one to a local Catholic, in the catchment of the town. There is a type of ‘monument index’ to suffering, replicating the dramatic graffiti in urban areas as a mark of territorial ownership. Monuments are potent symbols of hurt, community defiance and security and are locked deeply into the social consciousness of both communities as well as the landscape of Armagh.

Catholics are nearly twice as likely to feel that community relations are better in the town (41.3%) than Protestants (22.4%) (Table 4g). Moreover, the divisions are greater when it comes to the contentious issue of marching. Ninety eight percent of Protestants feel that traditional parades should have complete freedom to march in the area compared to only one-third of Catholics (33.8%).

The issue of ownership of territory is also highlighted when the sale of land is analysed. Sixty one percent (61.2%) of Protestants feel that people generally don’t sell their land or property to members of the opposite religion. Only 38.8% of Catholics believed this although 37.5% were not sure. In short, about one-quarter (24.5%) of Protestants and just over one-third (3 8.8%) of Catholics felt that the land market operated as a recognisable economic exchange system in the Oldville area. As Connaugton (1995) noted in his ‘Border Diary’, ‘Property doesn’t often change hands round here. Religion runs in the soil’ (Connaughton, 1995, p. 45).

4.6 Potential Actions
Catholics were generally more likely to think that specific cross community relations projects (72.6%). more sensitive policing (76.3%) and action on local parades were the key methods to improve community relations in the town. The Protestant ‘agenda’ emphasised paramilitary decommissioning (8 1.6%) and economic regeneration (83.6%). Protestants were less likely to emphasise the role of the police (46.9%) or parades (10.2%) as central issues in cross community relations in Oldville (Table 4h).

4.7 Conclusion: a Town of Paradoxes
The analysis has highlighted a number of important contrasts in community relations attitudes and behaviour, which collectively underscore the complexity of community dynamics in small town Ulster.

Whilst most people regard community relations as good in the town, a high proportion of people have little or no contact with the out-group, are less accommodating to the out-group than the population generally and have vastly different perceptions of violence and how conflict has affected townspeople.

The primacy of rural territory is reflected in the widely heldview that sectarian cleavages significantly distort the land market in a way that protects ownership within each community.

The two communities have contrasting community relations priorities. Catholics want action on policing and parades whilst Protestants regard decommissioning paramilitary weapons as crucial.

5.1 Introduction
This section explores the importance of territory at the micro-community level. Two villages, one mile apart, and located to the east of the study area were selected for focused analysis that employed small scale surveys, structured interviews and observational analysis. The villages are 7 miles north west of Newry and 9 miles south east of Armagh city. Glendale is predominantly Protestant and has a population of 140 persons whilst Whiteville is predominately Catholic, and has a population of 170 persons. The first part of the section provides the context for the analysis by setting out the results of a household survey in the village. The attitude and behaviour of villagers revealed in the survey is explored by tracing the respective and very different histories of the villages during the period of the ‘Troubles’. The importance of territorial activity is highlighted in the way villagers consume local services and facilities and this is further developed by examining the importance of land ownership in community identity and as a method of community survival.

5.2 Contact and Attitudes
A small scale survey was conducted in both villages to set a broad framework for the qualitative analysis. A total of 33 surveys (out of 65 or 51%) interviews were conducted with residents in Whiteville and 22 (out of 45 or 47%) in Glendale. All households in both villages were interviewed with the head of household taken as the respondent. The survey gave a useful, if incomplete picture of segregation in the two villages. According to the data, Whiteville is 96% Catholic while Glendale is 92% Protestant. Moreover, people in Whiteville were most likely to describe the village as Catholic (96%) while people in Glendale were strongly perceived their village to be Protestant (86%).

The survey confirmed the long established attachment to place in both villages. Nearly two-thirds (64%) of people in Whiteville and half (50%) of those in Glendale had lived there all their lives. Despite this, there were comparatively few close friendship or kinship ties across the religious divide. As Table 5a shows more than three quarters (77%) of people in Whiteville had most or all of their friends and relatives of the same religion compared to 86% of people in Glendale.

There is a sharp contrast in the perceptions of respondents when attitudes to community relations within, and between, the villages are examined. Table 5b shows that community relations in the two villages are perceived to be good. In Whiteville, 84% described relations between the two communities as either ‘very good’ or ‘quite good’. The figure for Glendale was 93%. However, only 45% of people in Whiteville and 29% in Glendale felt that community relations between the two villages could be described in that way. Indeed, more people in Glendale (36%) described community relations with Whiteville as ‘poor’ rather than ‘good’. It will be shown latter how these differences in attitude are conditioned by recent experience and fear about the long term sustainability of the Protestant population of the village. When prompted for specific explanations of this pattern some of the more incisive comments rooted the cause as a lack of contact with village neighbours:

One other respondent commented that ‘we have probably found ways of not meeting and mixing’ (Whiteville resident), highlighting the importance of ‘avoidance’ as a key variable in community relations attitudes and patterns of segregated behaviour.

The lack of contact is borne out by the statistics, as almost two-thirds (64%) of people in Whiteville said they never visit Glendale. The picture is slightly more complex for Whiteville as residents are more likely to travel through Whiteville to access other transport and service centres. Even here, however, 29% of villagers said they never visit Whiteville.

Attitudes to the future religious composition of the villages also reveal interesting contrasts. As Table Sc shows, half (50%) of the residents of Glendale wanted the area to remain all or mostly comprised of people of the same religion as themselves. Thirty six percent stated that they would prefer to see the village religiously integrated. In Whiteville however, 62% of residents said they would like to see the village mixed although a sizeable proportion (29%) wanted the dominant balance of the Catholic community maintained.

5.3 Activity Analysis
At the time of writing, there had been considerable debate about mutual boycotts of Protestant and Catholic shops in border areas and, in some areas, poster campaigns have been used to encourage people of both communities to support ‘their own’ community infrastructure at the expense of the out-group. However, in his seminal paper on the Shankill/Falls divide, discussed earlier, Boal (1987) identified the impact on movement and community activity as a key consequence of territory. In short, sectarian activity in the use of services, shops and community facilities has always been a feature of community life in highly segregated parts of Northern Ireland. As the figures overleaf show (Figure 5.1), this part of rural Armagh is no exception.

The residents of Glendale mainly look North to largely Protestant urban towns such as Armagh, Markethill and Portadown for comparison goods, convenience goods and services. It is interesting that more people go to Markethill for these goods than Newry despite the latter’s more dominant hierarchical settlement status offering a wider number, range and quality of services than Markethill which, while three miles closer, is a smaller order settlement in the sub-region. When the activity profile for Whiteville is examined (Figure 5.2) an almost mirror image emerges. Here the main population is drawn South to the mainly Catholic towns of Newry, Keady and even across the Border to Dundalk. The demographic analysis and in particular the greening of border areas explains part of the process but close inspection of local history reveals more immediate push factors on the Protestant population.

5.4 Death of a Village?
The recent past of Glendale represents a catalogue of community body blows which has created collective uncertainty, fear and, at times, hostility among local people. A number of incidents highlighted consistently in discussions with locals reveal the extent to which the basic economic, commercial, cultural and security infrastructure that any community relies upon for survival and development has been progressively eroded in a short period of time. When these specific incidents are placed in the context of wider demographic shifts the Northward orientation of the community is hardly surprising.

The first major incident occurred in 1976 during a period of sectarian murders, high paramilitary activity and a strengthening of security force presence with the development of an Army base two miles from the village. The main employment in the village was a small textile factory whose labour force was drawn mainly from surrounding towns and villages.

After a period of tit-for-tat sectarian murders in the area, 10 of the employees (all male and Protestant) were taken from the works bus and shot dead three miles outside the village. All the victims came from the nearby town of Bessbrook but the impact on community relations in Glendale was devastating.

Speculation about the role of local nationalists in the murders fuelled sectarian fear and in particular the feeling of the ‘enemy within,, difficult to define and harder to censure cumulated the sense of fear and insecurity. The murders were claimed by the Catholic Reaction Force who have claimed responsibility for a number of murders of Protestants across Northern Ireland. The incident was clearly a benchmark in local relations and is deeply ingrained in the Protestant psyche:

Another violent attack, a bomb rolled into the local UDR base (mentioned above), killed three soldiers and destroyed the base which never re-opened. The attack on basic security as well as the economy of the village was acutely felt by the Protestant community:

The sense of physical and psychological exposure, the location of the village in wider territory and the implied fear seemed very real after the base closed. All of this had a ‘knock on’ effect on local commerce and in the early 1 990s the post office and local grocery shop both closed and what passed for the commercial core of the village fell into physical neglect.

As noted previously census data showed that the Protestant population is declining, ageing and, with fewer numbers in the younger age cohorts, it has less opportunity for self-renewal. In contrast the local Catholic population is characterised by higher than average household sizes; 3.7 for Catholics compared to 3.0 for Protestant households in the area, more people in the younger age cohorts (21% for Catholics compared to 7% for Protestants) and equal proportions in the age range 65 plus (6 % compared to 5% for Protestants). These demographic trends have had an impact on the local Primary school which closed in 1995 and the remaining pupils were bussed to a school five miles away. A community activist was well aware of the implications of closure for the village:

Finally, the local orange hall was destroyed by fire in the same year. Indeed 4 out of 5 Orange halls in the area had been attacked from mid-1994 to the end of 1995. As a pivotal socio-cultural institution the practical and emotive implications were again well recognised:

5.5 Conclusions: the End of Critical Mass?
In summary, the local community has faced critical attacks on security, safety, education, the economy and commerce. The atmosphere in the village is one of neglect and, in places, abandonment. The long term future is uncertain as the northward orientation for shopping may translate into a decision to move. Central to such an event will be the ‘critical mass’ of the population. Fischer (1976) defines a critical mass of the population as a size large enough to permit what would otherwise be a small group of individuals to become a vital, active subculture’ (Fischer, 1976, p. 116). Breton (1964) defines the related concept of ‘institutional completeness’ as that stage ‘whenever the ethnic community could perform all the services required by its members’ (Breton, 1964, p. 194). Boal (1987) summarises the implications of both concepts for Northern Ireland,

The short review of conditions in Glendale seems to suggest that a dual force is at work. At one level the critical mass is being eroded as the population has declined naturally and, at another level, institutions are being fundamentally attacked in a way that is having a detrimental impact on self-perception of group critical mass. It is interesting to note that residents in Glendale felt strongly that the village was becoming increasingly Catholic. It is the vicious interaction of these forces, each feeding from the other, that has most serious implications for the long term sustainability of the population of the village.

In an exhaustive study of land transfers in Glenravel ward, Co. Antrim, between 1956 and 1987, Kirk showed how maintenance of land ownership patterns involved limited ‘seepage’ (p. 367) between dual land markets in the area. The duality of the land transfer system is analysed through case study material in this section. Kirk’s empirical analysis sets the context for the description of one exchange of a farm in the study area called, for the purposes of this paper, Pine Hill. The instruments operating dual and often symmetrical land exchange systems are described and some outline comparisons with analogous urban systems suggested. The core of the section is a discussion of the implications of rural territoriality for community relations practice at the micro-spatial scale.

6.2 Dual Land Markets: Preliminary Evidence
Table 6a is drawn from Kirk’s research in Glenravel. It shows the limited extent of cross-religion sales in the land exchange system. For example, of all the sales between 1958-87 only 12.8% happened between Catholics and Protestants. Most exchanges were within-religion sales leading Kirk to conclude that, ‘there seems to be generally accepted rules defining situations which pertained only to personal relationships as distinct from where group interests were involved’ (Kirk, 1993, p. 480) and specifically that personal interests were relegated when decisions about the sale of land would negatively impact on group interests.

In short, ethno-religious cleavages are the key determining variable when attitude to land and retention of territory is assessed. Preliminary interviews with local community leaders supported the operation of similar processes restricting land transfer in Armagh.

Table 6a Land transfers between Protestants and Catholics in Glenravel Ward 1958-1987

* (Source, Kirk, 1993, p. 496)

This latter point is significant because it suggests a measure of outside control on land exchange. Reports of threats, abusive phone calls and, what one interviewee referred to as, ‘friendly advice’ are all used to ensure that land stays within a particular ethno-religious group. While all of this is difficult to verify empirically, it is clear that some element of control is used. One Protestant interviewee revealed how he inherited a farm from a relation but as he lived and worked in Belfast he wanted to sell it quickly at the best price. The interviewee received what he clearly perceived to be threatening phone calls urging him not to sell to a Catholic. (The farm was sold to the highest, and in this case, a Protestant bidder).

The system of land exchange in rural Northern Ireland seems more complex than in urban locations. There are definable gatekeepers, principally, the auctioneer, solicitor and to a lesser extent the estate agent. Often there is a Protestant property infrastructure and a Catholic infrastructure each serving the respective communities. However, behind the scenes negotiations at a personal level characterise their style of mediation in land exchange in rural areas. Because the system is closed and not open to public scrutiny, suspicion and rumour about corruption and sectarian practice have room to grow. In some cases this has lead to intense and long standing bitterness over the sale of individual property.

6.3 The Battle for Pine Hill
This micro-case study illustrates the complexity of the process of interest mediation and sectarian outcomes. A small farm of about 8 acres with a dilapidated farm house came onto the market in 1987. The farmland itself was not of primary agricultural value or use. However, the house carried with it the potential for planning permission for a replacement dwelling. In an area of relatively strict planning control and housing shortage, the farm was, therefore, valuable for economic, housing and, to a lesser extent, agricultural reasons.

The property was on the market initially for 9 months during which three bids, all from Catholics, could be identified in the research. The respondents complained about delays in the response to their bids, lack of information and lack of access to the selling agent. Those interviewed remained convinced that, because they could be identified as local Catholics, the solicitor deliberately placed obstacles in their way. Each applicant was informed that their bid was inadequate and the property was taken of the market for a year. One of the applicants believed that he had offered above the realistic local value for the property because he wanted the farm for his son who was intending to get married and stay in agriculture. The applicant, a local farmer then explained that the next he knew of the farm was that it was sold to a Protestant farmer with a large farm in the area, The respondent was convinced that the land was taken of the market until ‘one of theirs could be found’ to buy it. He noted that the sale was not the product of open negotiation.

In the case study most evidence related to the retention of Protestant land and protectionism from potential Catholic buyers. This has caused some problems for the local rural economy. In the County Armagh area the Catholic population is growing, has a younger age structure and has higher than average household size. The demand for agricultural land to provide employment for the children of Catholic farmers is increasing. The inability of the land exchange system to release land in response to demand has, some community leaders would argue, resulted in inefficient and uneconomic farming, loss of population and often intense family friction due to competition for scarce land resources. This problem has to some extent been eased by the process of letting land through the ‘conacre system’. This system involves leasing fields on a long terms basis to individual tenants and it highlights some of the contradictions revealed in the survey data. While Protestant land owners are reluctant to sell their land to Catholics they will let their land on long leases and this further extends to sharing machinery and labour at peak times of the agricultural year.

6.4 Cross-Religion Sharing
The case study revealed three types of sharing and exchange that highlighted the complexity of relationships among rural communities. The first is about the letting of land. The process of conacre is well established in rural Ulster and affords both the land owner and tenant significant advantage. The tenant can use the resource in a flexible way, disposing of it as he or she no longer needs it or if they change farming practice or type. The landowner retains the capital value of the property, makes an economic return and keeps ownership within the ethno-religious group.

Social relationships can also be found in labour and machinery exchange between farmers. There was a degree of pride among those interviewed in the system of mutual help and support among the farming community. The lack of monetary exchange, mutual friendship and shared interest in weather, European Union grants or the effects of BSE bond farmers in a meaningful and clearly sincere way.

While the land exchange system does not work effectively, trading in stock and produce operates as a recognisable market. Markets or local ‘marts’ are places to buy and sell livestock but they are also a setting for social interaction between farmers. In these environs religion plays no significant part in exchanges either social or economic. The farming issues dominate and the distinctive farming culture is obvious in mart life. There is a sense of overt satisfaction in the way religion plays no part in such relationships:

6.5 Discussion
Vasquez has reviewed the role of territoriality in neighbourhood disputes at a national scale. He concluded that,

One of the central features of the empirical research has been the extent to which territory can not be de-coupled from community relations and contact in the study area. Control and ownership of land is the pivotal defining feature of the Northern Ireland conflict. Maintaining territory involves a complex interaction with the local market system, closed negotiation, influence over key agents in that system and at times, intimidation of both seller and buyer. In Belfast, territorial boundaries are sharpest and segregation most intense in public sector housing. The highly regulated system of public sector housing planning, development and allocation facilitates the maintenance of ethnic geography. The Housing Executive have rarely attempted to redraw long established territorial boundaries and respect for the territorial imperative has helped to diffuse violence in contentious areas of the city (Murtagh 1994). In rural areas the private exchange system which involves a small number of key gatekeepers takes on crucial importance. The potential of these gatekeepers for independent action is limited in the same way as the Housing Executive in Belfast have limited potential to re-draw an ethnic map that has less to do with ownership of parcels of land and more to do with the way that ownership expresses the legitimacy and permanence of ethic identity.

Recognition that territory, and how it is protected, is a feature of day to day interaction between the two communities emphasises the need for more realistic strategies to manage conflict at a strategic and local level. Approaches that ignore deep structural factors in the configuration of the Northern Ireland conflict will have limited currency. In a comprehensive review of community relations projects at District Council level, Knox (1994) divided activity into five types, only one of which involved what he called ‘focused community relations’ aimed at tackling, head on, controversial community relations issues. ‘This approach suggests that such issues, if left unresolved, compound insidious sectarianism and bigotry’ (Knox, 1994, p. 603). The other categories relate to the funding of high profile events such as festivals; mainstream community development work; cultural traditions that reinforce the identity of each tradition; and substitute funding, through which community relations resources are used to fund projects that would have been supported anyway. The District Council’s Community Relations Programme for Newry and Mourne emphasised the role of reconciliation and festival based events,

The Community Relations Officer points out that it is difficult for community relations work to get beyond local religious divisions which often run deep and have been sustained over a relatively long period of time. Therefore, community relations activity in the District has centred around the softer issues where contention and conflict are unlikely to surface or be addressed.

In broad terms Bloomfield (1995) locates this type of work within ‘resolution’ theory. In short, this approach is,

On the other hand, the ‘settlement’ approach is,

The empirical research presented here suggested that the reality of conflict at the local level requires a more realistic and pragmatic response than that offered by initiatives predicated on an analysis that suggests resolution is a realistic alternative. Hard issues, such as the system of land exchange and its role in reinforcing territory, need to be understood and addressed. They are not reducible to resolution based measures that characterise much of community relations activity in Northern Ireland. Centuries old practices and beliefs are locked into the rural socio-cultural and economic system and to talk about ‘resolving’ them is to miss the complexity of the conflict and how it is acted out in small communities every day. The best we can hope for in the short-term is to minimise their negative consequences and to hope for settlement of contentious territorial issues at the micro or local level.

6.6 Conclusions
It does not imply that the system of land exchange in rural Armagh is transformable, even in the long-term. But it does imply that a serious engagement between community relations practice and the reality of community division as it is acted out across communities all over Northern Ireland must occur. In particular, the balance of activity needs to reflect the limitations of resolution based activity and the need to pursue what Knox has described as ‘focused community relations work’. These implications are addressed in the final section.

7.1 Introduction
This section sets out the implications of the research for elements of policy and practice. The research highlighted an under-looked issue in territorial and segregation analysis which has important effects on the type, purpose and impact of community relations practice at the local level. This is the process of residualisation whereby sectarian population shifts have left behind some communities who are experiencing abandonment and isolation which itself conditions community relations values and attitudes. This provides a context to review the strategic implications of territory and the need to address its consequences comprehensively. The role of policy and planning are then discussed before highlighting the options open to community relations practitioners.

7.2 Territoriality and the Process of Residualisation
The processes of residualisation are already well established in the field of public sector housing. Saunders shows how the younger, more mobile and employed population are leaving the public rented stock for owner occupation with council renting being left to a ‘residuum of old, poor and unskilled people who can not afford anything else, while the rest of the working class makes good its escape into the home owning middle mass’ (Saunders, 1990, p. 319). An analogy can be drawn with the processes of change among Protestant communities close to the border.

For example, the research identified a number of communities experiencing decline related to selective ‘exiting’ of its younger and more able members. In places such as Protestant Glendale, direct and perceived threat has had a cumulative effect on community relations attitudes. These are best expressed in the activity segregation between it and Whiteville, in that, the Protestant village oriented itself north for services and facilities, whilst the Catholic village looked south to ‘Catholic’ towns to serve their needs. The sale of land and the asymmetrical institutional infrastructure that supports in-group sale, highlights the way in which territory is ingrained into social and economic as well as physical aspects of life in rural Ulster.

The establishment of a review body on parades in Northern Ireland was an illustration of a macro-level response to a crises territorial issue. However, by focusing on the parades issue specifically, the scope of the inquiry is limited to only one of a number of related manifestations of territoriality. Shop boycotts, church demonstrations, traditional parades, avoidance behaviour and the experiences of residual communities are all aspects of territory which run central to the Northern Ireland conflict. An opportunity was perhaps missed when, instead of commissioning a single issue review, a government supported standing commission on territory was not constituted to review all aspects of territory, its impact on community relations and the potential role played by a wide range of policy actors, As the empirical analysis illustrated, the control and ownership of territory is one of the clearest expressions of the nature of conflict and division and needs to be researched, analysed and responded to in a comprehensive manner. Such a commission, could embrace a wider set of interests than the parades review body and could undertake further analysis of the issues and relationships between territorial behaviour, its consequences for communities and possibly agreed principles about the use, of and respect for, group territory. As well as providing a framework for research, analysis and discussion, such a commission could sponsor pilot projects to explore best practice to a range of territorial related issues and problems.

7.3    Policy Implications
The Special Support Programme for Peace and Reconciliation in Northern Ireland and the Border Counties of Ireland (EU, 1995) described how mainstream policy concerns including rural and urban regeneration can be developed with community relations as a central objective. For example the sub-programme dealing with Rural Regeneration (2b) highlighted ‘reconciliation impact’(EU, 1995, section. 2B) as a key criteria in project assessment. There are signs that community relations issues are finding their way into the policy vocabulary of land use planning. In the Regional Strategy for the Belfast Sub-region, planners recognise that,

‘the social characteristics of Protestant and Catholic areas of Belfast’s inner city vary significantly, and have been determined by the sectarian geography as much as by the sorts of economic and social factors normally at play elsewhere. The housing stock, almost totally public, is occupied by an increasingly dependent community, within which are found the sharpest territorial perceptions and sectarian divisions manifested in a series of peacelines...’ (DOE (NI), 1996, p. 7).

The Discussion Paper goes on to recognise the need to move ‘beyond land use planning’ (DOE (NI), 1996, p. 3) and to view the city as a complex socio-spatial system where the consequences of sectarian division should be a primary policy concern.

This has resonance in recent literature on planning theory and practice. Heffley, for example, argues the need for planning to explore what is relevant as defined by individual communities and to recognise the power of argument and negotiation in resolving land use problems,

‘such interaction involves respectful discussion within and between discursive communities, implying recognising, valuing listening and searching for translative possibilities between different discourse communities’ (Healey, 1992, pp. 154-5).

She encourages the use of negotiation counsels to encourage individuals and groups to articulate their interests together in an environment where interaction is not simply a form of exchange or bargaining around predefined interests but is a genuine dialogue between competing interests. In Northern Ireland, the need to locate community relations issues at the centre of such a planning dialogue is crucial. This has implications for planners as well as community relations practitioners. A fusion of the substantive understanding of socio-spatial dynamics, area degeneration and demographics with conflict negotiation and mediations skills is an important training challenge. For example, Healey recognises the implications of community dialogue for traditional planning research skills,

‘planning processes should be enriched by discussion of moral dilemmas and aesthetic experiences, using a range of presentational forms, from "telling stories" to aesthetic illustrations of experiences. Statistical analysis coexists in such processes with poems and moral fables’ (Healey, 1992, p.154).

Donnan and Mc Farlane’s work shows how ethnographic methodologies can be powerful in unpacking communities and the tensions and conflicts laden in their daily activity and interaction. The perspectives of the people of Glendale on what might help regenerate their village must be preconditioned by an understanding of the different demographic histories of the two communities and the sustained decline of the Protestant population in particular. Helping marginal communities to articulate their concerns in a way that will impact on the prospects for maintaining a viable and sustainable future is a key challenge for both community relations practitioners and planners. June Manning Thomas (1994) has made a similar call for the role of planning in responding to the problems of marginal black inner city communities in the United States,

‘It means fighting market forces that are draining the inner city and older suburban neighbourhoods which have been tipped or will tip away from non black or middle class residence’ (Thomas, 1994, p.6).

Recalling Frieden’s argument to extend planning ‘beyond the physical environment’ Thomas argues the need for comprehensive community action on ill-health, poor housing, drug abuse, low self esteem and economic desperation (Thomas, 1994, p.7). Those acting on such an agenda will face formidable obstacles, not least the professional value system of planners and administrators. Hoch recognises this in the context of planning and race,

‘professional norms are not a source of moral guidance here. Instead, planners must turn to examples of democratic citizenship and civility exhibited by those who resist racial injustice and who find practical ways to make the boundaries of conventional American life more inclusive’(Hoch, 1993, p. 460).

Likewise Peach emphasises the policy implications of deeply segregated communities in a Northern Ireland context,

Mention has already been made of the inclusion of sectarian realities in local planning strategies including the Belfast sub-regional planning discussion document. The DOE (NI) Planning Agency has introduced training on Community Relations and Planning for all its staff, recent consultation documents on strategies to tackle deprivation in Belfast and Londonderry have highlighted the problems of territoriality in community regeneration and local Partnerships created under the Peace and Reconciliation Programme have built concerns about segregated areas into their action plans.

7.4    Community Relations
However, while there seems to be evidence of progress on the part of the statutory sector, community planners and those in community relations practice have been slower to respond to these issues. It has been argued previously that the emphasis on reconciliation as opposed to settlement based activity in community relations is part of the problem. Darby highlighted the need for a shift in emphasis,

‘moderate opinion, expressed through reconciliation groups, the churches and political action, has played a part in keeping this society together during twenty five years of violence. It will have a major role in conflict management in the future. But now, it must move over and make room for unsavoury bedfellows. Altruism is an unreliable basis for conflict resolution. Better to rely on hard negotiation firmly based on self interest’ (Darby, 1991, p.12).

A spatial focus on, what in community relations terms, would be the most serious cases - those feeling forgotten, abandoned and powerless - would seem logical for the community relations practitioners. Developing relevant practice in these areas would add greatly to the emphasis in planning on addressing problems through argument, dialogue and interest mediation. The failure to gel the related discipline in that way could marginalise community relations practice in a way that diminishes its relevance. Gallagher quotes the first chair of the Community Relations Commission in terms of its narrow policy brief.

‘The problem with community relations, of ensuring that people can live together in a society [is the] central problem of government ... It fails in that, government has failed. Neither is there such a thing as a separate "community relations " policy All the decisions of government, whether about education. location of industry, housing or law enforcement, have an effect on relationships between the communities’ (Hayes 1972 quoted in Gallagher, 1995).

7.5    Conclusions
This research has presented the conclusions of a pilot research study into the nature of conflict in a divided rural community in Northern Ireland. By its definition the extent to which broad based policy recommendations can be made are therefore limited. However, three broad ideas have been set out for discussion:

The potential role of a commission to address the issue of territoriality, how it is manifest at the micro-spatial scale and the relationship between events at locality level and the nature of conflict.

Territory is clearly fundamental to the nature of conflict and the need to respond with comprehensive policy measures, particularly in the field of area planning and rural regeneration initiatives is vital to the sustainability of marginalised communities.

Policy initiative alone can not address the problems of residual communities and the need for community relations practitioners to engage, articulate and positively approach their problems is equally important.

References

Adams, L. (1995) Cashel: A Study in Community Harmony, Cashel, Cashel Community Development Association.

Blacking, et. al. (1978) Situational Determination of Recruitment in Four Northern Irish Communities, SSRC, British Library.

Bloomfield, D.(1995) Towards Complementarity in Conflict Management: Resolution and Settlement in Northern Ireland, Journal of Peace Research, Vol. 32, No. 2, pp.151-164.

Boal, F. (1969) Territoriality on the Shankill-Falls Divide, Belfast,Irish Geography, Vol. 6, pp. 30-50.

Boal, F. (1987) Segregation, in Pacione, M. (ed.) Social Geography: Progress and Prospects, London, Croom Helm.

Breton, R. (1964) Institutional Completeness of Ethnic Communities and the Personal Relations of Immigrants ,American Journal of Sociology, Vol. 70, pp.193-205.

CACI (1993) ACORN User Guide, London, CACI.

Compton, P. (1995) Demographic Trends in Northern Ireland, Belfast, Northern Ireland Economic Council.

Compton, P. and Coward, J. (1989) Fertility and Family Planning in Northern Ireland, Aldershot, Avebury.

Connaughton, 5 (1995) A Border Diary, London, Faber and Faber. Darby, J. (1986) Intimidation and the Control of the Conflict in

Northern Ireland, Dublin, Gill and Macmillan.

Darby, J. (1991) What’s Wrong With Conflict, Occasional Paper No.3, Centre for the Study of Conflict, University of Ulster, Coleraine.

DOE (NI) (1996) The Belfast City Region. Towards and Beyond the Millennium. A paper for Discussion, DOE(NI), Belfast.

Donnan, H. and McFarlane, G. (1983) Informal Social Organisations, in Darby, I. (ed.) Northern Ireland: Background to the Conflict, Belfast, Appletree Press.

Donnan, H. and McFarlane, G. (1986) You Get on Better With Your

Own: Social Continuity and Change in Rural Northern Ireland, in Clancey, P., Drudy, S., Lynch, K. and 0’ Dowd, L. (eds.)

Ireland: A Sociological Profile, Dublin, Institute of Public Administration in association with the Sociological Association of Ireland.

Dunn, S. (1995) The Conflict as a Set of Problems, in. Dunn, S. (ed.) Facets of the Conflict in Northern Ireland, London, St. Martin Press.

European Union (1995) Special Support Programme for Peace and Reconciliation in Northern Ireland and the Border Counties of Ireland, EU, Brussels.

Fischer, C. (1976) The Urban Experience, New York, Harcourt Brace. Gallagher, A. (1995) Community Relations, in Breen, R., Devine, P. and Robinson, G. (eds.) Social Attitudes in Northern Ireland, Belfast, Appletree Press.

Gallagher, A. (1995) The Approach of Government: Community Relations and Equity, in Dunn, S (ed.) Facets of the Conflict in Northern Ireland, Macmillan, London.

Hamilton, A., McCartney, C., Anderson,T. and Finn,A. (1990) Violence and Communities, Centre for the Study of Conflict, University of Ulster, Coleraine.

Harris, R. (1972) Prejudice and Tolerance in Ulster: A Study of Neighbours and Strangers in a Border Community, Manchester, Manchester University Press.

Healey, P. (1992) Planning Through Debate, Town Planning Review, Vol. 38, No. 2, pp, 143-162.

Heaney, S. (1975) North, London, Faber and Faber.

Hoch, C. (1993) Racism and Planning, Journal of the American Planning Association, Vol. 38, No. 4, pp. 45 1-460.

Keane, M. (1990) Segregation Processes in the Public Sector, in Doherty, P. (ed.) Geographical Perspectives on the Belfast Region, Geographical Society of Ireland.

Kirk, T. (1993) The Polarisation of Protestants and Roman Catholics in Rural Northern Ireland: A Case Study of the Glenravel Ward, Co. Antrim, 1 956-1 988, Unpublished PhD Thesis, School of Geosciences, Queens University Belfast.

Knox, C. and Hughes, J. (1994) Policy and Evaluation in Community Development: Some Methodological Considerations, Community Development Journal, Vol. 29, No. 3, pp. 239-250.

Knox, C. (1994) Conflict Resolution at the Micro-level, Community Relations in Northern Ireland, Journal of Conflict Resolution, Vol. 38, No. 4, pp. 595-619.

Leyton, E. (1975) The One Blood: Kinship and Class in an Irish Village, Social and Economic Studies, Memorial University, Newfoundland.

Massey, D. and Gross, A. (1991) Explaining trends in racial segregation 1970-1980, Urban Affairs Quarterly, Vol. 27, No. 1, pp. 13-35.

Measor, L. (1985) Interviewing: A Qualitative Research Strategy, mR. Burgess (ed.) Strategies in Qualitative Research : Qualitative Methods, London, Falmer Press, pp. 55-77.

McFarlane, G. (1978) Gossip and Social Relations in a Northern Irish Village, Unpublished PhD Thesis, Department of Anthropology, Queens University Belfast.

Murphy, D. (1978) A Place Apart, London, Penguin.

Murtagh, B. (1994) Ethnic Space and the Challenge to Landuse Planning: a Study of Belfast’s Peacelines, Research Paper No.7, Centre for Policy Research, Jordanstown, University of Ulster.

Murtagh, B. (1994) Land Use Planning and the Challenge to Landuse Planning: A Study of Belfast’s Peacelines, Research Paper No.7, Centre for Policy Research, Jordanstown, University of Ulster.

Peach, C. (1996) The Meaning of Segregation, Planning Practice and Research, Vol. 11, No. 2, pp. 137-150.

Pettigrew, T. and Riley, R. (1972) Contextual Models of School Desegregation, in King, B. and McGinnes, E. (eds.) Attitudes, Conflict and Social Change, New York, Academic Press.

Poole, M. (1983) The Demography of Violence, in Darby, J. Northern Ireland: The Background to the Conflict, Belfast, Appletree Press.

Poole, M. and Doherty, P. (1996) Ethnic Residential Segregation in Northern Ireland, Centre for the Study of Conflict, Coleraine, University of Ulster.

Saunders, P. (1990) A Nation of Home Owners, Unwin Hyman, London. Shuttleworth, I. (1992) Population Change in Northern Ireland, 1981

1991: Preliminary Results of the 1991 Census of Population ,Irish Geography, Vol. 25, No. 1, pp. 83-88.

Stockdale, A. (1991) Recent Trends in Urbanisation and Rural Population in Northern Ireland, Irish Geography Vol. 24, No. 2, pp. 70-80.

Stockdale, A. (1992) State Intervention and the Impact of Rural Mobility Flows in Northern Ireland, Journal of Rural Studies, Vol. 8, No. 4, pp. 441-421.

Stutt, C. (1996) University of Ulster, Northern Ireland Economic Research Centre, Giplen Black and Social Information Systems Baseline Study for the Special Programme for Peace and Reconciliation in Northern Ireland and the Border Counties of Ireland, Belfast, Department of Finance and Personnel, European Division.

Thomas, J. M. (1994) Planning History and Black Urban Experiences: Linkages and Contemporary Implications, Journal of Planning Education and Research, Vol. 14, pp, 1-11.

Tutt, N. (1994) A Review of Community Relations Research in Northern Ireland, Knutsford, Social Information Systems.

Vasquez, J. (1995) Why do Neighbours Fight? Proximity, Interaction or Territory, Journal of Peace Research, Vol. 32, No. 3, pp. 277-293.

Yin, R. (1994) Case Study Research, London, Sage.