These extracts are copyright Marie Therese Fay, Mike Morrissey and Marie Smyth 1998 and are included on the CAIN site by permission of the author and the publishers. You may not edit, adapt, or redistribute changed versions of this for other than your personal use without the express written permission of the author or the publisher, INCORE. Redistribution for commercial purposes is not permitted.

mapping troubles-related deaths
in Northern Ireland 1969-1998

First Published 1997
Second edition with amendments reprinted 1998
by INCORE (University of Ulster & The United Nations University)
Aberfoyle House, Northland Road, Deny Londonderry, Northern Ireland
Tel +44 1504 375500

© Marie Therese Fay, Mike Morrissey, Marie Smyth & The Cost of the
Troubles Study
Unit 14, North City Business Centre, 2 Duncaim Gardens, Belfast BT15 2GG
Tel/Fax +44 1232 742682 or Tel +44 1232 747470.

Cover Design by Belfast Litho
Graphic Design by Lyn Moffett
Printed by Belfast Litho Printers Ltd.

All Rights Reserved

ISBN 0-9533305-3-2

We wish to acknowledge the help of the following people:
Sarah Oakes, Catherine McGurk; Lyn Moffett; Dr Brian Tipping, Research Services Ireland; Dr John Yarnell, Department of Public Health Queen's University & The Health Promotion Agency; Dr Debbie Donnelly, NISRA; Wendy Hamilton, General Registry Office; Alison Hamilton, Central Services Agency; Ronnie McMillen, John Park, Social Services Inspectorate; Yvonne Murray, Linenhall Library; Andy White; Conor Barnes; Jonathan Blease; Alan Breen; Survivors of Trauma, Ardoyne; Greencastle Women's Group; Damien Gorman, An Crann/The Tree; Tony McQuillan, Northern Ireland Housing Executive; Staff in the Central Library, Belfast, Queen's University Library; David McKitterick; Professor John Darby; Dr Andrew Finlay, Trinity College Dublin; Dr Maggie Martin, Eastern Connecticut State University; Brandon Hamber, Centre for the Study of Violence and Reconciliation, Cape Town; Patrick Ball of the American Association for the Advancement of Science in Washington DC for providing some of the international data; The Community Information Technology Unit, Belfast.

The Cost of the Troubles Study is funded by the Central Community Relations Unit, the European Special Support Programme for Peace and Reconciliation through the Northern Ireland Voluntary Trust, Making Belfast Work North and West Teams, The Joseph Rowntree Charitable Trust, and private donation. Additional funding to support work with children and young people was provided by Bamardo's, Save the Children, the Cultural Diversity Programme of the Community Relations Council and the Community Relations Council. Additional funding to support exhibitions and dissemination of our work by the Belfast European Partnership Board. This publication was funded by the Community Relations Council.

ABSTRACT

This paper describes the construction of a database on deaths in the Troubles in Northern Ireland, which includes deaths which have occurred outside Northern Ireland and deaths due to Troubles-related trauma. Existing lists of deaths in the Troubles are reviewed. Analysis of the Cost of the Troubles database has confirmed many of the findings and trends identified by others in previous work, although the list here is longer. The peak levels of deaths occurred in the first half of the 1970's. The suggestion that deaths peak in the summer months was not supported, but rather October and November have emerged as the months with the highest death rates. Different cycles emerged among the organisations principally responsible for killings, with peaks of activity occurring at different points in the period 1969 - 1998. In terms of the distribution of deaths, the overwhelming majority of those killed in the Troubles have been male, with the death risk highest in the younger age groups in the 20-24 age group highest, and almost 26 percent of all victims aged 21 or less. The absolute number of Catholics killed is greater than Protestants killed, and the death rate for Catholics is greater than that for Protestants, as other researchers have found. However, if we include Northern Ireland security forces deaths in the analysis, and exclude Catholics killed by Republican paramilitaries and Protestants killed by Loyalist paramilitaries, the death rates become much closer: - 1.9 per 1,000 for Catholics and 1.6 per 1,000 for Protestants. Civilians are the largest category killed, and account for 53 percent of the total killed, with the British Army accounting for almost 15 per cent. Republican paramilitaries account for almost 13 percent, the RUC account for 8 percent of those killed and the other groups each account for less than 6 per cent. In relation to perpetration of killings, Republican paramilitaries account for almost 59 per cent of all deaths, Loyalist paramilitaries for almost 28 per cent, the British Army for 9 per cent, the RUC for almost 2 per cent and other groups each for less than 1 per cent. Republican paramilitaries have killed 74 per cent of all Protestants killed over 25 per cent of all Catholics, and almost 96 per cent of those who were classified as "Non Northern Ireland." Loyalist paramilitaries killed 19 per cent of all Protestants killed, almost 50 per cent of all Catholics and just 2 per cent of the "Non Northern Ireland" category. On the distribution of deaths, a death rate by ward was calculated and a concentration of deaths was found in Belfast, with only 15 of the 57 highest-ranking wards outside the Belfast area. Deny Londonderry and Armagh account for most of the remaining wards.

The distribution of deaths in the Troubles was examined in the light of the Robson deprivation indicator. Whilst the wards with the highest death rates also score high on the Robson index, no overall statistical association between death rate and deprivation was found. When security force deaths were excluded, a positive correlation was found between death rates and deprivation scores in wards. The percentage change in the yearly death rate was examined in the light of changes in the Northern Ireland Gross Domestic Product and the numbers unemployed. Whilst a negative statistical correlation was found between death rates and GDP, and a positive correlation between deaths and unemployment, this is probably spurious. Deaths in the Troubles was compared with the annual suicide figures for Northern Ireland, and a negative correlation found, suggesting that suicide rose as deaths in the Troubles decreased. This is consistent with evidence from elsewhere. Whilst the findings generally support the findings of others. the finding of others that the spatial distribution of deaths is associated with the spatial concentration of Catholics was not confirmed. Variations of between 5-10% were found in the total number of deaths found between the new database and previously existing lists. View of others that the intensity of the Troubles meant that they ranked among the most serious of world conflicts was not supported by our preliminary international comparative work.

Mapping Troubles^[1] -Related Deaths and Deprivation in Northern Ireland

Marie Therese Fay, Mike Morrissey and Marie Smyth
The Cost of the Troubles Study

Introduction

The work presented in this paper has arisen out of a larger study, conducted within a participative action research approach, and concerned with establishing the effects of the Northern Ireland "Troubles" on the Northern Ireland population as a whole. The work presented here - the construction and analysis of a database of deaths in the Troubles in Northern Ireland - is the second edition of the first of a series of publications arising out of the project. This second edition analyses deaths from 1969-1998, whereas the earlier version analysed deaths from 1969 - 4 February 1994.

Background to the work

There has been some disagreement early in the Troubles amongst psychiatrists and psychologists about the extent of the effect of the troubles on the population in general. On the one hand some, notably Fraser (1971: Fraser et al 1972) have argued that there are visible and marked effects of exposure to violence. Others such as Lyons (1974) or later Cairns and Wilson (1989) reported rapid improvement in traumatic symptoms after a violent event and successful coping amongst those exposed to violence of the Troubles, suggesting a lesser impact. Since that early debate, there has been remarkably little interest in the specific psychiatric effects of the Troubles on the population. Nor is there any generally recognised and reliable measure of the effects of the Troubles on the general population of Northern Ireland.^[2]

After the cease-fires from 1994 onwards, a group of people from all sections of the population in Northern Ireland who had direct experience of being bereaved or injured in the Troubles were brought together to discuss their contribution to the new political situation. The widespread determination to have violence permanently ended seemed to be based on the unspoken recognition of the damage done by the violence of the Troubles. This group formed 'The Cost of the Troubles Study', which became a limited company and a recognised charity. In partnership with academic researchers from the university sector, a study of the effects of the Troubles on the population was planned and initiated.

The research approach

The research is conducted in line with participatory action research principles. This means that the management structure involves a range of people with direct experience of the effects of the Troubles. There are ethical questions about researchers becoming involved in this field of research which led to the need to make researchers accountable to those with direct experience of bereavement and injury. One of the most devastating after-effects of trauma is the sense of disempowerment that it can bring. Working according to a principle of partnership is an attempt to avoid further disempowering those we research.

In order to enter a field of research such as this, where the research is concerned with the impact of violence on a population, the researchers must become involved in the most intimate and traumatic events in the lives of some of Northern Ireland's citizens. Yet these events, the deaths and injuries sustained as a result of violence are also matters of public concern and political consequence. The personal grief, anger, shock and fear of those directly affected by violence has often been subsumed into the collective sense of outrage, grievance, fear and victim-hood. This happens largely through the media coverage of the violent incident and its aftermath. Media coverage of violence in Northern Ireland has been the focus of substantial attention (see Curtis:1984; Rolston & Miller; 1996 for example) much of which has been concerned with the manner in which the media, directly or indirectly, supports one or other political position in Northern Ireland. However, the concern here with the function of news media coverage is more general. News media coverage of violent incidents summarises the violent incident and its effects within a few hours of its occurrence. Such accounts and summaries may be repeated for a short period after the event, after which time, the event ceases to be considered "news". News coverage must also locate the violent incident within a larger political conflict, usually by announcing the socio-political identification of the victim and the perpetrator. Footage and sound track collected at the time of the incident is kept in the archive, and treated as historical documentation. Similarly, the social researcher chooses the focus of the research, enters the field, makes contacts with people who become "subjects," collects data, analyses it and publishes. Like media coverage, research usually collects evidence to support or contradict pre-existing ideas about the subject of inquiry. In neither case does the interviewee or the "subject" exert much influence, if any, on the angle of the journalist or the analysis of researcher. Having given consent to being interviewed, filmed or otherwise being represented, usually the "subject" exerts no further control over the manner in which the footage, sound-track or data is used. This material may be used again, usually without consultation with those portrayed in it, when documentary media material is being compiled, or in further research.

Elsewhere (Smyth & Moore: 1996) we have documented concerns about, amongst other issues, the relationship of researchers to those who participate as "subjects." We wish to resist the practice of using informants or respondents simply as containers of data, which must be collected. Our training in research (or journalism) does not necessarily equip us to consider the rights of the respondent, nor does it demand that we consider the appropriation of information and the subsequent marginalization of the respondent from the process of analysis as problematic.

A complex and two-way relationship has developed between the media (and researchers) and those affected by Northern Ireland's Troubles, and we do not wish to oversimplify it here. In the qualitative work in this project, we have encountered people angry at media, people alleging payment from media to incite young people to riot or otherwise perform for the cameras. People have reported distress on being unexpectedly confronted with seeing old footage of their tragedy on television. Some people who are finally left to their own devices after an intense media interest report feeling neglected and used. People have reported being misrepresented in the media. Many of these accusations can also be levelled at researchers. As a result of public interest in their situation, a number of people have reported the loss of their identity and its replacement with a new identity. Tony Doherty describes how, after his father's death, his name changed from "Tony Doherty" to "Tony Doherty whose father was killed on Bloody Sunday." Such a change may perhaps be at the expense of the individual's private feelings, and may limit his or her capacity to move through and beyond the personal effects of the Troubles. People can feel or sense a responsibility, not merely for managing their own emotional realities, but also for representing the feelings of their family, neighbourhood or community. In order to convey these feelings to the wider public, people rely on the media and, to a much lesser extent, on researchers. In the absence of awareness of the emotional needs of those affected by the troubles, many people who have suffered loss in the Troubles had very little personal attention and support. This has left some people vulnerable to exploitation by anyone who offers to pay some attention to them. Researchers and journalists fall into this category, and we identified the need to deal responsibly for how we deal with the vulnerabilities of those whose experiences they seek to portray or understand.

Professional 'neutrality" or "objectivity", and the professional distance which some claim separates us as researchers from the details of people's lives (which become data) was also an area in which we decided to depart from the professional norm. There are a number of ethical concerns arising out of these practices which must, we argue concern researchers and others proposing to approach people who have been involved in the process of interpreting and presenting violence. In particular, the compilation of the database involved us in daily handling of the tragic and often heartbreaking details of people's deaths. We found that, even in the analysis phase, when we were scrolling through screen after screen of the list of deaths, that we would regularly realise the nature of the data we were handling, and have to deal with our emotional responses. The discipline of remembering that this is a list of human beings who have died has meant a more real and complete connection with the data. We all know people on the list who have been killed, some of them are friends, neighbours and members of our extended families. Contrary to the old models of scientific or professional distance, we have not denied this to each other, rather, we have discussed our personal responses to the material, and made it part of our analysis.

Attempting to democratise the research process, by involving individuals from the researched population at a number of levels in the research process is a strategy often used particularly in sensitive research fields. The term "participatory action research" has been applied to such, often rather diverse, strategies, which attempt to engage the researched population in this way. In Northern Ireland, this approach has been developed in previous work.^[3] Perhaps the best known proponent is William Foote Whyte.^[4] Participative action research means different things to different researchers (see for example Benson, 1996; Small, 1995; Bartunek, 1993; Greenwood, 1993; Chelser, 1996; Kennedy, 1989; O'Connor, 1987; Oliver, 1992 and Argyris, 1989.) In this project, it entails, for example, democratising the management structure of the project management, as described above. This involved lay management in monitoring the ethical aspects of research practices; the involvement of lay people in analysis by discussion and by reading drafts of papers; a detailed process of providing transcripts to all interviewees; discussion and agreeing of transcripts; collaboration with interviewees on issues such as anonymity, and presentation of findings.

The structure of the project

The project, therefore, contains three groupings: the Board of Directors which is the executive body and the fund-holders; the Board of Directors have legal and executive responsibility for the management of the project. The Board of Directors is composed of many of the people who met after the cease-fires of 1994 and are from both sides of the sectarian divide in Northern Ireland, and all have direct experience of being bereaved or injured in the Troubles. Two of the research team also sit on the Board. The advisory group is a non-executive group which meets regularly with the research team, and is composed of funders, policy-makers and experts in the field. The research team is composed of two full and one part-time staff members (two of whom are also directors). The research team are responsible for conducting the research, and are supported and advised by the advisory group, while retaining professional autonomy on research issues. The structure of the project offers the possibility of incorporating into the research design, management and analysis the perspectives of those in the researched population. The effectiveness of this structure in addressing concerns about the accountability of researchers will be the subject of a future evaluation.

The scope of the project

The task of the larger study is to document the effects of the Troubles on the population as a whole, and to elucidate any patterns or trends in the way the effects of the Troubles are distributed within the population. The project employs both qualitative and quantitative methods, and people who have been directly affected by the Troubles inform the direction of the research.

The aims of the survey are to establish the prevalence of emotional and physical trauma arising out of the Troubles in Northern Ireland, and to identify the needs (health, emotional, social, financial) of those affected. This involves administering a questionnaire to a representative sample of the population of Northern Ireland. No existing questionnaire was adequate to the task, and it was necessary to develop an instrument for this purpose.

To this end, a three-part research strategy has been employed:

• Phase one identified the full range of self-help groups established by people adversely affected by the violence of the Troubles. A directory of groups and services available to those experiencing physical or emotional after-effects related to the Troubles has been drawn up and made available to all interviewers who can pass it to respondents where appropriate. Phase one also produced a mapping of deaths that have taken place due to the Troubles, primarily to provide a sample frame for the survey. It is out of this mapping exercise that the work presented here has arisen;

• Phase two involved conducting interviews with approximately 65 people to generate in-depth accounts for qualitative analysis, and generated data to inform the design of a questionnaire for use in a field survey of a representative sample of 3,000 people drawn from the general population.

• Phase three consists of the conduct and analysis of this survey.

This paper focuses on the work associated with the phase one task of mapping the deaths associated with the Troubles

Ethical issues in working with, and making available, data on deaths

In the initial stages, it was overwhelming that the names, addresses, ages and other information held on over 3,600 people were not merely data, but personal information about people who had died, often brutally, prematurely, and that for each death a number of other human lives had been inalterably affected. The Board of Directors holds the view that, even though some of the personal information on those killed is already in the public arena, (e.g. Sutton, 1993) it would not make available any personal details lest the information be used to invade the privacy of families, or worse, that revenge or other motivations be facilitated. Accordingly, a coded data file (either in SPSS or Excel format) from which all personal details such as addresses have been removed will be made available to other researchers and local groups in due course. However, issues about personal information having been resolved, there remain other ethical dilemmas.

The information presented, whilst as accurate as can be made, may lead others who read it to conclusions - or support them in actions - which are not in the long term interests of establishing an end to the violence we study. We wrote the original paper at a time of heightened tensions in Northern Ireland, in the run up to Drumcree 3, and we write the second edition in the period leading up to the establishment of the new Northern Ireland Assembly. We are mindful of the myth of the "objectivity" of the scientist. The information we handle and present here is not objective or neutral. Behind the statistics we present here is the life-blood of our friends, enemies and those we cannot classify, who have lost their lives and the oceans of tears which have been shed as a result. Holding up to a community such hard facts and figures in such times as we are currently living is risky and fearful work. We worry that our work will lead parties to the conflict to entrench their positions, and more bloodshed and loss of life will indirectly result. We have tried to resist the "publish and be damned" instinct which is both an occupational hazard and based on a naive assumption that more information is necessarily a good thing. Yet we must believe that clearer views of our situation, which we have endeavoured to provide, and evidence about the awful cost paid by this community will support, inspire and motivate some people to pursue new ways in which we can successfully address our situation.

We are hopeful that some of our work can be useful in shedding more light on the nature of the problem, particularly within some of the areas and sub-populations worst affected by the Troubles, and thereby shed light on the routes towards new and effective solutions.

The Mapping Exercise

The Northern Ireland conflict has produced several accounts of the numbers and characteristics of those killed as a result of the Troubles. There are a number of problems with the available list of deaths. For example, some lists vary in length from around 3,400 deaths to over 4,000. We are also unaware of any database on deaths which is publicly available and which looks at the geographical distribution of deaths. For the purposes of this study we have attempted to compile a comprehensive and reliable database, inclusive of all Troubles-related deaths both inside and outside Northern Ireland from 1969-1998.

Our database of deaths in the Troubles was created using the following major sources who have also compiled lists of deaths:

• Isabel Hylands: Count the Cost exhibition
• Official/RUC Statistics
• Malcolm Sutton: An Index of Deaths From the Conflict in Northern Ireland 1969-1993
• Irish Information Partnership
• David McKitterick and the Deramore Group
  

To check our information and attempt to fill in other missing details we used a number of books which deal with the conflict in Northern Ireland. These are as follows:

• Holland and McDonald (1994): INLA; Deadly Divisions
• Steve Bruce (1992): The Red Hand
• Flackes and Elliott (1994): Northern Ireland: A Political Directory
• Ciaran De Baroid (1989): Ballymurphy and The Irish War
• Local newspapers
• Information was also provided from local communities. For example a self help group in Ardoyne, Survivors of Trauma, have compiled a list of deaths for the greater North Belfast area.
  

After checking the information the database which we have now contains the following information about each death: the name of victim, age, gender, cause of death, town of incident, religious and political affiliation, occupation, organisation responsible for the death, a complete address of where the death occurred, and where possible the home address of the person killed. We have collected information on a total of 3601 deaths due to the Troubles in Northern Ireland.

The difficulties in completing such a big task has meant we have had to address certain issues:

Inclusiveness: Who have we included?

Sutton's and the RUC's list both exclude certain types of incidents such as army vehicle accidents, accidental shootings or deaths due to trauma, brought on by a conflict related incident. The RUC database does not contain deaths which have occurred outside Northern Ireland. Our database includes such categories as fatal heart attacks, suicides and includes all trauma-related deaths known to us which can be proven to be Troubles-related. For example, if there is evidence to support that a fatal heart attack was a direct consequence of a bomb explosion or a shooting or took place on hearing of the death of relative, friend or neighbour injured or killed in the conflict, we have included it.

Decisions about inclusiveness were informed by information and reports available on each death. One difficult decision has been to exclude the Mull of Kintyre crash in which twenty nine people were killed. All of the dead were security force personnel. On balance however, deaths due to this incident were excluded on the grounds that such an accident could conceivably have occurred had the Troubles not been happening.

What to do about gaps in the data?

Table 1 shows the percentages of missing information in the database. For example we are missing details on the ages of 43 people who have died in the conflict.

As is clear from Table 1 we are missing a high number of religious identity of people killed. On closer examination, it was found that a high percentage of these missing values were the religious identity of RUC members killed. Using data from the Fair Employment Agency, we established that the composition of the RUC is 92.2% Protestant. For the purposes of analysis only, we assigned "Protestant" as the religious affiliation of 92.2% of the RUC officers in the database, for whom we did not have religious affiliation.

Difficulties arose when attaching post-codes. There were cases in which we did not have a full address of incident, but only the name of the town of where the fatal incident happened, e.g. Lurgan or Enniskillen. In such cases, it was impossible to attach a postcode. For these, the incident was attributed to the centre of the town. For example, if we only had the address as "Lurgan" we took the postal code for an address in the town centre, taking the postcode for a house in the centre of High Street, Main Street, or Market Street. We followed this method consistently in cases where we did not have a full incident address.

Other problems arose where areas had been redeveloped and where street names changed. To find these addresses, we used early editions of postcode directories. There remain some outstanding postcodes which were recorded as missing values. (See Table 1.) The majority of these are in rural areas throughout Northern Ireland Addresses and postcodes of victims who lived outside Northern Ireland have been listed as "GB" for Great Britain, "ROI" for Republic of Ireland, and so on.

Whether to make the database available in its fullest form?

When we completed the insertion of postal codes, the names and addresses on the list were eliminated on the advice of the Board of Directors, in order to protect the privacy of families bereaved in the Troubles.

The database is now as reliable and comprehensive as can be achieved within the time and resources available for its establishment. However, we would emphasise that there is no perfect database and work on the database is ongoing all the time in order to check and correct all information included.

[The following tables (table 11 and 12) were extracted from the main body of the report. Row totals were added for ease of use.]

• The analysis described here has confirmed many of the findings and trends identified by others. However:
• the suggestion that deaths peak in the summer months has not been supported but rather October and November have emerged as the months with the highest death rates;
• different cycles have been identified among the organisations principally responsible for killings, with peaks of activity occurring at different points in the period 1969 - 1994;
• the overwhelming majority of those killed in the Troubles have been male, with the death risk highest in the younger age groups in the 20-24 age group highest, and almost 26 percent of all victims aged 21 or less;
• the absolute number of Catholics killed is greater than Protestants killed, and the death rate for Catholics is greater than that for Protestants, as other researchers have found;
• however, if security forces deaths are included in the analysis of death by religion, and Catholics killed by Republican paramilitaries and Protestants killed by Loyalist paramilitaries are excluded, the death rates for Protestants and Catholics become much closer: - 1.9 per 1,000 for Catholics and 1.6 per 1,000 for Protestants;
• civilians are the largest category killed, and make up 53 percent of the total killed, with the British Army as the second largest category making almost 15 per cent, Republican paramilitaries accounting for almost 13 percent, the RUC accounting for 8 percent and the other groups each accounting for less than 6 per cent;
• Republican paramilitaries have killed 74 per cent of all Protestants killed, over 25 per cent of all Catholics, and almost 96 per cent of those who were classified as "Non Northern Ireland;"
• Loyalist paramilitaries killed 19 per cent of all Protestants killed, almost 50 per cent of all Catholics and just 2 per cent of the "Non Northern Ireland category;
• the overall analysis of deaths may be hard for people from centain areas to accept, since their own local experience of the Troubles is very different to the overall pattern;
• our findings on the distribution of deaths, calculated a death rate by ward and found a concentration of deaths in Belfast with only 15 of the 57 highest ranking wards outside the Belfast area. Derry Londonderry and Armagh account for most of the remaining wards;
• the distribution of deaths in the Troubles was correlated with the Robson deprivation indicator, and whilst the wards with the highest death rates also score high on the Robson index, no overall statistical association between death rate and deprivation;
• however, when the number of civilian deaths per ward was used, the correlation between deaths and deprivation scores just about reaches significance (.501), and it is clear that wards with high deprivation
• scores predominate amongst those with the largest number of deaths;
• similarly, the percentage change in the yearly death rate was correlated with changes in the Northern Ireland Gross Domestic Product (GDP) and the numbers unemployed. Whilst a negative correlation was found between death rates and GDP, and a positive correlation between deaths and unemployment, this is probably due to other factors;
• we compared deaths in the Troubles with the annual suicide figures for Northern Ireland, and found a negative correlation, suggesting that
• suicide rose as deaths in the Troubles decreased;
• the analysis of death rates from other conflicts internationally suggests that Northern Ireland is not in the 'first league' of intensive killing.
  

The next phase of the work will concentrate on the economic impact of the Troubles, and the relationship between economic trends and the Troubles will provide the focus of further statistical and qualitative work. New data on mental health, economic and other aspects of the effects of the Troubles will be generated by a survey of a sample of the population of Northern Ireland.

As we finish writing this paper, the prospects of peace and political settlement in Northern seem to be changeable. The work of counting the costs of violence in a society whilst violence has continued has been a heart-breaking one, and has seemed at times like an endless - and perhaps pointless - task. We conclude this paper in the belief that the cost of violence matters, irrespective of the identity of the victim or the perpetrator. Careful auditing of the effects of violence will, we hope, contribute towards the growing sense of urgency and understanding of the need to find peaceful means of change.

Notes