The report also draws attention to the legacy of practical and personal difficulties which continue to affect those who suffered human rights abuses in the 1980s.

1.SELECTION OF CASE STUDY AREAS
Archival material provided evidence that human rights abuses were widespread throughout Matabeleland North and South, and also at times in the Midlands of Zimbabwe. It was decided to canvas actively additional data, but time and funding excluded collection on a national scale. After consideration it was decided to concentrate data collection in two administerial districts only; Tsholotsho/Nyamandlovu in Matabeleland North and Matobo in Matabeleland South.

Data on record made it clear that the two parts of Matabeleland had qualitatively different experiences of the Government action, with Matabeleland North being subjected to a massive 5 Brigade onslaught in 1983, and Matabeleland South experiencing an extremely long and harsh food embargo, together with mass detentions, in 1984. The decision as to which administrative district to target in each province was made partly with practical criteria in mind: the two chosen areas are near to Bulawayo, and readily accessible from it. CCJP also already had a substantial number of interviews from Tsholotsho on their files. The presence of Bhalagwe Camp in the second chosen area, Matobo, was an important selection criterion.

The two areas targeted for the case studies were:
1. TSHOLOTSHO/ NYAMANDLOVU: in the early 1980s, Tsholotsho Communal Land north of Bulawayo, was administered together with the more sparsely populated commercial farmland of Nyamandlovu adjacent to it. (This adjacent commercial farmland has since been incorporated into an administerial district known as Umgusa: the map of Zimbabwe on page designates district boundaries as used in this report, which in a few cases do not coincide with district boundaries recognised in 1996). Atrocities by Government agencies were known to be severe in Tsholotsho in 1983: the adjacent commercial farmland of Nyamandlovu was known to have been hard hit by dissidents. Making Nyamandlovu part of the case study area allowed for the inclusion of data on dissident atrocities in the commercial farming and forestry resettlement areas of Nyamandlovu: there was almost no information on dissidents forthcoming from people based in the Tsholotsho Communal Lands.

2. MATOBO (known as KEZI District prior to the 1980s), a largely communal area south of Bulawayo, where atrocities were known to be severe in 1984. In particular, there was already substantial data on record of detentions, beatings and killings at Bhalagwe Camp, near Maphisa (previously called Antelope).

Further evidence of atrocities in other parts of the country came to light during this process, and tables showing known atrocities in all affected areas can be found immediately following the two main case summaries in Part Two of this report.

2.A SUMMARY OF DATA SOURCES
Reliable statistics [of human rights abuses] are extremely difficult to come by in Zimbabwe. It is often all but impossible to verify reports of army abuses. The reports one hears in Harare about atrocities committed by dissidents often sound indistinguishable from the reports one hears in Bulawayo about atrocities committed by the security forces; neither side acknowledges any legitimacy in the other's version of events.

This report has sought to overcome the difficulties in collecting data on human rights abuses by relying upon a variety of data sources. The nature and quality of these sources are very varied and, in the case of press reports, at times conflicting, but together the data provide a complex picture of the 1980s conflict, and probably as complete a record as there is now ever likely to be. An outline of main sources follows.
I. Catholic Commission for Justice and Peace in Zimbabwe (CCJP) archival material, collected in the 1980s.

II. i: Bulawayo Legal Project Centre (BLPC) archival material, including records of legal clients.
ii: BLPC current material: current paralegal clients with legal problems arising from the 1980s, and interviews conducted in the case study areas in 1995/96.

III. Human Rights Reports, including:

i) Zimbabwe: Wages of War - A Report on Human Rights, published by the Lawyers Committee for Human Rights, New York, 1986.(Referred to in this report as LCFHR).

ii) Zimbabwe - A Break With the Past? Human Rights and Political Unity: An Africa Watch Report, Richard Carver, October 1989.

iii) Amnesty International Reports and Memoranda.

iv) CCJP Report on Torture in Zimbabwe, presented to the Zimbawe Catholic Bishops Conference (ZCBC), January 1987.
IV. Media reports contemporary to the1980s, both local and international, including newspapers, magazines and video clippings. The most comprehensive source here proved to be The Chronicle, Bulawayo's daily newspaper. As well as detailing much dissident activity, The Chronicle provides useful insight into the "official view" of events, recording the opinions and pronouncements of Government office bearers as events unfolded.
V. Academic research, including most notably:

i) two conference papers written by Jocelyn Alexander and JoAnn McGregor: these are part of a broader collaborative research project undertaken with Terence Ranger, which will cover a wide range of twentieth century history in Matabeleland North.

ii). Richard Werbner, Tears of the Dead: The Social Biography of an African Family, Baobab, Harare, 1992.
VI. Selected interviews with CCJP officials, commercial farmers and others.
VII. Medical and other material evidence: medical records and evidence from 3 sets of exhumed bodies.

3.A DISCUSSION OF DATA SOURCES

1: CATHOLIC COMMISSION FOR JUSTICE AND PEACE

The CCJP provided invaluable archival files on atrocities, compiled when the 1980s disturbances were taking place. As data were being collected simultaneously with events occurring, CCJP accounts remain the most accurate and valuable source, particularly in terms of dates: they also capture the horror of those years in a way less contemporary accounts cannot. Priests and doctors were recording events and noting the broader picture as well as the details, such as the movement and numbers of troops, as well as civilian casualties. CCJP files provide a firm framework within which data from other sources has been placed in context.
Strict curfews prevented the movement of all civilians in Northern Matableland during parts of 1982 and in early 1983, and in Southern Matabeleland in early 1984. This meant that resident mission staff were among the few who observed closely and recorded the unfolding of events during these years. They also made strenuous efforts at the time to protect people and to bring an end to the atrocities.

TSHOLOTSHO has three Catholic missions: Pumula Mission in the southwest, Magama Mission in the east, and Gwayi Mission in the north. In addition, there is Regina Mundi Mission, which is on the Tsholotsho-Lupane border, and whose parishioners are all from Tsholotsho, as there is only forestry land on the Lupane side of this border. Reports on events filtered back from all these missions. St Luke's Mission, which is also in neighbouring Lupane, has a hospital, and recorded some Tsholotsho victims among its patients.

MATOBO has two Catholic missions: St Joseph's Mission in the south-west, and Minda Mission in central Matobo. In addition, there are several Catholic schools - Guardian Angel School, St Thomas School and St Mary's School, along the western border of Matobo and Bulilimamangwe. There are also Brunapeg, Embakwe and Empandeni Missions in Bulilimamangwe. Again, mission staff at all these missions monitored events in their regions and kept invaluable records.

Presentation of CCJP data is of various types, and includes the following:
1. Seventeen very detailed statements, sworn and witnessed in front of lawyers, which were prepared for the Government Committee of Inquiry into alleged atrocities by security forces in 1983 and 1984. These are each several pages long and are accompanied by copies of medical records in a few instances. In all instances they give full details of victims, times, perpetrators and places where events occurred. There are also other well-documented and prepared statements by civilians, which were not notarised, as they were not ultimately selected for presentation to the Committee.

2. Detailed hospital records from mission hospitals, recording precise name, age, date of arrival, village of origin and the nature of injuries suffered by hundreds of victims. Injuries include evidence of beatings, bayonetings, burnings and gun shot wounds. There is a long statement of events in early 1983, made by a doctor at St Luke's. In addition there is also a long written statement from a government doctor working at Tsholotsho District Hospital, sent in February 1983 to the CCJP and detailing information given to him by patients, as well as his own observations of events in the village of Tsholotsho itself. There are also details of victims beaten and shot by soldiers from a doctor at Embakwe Mission, in Matabeleland South in 1984.

3. A significant data base, known as "Matabeleland Case Files", listing names and other details of approximately 1000 victims. There are several thick interview files which contain some, but not all, of the source interviews for this data base.

4. Letters written by priests at the various missions, recounting their horror at what they were witnessing and appealing for intervention and help.

5. Many other letters from Catholic priests or parishioners appealing for help in locating missing family members, or detailing other atrocities. Some of these are written by priests resident in Bulawayo or elsewhere, who have had news of events affecting their friends or families in the rural areas of Matabeleland.

6. General reports which were submitted to the Government at various times during 1983 and 1984, giving evidence of human rights violations by both security forces and dissidents, and appealing for a more humane approach to the security problem.

7. Files with lengthy legal documentation concerning specific people detained without trial, including requests for information as to their whereabouts, requests for detention orders to be reviewed, requests for medical treatment for certain detainees. There are also other files on detainees listing page after page of people known to be in detention at Chikurubi, or other centres, at certain points in time.

8. Statements taken by CCJP members based in Bulawayo in the 1980s, made by refugees from the rural areas.

Taken together, the CCJP raw data amount to well over a thousand pages, providing a comprehensive record of what happened in those years.

SHORT-COMINGS OF CCJP DATA
1. Letters or accounts written when atrocities were ongoing frequently do not name victims or informants, in order to protect them from further harm should the evidence be intercepted. There is one recorded instance of a person being murdered subsequent to making a phone call to Bulawayo reporting atrocities, and other instances in which people were detained and tortured after making phone-calls, and told this was the reason for their detention: concern for the safety of informants was very real. However, it makes it difficult to decide whether events described, perhaps by 3 or 4 different sources in Feb 1983, are all referring to the same set of victims or different ones. For example, there are 4 accounts among CCJP records of 2 pregnant girls being bayonetted to death by 5 Brigade in Tsholotsho in Feb 1983. In all 4 accounts the victims are not named and the exact location is imprecise. This was treated as one case validated from several sources, probably the one given in great detail in BLPC interviews 1146-1168 incl. It is impossible now to try to validate such CCJP accounts independently.

A conservative approach has always been taken when trying to quantify atrocities: it is always assumed accounts overlap unless there is a very good reason for not doing so, such as clear difference in location or timing of the alleged events. For this reason, many brief accounts of atrocities had to be completely disregarded as they lacked the detail to enable their distinction from other atrocities on record.

2. There is often no follow up on file to a letter of inquiry about a "missing person". Many young men in particular fled the country for Botswana or South Africa, or moved into town with relatives, but were too afraid to write and inform their families, so it is possible at least some "missing" persons turned up, perhaps even years later. Many may have turned up in detention centres and been released, or may have joined a gang of dissidents, but there is no way of knowing from available evidence.

Again, a conservative approach has been taken, so that people are not presumed missing unless the report of their disappearance is substantiated by other evidence suggesting they remained missing. Numbers of actual missing may therefore be higher than numbers given in the case studies.

3. Information on those in detention is incomplete. Typically, all one can say is that a person with a certain name was in a certain jail during a certain month. Where that person was originally detained, how long he had already been in detention or remained afterwards in detention, and who originally detained him, are details that are usually not given. For example, there is a large file on Chikurubi detainees from 1985, merely listing names of those in Chikurubi at the time.

Some people were in detention for 3 or 4 years, and others for a few weeks. Many were tortured. Certainly, the vast majority of those detained never made an official report of their detention and release to an independent body such as the CCJP: figures of those in and out of detention between 1982 and 1988 run to thousands, according to some sources. "Detention" is therefore not a uniform experience with the same implications for every detainee, and the actualities of every individual case, or real numbers of detainees, will remain speculative.

4. The "Matabeleland Case Files" had some short-comings: many names were ultimately discarded as being accompanied by too little information to make them useful. In some cases, information consisted of a name only, with no clear indication of alleged offence against that person, or district or perpetrator. Other entries contained some of the relevant information, but not enough for this project. Entries had to be accompanied by details of at least offence and year to be entered into the HR Data Base. At the same time, many hundreds of entries contained full details, and 431 victims were added to the HR Data Base from the Matabeleland Case Files.

II : BULAWAYO LEGAL PROJECTS CENTRE
i) ARCHIVAL DATA
LEGAL CASES: The BLPC original data base consisted of approximately 100 legal cases. The bulk of these cases involved representation of people by lawyers working for private law firms, who made their data available to BLPC. Most clients were people who had been detained under the Emergency Powers legislation. Several involved "missing persons".

ii) CURRENT DATA
PARALEGAL CASES: Approximately another 100 cases, predominantly deaths, were brought to the attention of the BLPC by their paralegals who, from the time paralegal offices began opening in rural Matabeleland in 1990, started receiving requests from clients for help in obtaining death and birth certificates. These cases involved people from all districts in Matabeleland.

INTERVIEWS: CCJP personnel had already collected many interviews from Tsholotsho residents in 1993/94, and this data had been incorporated straight into the BLPC Data Base. This base was extensively increased by further interviews in 1995/96, using the combined resources of CCJP and LRF.
1. TSHOLOTSHO - data was collected in Tsholotsho on a ward by ward basis. Tsholotsho is divided into 16 administrative wards, and all were visited in the course of 1995. Twelve visits were made, each lasting two days and taking in one or two wards. In most cases only one person was available to record the interviews, although on a few trips, a second interviewer was able to dramatically increase the number of cases processed in the short time available. Interviews were conducted in Ndebele, and written up simultaneously in English. Arrangements were made in advance with the ward councillors, who were asked to inform the inhabitants of their ward that the interviewer would be attending a certain central point in the ward on a certain day. Councillors and people giving evidence were told that the interviewer wished to collect data relating to what happened in the 1980s, to document any injuries or losses suffered by people during those years, whether at the hands of security forces or dissidents.

All evidence was given entirely voluntarily, and without suggestion of reward or promise of future compensation. Speaking about those years was visibly traumatic for many of its victims. While those who came gave evidence freely, some told of other victims who were still too afraid to come forward and tell their stories. A number of key witnesses made appointments to speak to the researcher and then felt they could not do so, and stated that it was fear of possible harm to themselves that had made them reconsider.

An examination of the data base also makes it apparent that while some victims are reportedly too afraid to speak out, there are others who have now told their story to various different bodies in the last 13 years. The same interviewee names and details of events are, in a few dozen cases, on file in CCJP archives, on BLPC paralegal files, recorded in interviews conducted by CCJP personnel in the early 1990s and/or recorded in interviews in 1995/96.

In other instances, many different interviewees recount the same incidents, naming a constant list of victims, particularly in incidents involving substantial numbers of deaths, such as hut burnings. These collaborating accounts span more than a decade and are often collected from widely distanced parts of the country.

The number of people who turned up to give evidence varied from ward to ward: in certain wards, particular councillors were inefficient about informing residents about the impending visit in good time. In one ward of Northern Tsholotsho, virtually no information was forthcoming on the first visit, and this appeared to be owing to lack of information given to residents. In 1996, the interviewer conducted a final series of visits to all the wards to identify some of the people who had been unable to give evidence the previous year. This brief trip resulted in a further 160 named victims, and once again, the small area in northern Tsholotsho produced very little data. It therefore seems reasonable to conclude that 5 Brigade missed this area in their initial sweep through Tsholotsho, as the reported cases only refer to 5 Brigade passing through the area in pursuit of dissidents in August 1983.

However, data collection in Tsholotsho remains far from complete: those who gave evidence in the final round of sessions in 1996 spoke of yet others who had not come forward. It was also noteworthy that out of all the testimonies collected on this last round, fewer than a dozen of the named victims were already on record.

A total of 910 named victims in Tsholotsho was collected through these interviews, many of whom suffered more than one human rights violation. The interview data also indicated huge numbers of unnamed victims. A more detailed discussion of this can be found in "Methodology" (see section 4 of this chapter), and in the case studies themselves. While the data collection process was far from exhaustive, it helped provide a clearer picture of the scale and nature of the violations of human rights in the 1980s.

MATOBO - the process of data collection here followed a similar pattern to that used in Tsholotsho. However, time ran out before interviewing had been carried out in all wards. Only 10 weeks were devoted to data collection in Matobo, with most of this time being devoted to publicising the project and setting up sessions. The Matobo Case Study is therefore more of an extended pilot study than a complete record of events in all areas. Interviewing was limited to 9 one day sessions at 6 different venues. Local councillors were not always supportive of the exercise, and in some cases actively undermined it, ordering people not to come forward. The CIO also put in what was perceived by the interviewers to be an intimidatory appearance at some sessions. In spite of this, a total of 350 named victims were identified, and thousands of others were implied by witnesses.

SHORT-COMINGS OF BLPC INTERVIEW DATA
1. Inadequate interviews: of the interviews made by CCJP personnel in Tsholotsho in 1993, approximately 50 left serious gaps in their accounts. Interviewees assumed local knowledge of places, which were therefore not always named. Interviewees would also be primarily concerned with their own experience, and so fail to provide general details of events on a certain day. For example, an interview might read:

"They came and took everyone in the line to the school. They beat us and then they shot people dead, including my brother, named XX."
Such information produces more questions than answers, and only one named victim. Fortunately, these interviews all referred to events in the Pumula Mission area, an area which was well covered by other data sources, in particular File H. 40 names from BLPC sources coincided with more comprehensive accounts of events in File H, and many other names coincided with events in villages documented by CCJP. Cross referencing of these multiple data sources allowed for a clear picture of events in the case study areas.
A revised interview form devised by BLPC and used thereafter by CCJP personnel, provided more comprehensive data. This form required precise details of the perpetrator, including clothing, weapons etc, and precise details of where the alleged incidents took place and who else was involved or witnessed events, and caused a dramatic improvement in the quality of information collected. A further handful of interviewees nonetheless were unable to give adequate details, usually because they were now very old and forgetful, and in a very few other cases because interviewees were mentally confused: in these cases the interviewer always noted his assessment of the interviewee. For example, one old man whose child went missing in 1983 was only able to keep repeating: "I want my son."

2. The Time Lapse: The BLPC interviews were conducted a full 12 years after the bulk of atrocities occurred in early 1983. While people interviewed were very clear as to the nature of their loss or injury, other details were forgotten. A person might know that on a certain day, his entire homestead was burnt down, or that his son was killed, and remember the perpetrators clearly, but not know whether this event happened in February or March, or even what year it happened. While dates have been recorded as given, there is every likelihood that some are inaccurate. Fortunately, data collected closer to events (such as CCJP files) have frequently cross-referenced with data collected in 1995/96, and has helped clarify the timing of certain events.

3. Rape: this remains dramatically under-reported. While CCJP reports - and The Chronicle - referred to widespread rape at the time, people are not willing, 12 years later, to report it. This is understandable and reflects a general reluctance of women to report rape under any circumstances. Many victims will now be married with families and will have put the incident behind them: to probe too deeply would be counter-productive. Reading between the lines, some interviews pointed to rape having occurred, but when interviewees were asked directly by the interviewer if rape took place, this was denied. The following extract is one such instance:

"The 5 Brigade came after dark when we were sleeping. They forced their way into the house and asked if we had any daughters. When we said our daughters were only young and were sleeping, they went to the bedroom, and took our 2 daughters aged 12 and 14 to the forest, where they beat them for half an hour, then brought them home...."

This interview was coded in the HR Data Base as a beating, not a rape, in accordance with the interviewees' assessment of the event.
In Matobo, men referred to widespread rape, especially in Bhalagwe, although the number of women admitting to rape remained far smaller than the men's accounts suggested.

4. False Information: This of course cannot be entirely ruled out, but it seems improbable that many people would be motivated to bear false witness at this stage. People do not easily invent dead relatives, and were not led to believe they stood to benefit by doing so: interviewers were careful to point out that the data collection process was for the historical record only, and not for purposes of individual compensation.

There are often more than 30 interviews testifying to events in a small area, and on occasions, some of these reports are made many miles away from the concerned village, by somebody who has been resettled or married away from that village in the last decade. It seems almost impossible for such witnesses to have colluded, so many years later and at comparatively short notice. There is also the obvious distress - and fear - that many people show in recounting these times, indicative of real, as opposed to invented, suffering.

In addition recent interviews have often served to confirm events on record in CCJP files since the 1980s. People giving witness also provided full personal details, so knew they were not making statements anonymously. Some interviewees even submitted death certificates or medical records to the interviewer for photocopying and returning.

5. Dissidents: Information on dissident atrocities was barely reported in Tsholotsho. Yet other sources indicate that dissidents were indeed a menace in the area. In particular, dissidents coerced food from villagers, and also committed rape. For the reasons described above, rape was under reported: furthermore, 10 years after the event, people may not feel it is worth specifically reporting occasions on which they were coerced into killing chickens in order to cook for and feed dissidents. The degree of sympathy for dissidents during those years and the role this might play in under-reporting, is discussed at greater length under "The Dissident Problem" in Part One, III: on the whole, there was apparently little sympathy for dissidents.

Independent research in adjacent districts of Northern Matabeleland suggest dissidents did not commonly murder villagers, unless they were considered sell-outs, were ZANU-PF officials, or had informed on dissident movements. In Lupane, for instance, independent researchers estimated a minimum of 750 deaths during the 1980s, of which only 25 were thought to have been committed by dissidents: of these 25, some were considered to have been committed by Government agencies in disguise.

In Tsholotsho, among an estimated 1000 dead, a total of 18 murders by dissidents were reported to interviewers. In addition, 21 deaths were inflicted by dissidents in the commercial farming area of Nyamandlovu adjacent to Tsholotsho. There were, however, many other references to army members disguising themselves as dissidents and committing crimes. This phenomenon is reminiscent of the war for Independence, when the Rhodesian Selous Scouts used to dress and pose as members of the guerilla forces.

BLPC DATA: EVIDENCE OF ATROCITIES COMMITTED IN THE 1970s
While it was not the primary intention of this report to collect data on events relating to the 1970s war of liberation, some information on people who went missing during the late 1970s was reported both to paralegals and to those interviewing specifically for this report. A total of 23 such reports was made involving people who left the country for guerrilla training and never returned. The relatives of such "missing persons" are eligible for compensation under the War Victims Compensation Act (see final section of this report for more details), and these reports were accordingly dealt with by paralegals.

In total, BLPC data amounted to more than 5 000 pages of raw information.

III: HUMAN RIGHTS DOCUMENTS
1. Lawyers Committee For Human Rights: Zimbabwe: Wages of War, New York, 1986.
The Lawyers Committee for Human Rights (LCFHR) has served as a public interest law centre since 1978. The committee works to promote international human rights and refugee law and legal procedures in the United States and abroad. Their Zimbabwean report was compiled after two visits to Zimbabwe in 1985 and 1986, during which committee members interviewed a wide range of Zimbabweans, including a large number of Government officials. Wherever possible, information given in interviews was independently checked and verified. The final report was written in May 1986.

This report provides a well-documented account of the conflict in Zimbabwe during the years following Independence. Its findings coincide to a useful degree with those of the current report. In addition, it provides an overview of various aspects of those years which it has not been possible for this project to research independently, and which would now be difficult to research, a decade after the events. For example, the timing and magnitude of various mass detentions and events in the Midlands, in particular in 1985, were well covered by LCFHR. Their scholarship is thorough and their estimates conservative: this is now apparent in the light of the evidence used for the present report, which indicates far larger numbers of dead and injured people and destroyed homesteads in the case study areas than LCFHR suggested. This makes the consideration of LCFHR estimates in non-case study areas seem reasonable.
The main shortcoming of the LCFHR report is the fact it was written in 1986: the disturbances continued for a full two years after its publication, until the Amnesty in 1988. This means potential key interviewees were in detention, or were hesitant to come forward at the time: the committee therefore had to rely on Government versions of figures, for example of damage caused by dissidents, there being no other data source. It also means there is no information in the report on events during the last two years of the disturbances, including the second Treason Trial in 1986 and the wave of detentions that accompanied this.

2.Richard Carver, Zimbabwe: A Break With the Past? Human Rights and Political Unity:an Africa Watch Report, October, 1989:
3. Richard Carver, Zimbabwe: Drawing a Line Through the Past, Amnesty International, June 1992:
4. CCJP Confidential Report on Torture in Zimbabwe, January 1987.
5.Memorandum to the Government of the Republic of Zimbabwe, Amnesty International, May 1986.

All the above human rights reports contributed both to the data base and to the overall historical record of events in Zimbabwe in the 1980s. They added a limited but well authenticated number of named torture victims to the HR Data Base, and also provided names of prisons where torture and other human rights violations took place. Carver's reports also gave a useful insight into the human rights violations in Zimbabwe as being a partial consequence of Rhodesian personnel having been retained in government agencies after Independence.

IV: THE CHRONICLE

This report dealt with The Chronicle as a separate entity, with a separate data base of recorded victims and perpetrators. The picture resulting from this can be seen in Part Two, III and IV.
The Chronicle remains one of the primary sources of dissident atrocities during the 1980s. There was without any doubt a serious dissident problem at the time, although it is also now clear that there were several separate groups of "bandits", with varying motivations. [For a more detailed discussion of dissidents, see Part One, III].
A total of 562 offences, committed between June 1982 and March 1988 and involving mainly dissidents but also some Government agencies, were identified from approximately 1500 media reports extracted from The Chronicle. Those media reports which did not refer to offences contained a record of public statements by Government officials and running details on various trials of dissidents, politicians and government agents.

The Chronicle records many attacks by dissidents on civilians, tourists, Government construction projects, and Government resettlement programmes. There were also many robberies and rapes perpetrated by dissidents. However, certain aspects of The Chronicle's reporting suggested it was better kept as a separate entity: in particular, it was difficult to cross-reference the incidents it reports with other data sources.

GENERAL OBSERVATIONS:
1. Peasant victims are seldom named, but tend to be referred to as a number of victims:
eg. "5 peasant farmers in Tsholotsho were killed by dissidents since the beginning of the month." As names are not given, nor precise villages, it is impossible to cross-reference these sorts of statements with, for example, BLPC interviews or CCJP data.

2. The perpetrator is almost invariably given as "dissidents" or "bandits", with very few acknowledgements of atrocities by security forces. It is only in instances where individual members of the security forces were prosecuted, which were rare, that the newspaper reported such atrocities. Most references to security force atrocities take the form of vociferous denials.

3. When acknowledged, deaths of civilians at the hands of security forces are at times referred to as being "deaths in crossfire", implying the unintentional killing of innocents where dissidents were the target. This is reminiscent of the statements made by security force headquarters during the 1970s, where civilian deaths were invariably accounted for in this way. Of the approximately 3500 named victims on file from other sources, there are in fact only 7 interviews which refer to 5 people killed and 2 homesteads destroyed in genuine cross-fire.

4. Detainees were named only if they were prominent members of society, or white. Similarly, white murder victims were invariably named.

5. The political nature of the disturbances is very clear from The Chronicle reports. Speeches made by Government office bearers and quoted in the press, make it apparent that it was PF-ZAPU that the ruling party sought to destroy, as well as the handful of dissidents operating at the time. This issue of these two overlapping conflicts has been referred to above, and is further explored in the Historical Overview following: in general, there are many statements referring to supporters of PF-ZAPU and supporters of dissidents as being one and the same menace, deserving of one and the same fate - "to die or go to prison", as Minister Enos Nkala, put it. The LCFHR also makes a strong case for the perception of the problems as being primarily political.

6. The Chronicle lists atrocities in 2 ways.
i) SPECIFIC REPORTS: there are weekly or monthly news reports, detailing incidents during these short time-spans. These could be considered "Specific Reports", as there is often some accompanying detail as to location and events, such as precise date and value of property stolen or destroyed from a particular store or mine. In articles listing "bandit" or "dissident" activities, large and small incidents are often given almost equal coverage.
ii) GENERAL REPORTS: The second listing of atrocities occurs in reports of speeches made in Parliament, stating general totals of atrocities, usually for the previous six months. These were read out as evidence for the need to continue the state of emergency, which had to be renewed by Parliament every six months.

It is very noticeable that the numbers of atrocities announced in Parliament is always significantly higher than the sum of the Specific Reports for the same time-span [see Part VII, comparative Tables III and V page ***]. Particularly noteworthy here is the disparity for "murders" reported in 1986. "Specific Reports" record only 9 murders by dissidents in that year, while the "General Report" for 1986 refers to 116 civilian deaths. A further confusing factor, when Government statistics are considered, is the phenomenon of Government agencies committing crimes "disguised" as dissidents (see below). As all official information and sources for Government figures on dissident atrocities were state controlled, it is impossible to resolve these discrepancies now.

7. Incidents which occur in very different parts of the country are not always clearly distinguished from each other, but may be listed together in one article. In fact, there were atrocities being committed by the Mozambique-based MNR in northern and eastern Zimbabwe during the 1980s and an analysis of Specific Reports shows that 10% of atrocities were not committed in Matabeleland or the Midlands.

It is not always clear to a casual reader which events occurred where, and whether ZIPRA sympathetic or Renamo (MNR) dissidents, or ordinary criminals were responsible. This type of reporting seemed to confuse the foreign press at times: for example, in the Sunday Times of London, 6 March 1983, there is a report called "Timetable of a Massacre". In it, the murder of a white farmer in Chinhoyi, the raiding of an armoury in Mutare, and the murder of three British tourists in Nyanga are included by this foreign journalist in a list of "dissident" atrocities which he represented as giving some justification to the Government's decision to send 5 Brigade into Matabeleland. All the above events actually took place in northern and eastern Zimbabwe, and in fact, the murder of the three British tourists strongly implicated 5 Brigade itself, which was training in Nyanga at the time.

8. Once The Chronicle reports had been collated for all issues between June 1982 and March 1988, with victims' names (where possible), dates and perpetrators extracted, these were cross referred with names collated from other sources into the Human Rights Data Base. The Chronicle Specific Report data amounted to 562 entries, and other sources amounted to 3 534 entries. It was discovered that fewer than 40 names could be cross-referenced. [If time and money allowed, no doubt many more cases could be verified: the 40 coinciding cases are merely those that overlapped without every newspaper reference being actively pursued.]

Of the names and incidents that could be cross-referred, 21 involved murders by dissidents in the commercial farming area of Nyamandlovu. Here all data sources agreed the perpetrators were dissidents in every case. Approximately 10 other cross-references involved the detentions of prominent ZAPU leaders, some of whom were in detention for many years. Here all sources agreed on obvious aspects of the detentions, such as who detained the men and when, although the sources may have disagreed on other aspects, such as allegations of torture of victims in detention.

In the remaining 7 incidents, which included a bus burning, the murders of 3 chiefs in Matabeleland, a shoot-out at a rural shopping centre in Inyathi, and the murders of health clinic staff in a car ambush in Nkayi, there were glaring disparities between eye-witness accounts given to independent sources, and the official version of events as represented in The Chronicle. In every one of these cases, The Chronicle attributes events to dissidents, but eyewitnesses put forward convincing arguments that the perpetrators were in fact government agencies.

Usual arguments for concluding that Government forces were the perpetrators include:
i) the inability of perpetrators to speak Ndebele fluently (all dissidents were, by both the dissidents' and the Government's own definition, Ndebele speakers).
ii) the fact that victims were often known to be hostile to the Government or have other political significance. For example, the Inyathi shopping centre shoot-out involved a prominent opposition ZAPU party member. While he in fact survived, seven others died, including several from the party member's family.
iii) the police and CIO either did nothing to prevent events taking place even if they were on the scene of the crime, or showed no interest in solving the crimes, even when perpetrators were positively identified to them by witnesses.
iv) the perpetrator was personally recognised as a specific member of a Government agency, known to the witness due to prior contact. On occasions, for example, members of 5 Brigade would parade as dissidents, then appear as 5 Brigade the next day, and punish villagers for having failed to report their own "disguised" presence the previous day.

However, as previously mentioned, most of The Chronicle reports did not specifically name victims. If the reports which specify location of atrocity are totalled for the first case study area, The Chronicle attributes 50 murders to dissidents in Nyamandlovu and Tsholotsho as a whole, including murders on commercial farms. This is fairly similar to the total of 39 murders arrived at via the HR Data Base. However, as most of The Chronicle's victims in Tsholotsho are unnamed, specific cross-referring of victims is not possible.

Reports in The Chronicle do not always indicate where murders took place, and the official view was certainly that Tsholotsho was a hot bed of dissident activity, which does not correlate well with the mere 14 murders in Tsholotsho that The Chronicle specifically identifies. The impossibility of reconciling such disparities at this stage is a major reason for keeping The Chronicle data separate: the two sets of data results are presented in parallel in Part Two, III, and readers of the report must draw their own conclusions. While dissidents are seldom regarded as perpetrators of crimes by villagers interviewed, The Chronicle almost never acknowledges atrocities by the army.

In summary, it seems fair to say that while there is certainly much substance in The Chronicle's portrayal of the "dissident menace", there are also contradictions and apparent inaccuracies within its reports, which justify maintaining its data in a separate base.

V: ACADEMIC RESEARCH

There is very little published academic research dealing with the history of events in Zimbabwe in the 1980s. Most historical research still seems to be concentrated on the less politically contentious task of establishing a more complete picture of the War of Liberation and the colonial years that preceded the war. However, there are a few key documents on the 1980s which have provided invaluable background for this report.
1. Richard Werbner, Tears of the Dead: The Social Biography of an African Family, Baobab, Harare, 1992. This anthropological work provides a comprehensive history of one extended family, based on interviews conducted in 1960/61 and further interviews in 1989. The "family", which consists of almost 500 people in all, is primarily located in Matabeleland South, in an area immediately adjacent to the second Case Study Area. This document therefore provided an invaluable insight into how the arrival of 5 Brigade was perceived by those in the Bango chiefdom in 1984.

2. Key research is currently being conducted into events in Lupane and Nkayi. This research is part of a broader research project in which Jocelyn Alexander, JoAnn McGregor and Terence Ranger will document the social history of this region for the last one hundred years. Events of the 1980s are therefore a small aspect of their research, but it has produced two papers of particular interest. These are:
i)Jocelyn Alexander, Dissident Perspectives on Zimbabwe's Civil War, Seminar Paper, St Antony's College, Oxford, 1996.
ii)Jocelyn Alexander and JoAnn McGregor, Democracy, Development and Political Conflict: Rural Institutions in Matabeleland North After Independence, presented at the International Conference on the Historical Dimensions of Democracy and Human Rights in Zimbabwe, Harare, September 1996.

This research is based largely on first hand interviews with civilians, including those who were dissidents in the 1980s, and has been of key importance in reconstructing the history of those years.

3. Various other academic documents have contributed to the writing of the Historical Overview in this this report, including:
i)D. Martin and P. Johnson, The Struggle for Zimbabwe, ZPH, Harare, 1981
ii)D. Martin and P. Johnson, (Eds), Destructive Engagement: Southern Africa at War, ZPH, Harare, 1986.
iii)N Bhebe and T Ranger, (Eds), Society (Vol 1) and Soldiers (Vol 2) in Zimbabwe's Liberation War, UZP, Harare, 1995
iv)J Hanlon, Beggar Your Neighbours: Apartheid Power in Southern Africa, Indiana University Press, 1986.
v)K Yapp, Voices From the Conflict: Perceptions on Violence, Ethnicity, and the Disruption of National Unity, Paper from The Britain Zimbabwe Research Day, St Antony's College, Oxford University, 8 June 1996.

Other written sources were used for very specific information, for example in the chapters on "Legal Damages" and "Implications of Organised Violence": these references are cited in the appropriate chapters.

VI: INTERVIEWS
A few selected, in-depth, interviews were conducted in 1995/96 by the research coordinator, to answer specific questions which needed clarification after other data had been analysed. In particular, commercial farmers were approached, as it was hoped their evidence could shed some light on dissident activities in the case study areas. Remarkably little evidence of dissident presence or activities was apparent from other data sources, yet there were, without question, dissidents committing atrocities during the 1980s. Farmers were in fact able to confirm dissident atrocities in the commercial farming areas.
A few interviews were also conducted with CCJP officials to clarify aspects of troop movements, and some gaps in the chronicle of events. These interviews were for general background purposes.

Interviews were also conducted in Johannesburg in September 1996, with a few individuals who it was hoped might know details of the extent of South Africa's involvement in destabilising Zimbabwe in the early 1980s. These included two journalists, and two ANC officials, one of whom works for the Truth and Reconciliation Commission. To date the South African role in Zimbabwean events still remains largely shrouded in mystery, although some new details are gradually coming to light. Hopefully more details will surface as the Truth and Reconciliation Commission continues.

SHORTCOMINGS OF INTERVIEWS
As with the BLPC interview data, the time lapse has taken its toll on what people can now remember of events. The interviewer was, on occasion, more in touch with those events, having better cause to be so, than those who were more involved at the time. People also destroyed key documents, having felt such documents were endangering their personal safety during the years when house searches and detentions were commonplace. Other documents have been destroyed more recently, in the belief that they were no longer of interest to anyone.

These shortcomings notwithstanding, personal interviews with people with a "larger view" of events proved very enlightening.

VII:MATERIAL EVIDENCE
Project personnel established that corroborating evidence for claims of epidemic violence in 1983/84, now made by over a thousand victims, exists in bulk in some places: some rural hospitals have, on their admission records, listings of hundreds of civilians admitted to their wards during the 1980s, suffering from beatings, bayonetings, gun shot wounds, and burns. Some of these records have already been referred to under CCJP data above. Hospitals where such records are known still to exist include not only the Catholic mission hospitals but also other mission hospitals.
The Government hospitals in Bulawayo and rural Matabeleland and the Midlands are also known to have admitted such patients, some of whom were referred to these better equipped hospitals, such as Mpilo in Bulawayo, by mission doctors unable adequately to treat seriously ill patients. Doctors who were employed in Government hospitals during the 1980s have independently confirmed this. The orthopaedic surgeon who was at Mpilo in the 1980s has confirmed that from mid-1982 onwards, he saw patients suffering from gun shot wounds. The 1982 patients were army personnel and "dissidents" allegedly wounded in shoot-outs. The latter were kept under armed guard in the wards. Then in early 1983 this same surgeon became alarmed at the sudden influx to his wards of gun shot wound and assault cases affecting civilians: at the request of colleagues, in March 1983 he compiled a list of current patients including their names, injuries and treatments and submitted it to the Minister of Health. These included gun shot victims, and patients so severely beaten by soldiers that some later died of renal failure. He also took photographs of patients and submitted a set to the Minister, who insisted that the photographic negatives be surrendered too. A duplicate set of prints had been made by the surgeon, and these are still on file in the Nederlands, as is a complete duplicate set of these medical records.

There are a few individual cases well substantiated by medical records, notably victims whose cases were prepared for the Committee of Inquiry by CCJP in 1984.

Apart from these cases, there are currently on project files only a handful of medical records substantiating claims made by interviewees, although many other victims claimed to have such records, but did not bring them to the interviewing venue. In other cases, victims brought records with them, but there were no photocopying facilities in the rural areas where interviewing took place, and interviewers, having no medical background themselves, were not in a position to note relevant details from such records. They were also hesitant to take such records away with them, as the logistics of returning them to remote rural dwellers were daunting: in any case, such records were in some cases needed on a current basis, by people paying regular visits to clinics.

In many other cases, victims did not still have medical records, or had never had them, having been too afraid to seek medical attention at the time.

There were also very few post mortem or death certificates issued for the dead which acknowledged violent causes of death, although a handful of death certificates acknowledging violent deaths are on project files.

There has been to date no large-scale, co-ordinated exhumation of the bodies of those persons whom others claim to have been murdered, in order to conclude independently their cause of death. However, bodies were exhumed from mine shafts in the Midlands and Matabeleland South in the 1990s, with coins in their pockets dating their violent deaths to the 1980s: bodies exhumed at Cyrene Mission in 1984 showed clear evidence of recent gunshot wounds. There is thus a handful of cases which have forensic post mortem evidence to substantiate the types of atrocities claimed by many hundreds of people.

Other material evidence is the existence of many mass grave sites, throughout the curfew areas of 1983/84. Many such sites were indicated in the interviews in the two case study areas, and were also brought to the attention of report personnel by those doing independent research in Matabeleland North. A few such sites were actually visited by project personnel, to confirm their location. People in both Matabeleland North and South also refer to the way in which bodies were thrown down mine shafts by Government agencies, and the findings in the two mines mentioned above point to the probable truth of this claim, and also to the possiblity of many other shafts which still contain bodies not yet exhumed.

People who had homesteads burnt down have also often not rebuilt on the identical foundations to the missing huts: the floors and foundations of such destroyed huts are recognisable in the case study areas.

The lack of specialised examination of such material corroboration of claimed abuses is a shortcoming readily admitted to by this report, which operated under severe funding and personnel constraints. It would have been unethical for personnel involved in this report to have tried to conduct forensic investigations, and to have thus tampered with potential evidence: this report seeks merely to bring to the attention of properly authorised and qualified personnel, the existence of material evidence which could be used to corroborate or contradict the report's claims, if the State so decided.

Similarly, claims of psychological disturbances still experienced by victims of the 1980s upheavals has to remain inferential in this report, based on what victims themselves said in their interviews, where they frequently referred to insomnia, anxieties, dizziness, headaches and other possibly psychosomatic symptoms which they date as having onset after particular events in their lives. Inferences can also be made based on known psychological consequences, which have been forensically established in work with civilians who suffered similar types of trauma in Zimbabwe during the 1970s. That those who experience psychological and physical torture suffer recognisable types of stress in consequence has been widely established, but to date there are no studies forensically corroborating this for 1980s victims in Zimbabwe.

4.METHODOLOGY
INAMED VICTIMS - HUMAN RIGHTS DATA BASE

The names of victims were collected from all the above-mentioned sources. With the exception of The Chronicle data, names were collated in the Human Rights Data Base (HR Data Base), which included all named victims from all districts of Zimbabwe.
The bulk of the named victims in the HR Data Base is from BLPC sources, with CCJP archival material providing the next largest number of victims. Human rights documents and academic sources provided a small number of named victims, whichfrequently validated names from other sources. In approximately one thousand cases, names would ultimately be validated from more than one source, with 3 or more confirmations occurring for more than three hundred victims: additional sources on any name were noted on the file print-outs.

Each victim was categorised and had the following information recorded in a running table:

1.A NUMBER was allocated
2.The SOURCE of data was indicated by a set of letters, such as CC for Catholic Commission, or PL for paralegal: the initials of lawyers, authors, or persons conducting interviews were also used.
3.The OFFENCE was indicated by a further set of letters, with most serious offence listed first in cases of multiple offences. More than half of the victims suffered multiple offences, such as physical torture and detention, or death and homestead destroyed. A complete key for offences is given in Table One below.
4.The NAME of the victim, including his or her surname and first names, was recorded. If the victim's own name was not completely indicated, the name of his or her spouse or parent was included.
5.The DISTRICT in which offence took place.
6.The PERPETRATOR, as alleged by interviewee.
7.The YEAR and MONTH of the offence.
8.The AGE of the victim was recorded, but only if the victim was under 18 years of age.

The sex of the victim was not recorded in the running table, although the distribution of male to female victims was separately assessed, by returning to the raw data in the case study areas. (Sex is usually apparent from the names of victims in any case.)

Periodically, data were sorted by the computer alphabetically according to districts and names, including first names, to eliminate the same victim being listed several times from different sources. At times, more than one person with the same name was established as having died or suffered injury, but this was only concluded after returning to the raw data, to compare the complete circumstances allegedly surrounding each incident.

TABLE ONE - CODE FOR OFFENCES

XDead

MMissing

PProperty loss

ASPhysical torture: Assault with Sticks, gun butts or blunt object

ABPhysical torture or injury resulting from Burns

AByPhysical Torture: Assault with Bayonette, knife or sharp object

AGGunshot wound

TPhysical Torture: including electrocution, water torture and other tortures not covered by above categories.

DDetention

RRape

It will be noted that various types of physical torture have been differentiated: in the case study areas, the phenomenon of "mass beatings" is also dealt with as a separate entity. This is to draw attention to beating, and in particular "mass beating", as the preferred means of physical torture during those years, in particular by 5 Brigade.
SUB-SECTIONS OF HR DATA BASE
When it became apparent that the data base was going to run to several thousand victims, it was sub-divided.
1.HR.1 consisted of 2 152 entries, including all data collated up until February 1996, from BLPC and CCJP sources.
2.HR.2 consisted of 411 entries, including data collated from academic and human rights sources, and two files of CCJP interviews conducted in the early 1990s.
3.HR.3 consisted of a severely reduced version of the CCJP "Matabeleland Case Files", excluding all those names already listed from other sources and all those without sufficient details. Remaining names amounted to a further 431 entries.
4.HR.4 consisted of 540 entries, representing all data collected from interviewing from July 1996 to October 1996.
5.HR.5 was a temporary data base constructed by moving all named victims from Matabeleland South already listed in HR.1, 2 and 3 into a sub-section, to facilitate comparing of names coming in from interviews in the Matobo region in late 1996 and being filed in HR.4, with those already on file from Matabeleland South.

The HR Data Base, inclusive of sub-sections HR.1, 2, 3 and 4 consists of 3 534 names, inclusive of all sources and districts of Zimbabwe.

The data base was closed at the end of October 1996 in order to facilitate graphing of existing data. However, data continued to be submitted to the BLPC, through the paralegals. Within a week of the base being closed, a further 8 deaths were reported to BLPC. In 7 cases, 5 Brigade were allegedly the perpetrators and in 1 dissidents were blamed. In the same week reports came in of one gun shot wound caused by dissidents, 4 cases of property losses (2 allegedly caused by ZANU-PF Youth and 2 by 5 Brigade), and 2 cases of assault, allegedly by 5 Brigade. This serves to highlight once again both the continuing problems facing people in areas affected by the 1980s disturbances, who continue to seek legal help, and the fact that the data base collated for this report is far from complete.

IITHE CHRONICLE DATA BASE

All The Chronicle news reports relating to the 1980s disturbances were extracted, from June 1982 to March 1988. Information about alleged victims was entered into a data base separate from, but identical to, the HR Data Base, for reasons discussed already.
As previously mentioned, these reports could be referred to as consisting of either "Specific" or "General" information. Only "Specific Reports" were entered into the data base. "General Reports" were treated separately (see Part Two, III for comparative tables and graphs).

As victims were often not named, the given number of victims in a news report frequently had to be entered instead of names. The names of farms, stores and bus companies were entered, when these were available and names of actual persons were not given. The value of property lost was entered if specified.
The Chronicle Data Base consists of 562 entries.

IIIUNNAMED VICTIMS - HUMAN RIGHTS DATA BASE

Apart from named victims, there were vast numbers of unnamed victims evident, not only from the interviews, but also from CCJP archival material, where victims were more often represented as numbers than names. Certain other documents, such as the LCFHR account, also referred at times to numbers of people injured or detained, without naming everyone.
In addition, in all districts apart from the two Case Study districts, named victims on file were from unsolicited sources, either archival CCJP names, or the names of legal clients with problems pertaining to these years. Most districts are therefore considerably under-represented on the named data base.

It became obvious that while it was important to keep the data base of named victims running, additional ways of assessing numbers of victims had to be found, if a realistic picture was to emerge.

On the HR Data Base, a number of victims unsubstantiated by every name was therefore occasionally entered. This was only done when the collator was certain that those victims were not already on the data base as named victims, and where the source seemed reliable. For example, several CCJP archival files refer to "2 school teachers shot dead at Dete Road turn-off" in February 1983. No interviews of named victims on file described these conditions for any death, so it seemed reasonable to assume these were new victims, and to include them in the data base. On the whole, very few cases involving purely unnamed victims in the CCJP archives were included in the HR Data Base, because of the problem of double-counting victims.

Occasionally numbers from other sources were included, such as those from the LCFHR document. This report often uses broad numbers to indicate people detained or injured, or property destroyed in a certain city within a given time span. For example, in its account of the disturbances in Matabeleland South in 1984, there is the following statement:

An American doctor, Davee Boyd, reported that he had treated more than a 100 assault victims with broken bones and stab wounds at his mission hospital [in Gwanda District] between February and the end of April [1984].

The HR Data Base had no named assault victims from Gwanda, although it had named deaths from Gwanda on record. This above statement was therefore entered into the HR Data Base, as "100 assault victims, Gwanda".
Similarly, the LCFHR document refers to numbers of properties destroyed in the Midlands during the 1985 disturbances. Compilers of the LCFHR document actually visited some of the affected areas in the immediate wake of these disturbances, and were therefore in a position to comment reliably. The HR Data Base had comparatively few of the Midlands offences on record, particularly from Silobela, so these figures were also introduced into the HR Data Base.

The LCFHR document was well researched and substantiated, and only those figures which the compilers considered fair were included in the HR Data Base. If the compilers were not sure that a certain figure could be substantiated, they said so. For example, when commenting on the post 1985-election wave of detentions in Bulawayo, LCFHR states:

A Zapu Spokesman... said that 415 Zapu members had been detained during the month of August, but this number could not be independently confirmed. Repeated attempts to obtain the names of those whom Zapu claimed to be in detention were unsuccessful.

This figure was therefore not included in the HR Data Base. There are, however, some named detainees from other sources included under Bulawayo in the HR Data Base, supporting at least in part the contention that detentions took place at that time.
The LCFHR general figures were also not included for Tsholotsho and Matobo, the 2 case study areas, because of the very different and more detailed way in which these two areas were analysed.

IV UNNAMED VICTIMS - THE CASE STUDY AREAS
As mentioned in the discussion of data sources, BLPC interviews always included the names of victims, while CCJP records tended to deal in numbers of victims, rather than consistently naming victims However, both CCJP and BLPC records of victims tended to record "village" where events took place in the case of each victim. In the two Case Study areas it was therefore decided to use "village" as the common parameter across data sources. In this way, it was possible to integrate information on both named and unnamed victims, without counting the same victim twice.

THE "VILLAGE BY VILLAGE" SUMMARIES
This method involved going back to all the raw data in the case study areas, and re-arranging it in terms of villages where offences took place, rather than in terms of overall district, or type of offence.

The "village by village" summary of events proved to be a very productive strategy when analysing data on Tsholotsho and Matobo, and helped reveal broader patterns of events. The locations of army units at different times, in particular 5 Brigade, was also apparent with this approach.

The presence of dissidents was also indicated, but they were comparatively rarely referred to as perpetrators. Those statements indicating dissidents were therefore highlighted in the summaries by ****.

As villages were mentioned in source data, they were located on a map, and a section on every village was opened in the "village by village" summary. Interview data on each village was included in highly abridged form, and this data was added to as new details came to light.

Total offences were included at the end of each village summary, once all data had been processed in this way.

A conservative approach was taken when assessing numbers of victims. For example, if CCJP recorded 8 deaths in a given village in Feb 1983, and BLPC had 10 named victims for that village, BLPC's victims were assumed to coincide completely with CCJP's, and 10 deaths were considered the total. In such cases, the CCJP archival record served as corroboratory evidence of statements being made in the 1990s. A reading of the case studies themselves will illustrate more precisely how different sources were used in conjunction with each other. CCJP sources are indicated by **, while source interviews are indicated by their HR Data Base file number.

As there was a high level of corroboration between sources throughout the case study areas, CCJP numbered victims were included for villages where there had been no information gathered in the 1990s.

In many of the interviews conducted in 1995/96, witnesses often tended to concentrate on a few named victims, without specifying more general numbers of victims exactly. For example, an interview could include the comment: " besides my father, many, many people died that day". No attempt has been made to quantify such statements: they are merely indicated in the Total Offence summary at the end of that village as "1 known victim, plus others".

PHYSICAL TORTURE

Mass beatings of villagers was a significant phenomenon of 5 Brigade activity. Interviews and CCJP files refer repeatedly to its occurrence, but what this means in terms of actual numbers of victims is difficult to assess. Many interviews refer to "all the people in their line" being marched at gun-point to a certain point and then being beaten.

The term "line" can mean very different things, in terms of population. Generally speaking, it refers to the way villagers were made to lay out their settlements when they were forcibly resettled in Tsholotsho by the colonial Government in the 1950s and 1960s. Homesteads were literally arranged in long lines, along the dirt tracks in the area. A "line" can indicate anything from 3 "sabuku" areas, to an entire school catchment area, running for several kilometers. A "sabuku" is an official, sometimes elected, but usually inherited or appointed, presiding over usually 6 to 10 families. So a "line" could be from around 20 to 30 families, to at least treble this number. Each family could conservatively be estimated to have 5 members (2 adults and 3 children), although in reality most families are larger than this. This means numbers of people present at a "mass beating" could be anything from 100 to several hundreds.

The problem then still remains as to what is meant by "everyone" being beaten. In some cases, even the elderly were beaten, and certainly women were beaten: interviews will refer at times to the women being allowed to take turns holding the babies in between beatings. Children aged 12 and upwards were also frequently beaten.

The number of villagers forced to witness mass beatings runs to thousands, and includes all age categories. Everyone present at such beatings was a victim of torture - either physical, if they were actually beaten, or psychological, if they were forced to witness the beating of others. For a full discussion of this, see Part Three, I .

A conservative estimate of 50 present at such beatings has been made.

DETENTION
Detentions have proved difficult to quantify: at one level, anyone who is held at gun-point or translocated against his or her will can be said to have been detained, and to have experienced intimidation and trauma. At another level, there were many hundreds of people who were detained for long periods of time in police or army camps or buildings of one sort or another. Again, it is not easy now to quantify how many.