The Integration of Guerrilla Armies into Conventional Forces

Lessons Learnt from BMATT in Africa *


Maj Genl A W Dennis, CB, OBE (rtd)
from the UK, previously Director of Military Assistance Overseas


Paper presented at a conference on Changing Dynamics: Military-Strategic Issues for a Future South Africa, hosted by the Institute for Defence Politics in conjunction with the Hanns Seidel Foundation, CSIR conference centre, Pretoria, 6 August 1992.

Published in South African Defence Review Issue No 5, 1992



Let me first make clear exactly who I am and what my credentials are. I was for 34 years a member of the British Army. I retired in January 1985 and have had no significant links with the services or defence intellectual circles in the intervening 7 1/2 years: indeed I have been running the fourth largest fund raising charity in the UK during that time. However, my last 3 years in the Army, 1982, 1983 and 1984 were spent as the United Kingdom's Director of Military Assistance Overseas. As such I managed the United Kingdom's military assistance on a tri-service basis world wide. Though not intimately involved in any individual commitment (that was the job of the Team Commander in country) I was responsible for advising Her Majesty's Government (HMG) and the government of the countries involved on the assistance that could sensibly be provided and, once the team was committed, for ensuring that it was functioning effectively. Given that experience I suppose I have some qualifications for speaking to you today.

But in doing so I must make several things clear. The views I express are my own. In formulating my ideas I am drawing on my own experience, on reading of much literature, and on private conversations with erstwhile colleagues, but I have received no official help or support from HMG and nothing I say should be thought to reflect British government policy in any way. Next, in order to keep my subject within reasonable bounds, I am going to confine my comments to land forces and will not try to cover sea and air forces as well. And finally I must make it clear that I speak as a soldier not a diplomat. I hope I will not say things that offend but if I do I crave your understanding. I am not an expert on South African affairs. I can only do my best to express views that I believe to be in the general interest. Where, as in this country, opinions are sharply divided as to the general interest, I think it almost inevitable that some of what I say will prove unpalatable.

I am basing my presentation on a brief historical survey of the activities of the British Military Advisory and Training Team Zimbabwe (generally called BMATT), hoping to draw some principles out as I go along. Having outlined the history of BMATT I will examine whether any of the lessons learned could be applied to the South African situation and if the answer seems to be positive I shall conclude by making some suggestions as to how this might, in my view, be done. I will also touch briefly on the activities of BMATT Namibia even though I have no first hand experience of this work.

The Lancaster House agreement, signed in December 1979, is the obvious place to start. That agreement contained provision for the introduction of a Commonwealth Monitoring Force to supervise the cease-fire, to encourage the Patriotic Front (PF) forces to report with their weapons to nominated rendezvous points (RVs), to move those that did report to the RVs into Assembly Places in which they could more easily be looked after administratively in the longer term and, of course, to oversee the withdrawal of the Rhodesian Security Forces to their barracks. It was a truly remarkable operation which, before the event, was thought most likely to end in complete disaster and widespread bloodshed. These misgivings arose largely from the understandable distrust that existed on all sides despite the signing of the agreement at Lancaster House. The distrust was not just between the Rhodesian Security Forces and the Patriotic Front but also between ZIPRA and ZANLA as well. After all we had a situation where three armies had been involved in a long and bloody conflict which no one had lost in a military sense. Indeed all three probably felt that they could have won militarily, given time. And all three elements, for differing reasons, initially distrusted the Monitoring Forces; the whites because they were being "sold down the river," the Patriotic Front because they thought the predominantly white monitoring forces would be partial to white Rhodesian interest.

But despite the deeply ingrained distrust the operation worked. Why? First, because a genuinely joint command and control system was established at the outset with all three factions and members of the Monitoring Force being represented at equal ranks at all levels. Patriotic Front Commissioners were deployed with Monitor Team Liaison officers to supervise PF activities at RVs and Assembly Places. Second because it quickly became apparent to all that suspicion of the Monitoring Force was unfounded - they acted in a totally even handed manner. Third, because of the bravery of small detachments of monitoring force personnel and their ZIPRA and ZANLA Commissioner teams in the RVs who never overreacted even in potentially very dangerous circumstances. Indeed half way through the week long RV phase it became apparent that success was not to be achieved by sitting waiting for suspicious PF members to report to RVs and detachments bravely went into the bush in ones and twos to overcome PF fears. By D+8 no less than 15 730 members of the PF had reached Assembly Places. And finally, because the Monitoring Force successfully undertook a task that was completely unexpected, the feeding and support, including medical support of PF members in Assembly Places while belated arrangements were made by the Government to take on this task.

Some simple principles can be deduced from this operation. First, that an outside agency, in this case the Monitoring Force, can help to allay the natural antipathy of people who have seen each other as enemies for many years. Next, that joint and equal representation at all levels, assisted by a neutral umpire, can resolve problems. That professionalism, calmness, bravery and flexibility are all essential attributes for the neutral umpires. And finally that one way to men's hearts is through their stomachs; had the administrative needs of PF members in the Assembly Places not been met trouble would almost certainly have arisen.

The next milestone occurred in March 1980 with the election of ZANU and Robert Mugabe with an absolute majority in Parliament. The Monitoring Force then left and BMATT was established. At this stage the threats in the emergent Zimbabwe probably fell under three heads:
  1. Civil war or coup d'etat
  2. Political schism caused by Joshua Nkomo and ZAPU
  3. A general exodus of whites with all that that would mean for the economy
Certainly the risk of the first and third of these threats would be significantly lessened if properly integrated security forces could be created quickly and a stable security situation achieved. And stability was also essential if foreign investment was to be forthcoming. It was probably for reasons such as these that Mr Mugabe asked the UK to provide Zimbabwe with military assistance. An added attraction of a UK BMATT was probably the reassurance it would provide to the white population, who would perceive it as a moderating force in the country.

A significant problem in achieving integration is common to both Zimbabwe in 1980 and South Africa in 1992 - only the old established security force had the framework of administration, command and control upon which a single army could be constructed and those security forces were totally mistrusted by the guerrilla forces. Another significant problem was the impossibility of simply discharging large numbers of unnecessary soldiers - all had to be found places in the new army or elsewhere if trouble was to be avoided. And to compound the problem, all three armies had totally different military backgrounds - the RSF based on the UK model, ZANLA with Chinese backing and ZIPRA trained on Russian lines. There was, therefore, no natural "meeting of the minds" and progress had to be "brokered" by much difficult negotiation. In the early days the white dominated RSF were clearly endeavouring to preserve their position by, for example, insisting on the need to maintain so called "standards" that ZANLA and ZIPRA did not have the training to meet. The Patriotic Front were naturally frustrated, not least because black majority rule was now a fact and yet they did not seem to be "calling the shots". It is not difficult to envisage a similar situation arising in South Africa, is it? So, at the highest level, the principle task of BMATT was to act as an active catalyst to keep the factions together and to provide a neutral element respected by all. This demands the presence of experienced officers of appropriate rank - in the case of Zimbabwe an infantry Major-General has headed BMATT from its inception. If a similar role is envisaged in South Africa, my view is that nothing less than 3 star rank will suffice.

BMATT Zimbabwe was not originally established to provide a presence in all the training establishments in which integration was being undertaken. Indeed the original concept was to train a cadre of leaders and instructors only and for them to train other instructors who would then train the rank and file. In the event this proved unsatisfactory partly because lack of a "neutral" presence in some establishments allowed trouble to erupt and partly because the process was proving too slow - again a recipe for disaster. So the whole process had to be speeded up. With the help of BMATT, a large number of ZANLA and ZIPRA potential senior, middle-piece and junior leaders were selected and put through a standard one month course focusing on basic skills, discipline and administration. A battalion's worth of leaders were then "married" to some 450 men each from ZANLA and ZIPRA and with the assistance of a BMATT major and warrant officer for 6 months trained their battalions. The initial rate of progress was one battalion every 2 weeks. It had been intended that a large proportion of ZANLA and ZIPRA would become so called Soldiers Employed on Economic Development (known as SEED) but this apparently excellent idea foundered due to lack of commitment and by late 1980 it had become obvious that every PF soldier would have to be integrated and the process was further speeded up to create 3 battalions per month, with a consequent increase in the size of BMATT. However, serious interfactional violence occurred in early 1981, indeed 3 of the 14 battalions that then existed were actively involved. Once again the "neutral" members of BMATT were able to play a part in cooling things down, though by far the greatest credit for re-establishing control belonged to the old Rhodesian Army and Air Force. In the event that outbreak proved three important things:
  1. It became evident that neither ZIPRA nor ZANLA was strong enough to impose itself on the other;

  2. The Joint High Command and the Government held together; and

  3. The former Rhodesian Army and Air Force proved loyal and could be trusted.
By the time Robert Mugabe's Government was one year old the most pressing problem was to extend the unification process up the chain of command above battalion level. In February 1981 twenty seven potential Brigadiers and Colonels were selected and after special training assumed many senior appointments; indeed by August 1981 the Joint High Command gave way to a Defence Headquarters co-located with the Ministry of Defence with former PF officers in the posts of Army Commander and Deputy Commander, two of the three chiefs of staff and 4 brigade commanders. At the same date the white component, which had been 2 800 at independence, of which 680 were officers, had fallen to 1 135, many of them rather too old for comfort.

By early 1982 integration itself was complete. The army was some 65 000 strong, staff training courses were well underway, the Military Academy was producing platoon, company and battalion commanders and a start had been made at creating appropriate logistic systems. All in all a pretty remarkable achievement in less than 2 years since the election of Zimbabwe's first black majority government. It has to be acknowledged though that the force that had been created was far larger than any perceived threat warranted, that discipline and training still left much to be desired and that administration was only of a rudimentary nature. It is on these aspects that BMATT has largely focused in the ten years since then.

One reason for the apparently slow rate of progress in these ten years has been the involvement of the Zimbabwe army in operations in Mozambique. As early as 1982 the decision was taken to assist with the defence of the communications corridors through Mozambique to the Indian Ocean. A brigade was committed in the Biera Corridor to start with but more troops were later deployed to protect the civilian engineers undertaking major repairs to the railway which runs parallel with the Limpopo River and joins the south of Zimbabwe to Maputo. About two thirds of the Zimbabwe army was deployed on operations either on the Mozambique border or along these corridors in 1991. This operational role had an affect on BMATT in that the role of the Battalion Battle School at Nyanga was adapted to provide appropriate training lasting about 2 months for battalions before they deployed. By 1989, however, the rapid turnover of units meant that training time had to be limited to the provision of a short, sharp refresher course rather than the previous course which covered both conventional and low intensity techniques. I think Zimbabwe can claim with pride that its units have proved themselves effective and impressive on operations in Mozambique, but I have to acknowledge that their logistic and maintenance systems are still pretty chaotic and that many other major deficiencies exist. Nevertheless, by comparison with most new armies (remember it was only 2 years old when first committed to operations), they have proved themselves in operations to be of well above average standard.

As will probably be known to all of you, others also provided training from time to time - for example the North Koreans, Chinese and Tanzanians. In my opinion, this diversification was a mistake since the differences in military philosophy of these nations must create problems within a single army and in fact none of the other teams actually lasted very long. Perhaps the original reasons for diversifying had to do with the UKs inability or unwillingness to provide a Training Team of sufficient size to meet all Zimbabwe's perceived needs? Or maybe it was felt that too many eggs were in one basket? Or even that inducements in some form or other were being offered? I do not know which of these was the reason; but I am quite clear that if South Africa chooses to invite some other nation to provide assistance it should stick to one nation only, making certain that the provider is prepared to meet the assessed needs of the assistance programmes in both the short and longer term. And in South Africa's case it would not be unreasonable to look for a very substantial contribution to the costs of the providing nation.

BMATT Zimbabwe is to-day principally involved in 3 activities. First, advisory. Here I think the British have demonstrated great sensitivity and a willingness to adapt to local needs rather than trying to impose our own systems. The second role focuses on staff training both at the Junior and Senior Staff Colleges which the team originally commanded and provided all the instructors but now only deploys several members of the Directing Staff. The third area of activity is in logistics, particularly at the Logistics School in Harare. In my experience lack of appreciation of the need for properly functioning logistic and maintenance systems is the biggest single failing of most African armies. It still is in Zimbabwe as indeed it is in armies elsewhere in the world.

Turning now briefly to Namibia. You will no doubt recall that the Angola accords were signed in Luanda on 22 December 1988. In November 1989 SWAPO won 57% of the votes in the Namibian General Election and immediately requested the help of a British Military Advisory and Training Team following independence on 21 March 1990. The team, initially 55 strong, was duly deployed on 26 March 1990 and the first leaders cadre, for the 1st and 2nd Battalions, was run from 17 April to 2 June. By 1 July the 1st Battalion, about 1 000 men strong, accompanied by 5 BMATT Advisors, had deployed to the northern border. By November 1990, only four months later, the 5th Battalion had deployed and in early 1991 the 21st Guards Battalion had also been formed, four staff courses had been run, support weapons and logistics training was well advanced (indeed a logistics battalion deployed as early as July 1990) and an operational test exercise had been conducted. In addition the Ministry of Defence, a mixture of civilian and military personnel, was operating as a department of state. No-one would pretend that everything was working perfectly, nevertheless, a great deal had been achieved in the first year following independence. Most people would probably agree that at some 7 500 strong the Army is unnecessarily large, but sensible plans will need to be made for the employment of any surplus soldiers before they are discharged. Integration has not been easy to achieve, at least in part, because of the need to use several interpreters to cope with the wide variety of languages involved. Battalions are made up of approximately 70% ex-PLAN and 30% ex-SWATF. This mixture could have proved explosive but hounded by their BMATT instructors they united in a common task (or perhaps in the face of a common enemy!) and soon realised that they could work well together. At the higher levels, integration has been more patchy, at least in part because of the departure of most white South African and SWATF officers. But the Government's intentions seem clear in that it decided to split the four MOD directorates evenly, appointing two white and two black (ex PLAN) directors. In all this, BMATT Namibia has played a role remarkably similar to that of BMATT Zimbabwe.

Looking back with the benefit of hindsight I conclude that the most important achievements of BMATT Zimbabwe and Namibia were:
  1. To gain the trust of groups who initially distrusted each other and all of whom distrusted BMATT;

  2. Having gained that trust BMATT was able to play the role of "honest broker" and thereby enable progress to be made more easily between erstwhile enemies;

  3. By including people of appropriate seniority and experience the Team was able to suggest and develop ideas - in short to be a catalyst for change;

  4. By virtue of their own professionalism they were able to create, and where necessary run, training and other establishments;

  5. And, most important of all, to accept that their most prized objective is so to improve standards in their field of activity that they work themselves out of a job, and the sooner the better.
What then of the integration of the South African Defence Force, MK and TBVC forces? Here you must bear with me as I will inevitably display a degree of naivety about the affairs in your country. But as an outsider looking in I can see several similarities in the military spheres between South Africa in 1992 and Zimbabwe in 1980. However, we must be quite clear that in at least two important respects the situation is totally different. First the size and military capability of the SADF is far greater than that of MK, whereas the RSF, ZIPRA and ZANLA were more evenly matched; and second the scale and nature of the wars that proceeded integration were quite different. But dealing now with the similarities.

First the fact that the SADF, like the RSF, is an effective conventional force in being and with the necessary infrastructure to support it. As with ZIPRA and ZANLA, the same cannot be said for MK and TBVC forces. So, like it or not, if a conventional force is required, the most efficient way to achieve it is to base it on the SADF.

Next, a similar degree of distrust seems to exist between SADF and MK as existed in Zimbabwe. This distrust is compounded by the direct involvement of SADF in internal security. A significant reduction in the level of distrust would probably be achieved if the SADF could be disengaged from its internal security role - a tall order perhaps but a true statement in my judgment.

The sooner genuine integration can be achieved the better since only then will ingrained distrust be overcome. In achieving acceptable integration the existing SADF must recognise the need speedily to incorporate members of MK and TBVC into key and senior positions. To enable these positions to be filled effectively, special training requirements will have to be met.

Large numbers of MK, largely untrained for conventional operations, will have either to be absorbed speedily into the new Defence Force (which will impose significant training requirements) or be otherwise suitably employed. The latter is preferable since the former will result in a Defence Force of a size well beyond South Africa's needs with all that means for the economy.

The higher management of defence will need to be changed to accommodate political imperatives. At the same time the new Defence Force must be created as an apolitical one with loyalty to the government of the day and firmly under civilian control. A Ministry of Defence combining civil and military would seem essential.

A careful reassessment of force roles and strengths will be necessary. To an outside observer like myself, it is difficult to see any threat to South Africa's security that requires the state to maintain forces even on the present scale and with the heavy equipment of the SADF, let alone that which will result if MK and TBVC forces in their present numbers are integrated with it.

Indeed, integration would probably be easier to achieve if the role of the Army were to be changed significantly. Let me explain why I argue on these lines. If the existing Army remains and its role and equipment are unchanged, integration will simply be a question of incorporating MK and TBVC into it - a role for which they are untrained and in which they would be numerically outnumbered. Given a different role, all three elements would be involved in the creation of something new to which all could have a significant contribution to make despite, or even by virtue of their differing experiences and backgrounds.

With these rather broad comparisons made, you will not be surprised if I conclude that a BMATT type of operation might be helpful if integration of SADF, MK and TBVC forces is required in South Africa. There may be differences in detail but many of the problems are so closely paralleled that the experience of BMATT Zimbabwe could, in my opinion, be used as a useful role model. Much care would have to be lavished on getting the design of the team right to begin with and forecasting how requirements will change with time. Priority tasks should, in my opinion, be:
  • Providing an "honest broker" element at the top levels of defence management to ease tensions and assist progress. I believe an officer of at least 3 star rank will be required to perform this role properly. He will need appropriate support, including perhaps a civil service element to help with designing systems for civil control of the new defence force.

  • Improving quickly the expertise of members of the MK, particularly senior and middle-piece officers and instructors of all sorts, to enable a proper proportion of key roles to be filled by them. Assistance could be provided both in South Africa and in the country providing the Advisory and Training Team.

  • Providing a presence on the ground to advise and assist at training establishments. Some might advocate a small "roving" team. Drawing on the Zimbabwe experience, I would recommend a presence at all training sites being used for integration to act as "neutral umpires". This will be expensive, but not as expensive as a serious breakdown of trust could turn out to be.
Next, while not a lesson from Zimbabwe, I would recommend a major commitment to the improvement of the general acceptability of the Police. Until that is achieved it is hard to see how the Defence Force can be dissociated from internal security and if it is so involved, I find it very hard to see how proper integration can be effected.

Finally, two concluding observations. I repeat again my view that if a BMATT type operation is envisaged all its functions should be provided by one country. Diversification will result in the introduction of differing and probably incompatible military philosophies. This will impede, not assist integration. I am, of course, talking about integration. My comments about single nation assistance would not apply if peace keeping or situation monitoring was the task. In that case, multi-national forces have positive advantages to offer - not the least of which is a demonstration of concern by the international community as a whole.

And lastly, while I do not speak with any official authority, I would be surprised if the UK would agree to provide a BMATT unless the request came as a joint one from all interested parties.