Operation Vula, with whose destruction I began this dissertation and with whose genesis I have just ended the tenth and final chapter, is the climax of our story. Footnote It provides us with a kind of promontory from which we can now, in this conclusion, look back and make summary observations about the ANC's line of march.
Looking back from Operation Vula, we can see how little ANC operational strategy actually changed between 1976 and 1986 - indeed, from the 1960s. Vula would be the final attempt by the ANC - or, more precisely, by a section of its leadership which had been given an open operational mandate - to beget a revolution in South Africa. Vula was supposed to be the answer to chronic shortcomings of earlier ANC strategising. And its originators held to the old fundamental assumptions of ANC operational strategy: Revolution was necessary to achieve ANC policy goals; armed struggle was the central requirement for that outcome; revolution was not merely possible but `inevitable'.
The last assumption - about the `inevitability' of revolution - set the questions that ANC strategists asked themselves. Associated with this assumption in the ANC's case was a contradiction in terms sometimes evident in Marxist-Leninist thinking. Revolution was said to be `inevitable', provided the ANC devised and implemented strategies and tactics that were suited to the historical conjuncture. From this it followed that, if the ANC was not succeeding in bringing about revolution, it was because its actions were in some way inconsistent with the drift of history. In this case, there were, broadly, three types of error the ANC could be making to cause this. The first was the simplest: the ANC might be implementing the appropriate strategy and tactics but be doing so carelessly. The second was a little more serious: the strategy and tactics decided upon might be inappropriate. In this instance, the ANC might have divined the conjuncture `incorrectly' and so decided on the inappropriate operational strategy. Or, it might have identified the conjuncture `correctly' but wrongly determined the strategy to suit it. The third was grave: the ANC's entire divination procedure - its theory of revolution and the assumptions it contained - might be false.
Between 1976 and 1986, ANC strategists never seriously questioned their theory of revolution. But they often had reason to conclude that a particular strategy was not working. Yet they could seldom say with any certainty whether this was the result merely of careless implementation or of the inappropriateness of a strategy or set of tactics. One reason for this difficulty was that monitoring mechanisms and operational management in the ANC were abysmal. This was attributable to a number of shortcomings. There was no ANC leadership presence on the ground inside South Africa. Key officials often failed to attend meetings of the top operational organ, the Revolutionary Council (RC), and of its successor, the Politico-Military Council (PMC). Rank-and-file ANC members had almost no ability to reward good operational leadership and punish bad: members of the RC and PMC were appointed by the National Executive Committee (NEC); and ordinary ANC members were given only two opportunities to influence NEC membership - in 1969, the year the RC was set up, and 1985. Parallelism between political and military structures made the timeous exchange of mutually relevant information extremely difficult. Personality and inter-departmental rivalries undermined decisions that were taken from time to time to improve information flows between sections. Furthermore, long and poor lines of communication into and out of South Africa, as well as between ANC machineries in different countries, meant that good information on which changes to methods and strategies might have been based was often out of date by the time it reached ANC decision-makers in Lusaka or Maputo. Assessing strategy and deciding on improvements to it could, hence, be a highly speculative process.
Whatever the reasons were for the ANC's operational difficulties, there is no doubt that ANC operations suffered a high failure rate. This is so whether measured by the criteria and objectives which the ANC established for itself or according to those that I might choose. Often the failure was of the most basic kind: the ANC's failure to implement what it had decided upon, or failure to change what it had resolved to change. Equally important was the failure to achieve stated operational objectives when new strategies or tactics were decided upon.
Failure was particularly apparent in the case of armed struggle, the fulcrum of ANC operational strategic thinking. Umkhonto we Sizwe (MK) did not develop its armed activity beyond the sporadic, symbolic endeavour it was when the ANC first resumed attacks inside South Africa in 1976, despite numerous decisions and attempts to do so. The number of attacks which MK mounted did, indeed, increase considerably between 1976 and 1986, as Chart One (on page 461) shows. But, as Chart Two indicates, the number of guerillas killed and captured by state security forces usually rose apace. On average, slightly more than two ANC guerillas were killed or captured by security forces for every three of the 634 guerilla attacks between 1976 and 1986 Footnote (almost all of which carried out by the ANC) Footnote . The ratio of guerillas killed and captured to attacks, or the `casualty rate' as I call it, was an average of 0.73 for the 11 years. Chart Three shows this casualty rate fluctuated year by year, and that it was lowest (i.e. most favourable to the ANC) in two years marked by sudden upsurges in popular political activity, 1981 and 1985. But there was only the slightest improvement, from the ANC's point of view, in this guerilla casualty rate as the number of attacks increased over the 11-year period.
Since many, perhaps most, MK attacks can be classified as modest in their dimensions, this casualty rate indicates a high cost for what were minimal operational rewards. Until 1984, the largest single form of attack was sabotage and the largest single target category consisted of economic installations (of which railway lines, electricity transformers and power stations were among the most common), as Charts Four, Five and Six show. Sabotage predominated, notwithstanding repeated resolutions by the RC and PMC to move towards popular armed struggle, or `people's war', in which state forces would become the primary targets. It was only in the 1985-1986 period that there was a marked change in target distribution. As Chart Seven shows, there was then a sharp increase in attacks against security forces and against civilians, while economic sabotage declined. But, after a fall in the MK casualty rate in 1985, it returned to a quite high level in 1986 and remained there for the two subsequent years. ANC guerilla casualties between 1984 and 1986 totalled 294 killed or captured. On the other hand, 58 security force members were killed and 739 were injured in the course of popular unrest.
Other statistics illustrate more starkly the weaknesses of the ANC armed activity. These are the rates of capture of the most common armament types used by ANC guerillas. On average between 1976 and 1986, for each one of the 163 hand grenades thrown by guerillas, more than nine (1,549) were captured by security forces (see Chart Eight); for each of 150 limpet mines detonated, more than three (518) were captured (Chart Nine); and for each of 82 incidents in which a firearm was used, eight (670) firearms were captured (Chart Ten). Interestingly, the exception to this appalling capture rate is provided by land mines. For every two of the 33 land mines laid by guerillas and subsquently detonated about three (49) were neutralised by security forces (Chart Eleven). It is reasonable to speculate that this low loss rate is explained by the fact that many land mines were laid only a few kilometres inside South African territory and were thus not long, if at all, at the mercy of the ANC's ordinance department or internal underground. When combined, the statistics indicate that on average between 1976 and 1986, for every three insurgent attacks, which were usually modest sabotage actions, security forces killed or captured two ANC guerillas, recovered seven insurgent hand grenades, captured three firearms and neutralised more than two limpet mines.
These statistics - which must treated with some caution since they originate from one of the antagonists, the South African Police - are offered mainly for illustrative purposes. Yet they correlate substantially with the frustrations and disappointments described by ANC strategists and members throughout the body of this dissertation.
There is no evidence to suggest that ANC armed activity ever achieved any significant degree of self-reliance inside South Africa. Had it done so, ANC armed activity might have had a substantially different, and greater, effect on outcomes in South Africa.
Self-reliance was an obvious requirement if the ANC was to make any meaningful progress towards developing a popularly-based armed struggle; but self-reliance was also the likely result of involving more ordinary, internally-based people in armed activity. Internal military self-reliance is an obvious desideratum of almost any revolutionary movement. It was, however, doubly desireable in the South African instance because the ANC had no prospect of secure rear bases in neighbouring states. Yet MK guerillas evidently had little success in developing or linking up with the few support structures inside the country. Short-term cross-border incursions were the norm and guerillas' periods of survival appear, on the whole, to have been extremely short. Guerillas who survived inside the country often remained dependent upon external command, ordinance and intelligence.
The failure of its armed struggle to make real military progress defined almost all strategic questions the ANC asked itself between 1976 and 1986. Likewise, almost all the ANC's decision to change operational strategy and structures between 1976 and 1986 were designed to remedy this failure. And a high proportion of these decisions were never implemented.
Non-implementation was most evident in cases where it was decided that political and military structures should cooperate more closely in the hope of creating the kind of organised domestic political base in which armed combatants might reasonably hope to survive. Personal and departmental rivalries and vanities persistently undermined these decisions. Political-military parallelism persisted in one degree or another, no matter what changes were decided upon for operational structures. The area political military committee (APC) document of 1981-1982 was never implemented. Attempts to develop an internal underground leadership of seniority and genuine organisational power were repeatedly deferred or frustrated - or, in the case of `Operation Butterfly' and `Operation Vula' had to be conducted as a subterfuge against not merely the South African state but a portion of the external mission leadership as well. The result was that ANC underground remained fractured between political and military components, each comprising units that were usually isolated from each other; the only link these units often had with the `ANC' was with the ANC abroad, which was unable to maintain a dynamic relationship with them or provide them with tactical guidance in what was, certainly from 1984, a fast-moving situation.
A number of other important decisions also remained largely unimplemented. Revolutionary Council (RC) decisions, some pre-dating the 1976 uprising, to develop the concept of the `MK auxiliary' - an ANC militant who received a crash course in combat work under secure conditions in a neighbouring state and was returned to live legally and fight in South Africa before his absence was noticed by security forces - was barely developed. The concept's latterday variant, the `grenade squads' of 1983-1985, was never given a proper institutional framework, with the result that various ANC leaders fought to control it while it appeared to be succeeding and denied any link with it when it failed. In the process, two ideas which could have contributed towards bridging the gap between a largely externally-based MK and internal militants, and between political and military forms of struggle inside South Africa, were squandered. Decisions in 1980 and 1982 to smuggle large quantities of small arms and explosives into South Africa in case there was a new round of uprisings had not been implemented by the time new uprisings did break out in 1984. The lasting impression is of an ANC which is eloquent in its reasoning and its resolutions but hidebound or incompetent in implementation.
As ANC strategic debates sought the success of armed struggle so they treated other forms of struggle as subordinate to the imperatives of armed struggle. Even those ANC strategists who argued so hard for stepping up political organisation inside South Africa - as in the strategic review of 1978-1979, in the Area Political Committee (APC) debate of 1981-1982, in the planning of Operation Butterfly and at the Kabwe consultative conference in 1985 - intended that these political improvements should ultimately facilitate the development of popular armed struggle. When ANC strategists spoke of `politics', they usually meant `armed struggle'.
A variety of factors explain why armed struggle became, for the ANC, a kind of Sisyphian task to which all else had to be subject. In the first instance, the brutal armed violence of the South African state in 1960-1961, and particularly in 1976, not only appeared to call for an armed response, it also stimulated in successive generations a powerful desire to strike back. The ANC's first attempt at striking back, between 1961 and 1963, had, however, been disastrous and brought it to the brink of extinction. This had a profound effect on the ANC's attitude towards armed struggle. It did not, as might have been expected, persuade the ANC to seek a different strategic way forward. Rather it induced in the ANC a determination to succeed with armed struggle. In ANC thinking, the costly thesis of armed struggle, particularly the disaster it had led to in the 1960s, could be justified and rationalised only if the ANC turned the military tables on the South African government: only revenge, no matter how costly, could depict the early disaster as antithesis in a process producing the desired and prophesied synthesis, revolutionary victory in South Africa. Hence, from the early 1960s, ANC self-justification and pride, indeed identity, came to depend to a significant degree upon achieving evident success with armed struggle.
The ANC's strategic discourse also encouraged and justified this stress on military struggle. The Marxist-Leninist theoretical heritage on which it relied conceived of the state as founded upon violence and removeable ultimately only through violence. Strategies of revolutionary warfare derived from this heritage and upon which the ANC drew - from Lenin to Mao, Giap to Guevara - provided models for revolutionary struggle in which political organisation was ascribed the role merely of helping to redress the military a-symmetry between revolutionary forces and the state.
And finally, the special character of armed struggle created demands which made its domination of strategy highly likely. Unlike non-violent aspects of political mobilisation, armed preparations had to be conducted entirely clandestinely. More so than other forms of struggle, the costs of armed struggle in personnel and material were bound to be high. These costs of armed struggle demanded disciplines of those involved in it. But these demands could not end with those directly involved: cadres pursuing other forms of struggle had also to ensure that the clandestinity of armed struggle was maintained as best it could be and that its success was guaranteed so that its costs could be kept to a minimum.
These costs were relatively easily calculated: the number of bombs that had been detonated or attacks mounted could usually be ascertained. But measuring the success or otherwise of political mobilisation or, say, the extent of ANC influence in the emerging trade union movement was considerably more difficult. This arithmetic, or mechanical, quality of armed struggle made its successes and failures the main criteria in the ANC for assessing the progress of the entire liberation struggle. Likewise, ensuring the success of armed struggle became synonomous with ensuring the success of the entire liberation struggle.
Yet, by the mid-1980s, the evidence strongly suggested that not only was armed struggle a failure at a military level but that popular political mobilisation, including the activities of the militant trade unions, provided a more serious challenge to the state than the ANC's military campaign. Much of this political mobilisation and organisation had been initiated or conducted autonomously of the ANC. Its origins most often lay in a growing realisation among sectors of the black population that the legal and political `space' existed for popular organisation and that, repression notwithstanding, open organisation could deliver improvements to their material situations. Popular organisation was seldom, if ever, stimulated by revolutionary armed activity conducted by the ANC or anybody else.
This was an irony for the ANC. For, through much of the 1976-1986 period, it had behaved as if revolutionary armed activity not only would stimulate political mobilisation and organisation but was a precondition for those two processes. Moreover, to the extent that it had contributed to domestic political mobilisation by non-violent political means it had done so in order to enhance the prospects for armed struggle. Yet popular armed struggle was not the result. Rather, non-violent political struggle remained uncoupled from military struggle and its evident efficacy (compared to the armed campaign's drama but minimal achievement) undermined the prospects for a sustained and serious popular armed struggle.
The rise in importance of political mobilisation occurred in a context in which the state had embarked upon a body of reforms. These reforms, although initially motivated largely by counter-insurgency considerations, had nonetheless created political and legal space from which political and union organisation could, and did, wring further concessions from the government. Why, then, did the ANC persist with its military struggle? There appear to have been several reasons.
Those who made operational strategy in the ANC were not only, or even mainly, looking forward when they did so. They were not solely concerned with formulating their intentions. They were also looking backward. They needed to make sense of, and to justify, the past 20-odd years of commitment to armed struggle, and its failure. To denounce armed struggle as a failure would have meant rupturing a sort of umbilical cord to an important element of the ANC's identity.
A second reason was that the failure of the armed struggle was never so complete as to exhaust the ANC's capacity for rationalisation. Circumstances always allowed MK a margin for drama, although real military achievement remained highly improbable. The ANC managed to explain a number of key changes to operational strategy, which had been necessitated by failure, as innovations or changes demanded by progress or principle. When the ANC took its `turn to the masses' in the strategic review of 1979, it made a virtue out of a necessity. Over the three previous years, its armed activity had failed to take root inside South Africa or to create a revolutionary political base under the ANC's aegis. What popular political organisations and trade unions had developed inside South Africa had done so mainly independently of the ANC or of the influence of its armed activity. The ANC faced a choice - of irrelevance or seeking to coopt autonomous organisations in order eventually to base its armed struggle upon them. It chose to do the latter.
When, at around the same time, the ANC redefined its armed activity, as armed propaganda, it was as much an acknowledgement and rationalisationn of its military limitations, as it was a new strategic departure. The redefinition was also somewhat disingenuous when applied to armed activity in the 1976-1979 period, which had been intended as the opening phase of direct action and a people's war against the state.
Having abandoned the Guevarist detonator approach to armed activity in 1979 because of its failure, the ANC looked to Giap - to a doctrine of protracted people's war interspersed by spontaneous popular uprising - as the framework within which to develop its armed struggle. The new approach offered a perspective within which the ANC's military inadequacies could, quite respectably, be compensated by the actions of ordinary people at large as they confronted the state. But this approach, too, had by 1985 failed to deliver a sustained armed struggle.
The more innovative of the ANC's strategists then began suggesting that ANC strategy henceforth depend, in effect, entirely upon ordinary people's energies. The ANC should abandon any thoughts of leading a gradual build-up of forces as entailed in the people's war approach. Rather, it should hold out the perspective of decisive popular insurrection against the state, a sort of quick fix of one or another degree of spontaneity. Whereas some in the ANC (like Slovo) maintained the ANC could not hope even to provide significant leadership to such an insurrection, which would be entirely spontaneous, others (like Maharaj and Kasrils) argued it could and must be led in ways suggested by the MCW approach.
Operation Vula was intended at one level to try to assert some ANC leadership over the autonomous forces identified as the engine of insurrection. It was to do so by improving operational management. Building an underground from top down was to achieve this. Vula's authors considered that many among the ANC's most important operational leadership had to be deceived to achieve this. They feared, perhaps justifiably, that, if these officials knew of the operation, they would have to be involved in it and that they might undermine it by inertia if not by active sabotage.
One way of seeing the strategic phases between Guevara and Vula is that each represented an attempt to disguise and to dignify the failure of the preceding strategic framework for the ANC's armed struggle.
The most compelling reason for the ANC's persistence with armed activity regardless of its cost was the prestige, influence and popularity that armed struggle gave it among potential supporters both inside South Africa and abroad. This influence, combined with the ANC's inclusive nationalism and attempts to unite the widest possible spectrum of anti-apartheid forces under its aegis, gave the ANC unrivalled cache. The ANC's rivals - such as the PAC and black consciousness movement - did not have armed struggles worth the name and, largely for this reason, could not achieve the same profile. The political dividend the ANC derived from armed struggle made the resources spent on it and lives lost in it worthwhile in the ANC's calculations. Indeed, the heavier the ANC's losses, the higher that return often was: the more MK cadres being buried or marching off to jail, the greater the evidence that the ANC had dared to struggle against a brutal, powerful and internationally infamous enemy. Therein lies the explanation for the paradox of the ANC's trajectory: how it found its success in failure.