Dr. Gift Mandlenkosi Dlamini is a Swazi political leader, pro-democracy advocate, governance practitioner, and development project management professional whose public work focuses on constitutionalism, human rights, accountable leadership, and democratic transformation in Eswatini.
Executive summary
The proposed transfer of KaNgwane and the Ingwavuma district to Swaziland was not, in strict legal terms, a “return” of land that the Swazi state still owned in 1982. It was a proposed cession of South African territory by the apartheid state, justified publicly as the “reunification” of the Swazi under one king and supported by older Swazi historical claims, but framed in legal instruments issued by the South African state. The offer was already under diplomatic discussion by early March 1982, became publicly visible in June 1982 through proclamations and political statements, and would have transferred KaNgwane plus Ingwavuma, with the added strategic value of a route to the sea through Kosi Bay.
The transfer failed because legal, political, social, and strategic pressures converged at once. On the legal side, the South African courts held that Pretoria had acted beyond its powers in taking Ingwavuma from KwaZulu without the consultation required by the homeland constitutional framework, and the KaNgwane proclamation could not survive once a parallel challenge exposed the same defect. Politically, the two most important black homeland leaders in the dispute, Enos Mabuza of KaNgwane and Mangosuthu Buthelezi of KwaZulu, both resisted the deal, even though they disagreed on many other issues. Socially, many residents feared loss of South African citizenship, jobs, mobility, and income. Strategically, Pretoria appears to have realized by 1984 that it could obtain much of what it wanted from Swaziland, especially security cooperation and pressure on the ANC, without actually paying the political price of handing over territory.
The most important analytical conclusion is that the deal failed not because the claim lacked all historical resonance, but because historical memory could not overcome modern realities. By the early 1980s, KaNgwane’s people were economically integrated into South Africa, Ingwavuma’s politics were tied to KwaZulu and Natal, and both territories had active leaderships capable of mobilizing courts and public opinion. Swaziland itself had internal doubts about the social, economic, and diplomatic costs of absorbing a very large population. After the deal was shelved in June 1984, KaNgwane remained in South Africa, later gained self-government, and both territories were folded into democratic South Africa in 1994. Since then, the legacy has survived mostly in land restitution files, archival memory, and periodic border-restoration rhetoric from Eswatini rather than in any formal transfer process.
Historical background and the nature of the Swazi claim
The roots of the dispute lay in the nineteenth-century expansion of the Swazi polity and in later colonial border-making. Hugh Macmillan’s reconstruction shows that Swazi ruling elites in the twentieth century repeatedly revisited claims to lands north and south of the modern kingdom, especially after the hardening of colonial borders. The specific historical memory behind the 1982-84 proposal combined two strands: claims to Swazi-speaking areas in the eastern Transvaal that later became KaNgwane, and claims to Tongaland/Ingwavuma on the coast, which had once mattered because of the possibility of an outlet to the sea. Macmillan also notes that Pretoria had floated the idea of “ethnic” border adjustments much earlier: Hendrik Verwoerd spoke in 1959 of possible incorporation of Swazi areas if Swaziland accepted South African “guardianship,” and a South African minister repeated in 1966 that territory could be ceded to neighbors to create ethnic units.
Ingwavuma was historically important because of its relation to old Tongaland and the strategic attraction of Kosi Bay. Wolfgang Senftleben’s 1984 study explains that Ingwavuma lay in the far north of Natal/KwaZulu, bordered by Mozambique and Swaziland, and contained the potential port site at Kosi Bay. He also reconstructs the earlier colonial moment: Britain annexed Tongaland in 1895, ending earlier Transvaal dreams of using Swaziland as a corridor to the sea. That earlier imperial history gave twentieth-century Swazi claims a real emotional and historical vocabulary, but not an uncontested modern legal title.
KaNgwane, meanwhile, was a constructed apartheid homeland. Macmillan notes that in 1980, there were 786,835 people classified as “Swazi” in South Africa, about one-third of whom lived in KaNgwane. The territory consisted of two separate blocks in the eastern Transvaal; its de facto population had grown sharply because forced removals pushed people from “white” South Africa into it. About 80 percent of the population was classified as Swazi, while the rest included Tsonga/Shangaan, Zulu, and Sotho people. That demographic complexity mattered politically because Swaziland’s historical claim assumed a simpler ethnic map than actually existed on the ground.
Just as important, KaNgwane and Swaziland had diverged socio-economically. Macmillan records that in 1981, roughly 57,000 labor migrants and 40,000 commuters from KaNgwane worked in white South Africa, while Swaziland itself exported far fewer workers and retained a more viable internal economy. This divergence helps explain why “reunification” rhetoric failed to persuade many people in KaNgwane: for the majority, the core issue was not historical symbolism but livelihood, citizenship, and access to South African labor markets.
This is the point at which the common phrase “Swazi-owned land” becomes analytically misleading. In the sources closest to the dispute, the claim is overwhelmingly described as historical, ancestral, or political. Contemporary descriptions of Ingwavuma refer to “Bantu reserve” land and “unalienated State land,” while later land records show former KaNgwane homeland endorsements, provincial title, and private ownership on different parcels. The 1982-84 issue was therefore less a matter of restoring registered Swazi title than of deciding whether the South African state would cede territory to advance apartheid’s ethnic map.
The offer and the documentary trail
The surviving record shows that the South African “offer” was underway before it surfaced publicly in June 1982. At a press conference in Maputo on 5 March 1982, Oliver Tambo said the African National Congress had already held two rounds of talks with Swazi authorities about the South African offer of “the KaNgwane Bantustan and the Ingwavuma strip,” and that the ANC’s concern included the threatened denationalization of roughly one million black South Africans. A separate contemporary “Swazi View” account reports that King Sobhuza II had addressed the nation on the issue on 19 March 1982 at Lozithehlezi Palace, after which Richard Velaphi Dlamini insisted that only official statements should be made about border adjustment.
Publicly, the deal broke into view in mid-June 1982. A contemporary Washington Post report dated 14 June 1982 said the South African government had agreed to give Swaziland “two chunks of territory” totaling about 1,800 square miles. On 17 June, the official South African chronology recorded Minister of Cooperation and Development Piet Koornhof defending the contemplated border adjustment as a step toward the long-cherished ideal of uniting the Swazi under one king in one country. A contemporary analytical article also reports that P.W. Botha told a press conference on the same day that the areas were not being “given away,” because they would supposedly go to the people living there. The core documentary trail can be reconstructed as follows:
Date
Document or event
Principal actors
Why it mattered
5 Mar. 1982
Maputo press conference on the South African “offer.”
Oliver Tambo; Mabandla Dlamini
Proves the offer existed before the June proclamations and was already the subject of regional diplomacy.
19 Mar. 1982
Sobhuza’s official announcement in Swaziland; media restrictions follow
King Sobhuza II; Richard Velaphi Dlamini
Shows Mbabane had formally entered the process by March 1982.
18 Jun. 1982
Proclamation R108 dissolves the KaNgwane Legislative Assembly; the territory is placed under direct South African administration
South African state; KaNgwane authorities
The first decisive legal step to clear KaNgwane for cession. Later scholarship identifies the Gazette as No. 8269.
18 Jun. 1982
Proclamation R109/82 excises Ingwavuma from KwaZulu
South African state; Government of KwaZulu
The KwaZulu arm of the cession strategy.
25–28 Jun. 1982
Durban court blocks the first Ingwavuma move; Pretoria responds with Proclamation R121
Government of KwaZulu, a South African state
Shows Pretoria immediately trying to route around consultation requirements.
30 Sep. 1982
Appellate Division rules Ingwavuma proclamation ultra vires
Government of KwaZulu, a South African state
The decisive legal defeat on Ingwavuma.
24–25 Nov. 1982
Proclamation R108 repealed by Proclamation R232; out-of-court settlement restores KaNgwane institutions
Government of KaNgwane; Enos Mabuza; South African state
Pretoria retreats on KaNgwane after parallel legal pressure.
6 May 1983
The Gazette establishes the Rumpff Commission
South African state; former Chief Justice Frans Rumpff
Moves the issue from unilateral proclamation to a formal inquiry into “desirability” and the wishes of affected Swazi people.
31 Mar. 1984
The 1982 security pact between South Africa and Swaziland is publicly disclosed
South African and Swazi foreign ministers
Strengthens the argument that the land deal was tied to security concessions.
19 Jun. 1984
Koornhof dissolves the commission and says Pretoria will not proceed unilaterally
Piet Koornhof, South African state
This was the practical burial of the deal.
1959 Verwoerd signals that Swazi be consolidated with Swaziland 1966 South African minister again says territory might be ceded to create ethnic units 1977.
KaNgwane receives a constitution and legislative assembly1981.
KaNgwane executive and assembly formally oppose incorporation 1982-03-05.
Oliver Tambo confirms talks on Pretoria’s “offer” 1982-03-19 Sobhuza informs Swaziland publicly of the land-deal issue1982-06-18 Proclamations R108 and R109 published1982-06-28.
Pretoria issues Proclamation R121after KwaZulu’s first court win1982-09-30Appellate Division rules Ingwavuma move ultra vires1982-11-24R108 is repealed and KaNgwane authority restored 1983-05-06.
Rumpff Commission is formally established 1984-03-31 Secret 1982 South Africa-Swaziland security pact is disclosed 1984-06-19.
Koornh of dissolves the commission and shelves unilateraltransfer1984-08-31KaNgwane is grantedself-government1994-04-27.
Territories are integrated into democratic South Africa 2025-05-19 Eswatini appoints anew Border Restoration Committee Timeline of the KaNgwane–Ingwavuma transfer attempt.
The timeline above is drawn from South African official chronologies, later legal reconstructions, and contemporary archival reporting.
Actors, alignments, and opposition
The single most important opponent inside KaNgwane was Enos Mabuza. Macmillan shows that Mabuza had long rejected both “independence” for KaNgwane and any deal with Swaziland, later calling KaNgwane “a region of South Africa.” Senftleben quotes him as rejecting the “dubious benefits of Swazi citizenship” and declaring that KaNgwane had no wish to become part of “the medieval monarchy that rules by decree.” Another contemporary report quotes him warning that “KaNgwane is a powder keg” that could explode if Pretoria pushed ahead. His position was not anti-Swazi in a cultural sense; it was anti-cession, anti-denationalization, and ultimately South Africanist in political terms.
KaNgwane’s institutional opposition was broad, not purely personal. The Lawyers’ Committee memorandum to the Rumpff Commission records that the KaNgwane Executive Council told the South African minister as early as 12 June 1981 that unification with Swaziland was unacceptable; the Legislative Assembly passed a formal anti-incorporation resolution on 1 December 1981; 17 of 21 chiefs signed a petition against incorporation in June 1982; and KaNgwane requested a referendum, which Pretoria denied. Senftleben also records that thousands of members of Mabuza’s Inyandza movement pledged resistance at a meeting in KaNgwane’s capital.
KwaZulu’s opposition under Mangosuthu Buthelezi was equally important because Ingwavuma could not be transferred without first being taken from KwaZulu. Contemporary South African chronology records that the Durban Supreme Court cancelled the state’s June 1982 Ingwavuma move because Pretoria had not adequately consulted KwaZulu, and later that the Appellate Division ruled the presidential action ultra vires. Contemporary reporting also shows Buthelezi aligning tactically with Mabuza, warning of violence if Pretoria tried to steamroll the issue, organizing protest meetings, and pressing the case beyond South African courts into African diplomatic forums. The Natal Provincial Council, angry at not being consulted, called for a referendum, and contemporaneous legal advocacy reported more than 18,000 petition signatures by October 1982.
Swaziland’s official line was that the land was historically Swazi and should be reunited with the kingdom. Richard Velaphi Dlamini tightly controlled public discussion in Swaziland; the authorized spokesmen mentioned in the contemporary “Swazi View” piece were Dr. Sishayi Nxumalo and Dr. George Msibi, both of whom emphasized historical justification, with Msibi tentatively pointing to the sea-outlet advantage of Kosi Bay. The same source, however, is revealing because it documents internal doubt: businessmen worried about the economic burden of adding 800,000 people, students saw the proposal as a South African buffer-state strategy, and a near-monopoly of official messaging was used to suppress open debate. Macmillan further argues that even within Swaziland’s leadership, not everyone supported the bargain; he specifically notes doubts from King Sobhuza’s last prime minister, Prince Mabandla.
Not all Swazi elites in South Africa opposed the deal. Macmillan identifies Chief Mkolishi Dlamini as the most visible pro-incorporation figure among South African Swazi chiefs. As early as 1978, he and ten chiefs had petitioned Sobhuza to negotiate incorporation into Swaziland. Macmillan also records that David Lukhele, who broke with Mabuza in 1981, joined Mkolishi in supporting the deal in 1982 and argued that many of the opponents were not really Swazi. A 1984 strategic note adds that a rival power center inside KaNgwane included Johannes Dlamini, a former homeland chief minister and member of the Swazi royal family. This matters analytically because the failure of the cession cannot be explained by a simple “Swazis versus Zulus” or “Swazis versus South Africans” formula; it was also a struggle within Swazi politics itself.
Broader public responses broadened the anti-cession coalition. White farmers in the eastern Transvaal feared that border changes would isolate the Komatipoort farming corridor, and Senftleben records that they held protest meetings. The contemporary Swazi report also describes how Swaziland’s government used refugees from Ingwavuma, estimated at about 12,000 to 13,000 by the UNHCR representative in Swaziland, as evidence of KwaZulu oppression and loyalty to Sobhuza. The ANC’s public messaging in 1982 was at moments confused, but the archival record shows that its leadership opposed the denationalization dimension of the plan and tried to prevent the dispute from turning into conflict between anti-apartheid forces and Swaziland.
Legal and constitutional barriers
The legal architecture of apartheid unexpectedly became one of the strongest obstacles to the cession. In the Ingwavuma litigation, the question was whether the South African state president could, by proclamation, remove territory from KwaZulu without the consultation contemplated by the homeland constitutional framework. Official South African chronology records first that the Durban Supreme Court blocked Pretoria’s initial step on 25 June 1982 for inadequate consultation, and later that on 30 September 1982 the Appellate Division in Bloemfontein ruled that the proclamation restoring Ingwavuma to direct South African jurisdiction was null and void because the state president had acted ultra vires.
Later legal summaries describe the Appellate Division’s reasoning more specifically: powers under the Black Administration Act could not be used in a way inconsistent with the National States Constitution Act and the consultative scheme governing self-governing territories.
The KaNgwane side was legally similar. The ICJ Review’s 1984 summary states that KaNgwane had been granted a constitution and legislative assembly in 1977, that its leadership strongly resisted both independence and cession, and that Pretoria in 1982 used Proclamation R108 to dissolve the constitution and assembly.
The same summary adds that KaNgwane and Enos Mabuza challenged the step in the Transvaal Provincial Division, arguing that the 1971 National States Constitution Act limited the state president’s powers and required prior consultation with the legislative assembly before its legislative powers could be repealed. The Lawyers’ Committee memorandum gives the case number, Government of KaNgwane and Enos John Mabuza v. Government of South Africa, Case No. 9990/82, and explains that Pretoria then withdrew the proclamation and restored KaNgwane’s institutions rather than risk an adverse judgment.
Citizenship law made the issue more explosive. Under apartheid’s homeland citizenship regime, black South Africans were assigned to ethnic homelands even if they lived and worked elsewhere in the Republic. The Lawyers’ Committee memorandum, drawing partly on John Dugard’s analysis, argued that ceding KaNgwane would denationalize or at minimum radically downgrade the status of around 800,000 Swazi-speaking South Africans, residents inside KaNgwane. The CSIS strategic note written after the deal collapsed reached a similar conclusion in practical terms: a KaNgwane transfer would have allowed Pretoria to divest hundreds of thousands of black South Africans of meaningful South African citizenship with a degree of formal legality unavailable through another homeland “independence” fiction.
Pretoria’s fallback was the Rumpff Commission, formally established in Government Gazette Vol. 215 No. 8693 of 6 May 1983. The terms of reference, as reproduced in the Lawyers’ Committee memorandum, directed the commission to investigate the desirability of incorporating KaNgwane and associated areas into Swaziland, while taking account of the wishes and best interests of the affected Swazi people. That formula was important for two reasons. First, it implicitly conceded that unilateral proclamation had become legally untenable. Second, it moved the dispute onto the terrain of consent and political desirability, precisely where Pretoria was weakest. The commission, however, never produced a public report before the issue was shelved.
A separate legal critique came from international law advocates rather than courts. The Lawyers’ Committee’s memorandum argued that any cession without consent would violate self-determination and racial non-discrimination, and that denationalization without consent would likewise be illegal. That memorandum did not become binding law in South Africa, but it mattered because it translated local opposition into an international legal language that resonated with anti-apartheid diplomacy.
Why the transfer failed
The first reason the transfer failed was straightforward: Pretoria’s chosen legal method broke its own rules. The South African state tried to use executive proclamations to do something that homeland legislation and institutional arrangements made politically and procedurally difficult. Once KwaZulu won on consultation and ultra vires grounds, KaNgwane’s legal position became much stronger. Pretoria could still, in theory, have gone back to Parliament and legislated more explicitly, but doing so would have escalated a politically costly confrontation.
The second reason was the emergence of an unusually broad anti-cession coalition. Mabuza and Buthelezi were crucial because they represented the two territories Pretoria needed to reconfigure, but they were reinforced by chiefs’ petitions, local protest meetings, white farmers in the east Transvaal, sections of Natal’s provincial establishment, and the ANC. In other words, the transfer failed because Pretoria could not isolate opposition inside a narrow ethnic constituency; resistance appeared in KaNgwane, KwaZulu, Natal, white farming districts, and regional diplomacy all at once.
The third reason was social and economic implausibility. Macmillan shows that KaNgwane’s population was deeply dependent on labor migration and commuting into white South Africa. Senftleben states that the large majority of KaNgwane’s people rejected the transfer because South African citizenship, however degraded under apartheid, was still more useful than Swazi citizenship for employment and mobility. The “Swazi View” source shows the same problem from inside Swaziland: critics feared the addition of 800,000 people would transform the country overnight and lower living standards unless accompanied by some substantial “dowry.” Put differently, the historical claim was emotionally intelligible, but the proposed merger was economically unattractive to many of the very people meant to benefit from it.
The fourth reason was Pretoria’s strategic recalculation. Macmillan writes that by 1984 it had become clear that the land deal had been offered as an inducement for Swaziland to sign a secret non-aggression pact and clamp down on ANC activities; he adds that the February 1982 agreement, publicly disclosed only in March 1984, was a precursor to the Nkomati Accord with Mozambique. The CSIS note written in mid-1984 reaches a similar conclusion in more strategic language: Pretoria had already used the possibility of territorial adjustment to secure a peace treaty and a Swazi crackdown on the ANC, and now did not need to surrender land. This is one of the strongest explanations for why Koornhof could bury the plan in June 1984.
The fifth reason was instability within Swaziland itself after Sobhuza’s death in August 1982. Senftleben notes that the succession struggle and wider political uncertainty in Swaziland reduced the likelihood of carrying the transfer through. Macmillan adds that disagreement over the deal was one factor in the post-Sobhuza power struggle, and that many Swazi leaders, including Prince Mabandla, had doubts about the kingdom’s probable loss of international status and the danger of becoming South Africa’s policeman against the ANC. That internal hesitation reduced Swaziland’s capacity to keep pressing the issue once resistance mounted in South Africa and Pretoria’s strategic need declined.
The sixth reason was international legitimacy. Senftleben reports strong opposition from the Organisation of African Unity and notes that Buthelezi and the ANC appealed to African heads of state to oppose the deal; he also records that Libya supported Swaziland’s claim while U.S. officials largely treated the dispute cautiously as a bilateral matter. This mixed external environment mattered because Swaziland could not afford to appear openly aligned with apartheid at the exact moment when its government was already drifting toward South African security cooperation. The deal therefore became more expensive diplomatically for Mbabane and less necessary strategically for Pretoria.
Land tenure, aftermath, and present-day legacy
The land-tenure record confirms that “return” was an ideological description, not a clean legal one. Contemporary geographic description says that Ingwavuma consisted partly of reserve land and partly of unalienated state land. Contemporary activist summaries describe KaNgwane not as property of the Kingdom of Swaziland but as land earmarked under apartheid as the homeland for South Africa’s Swazi population. Post-apartheid restitution gazettes then show a more granular legal picture: numerous farms in present-day Mpumalanga carry title endorsements “in favour of Kangwane Government,” others are vested in the Provincial Government of Mpumalanga or the Republic, and still others sit in private or community-trust hands. The legal material therefore points to a patchwork of reserve, state, homeland, communal, and private tenure rather than a single recoverable block of “Swazi-owned” land.
After the cession project failed, Pretoria effectively reverted to the homeland track. South African chronologies record that KaNgwane was later granted self-government in 1984, a move widely read at the time as confirmation that the Swaziland land deal had been abandoned. Neither KaNgwane nor Ingwavuma was transferred. In 1994 both remained in South Africa and were absorbed into the new constitutional order: KaNgwane into Mpumalanga, and Ingwavuma through the KwaZulu/KwaZulu-Natal line into modern KwaZulu-Natal.
Since 1994, the legacy has shifted from interstate cession to domestic land restitution and symbolic interstate claims. In Ingwavuma and the Kosi Bay/Maputaland region, post-apartheid disputes have largely proceeded through South Africa’s restitution machinery, conservation settlements, and municipal planning rather than through any bilateral territorial process. Government notices continue to record restitution claims in Ingwavuma, including state land and named farm parcels. In former KaNgwane areas, gazetted claims likewise track individual farms and endorsements linked to the old KaNgwane government.
Recent developments show that the historical claim has not disappeared rhetorically. In May 2025, Eswatini’s attorney general publicly announced new appointments to a Border Restoration Committee, chaired by Chief Mgebiseni Dlamini, and local Eswatini reporting described the body as tasked with pursuing land taken during colonial times. Yet the publicly visible South Africa-Eswatini bilateral agenda in 2025 was centered on border management cooperation, customs, and movement control, not territorial revision. That contrast suggests continuity of symbolic claims on the Eswatini side, but no evidence of an active state-to-state transfer process comparable to 1982-84.
Several points remain uncertain and should be treated carefully. The exact cartographic scope of the proposal appears to have shifted: early press accounts spoke of roughly 1,800 square miles, later scholarship referred to about 10,000 square kilometers, and one contemporary activist paper says some early variants may have included game reserves and the Jozini Dam area before those elements were dropped.
Direct digital access to the original 1982 government gazettes is still patchy, so some Proclamation references are verified through legal scholarship, archive reproductions, and memorandum quotations rather than an easily retrievable official PDF of each gazette. Finally, although the public record is rich on Enos Mabuza himself, I did not locate separately attributable contemporaneous public statements by other members of the Mabuza family; later archival work instead points to the family’s importance in preserving his papers and memory. Digitized primary or near-primary records that are currently accessible online include the March 1983 conference paper on the Ingwavuma/KaNgwane land deal, the 1982 essay “KaNgwane and Ingwavuma: The Swazi View,” the Lawyers’ Committee memorandum to the Rumpff Commission, the 1984 ICJ Review summary, and Buthelezi’s June 1984 statement linking the deal to security bargaining.
The most defensible bottom-line judgment, therefore, is this: the land was not transferred because the proposal was an apartheid-era cession project that ran into a rare combination of judicial defeat, homeland resistance, local economic realism, Swazi elite division, and changing regional strategy. Historical Swazi claims gave the project a language; they did not give it enough law, legitimacy, or consent to survive.
