The Khosis Community at Lohatla and Others v The Minister of Defence and Others

Three communities (Khosis, Gatlhose and Maremane) claimed restoration of land rights over approximately 62,000 ha in the north-western portion of the Lohatla Army Battle School in the Northern Cape. The land comprised former native reserves historically occupied by the Batlharo tribe and people of Griqua extraction. The reserves were excised from the Native Trust in 1969 and transferred to the state. In 1976-1977, approximately 12,000 'black' community members were forcibly removed to Bophuthatswana, while 'coloured' members remained. In 1978, the area was reserved for military purposes as a battle school. The Khosis area (8,500-9,400 ha) in the middle of the battle school remained occupied by the Free group (127 persons). In 1990, the majority of the Khosis community agreed to relocate to Jenn Haven (14,000 ha) where infrastructure was provided at state cost, but the Free group refused to move. The Minister of Defence and the Premier of the Northern Cape applied under section 34 of the Restitution of Land Rights Act for a declaration that no part of the battle school would be restored to any claimant before final determination of the land claims. The Gatlhose and Maremane communities indicated willingness to accept monetary compensation for alternative land, but the Free group insisted on restoration of at least the Khosis area.

• Whether section 34(6) of the Restitution of Land Rights Act 22 of 1994 confers a discretion to refuse an order even if jurisdictional requirements are met
• Whether it is in the public interest that the land should not be restored to any claimant (section 34(6)(a))
• Whether the public or any substantial part thereof will suffer substantial prejudice unless an order is made before final determination of the claim (section 34(6)(b))
• What factors under section 33 of the Act must be considered when determining whether restoration is in the public interest
• Whether the entire battle school area is required for military training purposes or whether the Khosis area could be excised

Appeal dismissed. The order of the Land Claims Court was confirmed, declaring that when any claim in respect of the Lohatla Battle School land is finally determined, no part of the land shall be restored to any claimant. No order as to costs.

The binding legal principles established are: (1) Section 34(5) of the Restitution of Land Rights Act does not confer an overriding discretion to refuse an order once the jurisdictional requirements of section 34(6) are met - the word 'may' performs a predicative function defining possible orders depending on findings. (2) The two-stage jurisdictional test under section 34(6) requires: (a) it must be in the public interest that rights should not be restored; and (b) the public or substantial part thereof will suffer substantial prejudice unless the order is made before final determination. (3) Both requirements involve value judgments (wide discretion) based on proven facts, not narrow discretion. (4) In determining public interest under section 34(6)(a), courts must consider the section 33 factors including: desirability of remedying past violations; feasibility of restoration; current use of land; history of dispossession; hardship caused; social upheaval. (5) The general approach is that dispossessed communities are entitled to restoration unless restoration is trumped by public interest considerations. (6) There is no substantive right to any particular form of restitution - claimants have a right to 'claim' restitution but the specific form (restoration, alternative land, compensation) only crystallizes when a court makes a restitution order. (7) Public interest may include: national security requirements; economic impact on provinces and communities; safety of claimants; ability to provide constitutional services (education, health) to claimants; welfare of other affected communities awaiting settlement; finite nature of land resources and competing claims. (8) An appellate court must show deference to value judgments of the specialist Land Claims Court and may only interfere if the lower court was wrong, not merely because there is scope for disagreement.

The court made several non-binding observations: (1) Cost considerations (including saving of legal costs to the state) can 'at best be a makeweight' in determining whether to grant a section 34 order. (2) The statement that 'justice delayed is justice denied' has special application to displaced communities living on insufficient land awaiting compensation. (3) A decision by community leaders on behalf of children that their health, education and lives are of lesser importance than a land claim is 'not only paternalistic; it is reckless and wholly unacceptable.' (4) The Population Registration Act 30 of 1950 'artificially divided' communities for racial classification purposes, and apartheid legislation had 'creeping impact' causing increasing racial division. (5) 'One death is one too many' when considering civilian deaths from unexploded ordnance. (6) The terms of the order (declaring no restoration of land) do not exclude consideration of other rights such as access to ancestral graves or erection of memorials, which may be considered in determining equitable redress. (7) The option of recognizing bare dominium in claimants while leasing land indefinitely to the SANDF was mooted at a meeting with President Mandela but was not pursued and its viability was not established. (8) The Act does not prescribe that all section 33 factors are necessarily applicable in every case. (9) Customary law interests contemplated in the definition of 'rights' (such as servitudes, water and mineral extraction, ploughing, grazing, gathering wood and soil) would be inconsistent with the SANDF's required exclusive occupation and were not realistically viable in a battle school environment.

This case is significant for establishing the interpretive framework for section 34 of the Restitution of Land Rights Act, which permits advance rulings that land will not be restored before final determination of claims. It clarifies that: (1) Section 34 creates a mechanism to break deadlocks in land restitution where public interest considerations are paramount; (2) Courts must balance the strong connection between dispossessed communities and ancestral land against other public interest factors including national security, economic considerations, and the welfare of claimants themselves; (3) The public interest threshold is not limited to the interests of government but extends to prejudice suffered by claimant communities (such as inability to access services) and other affected communities; (4) Restoration of land is not an absolute right - claimants have a right to 'claim' restitution but the form of restitution (restoration, alternative land, or compensation) depends on all relevant factors including feasibility and public interest; (5) Courts may consider the practical realities of occupation including safety, access to services, and dignity when assessing prejudice. The case also demonstrates the hierarchical approach to land restitution: dispossessed communities are presumptively entitled to restoration unless trumped by public interest considerations. It affirms the role of the Land Claims Court as a specialist tribunal whose value judgments warrant deference on appeal.