Practice Relating to Rule 37. Open Towns and Non-Defended Localities
Section B. Establishment of non-defended localities
South Africa’s LOAC Teaching Manual (2008) states:
2.6 Localities and Zones under Special Protection – Demilitarised Zones, Non-defended Localities, and Demilitarised Zones
Non-Defended Localities (Article 59 [of the 1977 Additional Protocol I])
“Non-defended localities” are improvised protected inhabited areas from which military objectives and military activities have been removed, and which:
- Are situated near or in a zone where armed forces are in contact;
- Are open for occupation by the enemy.
Parties to the conflict may declare any inhabited place near or in a zone where armed forces are in contact and which is open for occupation by the enemy as a non-defended locality. Non-defended localities differ from demilitarised zones in that non-defended localities can be established through a unilateral declaration notified to the enemy, while demilitarised zones require an agreement between the Parties to the conflict. However, it is advisable to always make use of formal agreements, also for non-defended localities.
The following conditions must be met regarding such a locality:
- All combatants, as well as mobile weapons and mobile military equipment must be evacuated.
- Fixed military installations or establishments may not be used for hostile purposes.
- No acts of hostility may be committed by the authorities or by the population.
- No activities in support of military operations may be undertaken. However, the presence of specially protected persons and police forces retained for the sole purpose of maintaining law and order in this zone is not contrary to this condition.
The declaration of an area into a non-defended locality must be addressed to the enemy and shall define and describe as precisely as possible the limits of the non-defended locality. The Party to the conflict to which the declaration is addressed must acknowledge receipt thereof and must treat the locality as a non-defended locality unless the conditions, as described in the previous paragraph, have not been de facto fulfilled, in which event it shall immediately inform the Party making the declaration thereof. However, even if these conditions are not fulfilled, the locality shall continue to enjoy the protection provided by the other provisions of the LOAC [law of armed conflict].
Parties to the conflict may agree to establish non-defended localities even if such localities do not fulfill the abovementioned conditions. Such an agreement should define and describe as precisely as possible the limits of the non-defended locality and if necessary it may lay down the methods of supervision.
The Party, which is in control of the non-defended locality governed by such an agreement, must clearly mark it by such signs as may be agreed upon with the Parties involved.
A non-defended locality shall lose its status when it ceases to fulfil the abovementioned conditions. However, if that should happen, the locality shall continue to enjoy the protection provided by the other provisions of the LOAC.
In view of the provisions of  Geneva Convention IV articles 14 and 15 and Additional Protocol I articles 59 and 60, it is important that the different zones and localities are clearly distinguished from another as far as their physical location is concerned, but also with regard to their aims. Their aims can be summarised as follows:
- Hospital Zones. Provide permanent shelter to military and civilian wounded or sick.
- Safety Zones. Provide permanent shelter to certain specially protected categories of persons. These zones require special protection.
- Neutral Zones. Provide temporary protection in the combat zones to wounded and sick combatants, non-combatants and civilians not participating in hostilities.
- Non-defended and Demilitarised Zones. Provide permanent protection in and/or near the combat zones to non-combatants and civilians not participating in hostilities. A Party to the conflict establishes non-defended zones by means of a unilateral declaration, while demilitarised zones are established by a formal agreement between Parties.
All these zones have one thing in common, apart from the aim to protect persons, to wit, that they are all dependent on recognition by an adverse Party.
The idea with such zones started in 1870 with Henri Dunant, who suggested that certain towns be declared neutral and that wounded persons be collected there. It is difficult to declare such zones before a war, as it is difficult to ascertain the strategic situation before a war. However, nothing prevents States from establishing a number of such zones in time of peace and only utilising a few such zones (or all of them) in time of war.
It is possible to combine different types of zones in one area.
2.7 Special Protection: Occupied Territories
Responsibilities during Occupation
The Occupying Power has all the responsibilities of the legitimate State. Its responsibilities include to:
- Establish hospital and safety zones and non-defended localities in occupied territories, if the need arises. (Geneva Convention IV Article 14.)
The manual also states:
Establishment and Protection of Protected Zones
- Preplanned protected zones are established by agreement between Parties to the conflict. …
- Parties to a conflict can also establish improvised protected zones. Such zones are areas from which military objectives and activities have been removed, and which
- Are situated near or in a zone where armed forces are in contact; and
- Are open for occupation by the enemy.
- Commanders must ensure that they take all necessary steps for the respect of such protected zones.
- Appropriate advice must be given to the civilian authorities regarding practical aspects and conditions to be fulfilled pertaining to such zones, such as who is responsible for the management of the zone, delimitation and marking of the perimeter, the removal of military personnel and equipment, access control, maintenance of public order and policing functions, supply, hygiene, keeping the public informed etc.
- In cases where the perimeter of the protected zone does not correspond to the official boundary of the town, district, etc. the civilian authority over such zone must be clearly established. Such an ad hoc area will require ad hoc authority with corresponding responsibilities.
- The zone perimeter must be clearly visible from the air and the ground, e.g. a beach, edge of a built-up area or forest, a road, river etc. Where necessary, the zone perimeter must be marked by agreed signs of sufficient size and visibility.
- All armed forces must be given precise instructions for behaviour regarding such protected zones when
- Leaving the protected zone;
- Abandoning it without fighting;
- Taking it over;
- Being prohibited from extending military operations to the zone; or
- When engaged in combat action in the vicinity of the zone.
In 1987, in the Petane case, the Cape Provincial Division of South Africa’s Supreme Court dismissed the accused’s claim that the 1977 Additional Protocol I reflected customary international law. The Court stated:
The accused has been indicted before this Court on three counts of terrorism, that is to say, contraventions of s 54(1) of the Internal Security Act 74 of 1982. He has also been indicted on three counts of attempted murder.
The accused’s position is stated to be that this Court has no jurisdiction to try him.
… The point in its early formulation was this. By the terms of [the 1977 Additional] Protocol I to the  Geneva Conventions the accused was entitled to be treated as a prisoner-of-war. A prisoner-of-war is entitled to have notice of an impending prosecution for an alleged offence given to the so-called “protecting power” appointed to watch over prisoners-of-war. Since, if such a notice were necessary, the trial could not proceed without it, Mr Donen suggested that the necessity or otherwise for giving such a notice should be determined before evidence was led. …
On 12 August 1949 there were concluded at Geneva in Switzerland four treaties known as the Geneva Conventions. …
South Africa was among the nations which concluded the treaties. … Except for the common art 3, which binds parties to observe a limited number of fundamental humanitarian principles in armed conflicts not of an international character, they apply to wars between States.
After the Second World War many conflicts arose which could not be characterised as international. It was therefore considered desirable by some States to extend and augment the provisions of the Geneva Conventions, so as to afford protection to victims of and combatants in conflicts which fell outside the ambit of these Conventions. The result of these endeavours was Protocol I and Protocol II to the Geneva Conventions, both of which came into force on 7 December 1978.
Protocol II relates to the protection of victims of non-international armed conflicts. Since the State of affairs which exists in South Africa has by Protocol I been characterised as an international armed conflict, Protocol II does not concern me at all.
The extension of the scope of art 2 of the Geneva Conventions was, at the time of its adoption, controversial. …
The article has remained controversial. More debate has raged about its field of operation than about any other articles in Protocol I. …
South Africa is one of the countries which has not acceded to Protocol I. Nevertheless, I am asked to decide, as I indicated earlier, as a preliminary point, whether Protocol I has become part of customary international law. If so, it is argued that it would have been incorporated into South African law. If it has been so incorporated it would have to be proved by one or other of the parties that the turmoil which existed at the time when the accused is alleged to have committed his offences was such that it could properly be described as an “armed conflict” conducted by “peoples” against a “ra[c]ist regime” in the exercise of their “right of self-determination”. Once all this has been shown it would have to be demonstrated to the Court that the accused conducted himself in such a manner as to become entitled to the benefits conferred by Protocol I on combatants, for example that, broadly speaking, he had, while he was launching an attack, distinguished himself from civilians and had not attacked civilian targets. …
… I am prepared to accept that where a rule of customary international law is recognised as such by international law it will be so recognised by our law.
To my way of thinking, the trouble with the first Protocol giving rise to State practice is that its terms have not been capable of being observed by all that many States. At the end of 1977 when the treaty first lay open for ratification there were few States which were involved in colonial domination or the occupation of other States and there were only two, South Africa and Israel, which were considered to fall within the third category of ra[c]ist regimes. Accordingly, the situation sought to be regulated by the first Protocol was one faced by few countries; too few countries in my view, to permit any general usage in dealing with armed conflicts of the kind envisaged by the Protocol to develop.
Mr Donen contended that the provisions of multilateral treaties can become customary international law under certain circumstances. I accept that this is so. There seems in principle to be no reason why treaty rules cannot acquire wider application than among the parties to the treaty.
Brownlie Principles of International Law 3rd ed at 13 agrees that non-parties to a treaty may by their conduct accept the provisions of a multilateral convention as representing general international law. …
I incline to the view that non-ratification of a treaty is strong evidence of non-acceptance.
It is interesting to note that the first Protocol makes extensive provision for the protection of civilians in armed conflict. …
In this sense, Protocol I may be described as an enlightened humanitarian document. If the strife in South Africa should deteriorate into an armed conflict we may all one day find it a cause for regret that the ideologically provocative tone of s 1(4) has made it impossible for the Government to accept its terms.
To my mind it can hardly be said that Protocol I has been greeted with acclaim by the States of the world. Their lack of enthusiasm must be due to the bizarre mixture of political and humanitarian objects sought to be realised by the Protocol. …
According to the International Review of the Red Cross (January/February 1987) No 256, as at December 1986, 66 States were parties to Protocol I and 60 to Protocol II, which, it will be remembered, deals with internal non-international armed conflicts. With the exception of France, which acceded only to Protocol II, not one of the world’s major powers has acceded to or ratified either of the Protocols. This position should be compared to the 165 States which are parties to the Geneva Conventions.
This approach of the world community to Protocol I is, on principle, far too half-hearted to justify an inference that its principles have been so widely accepted as to qualify them as rules of customary international law. The reasons for this are, I imagine, not far to seek. For those States which are contending with “peoples[’]” struggles for self-determination, adoption of the Protocol may prove awkward. For liberation movements who rely on strategies of urban terror for achieving their aims the terms of the Protocol, with its emphasis on the protection of civilians, may prove disastrously restrictive. I therefore do not find it altogether surprising that Mr Donen was unable to refer me to any statement in the published literature that Protocol I has attained the status o[f] customary international [law].
I have not been persuaded by the arguments which I have heard on behalf of the accused that the assessment of Professor Dugard, writing in the Annual Survey of South African Law (1983) at 66, that “it is argued with growing conviction that under contemporary international law members of SWAPO [South-West Africa People’s Organisation] and the ANC [African National Congress] are members of liberation movements entitled to prisoner-of-war status, in terms of a new customary rule spawned by the 1977 Protocols”, is correct. On what I have heard in argument I disagree with his assessment that there is growing support for the view that the Protocols reflect a new rule of customary international law. No writer has been cited who supports this proposition. Here and there someone says that it may one day come about. I am not sure that the provisions relating to the field of application of Protocol I are capable of ever becoming a rule of customary international law, but I need not decide that point today.
For the reasons which I have given I have concluded that the provisions of Protocol I have not been accepted in customary international law. They accordingly form no part of South African law.
This conclusion has made it unnecessary for me to give a decision on the question of whether rules of customary international law which conflict with the statutory or common law of this country will be enforced by its courts.
In the result, the preliminary point is dismissed. The trial must proceed.
In 2010, in the Boeremag case, South Africa’s North Gauteng High Court stated:
In Petane, … Conradie J found that the provisions of [the 1977 Additional] Protocol I are not part of customary international law, and therefore are also not part of South African law.
Referring to the fact that in December 1986 only 66 of the 165 States party to the Geneva Conventions had ratified Protocol I, the Court [in Petane] stated:
This approach of the world community to Protocol I is, on principle, far too half-hearted to justify an inference that its principles have been so widely accepted as to qualify them as rules of customary international law. The reasons for this are, I imagine, not far to seek. For those States which are contending with “peoples[’]” struggles for self-determination, adoption of the Protocol may prove awkward. For liberation movements who rely on strategies of urban terror for achieving their aims the terms of the Protocol, with its emphasis on the protection of civilians, may prove disastrously restrictive. I therefore do not find it altogether surprising that Mr Donen was unable to refer me to any statement in the published literature that Protocol I has attained the status of customary international law.
Important changes with respect to certain aspects applicable at the time of Petane have taken place. The ANC [African National Congress] has become South Africa’s ruling party and in 1995 ratified Protocol I. The total number of States that have ratified it, is now … 162.
This last aspect forms the basis on which the First Respondent [the State] and the applicants agree that Protocol I forms part of customary international law as well as of South African law. As requested, this position is accepted for the purposes of the decision, without deciding on the matter.
Despite these changes, it remains debatable whether the provisions of Protocol I have become a part of South African law in this way.
The consensus of both parties to the conflict is required. See Petane … and Article 96 of Protocol I. …
Parliament’s failure to incorporate Protocol I into legislation in accordance with Article 231(4) of the Constitution in fact points to the contrary, and is indicative that the requirements of usus
and/or opinio juris
have not been met. See Petane
[footnotes in original omitted]
The Court also held:
If the [1977 Additional Protocol I] applies in South Africa as customary international law, the two requirements that form the basis of customary law must be met. It is arguable that the requirement of usus
has been met by the vast number of States that have acceded or ratified it. By ratifying Protocol I the Republic of South Africa has indicated its intention to apply the Protocol, thereby fulfilling the requirement of opinio juris