Communal huts took many people several days to construct but with careful maintenance they could last for a decade or more. The available evidence indicates that communal house developed relatively late. Each of the Andamanese main groups built and - except for the Great Andamanese - still builds communal huts in its own unmistakable style. Communal huts are likely to have developed from the traditional village layout. We do not know the details but can guess at the following rough-and-ready evolutionary sequence: the individual huts of a village moved closer and closer together until they could be linked and then merged into one large building of individual units and finally, among the Onge, into one large construction without individual sub-units.
The Great Andamanese had communal round huts that consisted of many individual huts built together in a circle. They could be as large as 18 m (60 ft) in diameter and 6 m (20 ft) to 9 m (30 ft) in height, giving shelter to 50 to 80 persons with the capacity to fit in many more during larger festive gatherings. An outer circle of low pillars formed the equivalent of the back of an individual hut with an inner and much higher circle of pillars representing the former front of the hut. In the center there was left an open circle in the roof for the smoke of many fires within to escape while the rain was kept out by allowing one or two of the roof mats to overhang the others at the top. The center ground was used as a communal and somewhat cramped dancing area. The space of one hut was left open to serve as entrance to the whole structure though usually covered with roofing mats down to a few feet from the ground. The roof itself was covered with large rectangular or tapered mats made in exactly the same way as the square mats used to cover individual huts. The last known intact Great Andamanese communal hut was seen and photographed in 1895 and none could be found by 1908.
The Onge are today issued by the Indian welfare organization in charge of them, Samiti, with housing that does not resemble the traditional Onge korale huts (from ko = 'where' and ale = 'children' - 'where the children come from'). They have not been an unqualified success with the Onge. The traditional Onge communal hut, beyra, protected its inhabitants from even the heaviest downpours. It was still widely built and used in the 1950s but is not often found today. A beyra consists of a ring of poles held together and bent towards the center by rings of cane, giving the whole the appearance of a beehive or umbrella. This skeleton is then covered with mats made of woven palm leaves. Inside the house there are many sleeping platforms of varying sizes arranged radially all around the periphery on which the Onge sleep with their heads towards the center of the house.
The Jarawa communal round houses are closer in construction to those of the Onge than to those of the Great Andamanese but are smaller: a round house of the Jangil was found to be only 4.5 m (15 ft) in diameter and 2.7 m (9 ft) high. It was thought for a long time that the Jarawa had abandoned the construction of round houses during the late 19th century. Certainly none were reported between then and the 1960s. The ceaseless hostilities between them and the Aka-Bea and later with the British they had certainly forced them into a life of hiding and camouflage. Large and permanent structures were thought to have no place in such an adaptation. It therefore came as a surprise then when Indian anthropologists finally established friendly relations with the Jarawa and discovered in 1974 round houses in full use. The inability, for nearly a century, to spot sizable buildings in the jungle may serve as a chastening reminder of what else might have gone unseen and unnoticed by outsiders' eyes.
The Sentineli have a different kind of communal house, one that is unique in the Andaman islands. It is not a round house but a long, straight, rectangular structure. Known as a "long shelter" this building can be 12 m (40 ft) long, 3.6 m (12 ft) deep and 1.8 m (6 ft) high in front. It looks like a row of many conventional Andamanese lean-to huts lined up side by side. Until 1903 only ordinary lean-to huts had been reported from the island but in that year the first long shelter was discovered. No house of this type has yet been seen by recent Indian contact parties but this is hardly surprising since permanent structures are likely to be hidden in the forests where no recent expedition has ventured.
Great Andamanese Aryoto groups tended to place their villages and occasional community house in the jungle immediately behind the beach from where they could observe any approaching vessel without themselves being seen. From what little we know of the Sentineli, some of their villages are positioned in the same way, as are those of the Onge shore-dwellers.
The villages and camps of the inland tribes, the Great Andamanese Eremtaga and particularly the Jarawa, were preferably placed away from large trees in an open space surrounded by jungle. Large trees could be dangerous during storms while the nearby jungle provided protection from the sun and wind. The availability of fresh water not far away was also an important consideration when siting camps. The Jarawa strategy for survival involved a larger element of hiding than that of other groups, their villages were often found hidden on the ridges of hills where they could not easily be surprised. Alone among Andamanese, the Jarawa routinely place lookouts and even clear trees to get a better view of all likely approaches.
Most main camps and many temporary villages were built on piles of ancient domestic refuse, sea shells, stones, bones and the like. Forty of these kitchen-midden are known in the Andamans today and a few have been archaeologically investigated. Heaps of fresh, rotting organic refuse in the hot climate of the Andamans have been called fly factories in full mass production, and very smelly factories at that. The original human group to settle in an area picked a protected dry spot with easy access to fresh water to build their first village. They then had to chose a place, somewhat removed from the living area, for their rubbish dump. The dump grew, generation after generation, century after century, at a rate that has been estimated very roughly at around 20 cm (8 inches) per century. One day, the little hill would have grown big enough to provide a convenient raised platform for a small village. Even with a village at its top, the midden would still keep growing albeit at a slower rate. Some refuse
A well-placed midden allowed the owners to build a village nearer to food resources, provided a raised dry spot in an otherwise wet swamp and gave some protection from flooding during the wet season. A large midden was also a source of pride, a marker of ownership and a link to the past for the owners.