Teniky is a geological and archaeological site located in the Isalo Massif, a mountainous formation in Madagascar's southwestern Ihorombe region. It is situated in relative isolation and seclusion, over from the nearest coast. The site is notable for its unique rock-cut architecture, which is unlike any other found in Madagascar and the wider East African coast. The enigmatic complex spans nearly , and contains precise stone walls, quarries, terraces, niches, rock-cut boulders and stone basins.
Radiocarbon dating of charcoal found at newly-discovered man-made structures at the site found them to date to the 10th–12th centuries CE. Shards of Chinese and Southeast Asian pottery found at Teniky date to the 11th–14th centuries CE. A 2024 study of the rock-cut niches at Teniky identified their closest architectural parallels as first-millennium Zoroastrian niches in Iran, particularly in the Fars region, suggesting that the site may have been a necropolis constructed by settlers of Zoroastrian origin.
Site
The Teniky site is accessible only by hiking across steep and rugged terrain—a journey that took researchers eight hours from the nearest village in 2024. It is situated in relative isolation and seclusion, over from the nearest coast. Its cirque is accessed from an unpaved road originating in Beroroha and passing through Ranohira, Andriamanero, and Tameantsoa, before becoming a footpath going up the mountains of Isalo to Tsimivositra, crossing the canyon of Kelivozana, descending to Ambatomãra and finally branching north before the abandoned village of Sahanafo. Access from Fandranarivo-Berenty is also possible.
The archaeological site features "clustered caves, the entrances to which are partially closed by walls made of cut stones, flat areas with foundation stones, artificially hollowed excavations at the base of the cliffs, carved blocks of stones, etc." The Sahanafo River runs through the complex, and the terrain is marked by geological formations, including a heart-shaped cirque, across plains and wooded areas. Bushpig and lemurs are "plentiful" in the area, and the river contains a large number of eel.
A 2024 study by Guido Schreurs, professor of geology at the University of Bern, and others descriptively divides the Teniky site into eight archaeological zones, with the Sahanafo River cutting between the site's eastern and western sides. In Zone 1 are rock-cut niches discovered in 2021. Zone 2, which contains Zone 1, is the westernmost zone of Teniky, comprising a "north-south trending hill with terraces and dry stone walls". Zone 3 contains dry stone walls, including some in a forested area. Zones 4 and 5 contain sandstone quarries on the east and west sides of the Sahanafo, respectively. Zone 6 contains a "nearly closed" circular dry stone wall with a diameter of approximately . Zone 7 contains terraces and dry stone walls on a ridge. Zone 8 is the previously-known site.
Zone 1
In Zone 1 are nine niches—eight quadrangular and one nearly circular—carved into a cliff under a rock shelter. The very eroded main niche, where most archeological structures are concentrated in the zone, is carved with benches and pillars. Two of the three pillars are thinner towards the bottom, flaring toward the ceiling in a quadrangular shape. The third, central column is not carved thinner or flared, giving it a “more massive, rougher appearance". Two stone structures sit on the bench: one long squared sandstone block flanked centrally by two smaller ones, and one pile of mostly flat and rectangular stones. The long block from the first structure was likely taken from the walls in front of the niche, while the rocks in the pile may have been part of the first structure before being moved. Optically stimulated luminescence on the sediment in this niche was not successful, but radiocarbon dating of charcoal pieces from the niche found them to be 986 ± 22 and 1161 ± 22 years old, suggesting a calibrated origin date of 891–1155 CE.
The seven other quadrangular niches and the single circular niche in this rock shelter were excavated but contained no further architectural features or artifacts. Schreurs et al. suggest that the morphology of the niches and pillars in the shelter at Zone 1 indicate two phases of construction, with possibly only a quadrangular niche being carved in the first phase before the addition of the adjacent parts of the structure.
In front of the niches is a conglomerate block carved with feet and a regular central depression—maybe a ceremonial basin used to contain fire or water.
In the northeast part of the shelter are two sandstone walls faced with a flat area partially lined with rock slabs placed vertically in the ground, as well as a stone basin close to the niches in the southwestern part of the rock overhang. The stone walls were apparently laid directly on the earth without a dug foundational trench, and piles of blocks downhill of the structures were likely part of the intact walls. A structure resembling stairs or an entrance sits at the end of the shorter wall. Excavation along the longer wall found charcoal flakes and ash in the layers of sediment, as well as another layer containing burned roots, indicating possibly natural fire. A stratigraphic excavation along the shorter wall found a man-made layer of dust made of demolished sandstone blocks mixed with sand to stabilize the ground. Beneath this layer is sandy loam with charcoal flakes and unidentified animal bones. The fire’s nature as man-made or natural is indeterminate, and there are no apparent traces of man-made cuts on the bones. Radiocarbon dating of charcoal from this site found the samples to be from 891–990 CE and 991–1138 CE.
Adjacent to the walls is a two-sided structure defined by stones placed vertically into the earth. Between this structure and the longer wall are two fragments of a stone basin.
A shard of pottery, determined to be of Southeast Asian origin and dated to the 11th–13th centuries CE, was found in excavation of Zone 1. The charcoal and pottery dating suggests that the niches and walls in Zone 1 date to the late first/early second millennia CE. Schreurs et al. consider the entire complex in Zone 1 to likely have been of ritual use, and unlikely to have been of domestic use.
Zones 2 and 3 Zone 2 covers a 1,500 meter-long and 500 meter-wide hill trending north to south. Its highest point stands 200 meters above the valley floor and seats a rectangular structure of walls surrounding a raised platform. Schreurs et al. opine that the platform was probably used for occasional ritual activities. The hill is flanked with terraces, some of which are bordered by dry stone walls.
Zone 3, in the valley east of the Sahanafo River, contains a wooded area delimited by a square dry stone wall, inside of which are some other segments of dry stone wall. Schreurs et al. find that these are the remains of dry, unhewn stone walls no higher than 1 meter. One wall in the north is 120 meters long, and another, in the eastern part of the zone, is 750 meters long and runs north-northeast/south-southwest. Also in the northern part of the zone is a pair of circular dry stone walls 1.5 meters each in diameter and height. Shards of Chinese celadon ware and Southeast Asian stoneware dated between the 11th and 14th centuries AD were found between the square wall and the 750 meter-long wall.
Zones 4 and 5 Zones 4 and 5, respectively west and east of the Sahonafo River, contain sandstone quarries. Mineral compositions of the partially extracted blocks found in these quarries are similar to those in the blocks used to construct the walls at the rock shelters in zones 1 and 8, between which the two quarries are roughly central (roughly from the quarries to each wall). Schreurs et al. consider it likely that these rock shelter walls were constructed with blocks sourced from the quarries in Zones 4 and 5.
Zones 6 and 7 An "imposing" dry stone wall arranged in a nearly-closed circular shape with a diameter of 150 meters sits in Zone 6. The wall is constructed from blocks of sandstone and conglomerate rock. The wall is discontinuous in its northern, eastern, and southern portions. There are no apparent signs of demolition. The geography of the site slopes upward to the north, obscuring the horizon when facing that direction. This, coupled with the circle's discontinuity, make it unlikely in the opinion of Schreurs et al. that the structure was an observation post. The lack of signs of demolition to explain the openings in the structure's perimeter make it also unlikely that it was some sort of animal enclosure. Schreurs et al. consider it enigmatic.
Zone 7, conversely, contains a likely observation post, considering the excellent views offered by the summit of the escarpment found in the zone, which also includes a number of terraces and dry stone walls.
Zone 8
600 meters south of Zone 7 is Zone 8, the previously-known heart-shaped cirque formed by fluval erosion where rock-cut architecture, terraces, and carved sandstone walls are situated. Man-made terraces in Zone 8 cover an area of roughly , with the largest terrace having an area of . The main archaeological structures of the cirque include dozens of rock-cut niches, sandstone walls, carved boulders, a rock-cut chamber called the Petit Grotte, and a walled rock shelter called the Grande Grotte.
Petit Grotte The Petit Grotte, so referred to by Ginther and Hébert in 1963, is situated in the northern face of the southernmost cliffs of the cirque. Its entrance is approximately high and wide, and its maximum internal height is roughly . The chamber's floor is covered with quartz grains, and three large, "beautifully carved" pillars—one broken and attached only at the top—are carved in the chamber's center. Benches with small pillars on their backs are carved into three sides of the chamber. Schreurs et al. consider the chamber to have likely served a ritual function. The first archaeological description of Teniky, published in 1950, reports that residents of Sahonafo refer to the cave as "the market", and elderly Sahonato resident Tsimangataka said that this was where the Vazaha foreigners had stored their belongings (see ).
Grande Grotte
A large rock shelter, referred to by Ginther and Hébert as Grande Grotte, is delimited by two parallel stone walls assembled without mortar from precisely squared and stacked sandstone blocks. The inner wall is long, high, and wide, and contains an opening—probably an entrance to the shelter. Bevelled blocks on either side of the apparent entrance's base suggest that the opening may once have held "some kind of closure". The left part of the inner wall has collapsed, and had degraded somewhat since a 1940 expedition (photographed by ) by 2021. Almost every block of the wall bears carved graffiti with names and dates going back to the late 19th century. The remains of the parallel outer wall are darkly weathered from exposure to rain and wind, but scratching the outer wall's blocks reveals the same light color as the inner wall's blocks'. The blocks' sandstone was likely sourced from the quarries in Zones 4 and 5.
The construction style of the parallel walls of the Grande Grotte is unique and unknown across Madagascar and all of the East African coast. A niche containing a bench is cut into the northern part of the rock shelter, with a recessed opening that suggests it could be closed off.
Niches
Carved in a cliff face about northwest of the Grande Grotte are ten rectangular niches, roughly high and wide and deep, separated by vertical walls which have pillar-like shapes in frontal view. Some lighter, less friable layers in the sandstone still bear tool marks. north of these rectangular niches are four circular niches cut into a cliff face above ground level. A circular depression at the same level but from the southernmost niche is possibly a heavily eroded fifth niche.