Course Home


Turkana Archaeology pages

1. Dating. Some of the pillars at Kalokol are asserted to form sight lines to certain stars when they rose on the eastern and southeastern horizon approximately 2,300 years ago. For this to be true, the site would have to have been built and used as that time.

There is no absolute, independent date for Kalokol, so the assertion that it is 2,300 years old is an hypothesis that requires testing. Comparative evidence suggests that Kalokol was more likely to have been created and used sometime between 3,800 and 4,500 years ago. This is based on the association of Nderit Ware and evidence from other Namoratunga having the same architectural structure as the Namoratunga at Kalokol.

2. Ethnographic Model. The selection of stars used for the sight lines is based on stars that are important in the Borana calendar. The Borana are speakers of an Eastern Cushitic language and practice mixed agriculture and pastoralism. Excluding groups to the far north and west, they are one about 10 Eastern Cushitic language groups. Most of these groups each contain a number of different tribal communities. Most of these rely heavily on agriculture.

Using the Borana calendar poses a number of problems. (1) We do not know what calendars are in use by the other Eastern Cushitic groups. (2) We know little about what the Borana calendar is actually used for. Commonly, such calendars are used to schedule agricultural activities. As far as we know, the people who built the Namoratunga did not practice agriculture. (3) If Cushitic speakers 2,300 years ago used a calendar similar to that of the Borana, then a number of historically related groups should use similar calendars today. This has yet to be demonstrated. (4) It is assumed that the stars and their position in the calendar have not changed in 2,300 years. (5) It is assumed that the functional significance of the star alignments are equally valid within the Winter Rainfall Zone (Ethiopian Highlands) and outside that zone (Turkana Basin). (6) It requires that Kalokol was occupied once every month during the period of its construction and subsequent use, but there is no evidence for a nearby Nderit settlement or periodic camps.

3. Horizon Factor. The calculations of the point on the horizon where stars would rise uses the "astronomical" horizon. To the east, the actual and "astronomical" horizon are the same, but to the southeast a hill defines the actual horizon. Since stars do not rise straight up from the horizon, using the calculation to the "astronomical" horizon will introduce a small, but significant error.

TOP       Home

Credit: photo copyright © Maria Helena Barreira, 1998; from


4. Replication. If the Nderit people used Namoratunga as observatories to observe the eastern and southeastern skies, then it is very difficult to explain their location at Jarigole or Lothagam, seen here to the left, where hills or mountains blocked this view.

The results from Kalokol, when demonstrated to be valid and accurate, need to be replicated at other Nderit Namoratunga.



5. Site Structure. Many of the details of Kalokol's architectural structure have been ignored in posing and testing the hypothesis that the pillars there are part of an observatory

The first component to be considered is the cobble collar that surrounds the platform. What is the function of this collar. Notice first that that the collar overlaps in the southwest corner. This arrangement is a common form of entrence among pastoral peoples in many parts of northern and eastern Africa and there are similar arrangements a Jarigole and the large Namoratunga at Lothagam. If this is an entrance, why was it needed? ...


6. Stockade ? This is a Karamajong village, ca. 1960, in northern Uganda, about 200 miles SW of Kalokol. Notice the high stockade that is used to protect the village. In general, East African pastoralists put stockades or thorn fences around anything that they value. If that occurred at Kalokol, the view of the horizon would have been obstructed by the fence.



7. Pillar Decapitation. The top of Pillar 12 at Kalokol is formed by a conchoidal fracture that was created when a corner of the pillar was struck with sufficient force and its top detached after the pillar was in place. Multiple partial cones of percussion indicate that several blows were required to accomplish this. After decapitation, the fracture surface was used as a platform to strike a long, sharp flake down one corner of the pillar. This flake was probably used as a tool. The use of pillars as cores from which to strike flakes occurs more commonly at Lothagam and Jarigole.

This means that Pillar 12 was originally taller by at least six inches. Since the pillar is leaning considerably, replacement of its crown would shift the position of point used to calculate the sight line. Hence, the three sight lines that utilize Pillar 12 are to some small degree in error.


Credit: photo copyright © Maria Helena Barreira, 1998; from


8. Pillar Displacement. It is reasonable to assume that all of the pillars at Kalokol were originally placed in fully upright, vertical positions. They are now all tilted to various degrees, so that the points used to calculate lines of sight have been displaced between 5 cm. and 30 cm. Since the pillars used for sight lines are relatively close together, the errors in projecting these lines to the horizon can be substantial.

9. Unused Pillars. If it is the function of the pillars to mark and predict astronomical events, then all of them should form sight lines of some sort. Many, however, do not appear to have such a function. This needs to be explained.

©2009, 2010 by Charles M. Nelson
All rights reserved.