Skip to main content Accessibility help

A New Framework for the Study of Early Pastoral Communities in East Africa

  • Peter Robertshaw (a1) and David Collett (a1)


This review provides a new interpretative framework for the ‘Neolithic’ in East Africa. A seriation of pottery assemblages is used to delineate several archaeological traditions, the implications of which include rejection of the use of the terms ‘Neolithic’ and ‘Pastoral Neolithic’, and the demise of previous attempts at archaeological—linguistic correlations. Evaluation of the dating evidence brings into question the validity of early dates for domestic stock and cultivated crops in the region. A new model for the development of specialized herd management strategies in the Central Rift Valley is outlined. This model rests upon the definition of pastoralism as an ideological system rather than as a subsistence strategy. Finally, the archaeological evidence for the antecedents of the early pastoral communities of East Africa is examined and the ascription of some of these assemblages to the ‘aquatic civilization of Middle Africa’ is questioned.



Hide All

1 Leakey, L. S. B., The Stone Age Cultures of Kenya Colony (Cambridge, 1931).

2 See, for example, Sutton, J. E. G., ‘The archaeology and early peoples of the Highlands of Kenya and northern Tanzania’, Azania, I (1966), 3757; and Onyango-Abuje, J. C., ‘Reflections on cultural change and distribution during the Neolithic period in East Africa’, in Ogot, B. A. (ed.), Hadith, VI (Nairobi, 1976), 1430.

3 See, for example, Cole, S., The Prehistory of East Africa (Harmondsworth, 1954); and Cohen, M., ‘A reassessment of the Stone Bowl Cultures of the Rift Valley, Kenya’, Azania, v (1970), 2738.

4 Bower, J. R. F., Nelson, C. M., Waibel, A. and Wandibba, S., ‘The University of Massachusetts’ Later Stone Age/Pastoral “Neolithic” comparative study in Central Kenya: an overview', Azania, XII (1977), 119–46; Bower, J. R. F. and Nelson, C. M., ‘Early pottery and pastoral cultures of the Central Rift Valley, Kenya’, Man, n.s. XIII, iv (1978), 554–66.

5 We are here using the term ‘culture-history’ following American archaeological usage as defined in the text.

6 ‘Industries’ and ‘traditions’ tend to be interchangeable terms in archaeological parlance, the former being used by Stone Age researchers and the latter by Iron Age specialists. The two terms are not in fact co-terminous: an Iron Age ‘tradition’ perhaps is best correlated with a Stone Age ‘industrial complex’, while ‘phases’ and ‘fades’ of traditions are more, but not entirely, comparable with ‘industries’. The terminology of ‘tradition’, ‘phases’ and ‘fades’ is preferred in this article; see Willey, G. R. and Phillips, P., Method and Theory in American Archaeology (Chicago, 1958).

7 Cf. Sackett, J. R., ‘Approaches to style in lithic archaeology’, J. Anthrop. Arch., I, i (1982), 59112; see especially p. 64.

8 The problem of recognizing stylistic as opposed to functional attributes on stone artefacts continues to hamper the establishment of Stone Age culture-histories; see Sackett, ‘Approaches’.

9 As summarized by Cole, , Prehistory, 193et passim.

10 Sutton, ‘Archaeology’.

11 Bower et al., ‘Later Stone Age’; Bower and Nelson, ‘Early pottery’; Ambrose, S. H., ‘Elmenteitan and other Late Pastoral Neolithic adaptations in the central highlands of East Africa’, in Leakey, R. E. and Ogot, B. A. (eds.), Proceedings of the Eighth Congress of Prehistory and Quaternary studies (Nairobi, 1980), 279–82; Wandibba, Simiyu, ‘The application of attribute analysis to the study of Later Stone Age/Neolithic pottery ceramics in Kenya (summary)’, in Leakey, and Ogot, (eds.), Proceedings, 283–5; see also various preliminary reports in Nyame Akuma (Newsletter of African Archaeology) (Calgary).

12 Wandibba, ‘Attribute analysis’.

13 Bower, and Nelson, , ‘Early pottery’, 564.

15 Its usage in East Africa is rather different from its normal meaning; see Shepard, A. O., Ceramics for the Archaeologist (Washington, D.C., 1956), 318–19 for definition and discussion of ‘ware’.

16 See Doran, J. E. and Hodson, F. R., Mathematics and Computers in Archaeology (Edinburgh, 1975), 160. A botanist, for example, uses a key to aid identification; however, a species is not defined by the attributes used in the key.

17 The analysis is presented in full in Collett, David and Robertshaw, Peter, ‘Pottery traditions in early pastoral communities of Kenya’, Azania, XVIII (1983) (in press).

18 See below.

19 Ambrose, ‘Elmenteitan’.

20 Ambrose, S. H., ‘Historical linguistic reconstruction in East Africa: the archaeological evidence’, in Ehret, C. and Posnansky, M. (eds.), The Archaeological and Linguistic Reconstruction of African History (Berkeley, 1982).

21 Lwanga-Lunyiigo, S., ‘The Bantu problem reconsidered’, Current Anthropology, XVII, ii (1976), 282–6.

22 Ambrose, as cited in n. 20.

23 See below for discussion of the supposedly very early domestic stock from southern Kenya and northern Tanzania.

24 This Oltome material, unfortunately poorly dated, had been provisionally labelled as the Lokabulo tradition. See David, N., Harvey, P. and Goudie, C. J., ‘Excavations in the Southern Sudan, 1979’, Azania, XVI (1981), 754; Robertshaw, P. T. and Mawson, A., ‘Excavations in Eastern Equatoria, Southern Sudan, 1980’, Azania, XVI (1981), 5595; Robertshaw, P. T., ‘Eastern Equatoria in the context of later Eastern African prehistory’, in Mack, B. J. and Robertshaw, P. T. (eds.), Culture History in the Southern Sudan (Nairobi, 1982), 89100; for the problems of dating see Robertshaw, P. T., Collett, D. P., Gifford, D. P. and Nubi, M. B., ‘Shell middens on the shores of Lake Victoria’, Azania (in press).

25 The results of work by the British Institute in Eastern Africa and others are summarized in the archaeological contributions in Mack and Robertshaw, Culture History.

26 Bower and Nelson, ‘Early pottery’.

27 See n. 15.

28 Bower and Nelson, ‘Early pottery’.

29 Ibid. 559.

30 It seems to be almost universally agreed that radiocarbon dates on charcoal are more reliable than those on bone. However, there remains always the question of the validity of the association between the charcoal sample and the archaeological material to which the date is believed to refer.

31 Details of this analysis are presented in Collett, David and Robertshaw, Peter, ‘Problems in the interpretation of radiocarbon dates: dating the Pastoral Neolithic of East Africa’, African Archaeological Review, I (1983).

32 C. M. Nelson, pers. comm. and cited in Leakey, Richard, The Making of Mankind (London, 1981), 197.

33 Bower, J. R. F. and Gogan-Porter, P., ‘Prehistoric cultures of the Serengeti National Park, Tanzania’, Papers in Anthropology, No. 3, Iowa State University (1981).

34 Ibid. 49.

35 For the same reasons we reject a date of 5805 ±185 b.p. (GX - 8536; C13 corrected) on bone apatite associated with cattle remains at Gogo Falls; see Robertshaw et al., ‘Shell middens’.

36 Owen, R. B., Barthelme, J. W., Renaut, R. W. and Vincens, A., ‘Palaeolimnology and archaeology of Holocene deposits northeast of Lake Turkana, Kenya’, Nature, CCXCVIII (1982), 523–9; see also discussion in Collett and Robertshaw, ‘Problems’.

37 Phillipson, D. W., ‘The excavation of Gobedra Rock-shelter, Axum’, Azania, XII (1977) 5382.

39 Bower and Nelson, ‘Early pottery’.

40 See above, and Collett and Robertshaw, ‘Pottery’.

41 Collett and Robertshaw, ‘Problems’.

42 For the definition of ‘Narosura ware’ see Bower, et al. , ‘Later Stone Age’, 135; Wandibba, ‘Attribute analysis’.

43 Gramly, R. M., ‘Analysis of faunal remains from Narosura’, Azania, IX (1974), 219–22; Onyango-Abuje, J. C., ‘A contribution to the study of the Neolithic in East Africa with particular reference to Nakuru-Naivasha Basins’ (Ph.D. thesis, University of California, Berkeley, 1977); Gifford, D. P., Isaac, G. LI. and Nelson, C. M., ‘Evidence for predation and pastoralism at Prolonged Drift: a Pastoral Neolithic site in Kenya’, Azania, XV (1980), 57108; D. P. Gifford and J. K. Mengich, pers. comm.

44 F. B. Marshall and D. P. Gifford, pers. comm. These researchers are also doing valuable work on the age-structures of prehistoric herds.

45 Historical linguists, particularly Ehret, have argued that PN peoples would at least have known of various domestic crops.

46 Carbonized seeds have been recovered by froth- flotation from the site of Ngamuriak (Marshall, Fiona and Robertshaw, P. T., ‘Preliminary report on archaeological research in the Loita-Mara region, S. W. Kenya’, Azania XVII (1982) (in press)) but these await identification.

47 Robertshaw, P. T. and Collett, D. P., ‘The identification of pastoral peoples in the archaeological record: an example from East Africa’, World Archaeology, xv, i (1983).

48 Ibid. It is argued, inter alia, that some of the ‘axes’ are hoes.

49 Monod, Th., ‘Introduction’, in Monod, Th. (ed.), Pastoralism in Tropical Africa (London, 1975), 134.

50 D. P. Collett, unpublished data; Soper, R., ‘Iron Age Archaeology and traditional history in Embu, Mbeere and Chuka areas of Central Kenya’, Azania, XIV (1979), 3159.

51 M. C. DiBlasi, pers. comm.

52 If pastoralism is defined by ideology, then one may allow for economic and subsistence variation within a single cultural system of shared values and beliefs.

53 Ambrose, as cited in n. 11.

54 As has been documented, for example, in Maasai-Kikuyu relations in the nineteenth century; Waller, R., ‘The Maasai and the British 1895–1905’, J. Afr. Hist., XVII (1976), 429–53.

55 Oliver, Roland has written on ‘The Nilotic contribution to Bantu Africa’, J. Afr. Hist. XXIII, iv (1982); in a sense, what we are proposing here is a Bantu contribution to Nilotic Africa (though we would prefer not to make the archaeological correlation with Nilotic). Whereas we attribute changes in PN economy to the possibilities raised by the appearance of Iron Age farmers in the Eastern Highlands, Oliver (ibid.) attributes them to an influx of pastoral Nilotes replacing earlier Southern Cushitic food producers around the end of the first millenium A.D. The archaeological evidence for this period is all too scanty, and we put little faith in historical linguistics and less in glottochronology. However, even given population influx, specialised subsistence strategies of herd management are only possible where there is access to farming groups. Thus it is, in a sense, the Bantu, rather than Nilotes, who create the conditions for pastoral expansion.

56 Some of these predictions and the available archaeological evidence are outlined in Robertshaw and Collett, ‘Pastoral peoples’. The model will require elaboration and review as archaeological data accumulate.

57 Participants at a one-day seminar in Nairobi, November, 1982.

58 Salzman, Philip Carl, ‘Ideology and change in Middle Eastern tribal societies’, Man, n.s., XIII, iv (1978), 618637, at p. 635.

59 F. H. Brown, H. V. Merrick and J. P. S. Merrick, pers. comm.

60 Robertshaw and Collett, ‘Pastoral peoples’. Similarly, Maasai subsistence in the nineteenth century in southern Kenya was a reflexion of cultural values, environmental constraints, and economic ties with farming peoples.

61 Owen et al., ‘Palaeolimnology’; Barthelme, J. W., ‘Holocene sites north-east of Lake Turkana’, Azania, XII (1977), 3341; Butzer, K. W., ‘The Holocene lake plain of North Rudolf, East Africa’, Phys. Geogr, I, i (1980), 4258.Brown, F. H., ‘Barbed bone points from the Lower Omo Valley, Ethiopia’, Azania, x (1975), 144–9; Robbins, L. H., ‘Archaeology in the Turkana District, Kenya’, Science, clxxix (1972), 359–66; Robbins, L. H., ‘The Lothagam site: a Late Stone Age fishing settlement in the Lake Rudolf Basin, Kenya’, Michigan State Univ. Museum Anthropol. Series, I, ii (1974), 149216; Phillipson, D. W., ‘Lowasera’, Azania, XII (1977), 132.

62 Arkell, A. J., Early Khartoum (Oxford, 1949).

63 Sutton, J. E. G., ‘The aquatic civilization of Middle Africa’, J. Afr. Hist., xv, iv (1974), 527–46; Sutton, J. E. G., ‘The African aqualithic’, Antiquity, LI (1977), 2534.

64 Leakey, Stone Age Cultures; Cole, Prehistory; Ambrose, S. H., Hivernel, F. and Nelson, C. M., ‘The taxonomic status of the Kenya Capsian’, in Leakey, and Ogot, (eds.), Proceedings, 248–52.

65 Ambrose, et al., ‘Taxonomic status’.

66 Leakey, Stone Age Cultures; Oakley, K. P., ‘A bone harpoon from Gamble's Cave, Kenya’, Antiquaries' Journal, XLI (1961), 86–7.

67 Sutton, , ‘Aquatic civilization’, 533–4. He is supported by Phillipson who describes Eburran sites as ‘specialized fishing settlements’ (p. 435), which ‘presumably encouraged a more sedentary existence than was practicable elsewhere in the region at this time’ (p. 433): Phillipson, D. W., ‘The Later Stone Age in sub-Saharan Africa’, in Clark, J. D. (ed.), The Cambridge History of Africa, I: From the earliest times to c. 500 b. c. (Cambridge, 1982), 410–77.

68 S. H. Ambrose, pers. comm.

69 Pers. comm.

70 ‘Gumban A’ was renamed ‘Nderit ware’ by Bower et al., ‘Later Stone Age’. ‘Nderit ware’ has in turn been superseded by the ‘Olmalenge tradition’ (Collett and Robertshaw, ‘Pottery’).

71 Leakey, M. D., ‘Report on the excavations of Hyrax Hill, Nakuru, Kenya Colony’, Trans. Roy. Soc. S. Africa, xxx, iv (1945), 271409.

72 Sutton, , ‘Aquatic civilization’, 541.

73 Sutton, J. E. G., ‘Aquatic reflections’, in Leakey, and Ogot, (eds.), Proceedings, 321–2.

74 S. H. Ambrose, pers. comm.

75 Owen et al., ‘Palaeolimnology’.

76 C. M. Nelson, pers. comm.; Nelson, C. M., ‘Continuing work at Lukenya Hill’, Nyame Akuma, XIV (1979), 2830.

77 Dates for Olmalenge pottery from Turkana District (Robbins, ‘Archaeology’) are questionable on the grounds of dubious association between the samples dated and the pottery; see Collett and Robertshaw, ‘Problems’.

78 Leakey, L. S. B., Stone Age Africa (London, 1936).

79 See Phillipson, , ‘Later Stone Age’, 432.

80 Deacon, J., ‘Wilton: an assessment after fifty years’, S. Afr. Archaeol. Bull., XXVII (1972), 1048.

81 Mehlman, M. J., ‘Excavations at Nasera Rock, Tanzania’, Azania, XII (1977), 111–18.

82 Mehlman, M. J., ‘Mumba-Höhle revisited: the relevance of a forgotten excavation to some current issues in East African prehistory’, World Archaeology, XI, i (1979), 8094.

83 Collett, D. P. and Robertshaw, P. T., ‘Early Iron Age and Kansyore pottery: finds from Gogo Falls, South Nyanza’, Azania, xv (1980), 133–45.

84 Robertshaw et al., ‘Shell middens’.

85 Originally described by Chapman, S., ‘Kantsyore Island’, Azania, ii (1967), 165–91. The distribution, affinities and dating of Kansyore have been reviewed by Collett and Robertshaw, ‘Early Iron Age’; Kansyore assemblages now fall within the Oltome tradition (see Collett and Robertshaw, ‘Pottery’).

86 Chapman, ‘Kantsyore Island’.

87 Sutton, , ‘Aquatic civilization’, 543–4; Sutton, ibid, states that Kansyore should be seen against a background of ‘the balkanization of a once prestigious, self-assured Panafrican civilization’, of which Kansyore may be a local development.

88 Collett and Robertshaw, ‘Early Iron Age’.

89 Robertshaw and Mawson, ‘Excavations’.

90 Robertshaw, ‘Eastern Equatoria’.

91 Robertshaw et al., ‘Shell middens’.

92 Even the date of 2640± 120 b.p. (N-493) from Nyangʾoma (Soper, R. C. and Golden, B., ‘An archaeological survey of Mwanza region, Tanzania’, Azania IV (1969), 1579) accepted by Collett and Robertshaw, ‘Early Iron Age’, is now considered unreliable; see discussion of this and other dates in Robertshaw et al., ‘Shell middens’.

93 See, for example, part of a hut-floor revealed at Ngamuriak; Marshall and Robertshaw, ‘Preliminary report’.

A New Framework for the Study of Early Pastoral Communities in East Africa

  • Peter Robertshaw (a1) and David Collett (a1)


Altmetric attention score

Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed