Revisiting the irrigated agricultural landscape of the Marakwet, Kenya: tracing local technology and knowledge over the recent past

Abstract

This paper revisits previous work on the irrigated agricultural landscape of the Marakwet of Kenya and presents this alongside the results of new interdisciplinary archaeological and anthropological analyses that address the temporal dynamism of Marakwet farming technologies in relation to both endogenous and exogamous developments. In particular, it presents a major new re-mapping of the Marakwet irrigation system and higher-resolution analysis of irrigation practices at the village level. The paper argues that irrigation and farming shift across the landscape through time in relation to ecological and social parameters (and their associated timescales) and suggests that rural African farming systems are more complex than how they are represented in current literature. It shows how the Marakwet system has dynamically incorporated small-scale technical changes and increased in scale over the last 30 years while resisting wholesale technical and managerial change and spatial reorganisation. It further argues that historically contextualised interdisciplinary analyses, combined with extensive community collaboration, are essential not only for obtaining more complete understandings of specific farming systems past and present, but also for learning more broadly about the intersection (and regular mismatch) between local contextual knowledge and the approaches of external developers.

Cet article ré-examine les travaux antérieurs qui ont été faits sur le paysage d'agriculture irriguée du pays Marakwet (Kenya), et présente en parallèle les résultats de nouvelles analyses archéologiques et anthropologiques. Ces dernières ont été menées de manière interdisciplinaire et ont eu pour but d'examiner le dynamisme à travers le temps des technologies agricoles Marakwet, considérant à la fois les développements internes et ceux venus de l'extérieur. En particulier, nous proposons une nouvelle cartographie du système d'irrgation Marakwet et des analyses à plus haute résolution des pratiques d'irrigation au niveau des villages. Cet article avance l'argument que l'irrigation et l'agriculture se déploient à travers le paysage au fil du temps en réponse à des paramètres écologiques et sociaux (et aux canevas chronologiques se rapportant à ceux-ci), et propose que les systèmes agricoles ruraux en Afrique sont plus complexes que ne le représente la litérature actuelle. L'article montre que le système Marakwet a incorporé, de manière dynamique, des changements techniques de petite envergure et qu'il a par ailleurs pris des dimensions plus importantes sur les 30 dernières années, résistant, cependant, à des modifications globales au niveau de la réorganisation spatiale, de la gestion ou des techniques. L'article avance que les analyses interdisciplinaires historiquement contextualisées, combinées avec une proche collaboration avec les communautés, sont essentielles: non seulement pour atteindre une compréhension plus complète des systèmes agricoles dans le passé et dans le présent, mais aussi pour mieux cerner l'intersection (et souvent, le manque d'intersection) entre les savoirs locaux et les approches des développeurs externes.

Keywords:

• African farming
• irrigation technology
• useable archaeology
• applied anthropology
• Indigenous knowledge
• landscape
• Marakwet
• Kenya

Acknowledgements

We especially wish to thank the Marakwet community, particularly at Tot-Sibou, and the staff of the Marakwet Research Station (http://www.biea.ac.uk/research/marakwet-research-station/) who collected and interpreted many of the research data used here: Helena Cheptoo, Willy Chukor, Nelson Bailengo, Noah Kiplagat Rutto, Felix Krellkut Kiptoo, Sammy Kimwole and Florence Cheptum. We should also like to thank the Kenyan National Commission for Science, Technology and Innovation (NACOSTI) for research permission and the National Museums of Kenya for support and affiliation; the McDonald Institute for Archaeological Research, University of Cambridge, and the British Institute in Eastern Africa for financial and logistical support; and Mr Benson Kimeu, who processed some of the initial GIS map data. Finally, we should like to thank all of the other numerous students and researchers who have assisted and participated in the development of this research.

Notes on contributors

Matthew Davies is Leverhulme/Newton Trust Early Career Fellow at the University of Cambridge. He completed his DPhil. on Pokot farming at the University of Oxford in 2009 before taking up the post of Assistant Director of the British Institute in Eastern Africa between 2008 and 2010. He has been involved in a range of Eastern African research and currently co-directs the Marakwet Heritage Project (www.marakwetheritage.com) with Henrietta Moore. See www.md564.wordpress.955.com for more information.

Timothy Kipkeu Kipruto is Director of the Marakwet Research Station (www.biea.ac.uk/research/marakwet-research-station). He is a trained forester with over a decade of experience working on anthropological research in Marakwet. For the last four years he has been the focal point for the day-to-day operations of the Marakwet Heritage Project, as well as a major contributor to project design and data collection.

Henrietta Moore is a social anthropologist and Director of the Institute for Global Prosperity at University College, London (www.bartlett.ucl.ac.uk/prosperity). As well as numerous contributions to anthropological theory, gender studies and African studies, she has worked with the Marakwet of Kenya for over thirty years. See www.henriettalmoore.com for more information.

Notes

1. These also often assume a ‘punctuated’ rather than diachronic temporal model of historic change driven by certain prime movers (population growth, climate change etc.). See Davies (2014: 189–193) for greater discussion of this point.

2. Itself internally diverse.

3. See Adams and Carter (1987: 1) for a broader discussion and references relating to the failures of large-scale irrigation schemes across Sub-Saharan Africa.

4. Such trends are rarely properly quantified and partially relate to popular trends and assumed models of change (see Note 1 above). To give a short example, in the late 1990s the environmental concern in Marakwet was soil erosion caused by irrigation (Adams 1996; Adams and Watson 2003). More recently the concern seems to have shifted to crop failure caused by a lack of irrigation (Kipkorir and Kareithi 2013), while our own, admittedly anecdotal experiences of environmental concerns as perceived by the local community largely hinge on climate change, water and deforestation.

5. We have deeper concerns about the scale of ecological changes being implemented by this and other proposed development schemes that may adversely affect the indigenous system. We shall explore the ongoing history of outside development interventions in Marakwet in future articles.

6. Exact figures are difficult to obtain, Elgeyo-Marakwet District (shared with the neighbouring Elgeyo) held 380,000 people according to the 2009 census, but many Marakwet live in neighbouring districts.

7. Use of the term ‘furrows’ (or ‘hill furrows’) has a long history in Marakwet going back at least to Henning (1941), who used the term with a rather dismissive colonial attitude. Nevertheless, the term has been used by multiple researchers since and is well established in the literature on Marakwet and for pre-colonial irrigation in Eastern Africa more broadly. Elsewhere (especially in the Americas) similar water channels might be referred to as canals (see Widgren 2014 for a fuller discussion of these points). The term ‘furrow’ is retained here since this is the English term used by most Marakwet to refer to these channels. The term furrow (and ideas of ‘channelling’ water) also suit our understandings of the flexibility and impermanence (yet resilient) shifting nature of the water management system (see below). It should also be noted again that these human-made water channels are not used solely for agricultural irrigation but are also employed for a wide range of domestic purposes.

8. Future research will detail the great variety of crops grown including some fifteen different cereal varieties, among them finger millet and sorghum (Jones and Shoemaker, pers. comm.).

9. All team members (see our Acknowledgements) have worked on or around anthropological research in Marakwet for a decade or more and are consequently experienced in and motivated by obtaining and communicating knowledge about the community. In the most recent phases of research the team has been working on full time salaries for 6–8 months each year with team supervisors employed year-round. At the same time the team's members remain active local farmers and are offered flexible working hours. We believe that flexible but guaranteed and largely full-time occupation (not just when the ‘researchers’ happen to be there) is paramount in terms of generating a sense of local research ownership. We hope to continue to develop and improve the nature of these working relations and expand upon them in future articles on research methods.

10. Numbers in brackets after furrow names refer to the corresponding numbers on Figure 3 and in Table 1.

11. There are at least thirteen distinct ways of obtaining access to rights in land, not all of them through the patriline. These will be discussed in later papers.

12. Although there are numerous ways to loan or ‘purchase’ water from neighbours, something especially important for female-headed households (Adams et al. 1997).

13. Watson et al. (1998) put the total area of irrigated land at 4000 ha based on their analysis of satellite imagery.

14. Prior to the colonial period it cannot be said that the Marakwet were a single community and today the Marakwet clans retain detailed memories of separate and divergent histories leading to their settlement in their present locations. The term ‘(proto-) Marakwet’ is used here as a shorthand for these diverse communities who lived in the region in the century or more proceeding colonial rule (see Moore 1986).

15. Each of these branches is intricately named based on a range of criteria including features of the local area, the names and characteristics of those who use the branch and other stories related to the people or area. Such naming hints at an extremely detailed knowledge of the local environment and a high degree of place-making through naming and associated stories and narratives.

16. If this figure is scaled-up across the entire Marakwet irrigation system then we would estimate the total network of irrigation at any one time as being around 3000 km long.

17. By ‘semi-permanent’ we here mean that both field boundaries and associated secondary irrigation branches hold a certain degree of permanence over many years, probably extending to decades. However, we also recognise that these branches and field boundaries can and do change over longer timescales and that these changes are at least partially governed by human life and generation cycles as families and lineages grow and decline. See Davies (2014) for a discussion of these inherent temporalities among the Pokot.

18. Through extensive discussion and on-the-ground mapping with community elders we have produced an outline of these shifting patterns of cultivation back to 1960. These will be discussed in more detail in future articles.