Theorising the Majimaji – Landscape, Memory and Agency

The Majimaji was a war of resistance against German colonial rule in Tanzania which occurred between 1905 and 1907. The war is largely known from historical sources which include the German observers of the war, African historians and Africanist historians. Very few archaeological researches inform the Majimaji war. Although the materiality of the war exists, the landscape and memories of the war create a potential database for archaeologists. This paper theorizes the Majimaji war from landscape, memory and agency perspectives. In a broad sense, the paper delineates concepts that define conflicts and the landscape of conflict; agency as a broad theoretical framework on which the paper is grounded; and the processes of memory, memorialization and the creation of the war memorials.

Different regions have various criteria for differentiating battles from ambushes, skirmishes, riots, massacres or other kinds of violent events or encounters.For instance, military historians tend to discuss battlefields in terms of terrain (Doyle and Bennett, 2002:1).To them, the battlefield is composed of hills, ridges, rivers, and forests all of which may influence the disposition and movement of troops and in doing so play a vital role in determining victory or defeat (Pollard, 2002:123).On the other hand, the USA battlefields are deemed suitable for preservation on the basis of their historical importance while the English Heritage has determined that for a battle to be included in its official Register of Historic Battlefields should fulfil the following criteria (Carman, 2005:217):

•
It must not represent a lesser form of engagement, such as a skirmish; • it must have involved recognized military units, and thus not riotous crowds or other disorganized groups; • it must have had political, military, historical or biographical significance; and, • it must have been fought over a definable geographical space.
The aforementioned aspects of battle seem to be grounded on obvious rationality and accumulated historical focus of the battles as epitomes of war and even philosophical attention.The same formalities were used by the German administrators (e.g.Götzen) to label the Majimaji war a 'rebellion', a 'revolt' or an 'uprising' (Gwassa, 1973).While Tanzanians on the other hand, have regarded the Majimaji as a war, an active resistance and actual battle that occurred on Tanzanian soil (Lawi, 2009).The non-recognition of an event as a 'battle' is of minor importance here.It does not impair historical or archaeological studies to be conducted on it.
This paper regards southern Tanzania as an historic Majimaji battlefield and a landscape of resistance by virtue of being a place where the war was fought.War is an ultimate place maker (Snead, 2008:137).The battlefield landscape is fashioned by human agency and is constantly in a relationship with them (Tilley, 1994:23).The landscape is construed here as a cultural construct, inhabited and given meaning by people (Pollard, 2002:123).It is conceptualized, experienced and remade by those who inhabit it.The argument that Southern Tanzania is the Majimaji battlefield landscape considers the entire cultural history of the landscape, not simply the few hours or days during which the battle was fought.This is because the Majimaji is not the only human activity that happened in Southern Tanzania.Thus, landscape represents a certain aspect of history and mutes the rest, especially in the absence of multi-vocal interpreters (Snead, 2008:140).

The Concept of Battlefield
This paper refers to concepts of 'battlefield' 'battlefield site' and 'battlefield landscape'.There are qualitative differences between these three terms.A battlefield is a place where a battle is being (or has been) fought (Pollard, 2002).Battlefields can be categorized into two types: battlefield site and battlefield landscape.The manner of marking a battlefield site is also quite different from that of a battlefield landscape.Traditional battlefield sites interpretation is often limited to historical data augmented occasionally by knowledge based upon haphazard relic collecting and providing largely static or synchronic perceptions of battlefield sites behaviour.A battlefield site is most often considered to be a 'historic place', rather than an 'archaeological site' (Carman 2005:215).
Archaeologists define an archaeological site as a locus of culturally interpretable material (see Pikirayi, 1993:48).When this definition is employed on a battlefield site, the principle characteristic would be the availability of material across an area of land, constructed features or places which can be interpreted.Place is very fundamental because some battlefield sites lack artefactual materials but the 'place' in its own can be interpreted archaeologically (Snead, 2008:138).Similarly, a battlefield site boundary can either be a physical barrier or merely a conceptual border.A physical battlefield site border can be marked in by placing an object near it.
Battlefield landscape is an expansion beyond an individual battle site to a more comprehensive distribution of human traces in and between specific places of special interest.Employing battlefield landscape approach addresses the complex issues of the ways that people have consciously and unconsciously shaped the physical environment around them (Fennell 2010;Clark et. al. 1998;Tilley 1994).These include routes that brought soldiers into the battle space and the movements of troops, both victorious and defeated, to and from the place of battle (Carman 2005:219).Battlefield landscape boundary is merely conceptual and usually intangible (Fennell, 2010:3-4).Landscapes flow and are not bounded except by the barriers to human vision.As landscapes, battlefields cannot be conveniently bounded and cannot have edges placed to them (Carman, 2005:219).The landscape has long been central to archaeology as the context within which sites and monuments are preserved and as a long lived dynamic entity deserving explanation (Clark, et. al.1998).Terrain and geology which underlies battlefield landscape have helped to shape the cultural iconography of battle (Doyle and Bennet, 2002).The landscape has a profound effect upon the strategy and tactics of any military engagement and has consequently played an important role in determining the battle history (Kimble and O'Sullivan, 2002).Consequently, battlefield landscape is not only germane to a better understanding of the phenomenon that took place over southern Tanzania during the Majimaji war with a maximum amount of detail but also what happened after the battle in recent times.

Theoretical Formulation
Within a framework for a contested African colonial battle like the Majimaji which is conceived by the coloniser as a resistance, heritage and memory can be very complicated.This is, firstly, because the making of memories is dynamic, and involves individual and collective remembrance.Secondly, battlefield heritage involves historicizing power and agency in formulating memorial heritages (Moshenska, 2009a:35).Lastly, a significant part of resistance and warfare behaviour is intertwined with questions of political power which does not immediately and directly generate material remains (Vencl, 1984:117).This makes it necessary to discuss the theoretical underpinning of the fieldwork methodologies, the diverse landscape and the conceived memory scale of the Majimaji battle.

Theorising Fieldwork
A successful archaeological fieldwork should be well informed by theory (Lucas 2001: 3).Theorized fieldwork contemplates the aims, methods and hypothesis about what the fieldwork should do and why (Shennan, 2007:39).When fieldwork is kept separate from the operations of theorizing and write up, the transformation of raw data into interpretation creates neglect and some data loss in the process (Lucas, 2001:13).Nevertheless, the understanding of the major theoretical trends, perspectives and approaches contributes to a well-informed fieldwork.Different parts of a region have different kinds of sites and different activities have taken place in different parts of sites, thus strategies of studying sites must be guided by theories to gain a view of internal variability (Shennan, 2007:38).
Theorizing fieldwork sees the whole fieldwork enterprise as part of the process of interpretation, not simply as data collection.When fieldwork is theorized, excavation becomes interpretation, reflexivity, multi-vocality and a blend of feature descriptions, diary entries and narrative discussion (Hodder, 1997).Theorizing the excavated data alone misrepresents the complex historical and social political interactions that condition every stage of archaeological research, from the naming and classifying of archaeological data through the writing of the interpretive explanations (Gero, 1996:254).It was very important to theorize the fieldwork processes for the Majimaji war research.While the methods used ranged from excavation, survey, ethnography and consultation of archival sources (written sources, photographs, pictographs and cartography) the interpretation of battlefield sites depended on two different authorities; archaeologist (researcher) and communities.The presentation of the communities' point of view, for example, was theorized because communities give meaning to sites in the absence of the archaeologists' interpretation, thus, archaeologists have something to learn from the communities' perspectives (Shannan, 2007:42).This necessitates that alternative interpretation be well grounded, not only on data from the site, but also the theories that underpin the research.Communities' are the core of any archaeological research and thus they must be included in the research design.The objective here is to bring the research institutions and the communities together and share the experience of uncovering the past (Pikirayi, 2007:307).

Theorising Landscape
In studying the Majimaji resistance, the war landscape offers an opportunity to examine both the physical and symbolic aspects and how environment and landscape were a socially construed phenomenon.There is a connection between the landscape, the person and the place, the past and the present (Brown and Bowen, 1999:259).Given the fact that the Majimaji war covered a vast and diverse region, with different landscapes and environments, it is important to understand the range and scale of human social formations and agency over the landscape in accelerating the Majimaji war.In looking at any landscape, it is clear that some places are or have been more important than others (Aston, 2002:44).This suggests that there is a hierarchy of places; meaning that a particular place has a certain status in the local community than the other (Pikirayi, 2009).
Two places were very significant agents in terms of the organization of the Majimaji war: chiefs' residence and traditional healers/shrines.In principal, the locally or regionally important centre fulfilling a number of functions for the settlement around is always of high status in the local settlement hierarchy and usually occupies a central position geographically (Aston, 2002:44).However, these landscapes are foci for individual activities rather than a collective range of functions.The chief's residence, for example, was an agent and a venue for the distribution of the Maji (medicinal water).Planning and strategies for the war were also made on this residence.The choice of the chief's residence was very important so as not to raise suspicion with the Germans.As the Germans knew that the locals had to pay tribute, offerings and held celebrations at the chief's residence, they were unable to detect the preparation of the war.
Other significant agents over the Majimaji landscape were the ritualists' residence and shrines.Although Kinjekitile Ngwale was the principal ritualistic leader and his residence the centre for the ritualistic activities, communities subscribed to their own religious leaders and shrines during the Majimaji.This was mainly due to two reasons; firstly, Kinjekitile was arrested and killed shortly before the beginning of the war; secondly, the war spread over a too large an area with differences in culture and religious practices.Throughout, the war was guided by religious ideologies and thus the spiritual leaders and shrines were important places over the Majimaji landscape.At the end of the war, however, the Germans carried out military expeditions to round up healer 'zauberers' and chiefs and even offered to those who brought a spiritual leader or chief to their military stations (Monson, 2010:36).Theorizing the landscape, therefore, brings into observation these local centres of power as agents of the war and the role of individual places such as the chief's residence or religious/ritualistic places in order to differentiate them from the collective battlefield landscape.
Theorizing landscape cannot be done outside the local knowledge system (Lightfoot et. al. 1998;Stewart-Abernathy, 2004).Worldwide, the incorporation of indigenous view of past landscape uses has often been used to verify existing theories based on objective observations of the archaeological record (Spencer-Wood and Baugher, 2010).Traditional knowledge provides archaeologists with essential information for locating and interpreting both individual archaeological sites and the larger social, settlement they reflect (Dornan, 2002).Schmidt (2010:256) argues that mnemonic devices such as royal drums, landscape features and ancient shrines are crucial for understanding the archaeology of Africa.They aid in understanding the rhythm of daily life, local technologies, oral traditions, traumatic memories, cosmologies and language.On the other hand, paying closer attention to traditional knowledge may lead to challenges of those theories or at least offer alternative explanations about landscapes.This means that landscape-scale studies are not only a useful heuristic tool for archaeologists but also archaeological sites cannot be divorced from their larger environmental and cultural settings.

Theorising Agency
There is little consensus about what agency actually means (Dobres and Robb, 2000:1).Anthropologists consider the concept of agentive practice as neither a theory nor a method in itself, but a symbol, in the name of which a variety of theories and methods are being developed (Dornan, 2002:304).Within an archaeological context, agency as being accepted to be the way in which societies structures; inhabit and empower agents; those agents' aims, ideals and desires; and the material conditions of social life (Dobres and Robb, 2000:8).Agency was employed under this study basically to articulate the roles of individual and collective agents in the making of the resistance.Social agents are viewed not as omniscient and often impractical people but interactive between structures in which agents exists and paradoxically which they create (Dobres and Robb, 2000:3;Jones, 2002:176).
Agency emphasizes the human scale.Contradictions and conflicts are worked out, lived through and resolved at this level (Hodder, 2000:26).Human agents shaped past social and natural environments and were also constrained by the same (Pauketat, 2000:130).The use of agency theory allows articulation of the role of individual agents in creating their material world and expressing their discontent or resistance to the dominant culture (Shackel, 2000:243).Paying careful attention to the material practices and agents of the Majimaji battle constitutes an interpretive strategy.It provides room to studying individual's actions and provides deep meaning of the war actions, people and landscape.Ethnographies and photographs, for example, constitute evidence in the hands of people; who in turn could speak for the individual and collective memories of the war among the Majimaji communities today.
Another important war agency was the war medicine (maji) for which the conflict is named.Historians have referred the medicine to have both instrumental and symbolic efficacy of ritual actions that promised invulnerability and as a means to bring people together into a mass movement (Monson 2010:36).In their efforts to identify and punish those who had supported the uprising, the Germans used medicine as the primary source of evidence: the determination of guilt and punishment by death was based on whether an individual or a group had accepted maji (Hassing 1970:387).The archaeological ethnography conducted during this research among local communities in Nandete emphasized that maji was the source of the war prowess; meaning no maji, no war.The association of agency with maji is very essential.Understanding of the Majimaji events requires broad-based information about the actual dynamics of social agent's practices, shared values, and those of historically constituted agents.Thus archaeology must attempt to provide the broadest possible context based on multiple sources of data that looks at as many spheres of life as possible and thus draws interpretive force from the contestation of these domains (Dornan 2002:325).

Memory, Memorial and Memorialisation
Memory, memorial and memorialisation are three important terminologies attached to memory (Keremedjiev, 2013;Portelli, 2006;Hodgkin and Radstone, 2003;Kansteiner, 2002).Although they may sound confusing and often times standing for the same term 'memory', each of the three words present a unique form of meaning.Memory is concerned with the ability to remember information, experiences, events and personalities (Moshenska, 2009a;Monson, 2000b;Nora, 1989).It is a cognitive device that, while used by particular individuals, can be understood only as a social process that catalyse emotions, senses, participation, pain, joy and togetherness (Aguilar, 2005:60).A memorial is an object created for purposes of remembering a person or people who have died in a particular place.It is normally a statue or a stone or a building or structure constructed in order to remind people who died in an important past event or a famous single person's death (Castell & Roura, 2002:260: Battaglia, 1992:3).A memorial can also be a landscape without any human-made feature but recognized by the community as a marker of a particular event.Memorialization is the process of remembering (Green 2004;Lucas 1997;Roberts and Roberts 1996).The act of creating agents that will continue to exist and remind people of somebody who died or something that has gone.An example of memorialization is commemoration ceremony (Domansky, 1992).

The Majimaji Memory
The Majimaji memories can be assigned to three levels of agents: the individual, the social group and the collective memories.The individual Majimaji memories are those formed by remnants of families of the leaders of the war, the known warriors and Maji (the Majimaji war medicine) agents.The social group's memories involve villages, districts and the de facto regional scale.At this scale, memories have also involved Islamic and Christian religions fighting over remembering and forgetting of the heroes of the war.Collective memory involves the teachings of the Majimaji as part of the national history project, creation of commemoration days, activities and rituals.The evaluated war memories under this research were two folds: pedagogy and 'out of pedagogy'.The pedagogy memories involved interacting with secondary school students from Luegu Secondary School while out of pedagogy memories were collected from Nandete where the war originated.
Luegu Secondary is a ward school located in the Luegu ward of Namtumbo district, Ruvuma region.It is within the vicinity of the famous Majimaji battlefield -Namabengo.The Form Three class which had 24 students; 13 male and 11 female, was used for assessment of the formal-classroom Majimaji memories.Among them, 22 were Ndendeule, one Ngoni and one Matengo by ethnic group.In Tanzanian history pedagogy, the Majimaji is one of the topics in both primary and secondary schools syllabi.The learner centred approach was used to evaluate the student memories of the Majimaji war.The discussion centred on the meaning of the war according to student memories.The students defined the war as a resistance against colonialism, however, a number of shortfalls occurred in their definitions.The most serious ones were like Majimaji was the war of the Matumbi (instead of majority of the ethnic groups of southern Tanzania and some of the eastern ethnic groups); Majimaji occurred in Southern Africa (instead of east and southern Tanzania); and Kinjekitile used water from Ruvuma River to make maji (instead of water from a confluence of Ngarambe and Namang'ondo rivers).The Ndendeule participated in the war but only 2 students out of 24 had an idea that they did.Moreover, the Majimaji Museum is found in the Ruvuma region where Luegu Secondary is located.Yet, only five students out of twenty-four have ever visited the museum.
Nandete is predominantly inhabited by the Matumbi people.Historically, the Matumbi were unique in how they preserved historical legends of their communities' through makolo (literally history).An expert whose title was also makolo derived much pride and satisfaction from a correct production of historical accounts.This was a challenge narration usually important at informal gatherings as part of entertainment (Gwassa, 1970:20).Currently, the makolo culture is dying.What has replaced makolo in preserving the Majimaji memories in Nandete is the Baraza la Majimaji (Majimaji council).Baraza la Majimaji comprises of old men and women as well as the young girls and boys under the leadership of the village chairman or Ward Executive Officer (WEO).The local participants of the council are chosen by merit; the knowledge of the war or the grandchildren of the prominent warrior and original makolo are preferably used as council members.The village chairman or ward executive officer are government employment posts.While keeping of memories under the local government level brings a form of centralization and official record keeping, the majority of these government employees come from outside the Majimaji zone and do not have enough knowledge of the war and the cultural history of the people they are entrusted to lead.The strength of the Baraza often depends on these officials.If they like history, they acquaint themselves with the culture of the place and even support cultural activities in the area they are leading.Lack of historical interest often result in weak Baraza as well and this has affected the Majimaji memory keeping processes.

The Majimaji Memorials
There are very few Majimaji memorials in the region.A survey of 26 Majimaji war battle sites in Ruvuma, Lindi and Mtwara indicates only seven memorial obelisks of the war in Nyangao, Mikukuyumbu, Nandete, Peramiho, Songea, Kilwa Kivinje and Yakobi.Yet, the available memorial obelisks pose a great challenge in terms of their interpretation.With the exception of the Songea and Nandete obelisks, which are national monuments, the rest are missionaries' obelisks and pilgrimage sites for the purpose of commemorating individual or groups of missionaries who died in those particular areas.The obelisk at Nyangao, for example, was installed at the site of Sr. Walburga's death; Peramiho for Rev. Fr.Fransiskus and Mikukuyumbu for Bishop Cassian, the two reverend sisters and brothers who were all victims of the Majimaji war.The missionaries' obelisks single out particular people of 'importance' rather than all the people who died there.With these memorials, the complete picture of the war is obscured through the erasure of other participant and places significant to the war.
Further to this, the obelisk in Kilwa Kivinje is controversial.It is presented at an exact point where there was a hanging mango tree (mwembe kinyonga) during the German colonialism.The tree was cut to allow the construction of the memorial instead of keeping both the tree and the obelisk (see Laband and Thompson, 1983 on the Zulu war memorials).In addition, the names of the people under the list on the obelisk include those who died during the coastal resistances prior to the Majimaji war.One of them is Chief Hassan bin Omari Makunganya who was hanged in 1894.These errors are present at other sites too: the memorial obelisk at Kilwa is selective on who is celebrated and what is remembered.Muslims have secured the land around the memorial in order to build a commemorative mosque which may exclude non-Muslims.Whatever memorials will be constructed on this land, will be read as an official interpretation of the landscape.

The Majimaji Memorialisation
The Majimaji memorialisation is performed through the Remembrance Day ceremonies conducted in the Ruvuma region on the 27 th of February each year.The event is organised in remembrance of more than 68 leaders and warriors of the Majimaji war in killed in Ungoni by the Germans in 1905.The Songea Museum and obelisk is one of a strong evidence of the war and well commemorated not only by the state but also by the communities around it who have deep connections to it.As the most prominent memorial, the majority of Tanzanians have also narrowed the Majimaji commemoration to Songea.Nandete has a well-built obelisk but the remoteness of the area and poor infrastructure has meant that it not part of the Majimaji memorialisation.Ngarambi which is the site of Kinjekitile's home and source of the maji (the war medicine) has no memorial obelisk even though it is highly regarded by communities around it.Another form of memorialization involves the use of songs in commemoration of battlefield sites.Archaeological monuments and battlefield landscapes have attracted folklores and songs which are important sources of history in the monuments' continuing histories (Brown and Bowen 1999:255).Some of the Majimaji battlefield sites are remembered through songs that record the events and personalities.The also bring out the emotions associated with war.An example is the Muhuru battlefield site in Mweraland that is commemorated in the Mwera dance songs.(Symonds, 1999:123).In African societies, narratives like folklores, songs, idioms and riddles carry the structure of people's lives, actions, movements and use of the landscape (Schmidt and Walz, 2007).Songs can be deduced to the individual and collective agency coping with a particular landscape (Symonds, 1999:124).Although individual participants change and the song can be modified, they provide the structure of commemoration that has the ability to transcend generations (Symonds 1999:123) while emphasizing the past as an example of how people should live (Shackel and Gadsby, 2008:230).As Jamie Monson (2010:33) writes, the legends of the Majimaji are impressive, but the truth that can be gleaned from these legends is even more interesting.
My argument here is that communities involved in the Majimaji war are positioned to influence the landscape of the battlefield.These communities were the agents of activities over the battlefield landscape.The Majimaji battlefield is interpreted variously as a tomb, a shrine, a pilgrimage, heritage or military training landscape.The framework depicts that communities' agents and practices affect the status of the Majimaji battlefield to acquire this range of meanings and cultural significance.However, the landscape may also be affected by communities' memory, the creation of memorials and even memorialization of the events of the various battles fought.

Conclusion
The study of battlefield memory and memorials in the archaeology of different places, periods and social context requires different methodologies tailored to specific circumstances.This is because each war differs from the other in terms of participants, landscape and nature of the conflict.Battlefield archaeology needs a more intricate appreciation of memory and utilisation of the advantages of the public memory.In societies where monuments and sites are as readily accepted as symbols of traumatic memory, doing archaeological research will inevitably be regarded as reopening old wounds (Moshenska, 2009b).Therefore community engagement is of paramount importance in telling of which site to preserve, which ones to hold on to as sacred and which ones to utilize for purposes of memory and memorialization.Such approaches recognize the presence and value of both the tangible and intangible attributes of sites and landscapes (Munjeri, 2000).
The theory proposed here to interpret the Majimaji war archaeologically with the intention of reconstructing its identity emphasizes three basic components; agency, landscape and memory.Critical to all the three, are communities who have always engaged with the past in the process of establishing meaning in the present and routinely incorporate objects and places associated with remembered or imagined past events into the narratives that create and sustain them (Marshall, 2002).Such actions have been integral to people's lives long before archaeology was invented.Such community-based approach is important in integrating indigenous epistemologies into archaeological research in order to understand local perceptions of the world view and how things came to be.This is vital in the identification and elimination of historical bias in archaeology while making archaeology more relevant, to both its practitioners and to the communities where it is conducted.