Material assemblages and percolating pasts in Zigua households,north-eastern Tanzania

Abstract

The material culture and dwellings of the Zigua villages of Kwa Fungo and Kwengoma, in north-eastern Tanzania, bear the traces of complex social and historical dynamics. In this paper, we analyse household inventories and ethnographic interview data originally collected in 1991 by a team from the University of Dar es Salaam and the National Museum, Tanzania. We rely on oral histories as well as on Zigua epistemologies and ideas of percolating pasts to historicise and contextualise the processes that shaped the material world of these two village communities. The paper focuses on investigating the shift from round (msonge) to rectangular (banda) house-types, the household material changes generated by labour migrations and Nyerere’s Ujamaa, the materialisation of healing practices, and the formation of specific aspects of identity.

Introduction

In 1991, a team from the University of Dar es Salaam and the Tanzania National Museum collected archaeological and ethnographic data in Zigua villages in Tanga Region, NE Tanzania, as part of a study aimed at tracking transformations in material practices and household and settlement spatial organisation during the 19th and 20th centuries. The ethnographic data, which have not previously been analysed in detail, ¹ consist of household inventories, scale plans of house interiors and external compounds, records of craft activities, especially pottery production, and qualitative interviews. These data are supplemented by extensive records concerning house building techniques and materials, processes of mud-and-wattle wall maintenance, and observations on different stages of wall collapse and house abandonment as evident in three villages. The team also collected data on the history of settlement in the area, especially during the 19th century, and identified several former settlements, of which a few were subsequently excavated over a decade later (see Biginagwa and Lane, 2022). The guiding research objectives of the project were initially to understand changes and continuities in the organisation of household space and material culture over the last few hundred years, paying due attention also to processes of site formation and abandonment, so as to better link the ethnographic (source) and archaeological (subject) sides of the analogical equation. In conception, the original study can be best described as following a ‘direct historical approach’ informed by early post-processual ethnoarchaeological studies, including Paul Lane’s doctoral research on the intersection of space, time, and materiality in Dogon (Mali) households in the early 1980s (Lane, 1986).

Whereas the original project was distinctly ethnoarchaeological in orientation, this paper offers an analysis of the ethnographic data on households and household inventories that draws on more recent theoretical discussion of relationality and material agencies. Focusing on two Zigua communities, our analysis seeks to historicise their household material culture and architecture in light of the oral histories told by the project’s original interlocutors and relational Zigua and Shambaa epistemologies, as materialised in healing practices (Thompson, 1999; Walz, 2009, 2016). We draw also on insights taken from assemblage thinking and the significance of the percolation of time in the present.

Assemblage thinking (Braidotti, 2013; DeLanda, 2006, 2016; Deleuze and Guattari, 2014) contends that the world is never fixed but rather always emerging from a flow and flux between things, people, materials, plants and animals, and/or ideas. This concept has gained considerable popularity in archaeology in recent years (e.g., Crellin, 2020; Hamilakis and Jones, 2017; Harris, 2018; Hodder and Lucas, 2017; Lucas, 2012, 2017; Richard, 2019). While less used by archaeologists, the concept of the ‘percolation of time’ refers to how objects physically hold past temporalities and how multifarious pasts can therefore percolate in a single present (Olivier, 2000, 2015; Serres with Latour, 1995) and can aid understanding of the social and symbolic significance afforded to the coexistence of artefact types and forms of different ages within household assemblages (González-Ruibal, 2006; see also Lucas, 2015; Ray, 1987, for analogous theories about deep histories and multi-temporality). Aspects of both concepts resonate with various Indigenous epistemologies (Rosiek et al., 2019; Todd, 2016), including Zigua ones (Thompson, 1999; Walz, 2009, 2016).

Indigenous Zigua ideas about temporality as well as the concept of percolating pasts, highlight, in particular, how the reified division between past and present in Western thought (and much of modern archaeology) obscures how present configurations are the legacy of past relations. The ahistoricism of many older ethnoarchaeological studies derives from a similar distinction between past and present (Gosden, 1999; Gosselain, 2016), often concealing the source of past relations of exploitation or domination (such as colonial and neocolonial ones). The division also obscures the agency of local communities in negotiating these power relations and producing the contemporary material assemblages and entanglements in which they live. To illustrate these points further, this study endeavours to demonstrate how the household assemblages analysed here were all infused and informed by the histories and agencies of Zigua communities, individuals, and their material world at the time of the study.

Study area and methods

The data on Zigua household assemblages and material culture discussed in this paper were collected by a team from the Archaeology Unit and Department of History, University of Dar es Salaam (UDSM), the National Museum, Tanzania (NMT), and the British Institute in Eastern Africa (BIEA) in July and August 1991, led by Paul Lane. Fieldwork was conducted in an area of the Pangani Valley between Korogwe and Muheza, in Tanga Region, NE Tanzania (see Figure 1). Field research focused on three localities: Kwa Fungo and Ngombezi—both selected as settlements known from documentary and oral sources to have served as caravan halts in the latter part of the 19th century and where archaeological traces of these had survived (Biginagwa and Lane, 2022)—and Kwengoma, selected because it was identified as one of the few localities in the general study area where a ‘traditional’ Zigua type of cone-on-cylinder house (Lane, 1997) was still in use. The original intent to compare the ethnoarchaeological household forms and inventories with those recovered from excavation of the 19th-century settlements at Kwa Fungo and Ngombezi was curtailed, unfortunately, by a temporary ban on all archaeological fieldwork in Tanzania at the time of the original fieldwork. More recent work on archaeological materials from both sites (Croucher, 2006; Lane, 2011, 2015) and the wider region (Walz, 2010; Biginagwa and Ichumbaki, 2018) have partly rectified this, but larger area excavations at a selection of sites aimed at revealing complete house plans have yet to be undertaken.

Figure 1.

Map of the study area and location within Tanzania.

Here, we focus specifically on just two of these localities: Kwa Fungo and Kwengoma. Kwa Fungo is located in Muheza District, while Kwengoma—sometimes spelled Kweingoma—is in Handeni District. At the time of the survey, both were nominally Zigua villages, although Kwa Fungo is located outside of Zigualand (the historic area of the Zigua), in what would have been considered Bondei territory in the immediate precolonial era (for discussions of the fluidity of such ethnic labels, see, inter alia, Willis, 1992, 1993a). Twenty compounds in Kwa Fungo were mapped and fully inventoried, representing c. 18% of the total number of occupied compounds, and six in Kwengoma, representing the entirety of the settlement at the time of the survey. In total, more than 3000 items from the handwritten room inventories of Kwa Fungo and Kwengoma were digitised, categorised, and tabulated to form a room-by-room inventory. Twenty-nine oral history interviews were then consulted to transform these static inventories into richer, diachronic, and historicised accounts of the Zigua and their material culture.

A survey of all types of containers in 30 households in Kwa Fungo was also conducted, supplementing records for a similar number of households in Ngombezi, where an additional 20 compounds were also mapped and inventoried. These latter datasets have yet to be fully analysed but would offer considerable scope for comparison with the more recently excavated assemblages from Ngombezi, Kwa Sigi, and other sites along the Pangani (Biginagwa, 2012).

Zigua and village histories

The Zigua are a Bantu-language speaking people mostly found in the space historically referred to as Zigualand, although the sites of Kwa Fungo and Kwengoma are located in a borderland area. In the 19th century, Zigualand underwent rapid socio-political changes in response to the expansion of the caravan trade in elephant ivory and enslaved people (Giblin, 1992; McInneshin, 2009; Lane, 2011). The region was at the time under the overlapping spheres of influence of local polities, until it was brought under the colonial authority of the Deutsch-Ostafrika Protectorate (1885-1919). The period was one of widespread religious conversions through the enterprise of Christian missions (Willis, 1993b), while Islamisation also spread inland as a form of resistance to German colonialism (Glassman, 1995; Sperling and Kagabo, 2015). After the defeat of German forces in 1916 during World War I, the British assumed control and maintained their colonial authority via processes of indirect rule until 1961 (Willis, 1993c), when Tanzania gained its independence and Julius Nyerere was elected as the first independent leader (Ekemode, 1973; Giblin, 1992).

Nyerere’s socialist program, Ujamaa (‘familyhood’), was aimed at improving socio-economic development by promoting a communal way of life. One of the policies constituting Ujamaa was ‘villagization’, a large-scale resettlement of rural populations into villages (Schneider, 2004; Scott, 1999). Starting as a voluntary scheme, it fell short as people were often unwilling to move away from their existing homes, prompting the government to transform villagization into a compulsory mass re-settlement scheme conducted between 1973 and 1976 (Coulson, 2013; Nursey-Bray, 1980; Schneider, 2004; Scott, 1999). Ujamaa was eventually abandoned in the mid-1980s by Nyerere’s successor, Ali Hassan Mwinyi (Sanders, 2001; Tripp, 1997). It was in this context that the team from the University of Dar es Salaam conducted its ethnoarchaeological survey of Zigua villages.

The development of the settlements of Kwa Fungo and Kwengoma is deeply tied to this regional history. Kwa Fungo was founded in the 1860s-70s when Fungo, a Zigua from Handeni, settled in the area followed by some relatives (hence why the settlement is considered Zigua despite being in Bondei territory) (JOF, 1991; interview). The site, which reportedly was surrounded by a wooden palisade, grew to become a small caravan halt of which Fungo became the jumbe (headman or chief) (Ekemode, 1973; Lane, 2011). After the death of Fungo and partly because of nuisances and crop damage caused by monkeys, the site was abandoned in the 1920s. It was repopulated again in 1935 as the plantation of rubber started in the region. However, the villagers that returned in 1935 were not related to Fungo, and were mainly labourers seeking work on the rubber and sisal estates (JOF, 1991; interview). In the first half of the 20th century, cash-crop estates for rubber and sisal production were established in the Pangani Valley, coinciding with the promotion of wage labour by British colonial administrators, encouraging intra-Tanzania labour migration (Giblin, 1992; Sabea, 2008). The old site of Kwa Fungo was once again abandoned in 1976 during villagization, to be moved to its present location (JOF, 1991; interview; Lane, 2011). Similarly, Kwengoma was also probably created during the 19th century, given that one family had lived on the site for five generations (SS, 1991; interview). However, except for this family, all the other households in Kwengoma reported arriving in 1976 during Ujamaa. As with many other villages created during Ujamaa, Kwa Fungo and Kwengoma housed communities displaced from both neighbouring and more distant areas, and were mostly oriented toward agricultural production.

In 1991, farming and casual wage labour on the neighbouring sisal plantations, sometimes complemented by craft activities or beer (pombe) brewing, were still the main occupations of most inhabitants of Kwa Fungo and Kwengoma. This was also reflected by a strong relative equality among the inhabitants. No data on household incomes is available from that period, but a Gini coefficient—a statistical measure of dispersion ranging between 0 and 1 used to represent wealth inequality within a population (Gini, 1921)—was calculated based on the number of items per conjugal unit. This was chosen as a proxy for wealth instead of the size of the house, in line with comparative literature (Cutting, 2006: 241). The Gini coefficient obtained for Kwa Fungo was 0.17 and for Kwengoma 0.19. These values indicate a high degree of wealth equality between households, as coefficients under 0.2 are generally considered as showing high or perfect equality (UNICEF, 2018: 35). The low Gini values of Kwa Fungo and Kwengoma are not surprising at the level of a settlement, especially in the wake of Ujamaa and given the similar occupations of the inhabitants.

The impact of migratory labour and villagization

By the early 1990s, two key political-demographic processes had left a long-lasting mark on the settlement organisation and demographic composition of Kwa Fungo and Kwengoma: the labour migrations of the 1950s and villagization in 1976. These historical events led to, or accelerated, widespread material changes, including some of an architectural nature.

These mobility dynamics were indeed catalysts of a shift in house-form that happened throughout Zigualand. The main pre-colonial house-type of the Zigua was the msonge: a round, cone-on-cylinder type of house composed of an inner and outer circle. The inner circle (kugati) was used as a sleeping and cooking area, while the outer was often divided into distinct sections used as additional bedrooms, lounge rooms, or to keep sheep and goats (Lane, 1997) (Figure 2). However, by the time of the survey, these had widely been replaced in Zigua villages by banda types, which are rectangular rather than circular in form, and only a single extant msonge was recorded in the two villages (at Kwengoma). The term banda includes ‘modern’ Swahili-style houses—the most frequent house-type in Kwa Fungo in 1991—rectangular in shape and derived from a style of Swahili house built on the East African coast since the 10th century (Fleisher, 2015). They are characterised by a central hall with rooms on each side, and a veranda or kitchen adjoining the entrance (Kruse and Torstensson, 2011; Nguluma, 2003; Vestbro, 1975). Large, L-shaped or U-shaped houses, sharing many organisational similarities with the Swahili banda, were also present in Kwa Fungo (Figure 3). Other types of banda present in both Kwengoma and Kwa Fungo (Figure 4) included simple rectangular houses with, in Kwengoma, a variant of these, the kimbingo (rectangular with round corners, see Figure 4(b)), being seemingly very popular and perhaps representing a transitional form between circular and rectangular types. Previously, circular single-room dwellings (tutu or bweni) were also built next to the msonge to house unmarried offspring, or were built by young men moving out of their parental home (Lane, 1997). Although a single circular bweni was still present in Kwengoma, this form had mainly been replaced in both settlements by rectangular-shaped outhouses, still called bweni, which comprised a single bedroom and an entrance.

Figure 2.

Example of one of the house plans from Kwengoma. Drawing credit: Birgit Uenze.

Figure 3.

Examples of the main house types found in Kwengoma and Kwa Fungo. Drawing credit: Elias Michaut, re-drawn from original field plans.

Figure 4.

(a) Two rectangular houses in Kwengoma. Photo credit: Paul Lane. (b) A kimbingo in Kwengoma in the process of being built. Photo credit: Paul Lane.

In 1991, banda houses were the predominant shape of dwelling in Kwa Fungo and Kwengoma, but their overwhelming presence bore testimony to a variety of recent processes. This shift from msonge to banda has been noted previously by other authors, and presumed to be linked to villagization (Biginagwa, 2012; Lane, 1997). While both msonge and banda coexisted in Kwengoma, only banda were present in Kwa Fungo. However, this had not always been the case, and at the end of the 19th century, there were only msonge in Kwa Fungo (JOF, 1991; interview). Some informants held that the banda style was introduced by the Nyamwezi (AS, 1991; interview), many of whom served as porters in the caravan trade (Rockel 2000), while others said that it was by Swahili from Pangani after World War I (YM, 1991; interview). In either case, it appears from the interviews that the process of replacement of the msonge had already started in the first half of the 20th century, but took a faster turn following labour migration and villagization. As JOF (1991; interview), a grandson of the headman Fungo, put it, the decline in msonge accelerated after 1947 was because ‘people were moving around in search of work on estates and so were exposed to new ideas’.

The multi-ethnicity and multilingualism of Kwa Fungo in 1991 were testimonies of this influx of new populations. According to the oral history interviews, 15 ethnic groups were present in the village in 1991: the main ones were Zigua and Bondei (the two ethnic groups native to the area), followed by families from around the Pangani Basin (Shambaa people) and from southern Tanzania (from the Bena, Pangwa, Yao, and Ngoni people). At least six languages were used in Kwa Fungo, with a clear phenomenon of diglossia (with Swahili being used in public circumstances, while local languages were mainly spoken at home, for instance when informants named objects around their house). Most of the families had settled in the village either in the 1950s, when labourers from around Tanzania came to work on the rubber and sisal estates, or in 1974-76 during villagization. The ethnic diversity of Kwa Fungo was therefore directly tied to internal migrations and was probably one of the main causes behind the change in architectural taste. On the other hand, in Kwengoma where a msonge was still present and in use in 1991, all the inhabitants declared themselves Zigua, although the majority still arrived in 1976 from other localities in the lower Pangani Basin and Handeni District.

The shift in house-forms was therefore clearly accelerated by the effects of the 1976 villagization—which, by re-settling populations, also forced people to abandon old dwellings, potentially favouring new styles. In Kwengoma, the generational character of this change was emphasised, with an interviewee noting that ‘old people prefer this style [msonge] but the young generation do not like it’ (EP, 1991; interview). This perhaps was truer in all-Zigua villages like Kwengoma than in others like Kwa Fungo where many non-Zigua inhabitants had never lived in msonge.

Informants also advanced material considerations to justify this change, or conversely to explain some of its obstacles. Many noted that banda are simpler to build, and require less building material and maintenance (perhaps favouring their building during the Ujamaa resettlement). On the other hand, the persistence of a msonge in Kwengoma was linked to previous building knowledge and skills: the msonge inhabitants had lived in a msonge before, and therefore possessed knowledge about its construction.

Finally, this shift in house-forms almost certainly generated changes in space use. Obvious examples are the gendered eating-spaces, which moved from separate rooms in the msonge to designated spaces at opposite ends of the halls in Swahili-type banda. The devices for storage changed too: lofts being less common in the banda, their inhabitants resorted to cereal storage in baskets or on elevated wooden platforms. These examples thus point to how the abandonment of the msonge led to the spatial reorganisation of daily activities conducted within the household, but often without directly changing underlying behaviours (e.g., eating spaces remained distinct).

In summary, villagization and previous labour migrations, with the widespread mobility they brought about, were crucial in setting off the architectural shift from msonge to banda styles. This happened according to personal and generational tastes but was also limited by building skills. In 1991, villagization and labour migrations had impacted and durably shaped the assemblages that people formed with their houses, daily activities, use of language, and everyday material world.

Percolating pasts in substances and ‘modern’ items

Substances such as plastic, aluminium, or enamel, or ‘modern’ items such as radios, were all present in the 1991 household inventories. However, their adoption was a recent fact, tied to the labour migrations, villagization, and the increased incorporation of Zigua households into global markets.

The vast majority of the crops cultivated in the compounds of Kwa Fungo and Kwengoma (coconuts, tomatoes, papayas, pumpkins, or sweet potatoes) were all of exogenous origin, introduced either via the long-established transoceanic Indian Ocean trade, or by European settlers and colonial authorities. Maize was the main staple cereal cultivated in the local shamba (farms) in 1991, having replaced African domesticates such as sorghum and pearl millet during the 19th century in the Pangani basin (Dale, 1896; Maundu and Imbumi, 2003; Raschke and Cheema, 2008). Similarly, sisal, used from the 1940s as the main cash crop on the estates surrounding Kwa Fungo, was brought from Mexico to Tanzania in the 1890s by German planters (Hitchcock, 1959), who established sisal plantations along the lower reaches of the Pangani at an early date.

As a result of their ‘indigenisation’—the process by which selected alien products become staples of local cuisines (Dietler, 2006, 2010)—many of these exogenous plants are today central to East African foodways. The main example of this is ugali, perhaps the most common dish in Tanzania, which is prepared from maize flour. This large-scale crop transition also altered the material culture found in the houses of Kwa Fungo and Kwengoma, from cooking utensils (ugali stirrers, coconut ladles, iron sheets for frying cashews, etc.) to other items such as sisal strings for beds or the mbuzi (coconut-grater stools). Many of these utensils and items might also have previously been used for other purposes (e.g., stirring sorghum porridge, frying other staples, etc.), and these changes may therefore simultaneously be examples of cultural continuity. The strengthening of assemblages of alien crops and daily material culture was likely fundamental to the indigenisation of the former.

While long-distance trade networks and later market capitalism led to the importation and cultivation of foreign crops in the lower Pangani Basin, they also favoured the adoption of new ‘substances’ (Hahn and Soentgen, 2010; Tilley, 2004) by Zigua communities during the 20th century such as plastic, aluminium, and enamel.

The most evident example of the adoption of new substances comes from containers. Aluminium sufuria (cooking pots or saucepans) were present in every kitchen in Kwengoma and Kwa Fungo in 1991 (Figure 5). However, an interviewee from Kwengoma recalled that before World War I everyone used only clay pots (SS, 1991; interview). The reason for this widespread adoption of aluminium can be found in its properties. Informants commented that although food cooked in clay pots tasted better, sufuria were now favoured because they are more durable and enable faster cooking (hence also necessitating less firewood). Aluminium indeed is less breakable than ceramic and has better thermal conductivity (see Derbyshire et al., 2020; Jégourel, 2015; Romainville, 2009 for discussions of the adoption of aluminium in different African contexts). Enamel, present in many Zigua households, may also have followed a similar path. Aluminium and enamel, made accessible by globalisation, were therefore adopted and re-appropriated by the local communities mainly because of their positive material properties. The adoption of these substances may have generated a variety of changes in the domains of consumption and taste, but such issues require further study.

Figure 5.

(a)–(c) Examples of kitchen interiors. Ceramic pots, sufuria, and other plastic, enamel, and aluminium vessels are visible. Photo credit: Paul Lane.

Recent processes of globalisation also percolate within households in mass-produced items that have undergone processes of re-appropriation and transformations in usages and meanings through the agency of local communities (see Hahn, 2015; Platte, 2004; Rowlands, 1996). A variety of items from Kwa Fungo and Kwengoma qualify as such, from radios and photographs to soda bottles, sewing machines, or automobile parts. The absence of radios and photographs in Kwengoma perhaps points to a better integration of Kwa Fungo than Kwengoma into wider regional trade networks. Moreover, some of these items were obtained through personal relations: for instance, a radio was gifted to a Kwa Fungo inhabitant by a Greek kinsman (EB, 1991; interview)—the Greek diaspora was indeed present in many East African countries since the late 19th century (Ekemode, 1973). This category of items can therefore reveal some of the networks of consumption, acquisition, and exchange in which the households of Kwengoma and Kwa Fungo were positioned. Zigua households therefore adopted some of the crops, substances, and items made available or imposed by global processes, from colonialism to market capitalism. This absorption into local assemblages, sometimes facilitated by the intrinsic qualities of the material, was often dependent on personal taste, which paradoxically resulted in making many of these new items potential examples of cultural continuity (such as items made of plastic serving identical purposes as items previously made of wood, or items used for specific food now used for preparing newly introduced food).

Constellations of charms, medicine, and ritual objects

The analysis of household inventories from Kwengoma and Kwa Fungo reveals the constellations formed by communities, their material culture, architectural space, subsistence strategies, and beliefs. Traditional medicine and religious and ritual items comprise a particular category of material culture that can be studied through this lens.

Charms and traditional medicine (dawa) are central to Zigua healing and protective practices. Virtually every house in Kwengoma in 1991 was protected by its gobwelo, a charm of maize leaves. The use of maize in ritual ceremonies was already attested in the 1890s (Dale, 1896). More generally, in the 1990s in Kwa Fungo and Kwengoma, dawa and charms were kept in containers or hung on walls. The list includes calabash charms against illnesses, Achatina shells to protect animals, elephant dung to be burned against bad spirits, and medicines against stomach ache, to prevent sleepwalking, or against snakes and insects. The zindiko, a glass bottle buried in the house entrance by the healer (EP, 1991; interview), was one of the most frequent types of dawa in both settlements.

All these charms and dawa fit within broader healing practices, which have been extensively studied by Walz (2009, 2016, 2017) among the Zigua and by Thompson (1999) among the Shambaa. Dawa is prepared by the mganga (healer), who collects for this purpose natural elements (tree bark or plants) and human-made items associated with local memories, such as beads retrieved from sites along former caravan routes. According to Walz, many of these items relate to memories of past social traumas, such as the ivory and slave trade. The use of these components from the past materialises a process of coming to terms with past social ruptures. The mganga mediates these past tumultuous experiences and ‘repurposes’ them through the making of dawa, in order to heal and forge future possibilities. From this, Walz (2009, 2016) argues that the Zigua worldview is couched in terms of healing and harm enchaining together the past, present, and future. This coexistence of different temporalities in specific objects, here the ones holding ritual or medicinal properties, can be compared with the Western idea of the percolation of time in material culture, although the two belong to radically different understandings of the world. These healing practices should however also be historicised, and while they were recorded in the 19th century (Dale, 1896), they were likely revived as a reaction to missionisation and European colonisation (Ekemode, 1973), and have undergone a process of commodification in recent decades (Sanders, 2001). Nevertheless, this worldview enchaining past and present took a very material and tangible form in the households of Kwengoma and Kwa Fungo, with the presence in each house of charms made by a mganga.

In the household material assemblages, Bibles, rosaries, and envelopes with verses from the Quran were frequently kept next to bottles of dawa and other charms, showing how healing ontologies easily accommodated in daily life to the ‘world religions’ (Islam and Christianity) that had penetrated the Tanganyika interior by the 19th century. These were also kept near ritual items tied to singular heritages such as the leather shield, sword, and whistle that belonged to Fungo, the founder of Kwa Fungo, and were inherited by his descendants (Figure 6). These objects still held a particular function in the 1990s both as objects of memory but also as objects of ritual and display of power. At the time of the survey, Fungo’s senior surviving descendant would bring them to ritual gatherings, where he was called to perform ceremonies to call for rain or after unusual things had happened—a role inherited from his ancestor and tied to these singular inalienable objects (JOF, 1991; interview). These objects, markers of the importance of Fungo’s lineage, thus represented a continuity with a glorious past, entangled with local memory and ontological systems expressed in rituals and ceremonies.

Figure 6.

(a) Fungo’s shield and club, which in 1991 were still owned and used in rituals by his descendants. Photo credit: Paul Lane. (b) Fungo’s sword and scabbard. Photo credit: Paul Lane.

In summary, these objects (charms, dawa, religious items, inalienable possessions) formed part of a wider assemblage interlinking beliefs and religions, materials, crops, labour and subsistence, and memory; overall, they drew together the human (communities and individuals) and nonhuman (objects and crops), the material (material world and material conditions) and immaterial (beliefs, ideologies, or religions), and sacred or traumatic pasts with the present.

Object constellations make subjects

In the 1990s, Congolese linguist Ntole Kazadi (1996) noticed how most African ethnographic studies failed to address the role of objects—and especially everyday objects—in the construction and strengthening of individual and social personalities. This resonates with many debates in archaeological theory today that emphasise how material culture, rather than simply reflecting beliefs, social organisation, and identities, also becomes entangled with these and can actively participate in forming them (Fowler, 2010; Harris and Cipolla, 2017: 87-108; Hodder, 1982; Miller, 1987).

For instance, among Zigua, young adult men traditionally moved out of their parents’ house around the age of 15 to live in outhouses that they built themselves. From the household inventories, it transpires that these outhouses tended to have very similar groupings of items, differing from other dwellings. These included newspaper pages on walls, musical instruments such as xylophones and drums (some young adults had started a music band in Kwa Fungo), or paper chains for decoration. They also generally had fewer objects—outhouse bedrooms had on average the lowest number of objects of all types of rooms—and these few possessions were often collected or repurposed items such as half plastic bottles used as dust-pans or upturned broken mortars used as tables. Young adult men were therefore surrounded by a constellation of material culture specific to them, different from the intergenerational material culture of the main houses and therefore likely materialising their identity as young adults.

Another example of the entanglements between objects and identities is pots and ceramics seen as markers, and in turn makers, of specific ethnicities. In archaeology, this recursive relationship has been notably theorised in the work conducted in the 1980s by Hodder (1982) and his students. Zigua and Bondei ceramics were for instance looked at in relation to changing notions of ethnicity through time (Lane, 2015), as ethnic identities in East Africa were for the most part fluid and porous, better understood as toponyms, before European colonisation (Beidelman, 1978; Gartrell, 1983; Willis, 1992). In Kwa Fungo in 1991, the multi-ethnic situation resulted in a wide diversity of pots. Pots strongly associated with the Zigua and Bondei (who seem to have had very similar potting traditions)—such as the biga or nyungu vessels—were some of the most frequent in Kwengoma and Kwa Fungo, but in the latter they shared their place with pots mentioned by the interviewees as belonging to Ngoni (kitereku or kiteneko vessels), Shambaa (the kisumba), Pangwa (the sipwalimwene), or Bena traditions (the ibibili). These pots were often named in the native language of their owner. There existed stylistic variations between these ceramic traditions: some informants noted that Zigua and Bondei pots tended to not be decorated, while they recalled that most of the pots in their region of origin were. These containers, by virtue of being explicitly tied to a particular ethnicity and thereby rendering the latter tangible, took part in stabilising and re-inscribing these ethnic identities. Some of the 1991 data about containers, not directly consulted for this article, could form the basis for a distinct study of such matters.

Items other than pots were also markers and makers of ethnicities, such as the three-legged stool of the Zigua, present in both villages. This stool was associated with traditional marriage practices, and many Zigua interviewees regretted its gradual disappearance. Other examples included a Pangwa spear for hunting warthogs, or a particular winnowing basket associated with the Zigua. Most of these examples came from Kwa Fungo, perhaps because the multi-ethnic context brought ethnic differences in material culture more readily to light.

Finally, the interaction between objects, space, and identities can also be explored with regard to gender. The oral history interviews reveal that taboos and prohibitions related to gender were frequently associated with objects. Men were not supposed to carry hoes and simultaneously pick vegetables while women could (MR, 1991; interview); Zigua women could not sit on mortars (AJ, 1991; interview); and touching pestles was taboo for Zigua men because pounding was associated with reproduction and hence would make men impotent (BJ, 1991; interview). These prohibitions seem to have often derived from the gendered division of labour—for example, men were prohibited from weaving or making pots because it would attract evil spirits (RF, 1991; interview).

Moreover, household space also tended to be gendered: as discussed previously, in the halls of the Swahili-type houses in Kwa Fungo, distinct areas were allocated to each gender for eating (with men on one side and women and children on the other). One interviewee from Kwa Fungo noted that the introduction of the ‘world religions’ (Christianity and Islam) had exacerbated gender differences, excluding women from many rituals and occupations (MM, 1991; interview). These conclusions however are not new, and the role of European colonialism (and religious conversion) in imposing modern/colonial Western gender categories in African societies has been noted by many postcolonial feminist historians (e.g., Lugones, 2007; Oyewumi, 1997; Schmidt, 1992), among other scholars.

Different facets of the inhabitants’ identity, from their age to their ethnic identities or their gender, therefore came to form constellations with their material world. While being reinforced by this materialisation, as shown in the examples from Kwa Fungo and Kwengoma, the entanglement of the material world with these discourses also transformed the former into a repository of these social identities, whose articulations and history can then be shown through analytical work. The conclusions presented here were nevertheless limited by the depth of the data collected in 1991, and future work could reveal more precisely how changing identities and material culture interacted through time.

Conclusion

The household material culture of the two Zigua communities discussed here, and more broadly their material world, was certainly shaped by a series of complex historical processes. This variety of overlapping social and material processes can be ‘read’ (Olivier, 2000) in the household inventories and oral history interviews from Kwa Fungo and Kwengoma collected in the early 1990s, in a manner that produces a vibrant and diachronic account of some of these material changes since the 19th century. For instance, the progressive shift in house type, from round to rectangular dwellings, initiated in the 19th century and that continued across Zigualand in the 20th century, was influenced by changes in generational tastes, material advantages, and previous building knowledge, entangled with the increased mobility brought about by the 1950s labour migrations and Nyerere’s villagization policies. During the same timeframe, the introduction of new substances, such as aluminium, was facilitated by global networks of exchange, but these were eventually adopted by the inhabitants because of their material qualities and affordances. The tangible form taken by the healing practices of the Zigua was evidenced by the high number of charms and dawa recorded in the households of both villages, materialising a worldview tying tumultuous pasts to the present and to future possibilities (Walz, 2009, 2016). Finally, these constellations of objects played crucial roles in reinforcing or negotiating certain aspects of identity.

Concepts of percolating pasts as well as Indigenous epistemologies such as Zigua ones, along with a greater attention to oral histories, therefore challenge conventional approaches to analysing ethnographic data. They provide an alternative to the centrality of analogies in ethnoarchaeology, and enable researchers to historicise the present-day material assemblages they study while avoiding simple narratives of acculturation and modernization. This highlights the potential of alternative theoretical models for producing different approaches to material culture and ethnographic or (ethno) archaeological data.