Frontiers of cuisine: persistence and substitution in the evolution of porridges in south Central Africa

Introduction

A steaming pillow of starchy boiled porridge is taken from the pot, torn by hand, and dipped into an oily relish. This is usually made up of leaves from one or several species of brassica, alongside other ingredients such as okra, onions, groundnuts, local fruits, and sometimes some meat or fish. In Zambia, this dish is commonly called nshima and it is the staple cuisine. This dish, or something like it, will be familiar across much of Africa, from Nigeria to Kenya, and from Sudan and Mali to South Africa, and throughout the world via the African diaspora. It has many different names and can be prepared using flour derived from diverse and somewhat interchangeable sets of ingredients depending on regional tastes, availability, and customs (Haaland, 2007). The essence of the dish is its physical form as a porridge, though there can be significant variability in preferred stiffness and texture, meaning cooks can innovate with new ingredients, mixing flours to add taste or fortify them with added nutrients (McCann, 2009; Osseo-Asare, 2005; Smith and Allen, 2004). Fermented beverages similarly share a relationship with the pot or calabash and can be produced from a range of ingredients including palm sap, sorghum, millets, and a range of wild fruits to produce alcoholic beers, or to produce sour fermented milk, commonly known in Zambia as mabisi (in BaTonga and Lozi languages) (Hedlund and Lundahl, 1984; Moonga et al., 2019; Young, 1949).

These practices of food preparation and consumption form cultural linkages across the African continent that transcend languages and language families, reflecting shared technologies—namely the “porridge and pot” based cuisine, a concept set out by Haaland (2007). Porridge in this context is not simply nutrition but within different environmental and cultural contexts can be an embodiment of taste, values, and cosmological beliefs (Bourdieu, 2013; Fuller and Rowlands, 2011; Goody, 1982). This creates what Stahl (2002) calls a “cartography of taste”, object worlds where continuity and change in practices of taste can be viewed comparatively in time and space.

The African tropics have seen major shifts in food systems and human ecology over the past 4000 years. The most archaeologically debated is the spread of Bantu-speaking mixed economy societies through migrations and cultural exchange associated with the “Iron Age” (see Figure 1). This process has been linked archaeologically and linguistically with the introduction of “porridge and pot” cuisine to southern Africa (Ashley, 2010; Haaland, 2007; Ricquier, 2014), inferred from the co-occurrence of pottery, iron working, and sedentary behaviours linked to agricultural production (the “Iron Age package”) (Bostoen, 2018; Crowther et al., 2018; Robertshaw, 2012). The Bantu Expansion is thought to have originated in the grasslands of the Nigeria-Cameroon borderland, from where Bantu-speaking communities dispersed across much of sub-Saharan Africa over several millennia (Bostoen, 2018; Grollemund et al., 2015).

Historical linguistics indicate that Bantu languages were present in South Central Africa earlier than the current archaeological evidence of the Iron Age, leading to hypotheses such as Vansina's “Slow Revolution”, which posits that food system activities changed through a cumulative process involving both migration and the diffusion of ideas, rather than a series of “packaged” events (de Luna, 2018; Vansina, 1994). As de Luna (2008, 2016) has shown in Zambia, regional ontologies for farming and horticulture are not distinguished from other forms of gathering and foraging, instead, these activities are regarded as different forms of “collection”. Still today, food production is practised alongside foraging, fishing, and hunting. As Vogel (1986) observed, the relative importance of each of these activities would have fluctuated across both space and time.

A significant question for the cultural history of communities in sub-Saharan Africa, is how and when did these cuisines become popularised across much of the sub-continent? Is porridge and pot cuisine a feature of the African “Iron Age” or does it have deeper or more diverse origins? How do environments and social and material culture mediate ingredients and preferences for ingredients? The aim for this paper is to consider how these different modes of food procurement, availability of foods, and different food technologies govern the possibility of porridge-type cuisine. We can consider questions about porridge as a cuisine at two scales. The first concerns its broader category as a culinary form, much as we might have a category for bread in Europe and Asia. The second concerns how porridge is made at regional and local scales, where ecology, social preferences, and political economy together shape what ingredients are used and how taste is expressed.

Following Haaland's (2007) work on “porridge and pot” cuisine and Stahl's (2002) “cartographies of taste” as a means of exploring cultural and ecological dynamics, this paper highlights the continuities and important variations of cuisine in South Central Africa. This paper expands archaeobotanical discussions beyond subsistence to consider how cuisine and cookery reflect the broader dynamics of plant materiality, environment, and cultural and social expressions over time. We begin by contextualising the archaeology of food systems in the region, before examining how environmental conditions and ecology can serve as infrastructures that shape food systems and cuisine across the Zambezian Bioregion. We then consider how research theory and methods can provide a more nuanced understanding of the historical geography of porridge as a cuisine in South Central Africa. We conclude by outlining the implications of treating cuisine and cartographies of taste as significant expressions of individual and collective values embedded into both regional and hyperlocal environments.

Culinary frontiers?

The “Iron Age Package” is often associated with the spread of Bantu speakers includes the cultivation of a suite of African cereals, particularly pearl millet (Pennisetum glaucum (L.) R.Br.), sorghum (Sorghum bicolor (L.) Moench), and finger millet (Eleusine coracana Gaertn.) alongside pottery as a primary culinary technology (Crowther et al., 2018). This is often treated as syllogistic with the use of boiling and the preparation of foods oriented around starchy porridges (Haaland, 2007). Similarities in pottery styles (grouped under the “Chifumbaze Complex” (Phillipson, 2005); and iron technology across large parts of equatorial, eastern, and southern Africa have reinforced this view, lending support to the idea of relatively uniform cereal-based food systems across the region (Lane, 2004). Three critical issues emerge from this framework. First, it is unclear whether all parts of the “Iron Age Package”, including porridge as a preferred cuisine, spread simultaneously (see discussion in Crowther et al., 2018). They may instead reflect a gradual accumulation of practices over time, as Vansina (1994) suggests, or some combination of both processes operating at different times and places. Second, cuisine shows considerable geographical and cultural heterogeneity in terms of ingredients and local value, at biogeographical macro scales and hyperlocalised micro scales. Third, the framework assumes that porridge requires ceramics, when it may be possible that earlier Bantu-speaking societies, as well as pre-Bantu speaking peoples such as Batwa and Khoisan speaking societies (noting the latter did have ceramics from ∼2,200 BP), may have used alternative methods (Lander and Russell, 2018; Russell and Steele, 2009).

A persistent framing of Bantu migrations involves two main streams: a Western Stream, travelling south from the central African rainforest, and an Eastern Stream travelling to the Great Lakes region and the Indian Ocean coast. Archaeobotanical evidence of pearl millet at sites in Cameroon and the Inner Congo Basin in the Democratic Republic of Congo (DRC), dating to the late 1st millennium BCE, are thought to reflect the Western Stream migrations. Notably, the pearl millet evidence from these sites was associated with pottery but not evidence of iron working, though this may be attributed to taphonomic conditions of wet tropical soils (Kahlheber et al., 2009; 2014). Sites associated with the Eastern Stream, characterised by Urewe ware emerge around Lake Victoria and in Kenya and Tanzania in the late 1st millennium BCE/early 1st millennium CE (see Leakey et al., 1948; Posnansky, 1961). The related Kwale wares are found along the coast and in the coastal hinterlands of Kenya and Tanzania from the early 1st millennium CE (Chami, 1992; Helm, 2000; Soper, 1967), as well as its variant Matola ware which has been found in Mozambique from the mid-1st millennium CE (Sinclair et al., 1993). Likewise, local variants of Chifumbaze pottery appears at sites in South Central Africa from approximately 2,000 BP (Huffman, 1989).

Systematic sampling for archaeobotanical analysis has only recently been applied in South Central Africa and aligns with evidence across the wider region that points to the introduction of cereals from the mid-to late first millennium CE. Early dates for cereals at Fibobe II, for some time cited as pivotal evidence for an earlier arrival, were revised on further analysis of the site to be some 500 years later than first reported (Goldstein et al., 2022). The earliest evidence regionally comes from pearl millet grain impressions in pottery from Silver Leaves in South Africa, dating to the 3rd–4th century CE (Klapwijk, 1974; Klapwijk and Huffman, 1996). The limitations and subsequent improvements in radiocarbon dating suggest these dates should be treated with caution. Ongoing studies from the Central Kafue and the site of Kalundu indicate the presence of crops from the 6th–7th century CE (Farr, 2024; Goldstein et al., 2021). There are currently no direct dates for cereals in South Central Africa associated with early Bambata ware (within the broader set of Kalundu Tradition/Western Stream ceramics – see Figures 2A, B), with the earliest evidence being found in Zimbabwe, poorly dated but suggested to be late 1st millennium BCE (∼350-150 BCE). Likewise, Early Ziwa wares (within the Nkope Branch of the Urewe Tradition) have yet to be directly associated with archaeobotanical evidence of cereals (Huffman, 1971, 1989, 2007; Lander and Russell, 2018).

Vansina (1994, 1995) uses linguistic evidence from Zambia to argue that the adoption of agriculture came about through a “slow revolution”, with the gradual accumulation of ecological knowledge combined with access to innovations in technology, plant breeding, and crop varieties through later migrations and cultural diffusion. Given the current paucity of archaeobotanical evidence from the region, the presence or absence of cereals before the mid-1st millennium CE remains difficult to confirm (Crowther et al., 2018; Farr, 2024). Current evidence attributing the spread of porridge and pot cuisine into South Central Africa to Bantu-speaking farmers remains unsettled. Likewise, the association between porridge and cereal cultivation should not be treated as self-evident.

The evidence suggests that pottery, grindstones, and the capacity for porridge-like food preparation existed independently of Bantu migrations. In Southern Africa, pottery is associated with herders and hunter gatherers well before the Iron Age, and ethnographic records document porridges made from wild seeds, insects, and milk among non-farming groups (see Figure 2C) (Stewart et al., 2024). Evidence from Namaqualand shows distinct forager and herder pottery traditions emerging in the early first millennium BCE (Lander and Russell, 2018). The use of grindstones by hunter-gatherers, herders, and farmers alike further complicates the picture; the capacity to produce flour or pulp from a wide range of plants provides the possibility of porridge-like cuisines without cereal-based agriculture (Nic Eoin, 2016; Shoemaker and Davies, 2019; Wojcieszak, 2018).

It is also important to consider whether boiling itself required pottery. Experiments with calabash show that sub-boiling temperatures could be reached over fire or using hot stones. Other technologies using tree barks, animal stomachs, animal hides, clay-lined pits, and ostrich eggshells offer additional possibilities for food preparation in pre-ceramic contexts (Langley et al., 2023; Nelson, 2010). Cold porridges and fermentation may have served a similar role. As noted by Fuller and Rowlands (2011), texture is often the most important dimension of taste in food cultures; if texture is what defines porridge, then neither pottery nor cereals were strictly necessary to produce it. In this regard, pottery may have been an important innovation for porridge-making but need not have been a culinary frontier in its own right. Similarly, cereals may have become the preferred ingredient in many contexts but may represent a substitution for earlier ingredients rather than the foundation of porridge cuisine, a possibility explored further below.

Biogeography as a driver of taste

Africa has highly diverse biogeography ranging from wet rainforests and dry savannas to montane zones, which have been inhabited by humans for almost our entire history as a species. The tropical rainforests of Central and West Africa have very different environmental conditions to the arid savannas and woodlands of the Sahel, Sudan, and Miombo regions (see Figure 3) (White, 1983). The structure of the environment plays an important role in shaping the possibilities of human behaviour, particularly in relation to food, with different ecosystems supporting different plant and animal communities (Farr, 2024; Farr and de Luna, 2024). Human mobility across these ecological systems has been high, with cultural practices both responding to and actively reshaping local environments (Boivin et al., 2016; Roberts and Stewart, 2018). Certain cultural continuities are expressed either through the cuisine, or through the translocation of new crops or plants from elsewhere into new locations, even as other elements change (Kirksey and Helmreich, 2010). This section examines how the climate and ecology of different biogeographic zones shaped the availability of foods, in turn influencing the routes and pace of Bantu-speaker migrations as they followed preferred ecologies or adapted to new ones through technology and ecological knowledge (Bleasdale et al., 2020; Grollemund et al., 2015, 2023; Koile et al., 2022).

Periodic climate change in the African tropics over the past 4000 years may have created conditions that promoted migration and networking among Bantu-speaking communities (Nash et al., 2016; Nicholson, 2017; Pages 2k Consortium et al., 2013; Tyson and Lindesay, 1992). Linguistic and material culture evidence suggest that people tended to migrate more rapidly within or along the rainforest ecotone typical of the Guinea-Congolian bioregion, though significant movements also occurred within the rainforest itself (Bleasdale et al., 2020; Bostoen et al., 2015; Grollemund et al., 2015, 2023; Koile et al., 2022). By contrast, archaeological and linguistic evidence indicates that initial migrations of Bantu speakers into the Zambezian bioregion, which is characterised by dry wooded grasslands and open savanna, appear slower and less intense until the early first millennium CE. The distinct ecological differences between the Guineo-Congolian bioregion and the Zambezian-Miombo ecoregions (see Figure 2) likely delayed the spread of intensified food production from equatorial Africa into southern Africa (Bostoen et al., 2015; Grollemund et al., 2015).

Evidence from archaeobotany, linguistics, and ethnographic analysis indicates yams, canarium, cola nut, and Bambara ground nut were important parts of the food package in the Guinea-Congolian bioregion for Bantu-speaking communities, with some studies also providing evidence of pearl millet use in the early 1st millennium CE (de Maret, 2025; Kahlheber et al., 2009; Neumann et al., 2012, 2022). The rich variety of wild edible plants in the rainforest biome suggests that crops such as pearl millet may not have necessarily been central to communities in this region. Contemporary preferences across these wetter tropical zones for cassava, which arrived after the 15th century through the Columbian Exchange, and yam porridges, may hint at a deeper practical and cultural preference for tubers. This contrasts markedly with the Zambezian bioregion south of the rainforest, where there is a greater preference for cereals, though yams are still important in many areas (Bahuchet et al., 1991; Bruschi et al., 2014).

The Zambezian bioregion, which is the largest phytochorion in sub-Saharan Africa in terms of plant species richness and degree of endemism (White 1983), is a less stable environment than the rainforest, characterised by unimodal rainfall, generally poor-quality soils, and high sensitivity to the effects of El Niño Southern Oscillation (ENSO) events. The variation of winds and sea surface temperatures can also exert a strong influence over the distribution of rainfall, creating somewhat unpredictable conditions. Long dry seasons and regular cycles of severe drought have shaped an ecological system resilient to these pressures, of which the Miombo ecoregion, a dry tropical woodland dominated by Brachystegia spp. Benth., Julbernardia spp. (Benth.) Troupin, and Isoberlinia spp. Craib & Stapf trees, is the dominant element (Ribeiro et al., 2020).

While maize (Zea mays L.) has become the predominant cereal in South Central Africa over the past century, historically the staple crops for farmers were sorghum, pearl millet, and finger millet. These African domesticates are well adapted to drought conditions, with many varietals bred and exchanged for different traits to suit different environmental conditions. Pearl millet is grown in particularly arid areas, sorghum is suited to tough clayey soils, and finger millet is well adapted to low-quality upland soils (Christiansen, 2008; Joshi et al., 2021; Mercuri et al., 2018). There is evidence that these cereals were grown as complementary crops in some regions (see Farr, 2024; Goldstein et al., 2021; Steyn and Antonites, 2019, 2025), while elsewhere only one or two are grown, often alongside other starchy staples such as yams, or more recently, cassava (Manihot esculenta Crantz) (Trapnell and Clothier, 1996; Zulu et al., 2020).

Wild edible plants have received significant attention in recent years owing to their importance to local cuisine and as buffering foods essential for maintaining food security, particularly during periods of drought (Daum et al., 2021; Gumbo et al., 2018; Polesny, 2022). Edible fruits found across Zambia's cooler plateau region include Azanza garckeana F.Hoffm. (English: Snot Apple), Diospyros mespiliformis Hochst. (English: Jackal Berry), Garcinia huillensis/buchananii Baker. (Bemba: Musongwa), Grewia flavescens Juss. (English: Donkey Berry), Landolphia pavifolia K.Schum. (BaTonga: Mutowa), Parinari curatellifolia Planch. ex Benth. (BaTonga: Mbula; Bemba: Mupundu; English: Mobola Plum), Strychnos cocculoides Baker and Strychnos spinosa Lam. (Bemba: Ifisongole; English: Monkey Orange), Uapaca kirkiana Müll.Arg. (BaTonga/ Bemba:Masuku), Schinziophyton rautanenii Schinz (BaTonga/Bemba: Mukusu; Common Name: Mongongo), Sclerocarya birrea (A.Rich.) Hochst. (BaTonga/Common Name: Marula), Berchemia discolor, Ximenia americana var. americana (Klotzsch) Hemsl. (BaTonga: Munyii; English: Bird Plum), and Adansonia digitata L. (Bemba/Nyanja: Mubuyu; Common Name: Baobab) (Gumbo et al., 2018; Kalaba et al., 2013; Ribeiro, 2015; Urso et al., 2016). Some variation in wild edible plants across the Miombo ecoregion can be attributed to rainfall and altitude, with hotter, drier river valleys home to a number of species rarely found on the cooler plateau (Scudder, 2011). Important porridge additives and bases of the Dry Miombo ecoregion include mongongo, baobab, lusala yams (Dioscorea hirtiflora Benth.), and water yams (Dioscorea alata L.), while the Wet Miombo ecoregion supports food species such as the fruit, Aframomum alboviolaceum (Ridl.) K.Schum. (Bemba: Ntungulu), giant ichikolowa mushrooms (Termitomyces titanicus Heim), yellow yam (Dioscorea cayenensis Lam.), white yam (Dioscorea rotundata Poir) and bitter yam (Dioscorea dumetorum (Kunth) Pax).

Persistence and substitution

Similarities in material culture between sites on the East African coast and southern-central Africa, particularly in pottery styles and decoration, have been used to infer a degree of cultural uniformity across these broader regions in the first millennium CE. While the significance of cuisine and ingredients as indicators of cultural distinction and evolution has been explored in detail in parts of West Africa (see Logan, 2016a, 2016b; Logan and Stahl, 2017) and East Africa and its islands (Crowther et al., 2016a, 2016b; Quintana Morales et al., 2022; Walshaw, 2010; Walshaw and Stoetzel, 2017), such questions have only begun to be explored in South Central Africa (de Luna, 2008, 2016; Farr, 2024).

Vogel (1986) highlighted that discussions of early food producing societies in South Central Africa should pay attention to local and hyperlocal environments. The mobility of people across different biogeographical regions and ecosystems can involve the translocation of familiar species and the adoption of new ones. As Stahl (2002) notes, the incorporation of novel foods is often mediated by how well they translate into existing practices (see also Gosden and Hather, 2004; Leach, 2005). The materiality of a food is key, either in terms of the plant's capacity to be grown in a new region, or its suitability for being processed into flour or pulp and cooked as a porridge. In the latter case, that food can be subsumed into an existing culinary ontology, functioning as porridge base or a sweetener. This is evident with the arrival of New World crops, maize and cassava, both of which provide close analogues to earlier food types and can be made into porridges, making them especially strong candidates for substitution (Cliggett et al., 2007; Kakeya and Sugiyama, 1985; Trapnell and Clothier, 1996). If the spread of cereals in South Central Africa did occur in the mid-late 1st millennium CE, it may represent an earlier episode of the same process; one in which new migrations, technologies, political economies, and climate conditions made cereals more appealing to food producing societies than existing alternatives. Whatever direction those processes took, porridge cuisine appears to have been a fixed point around which other elements were negotiated.

The diversity of ingredients used to make porridges makes clear that form and texture are themselves significant components of “taste” (Fuller and Rowlands, 2011; Stahl, 2002). The ingredient base of porridge flour can be varied for many reasons. Environment suitability sets the baseline of what is available, but cultural preferences are often the most prominent driver and can be a key marker of cultural distinction between groups. Other factors might include use for auspicious occasions, seasonality, or scarcity, where “famine foods” are used in place of preferred ingredients.

Ethnographic studies over the past century show regional variations in food systems that reflect a dynamic interplay between ethnolinguistic identities and agroecological conditions. For example, people who identify as Bemba have a deep cultural affiliation with finger millet, which is better adapted to the higher elevation plateau of their home regions in Katanga, the Copperbelt, and northern Central Province. Bemba clan names, which include mushroom, millet, pot, and porridge, form part of an ontology of food types, natural phenomena, and objects linked through a system of oppositions or complementarities known as Banungwe. This creates informal kinships through “joking cousins”, relationships that permit social leeway and obligations between members of different clans. Joking relationships are not confined to ethnolinguistic boundaries. Tonga, Ngoni, and Bemba groups engage in gastronomic teasing of each other's food habits, usually centred on preferences for game meat (Annear, 2004; Richards, 1937, 1950, 1995). Despite broadly overlapping food systems, this functions as a form of schismogenesis, where differences in taste are used to shape and maintain social and cultural distinctions that, in this context, make fertile ground for friendly banter between strangers.

Studies of the Gwembe Tonga in southern Zambia indicate that their preferred cereal is pearl millet, which is better suited to the drier regions of the Zambezi Valley and the Middle Kafue. The southern regions are also home to baobab and marula, which are important additives in porridge cuisine for people in that region, but are not found on the higher elevations of the central African plateau (Frost, 1996; Timberlake and Chidumayo, 2011). Ecological knowledge can also be highly parochial. It was reported that Gwembe Tonga people who were relocated during the construction of the Kariba Dam were poisoned by herbs that were a food source in their home landscapes but had acquired toxicity on the different soils at higher elevations (Scudder, 2011).

In Zambia, parallels have been drawn between the farming practices of people living in the past century and traditions extending back at least a thousand years. Scudder's (2011) extensive ethnography of the Gwembe Tonga in the Middle Zambezi Valley offers important caution against the over-application of ethnographic evidence to archaeological interpretation, highlighting that this approach can flatten or obscure the dynamics and complexity of cultures as they exist across time and space. This is particularly pertinent to central and southern Africa, where a great diversity of food systems, cultures, and social systems coexist in both adjacent and overlapping territories that have experienced significant change over the last 2000 years (de Luna, 2012b, 2016).

An important consideration is the labour involved in collecting and processing different food resources. Beyond the demands of cultivation itself, the collection, processing, and grinding of foodstuffs into flour for porridge is a laborious process (Fuller et al., 2014). The strenuous work involved in cereal cultivation compared with the collection of wild foods, may have delayed its uptake or eventual rise to prominence in food systems in South Central Africa (Scott, 2017; Vansina, 1994). De Luna (2012a, 2016) advances this argument in the context of the Middle Kafue, where a proliferation of iron working led to a cultural expansion in bushcraft, hunting, and collecting alongside cereal cultivation, which, it has been argued, was made more feasible by the availability of iron hoes.

Conversely, systems that offered greater storage capacity and increased sedentism may have opened new opportunities for social and cultural development. In this sense, crops can be considered as technologies that hold their own possibilities and co-create realities around them (Fuller and Boivin, 2009; Kirksey and Helmreich, 2010). This is echoed in the current prominence of maize, which despite being less well adapted to soils in the study region, is very low maintenance compared with other cereals (McCann, 2005; Smale and Jayne, 2003).

Archaeological methods and challenges

Archaeological evidence of cuisine shows a degree of consistency in terms of grinding stones, mortars, and pottery types, that could be taken as evidence of uniformity, but may well underplay the heterogeneity of ingredients, technique, and broader cultural significance visible in contemporary African foodways. Cartographies of taste across Africa show remarkable consistency in the form of porridge and pot-based cuisine, yet great flexibility in the range of ingredients used to produce it. Just as it would impoverish our understanding of European, North African, and West Asian food systems to describe them simply as a uniform wheat-and-barley-based bread complex, it would be equally reductive to remain incurious about the variation and consistency of food systems and cuisines across the vast and diverse geography of Africa.

Archaeobotany provides a means for obtaining direct empirical evidence of past plant use that can be complemented by data from historical linguistics to understand how people articulated themselves socially and culturally in different environmental contexts. The relative paucity of archaeobotanical research in Africa means there is a significant gap in the data available to reconstruct preferred cuisine. This challenge is compounded by the poor preservation conditions of soils across much of Africa, which tend to have high levels of acidity and extensive bioturbation (Crowther et al., 2018; Farr, 2024).

Even where archaeobotanical evidence is abundant, reconstructing specific cuisines can be challenging. The identification of charred remains of foods (as opposed to the raw ingredients) is an emerging method that has had some success in West Asia and Sudan (e.g., Arranz-Otaegui et al., 2018; Fuller and Gonzalez Carretero, 2018; Willcox, 2002). Residue analysis to identify lipids has also shown promise, with recent studies confirming dairying by Khoikhoi herders in Namaqualand (Hopper et al., 2023), Bantu-speaking agro-pastoralists in Mpumalanga (Becher et al., 2024) and Swahili societies at Songo Mnara on the East Coast of Africa (Quintana Morales et al., 2022). A particular challenge for identifying porridge through this approach is that biomarkers and lipids from cereals and fruits tend to occur at lower concentrations than those from animals and dairy products (see Colonese et al., 2017; Hammann and Cramp, 2018). Experimental work is needed to establish baseline references for local African flora, which remain largely absent from the comparative record. Boiling technologies that involved organic vessels such as calabashes, animal skins, tree barks, and eggshell containers would likely be archaeologically invisible. Further research on the use and significance of non-ceramic boiling technologies are needed (Langley et al., 2023; Stewart et al., 2024).

Deeper integration of ethnobiology and ethnobotany with archaeological research is important for understanding the specific and diverse approaches to foods and ingredients across different parts of Africa. While the limitations of projecting ethnographic records into the past must be kept in mind, ecological knowledge and environment-specific logics provide invaluable interpretative perspectives. Given that porridge and pot cuisine is the living heritage of communities across Africa and the African diaspora, connecting histories of cuisine to peoples and places is also integral to its salience as a research endeavour.

Conclusion

The steaming pot of nshima described at the outset of this paper represents far more than a staple meal; it is the product of thousands of years of ecological negotiation, cultural continuity, and culinary substitution. This paper has evaluated the evidence for how and when porridge and pot cuisine became established in South Central Africa, highlighting the diverse and cumulative process that likely took place over hundreds of years. We have also explored the possibility of porridge without pottery or, for that matter, intensive food production. While porridge may be an innovation associated primarily with pottery and the cultural diffusions of the Bantu expansion, pre-existing forms may have present in South Central Africa and Southern Africa, such as among the Khoikhoi, Batwa, and San speaking groups. Pottery is no doubt an important innovation in food technology, but the possibility of boiling technologies before ceramics, and what that means for our understanding of African culinary history, remains to be fully explored (Stewart et al., 2024).

Despite consistency in the fundamental form of porridge-type cuisines, biogeography has likely played an important role in shaping the tastes of different cultural groups. The concept of cartographies of tastes advanced by Stahl (2002) highlights how changes in food systems are mediated through existing social and cultural paradigms, informing the persistence and substitution of ingredients as a groups' access to preferred foods shifted, whether through movement across biogeographical zones, or through changing availability due to drought, disease, war, colonialism, or other disruptions. In South Central Africa this is particularly evident with regard to wild edible plants, which are known to have been part of ancient diets and continue to be used today as additives or bases for porridges.

Variations between communities underlines the important role of food as a mark of distinction within and between clans and ethno-linguistic groups. Schismogenesis, where the edges between groups become sharper, and associated taboos recorded among the Bemba illustrate how biogeographical logics of availability do not always govern social practice (Richards, 1937, 1950, 1995). Cultural preferences can lead to conservativism around favoured ingredients, or the selective adoption of new ones based on particular attributes such as texture or flavour. Cuisine can also carry hyperlocalised qualities that are unique to social or ecological conditions (Scudder, 2011).

The adoption and evolution of porridge cuisine raise important questions for political economy, labour, and social organisation. The transition from millets and sorghum to maize and cassava in South Central Africa over the past centuries has been driven in part by colonial imperatives but has also been sustained by their lower labour requirements and, in the case of cassava, drought resilience. Given that evidence for cereals currently postdates the arrival of pottery in the region by several centuries, recovery and dating biases notwithstanding, we must ask what the potential resistance to or gradual spread of cereals in South Central Africa tells us about different forms of social organisation, environments, and cartographies of taste, and what it means for how we understand the deep history of the food in the pot today.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Leilani Lucas, College of Southern Nevada, United States

Reviewed by: Claudia Moricca, Sapienza University of Rome, Italy

Hennie Fisher, University of Pretoria, South Africa