The moral economy of isolates: ‘Escape agriculture’ and warzone public authority

ABSTRACT

Edward Banfield was one of the first scholars clearly to describe the condition of peasant isolate social ordering, problematically terming it ‘amoral familism’. James Scott offered a more positive framing by referring to the potential of ‘escape agriculture’ to sustain political autonomy. The present paper returns to these debates about peasant moral economy. Norms concerning production and distribution are not exogenous givens but products of communal organization and social context. Banfield’s mistake was to refer to peasant autonomy as ‘amoral’. Here, isolate ordering has a positive moral valency, and that public authority in zones of war or disaster recovery should seek to accommodate it.

KEYWORDS:

• Marronage
• moral economy
• civil wars
• public authority
• escape agriculture
• West Africa

Introduction

Are human norms universal, or specific to various societies? Adam Smith (1767) thought morality was built into the human constitution and shared across time and space. This essentialist view of morals has a long history in the social sciences, and especially in economics, reflecting the hegemony of industrial capitalism and imperialism as frameworks for large-scale economic organization. More recently, situational perspectives have gained traction, spurred by the arguments of moral philosophers (e.g. Williams 1985) and internal developments within the social sciences – notably the greater centrality accorded to institutions by economists (North 1990; Williamson 1981).

An important aspect of this institutional ‘turn’ was the resurrection of older debates concerning moral economy by the English Marxist historian E. P. Thompson (1971; see also Thompson 1991), who argued that food riots in eighteenth-century England had a moral coherence, in that they sought to bring back into the public arena older assumptions about what constituted a just price for the essentials of life, in a world turned upside down by new-fangled market norms.¹ This is not far from the underlying logic of institutional development proposed by Douglass North, who suggests that institutions are rooted in custom but that successful institutions become formalized and that formal institutions tend to drive out informal ones (North 1992).

A different approach to institutions was advocated by Oliver Williamson (Williamson 1975; Bulte, Richards, and Voors 2018). Williamson steers clear of North’s successional (or evolutionary) account of institutions and focuses instead on an economic world of differently configured institutions, existing side by side. Firms may be organized hierarchically, but sales departments are an outlet into a world of competitive individualism. What happens at the boundary of differently ordered institutional worlds? Williamson’s answer is that organizational incompatibilities are negotiated through transaction costs (Williamson 1981).

The anthropologist Mary Douglas (Douglas 1986) commends Williamson for having recognized plurality of institutional ordering as an inescapable fact of social life but criticizes him for attempting to adjudicate institutional differences through transaction costs, since this means granting market values a superordinating role. This (to Douglas) is simply not realistic. No community is regulated in the last instance by market forces. When the global banking system was threatened by failure in 2008 states stepped in to rescue the banks (Bear 2019; Tett 2021). When the global pandemic caused by the Sars Cov-2 virus struck in 2020 every country on the planet overrode normal economic rules (De Waal 2021). Some had no compunction in closing markets temporarily.

So how, then, are these kinds of decisions to be understood? Douglas proposes taking a hard look at how institutional values are formed (Douglas 1986). For this, she argued we need a map of the terrain in which all actually existing institutional arrangements can be located (Douglas 1999a) but in which prior judgments about which institutions are good and bad, or formal and informal, are held in abeyance until we understand better what each set of arrangements tries to achieve, what circumstances cause them to be adopted, how transits across differently ordered institutional landscapes are best navigated (Williamson’s project), and how and why existing institutional arrangements persist, evolve, or implode.

Douglas’s map is based on suggestions made originally by Emile Durkheim ([1897] 1951), who had proposed to write, but never completed, a foundational study of the science of morality, a project more recently revived by Didier Fassin and colleagues (Durkheim [c. 1900] 1957, [1914] 1973; Fassin 2012; Fassin and Leze 2013; Karsenti 2012). Douglas’s variant on Durkheim’s map has quarters allocated to hierarchy and individualism, as well as what she termed enclave and isolate ordering (Douglas 1999a).

It is important to understand that this map is not a theory of institutional types, far less an account of institutional stasis. This misreading of Durkheim crept into the literature through textbooks but has been decisively corrected by more recent scholarship (Smith 2020; Stedman Jones 2001; Warfield Rawls 2005). Douglas’s scheme is simply a means of becoming aware of the full range of institutional types likely to be encountered and of some of the typical planes of cleavage between them.

For Douglas, this map was necessary to counter a rush to institutional judgment. In a late article (‘Traditional culture – let’s hear no more about it’) she lambasts the World Bank for pushing the argument that poor people were poor because they were in the grip of ‘informal’ institutions (Douglas 2004a).² This, she pointed out, was tantamount to blaming the poor for their poverty, something she also thought typical of the Thatcher government in Britain. So undoubtedly, she would have thought the proposed cure for all the world’s ills proffered by Acemoglu and Robinson (2012) – the creation of processes of inclusion within virtuous circles of superior institutionalization – far too far ahead of itself. How can we advocate abandonment of ‘inferior’ institutions before we understand what they are about?

The Ebola crisis in West Africa in 2014–2015 provides a clear example. Corpse handling procedures created infection pathways for this deadly virus. International responders thought the answer to the problem was simple. Ban funerals and dispose of corpses in a hygienic manner. In Sierra Leone (the worst affected country), this was a social disaster (Richards 2016). In a subsistence rice farming economy where labour mobilization was a major challenge, a funeral is an essential ritual in reigniting inter-family social cooperation in the face of death. Attendance is an obligation that permits no exception. Without participation in funerals, one young man told us, selfishness would become rampant. The social world and the economy would both collapse.³

In the end, the authorities had to come to terms with local institutional demand. Protocols for safe handling of the sick at home ‘while waiting for help’ were eventually introduced (Richards 2016, 116–117). Community burial teams were trained and equipped to carry out safe burial. Attendance by families and pastors or imams at funerals, even if with social distancing, was permitted. In short, institutional accommodation was contrived through ritual innovation (Mokuwa and Maat 2020).

The Douglasian map was useful in pointing out to the conflicted parties that local funerals and Ebola safe burial were institutional worlds apart – local funerals inhabited the enclave quadrant and safe burial the hierarchical quadrant. Thus, there was a major institutional gap to be overcome, but not (pace Williamson) one that could be overcome with money. The transaction costs of compromise had to be calibrated in a ritual currency.

The point of Douglas’s map was to demand respect for all the institutional quadrants. Let us take a look at the map again and take note there is one quadrant about which little is said – isolate ordering. It would be good to open up that quadrant to closer inspection. My aim in this paper is to build on the limited literature of isolated ordering.

My starting point is with Banfield’s book on what he termed ‘amoral familism’ (a degree of rural impoverishment that turns the peasantry in on itself). His book has been highly criticized, and perhaps rightly so. It is weak on context and (in particular) fails to take adequate account of prior mass migration out of rural southern Italy in damaging the social fabric of peasant life in that region.⁴

Even so, and despite these manifest weaknesses, it was among the first attempts to offer an explicit picture of isolate ordering and thus notably enlarged the picture of social possibilities, as was well noted by Wildavsky (1994). Later commentators do not question the verity of the social picture offered by Laura Fasano Banfield’s field interviews, even if the explanation in terms of ‘ethos’ is weak (Putnam, Leonardi, and Nanetti 1993, 88). Isolate ordering is something I shall argue we need to address if the social shocks administered to peasant populations by armed conflict or chattel slavery are to be fully understood.

In undertaking a critical re-thinking of isolate ordering, however, I find myself required to reflect on Banfield’s choice of the term ‘amoral’. Why was peasant introversion without morality (presumably Banfield meant neither good nor bad)? Here, I will seek to show (contra Banfield) that on its own terms isolate ordering can be regarded as highly moral. In some circumstances, such as self-emancipation from slavery, or in community recovery from disaster and war, social isolation may prove to be a crucial protective asset, beneficial not only to individuals but to the social fabric more broadly. Accordingly, I will argue that public authorities in zones of recovery from disaster and war should seek to offer isolate ordering respectful accommodation.⁵

The Banfields in Italy: the moral basis of a backward society

Peasant life in the commune of Chiaromonte, in the Province of Potenza in southern Italy, was (according to Banfield 1958) characterized by exceptional social and political inertia. This was not the case (the author noted) in many peasant communities across the globe. ‘Montegrano’ (Banfield’s pseudonym for Chiaromonte) was an extreme case summed up as follows:

In the normal course of their lives … Montegranesi have little or no direct contact with the world beyond their town. Most of the townspeople make no use of the highways and other communication facilities. They do not take the bus. They do not read the newspapers. They do not go into the bars to listen to the radio. They send no telegrams, make no telephone calls, and, except on rare occasions, write no letters … . (44)

In Chapter 5 Banfield then develops what he calls ‘a very simple predictive hypothesis’: that the Montegranesi ‘maximise the material, short run advantage of the nuclear family, and assume all others will do likewise’ (85). This generates an ethos he terms ‘amoral familism’, the first rule of which is that ‘no one will further the interest of the group or community except as it is to his private advantage to do so’ (85). A further rule (according to Banfield) is that ‘ … there will be no leader and no followers. No one will take the initiative in outlining a course of action … and if one did offer leadership, the group would refuse it out of distrust’ (180).

There is no peasant leader in Montegrano, and never has been, or so Banfield claims. The reason is made explicit in Chapter 7. ‘In the Montegrano view, a man is under the necessity of contending against brutal and capricious nature for the survival of his [nuclear] family’ (129). In other words, the local peasant has internalized a moral order in which a harsh nature bears down on all, squeezing out space for cooperation. Why cooperate beyond the immediate family when the neighbours cannot be trusted? All such efforts are doomed to failure by low social integration and strong environmental regulation.

Banfield’s pessimism is doubtless overdone but was influential. ‘Amoral familism’ is recurrently referenced in Robert Putnam’s widely-read book, Making democracy work (Putnam, Leonardi, and Nanetti 1993, 88, etc.) as something against which an opposed resource, ‘social capital’, could be defined, and for a time, lack of social capital was widely cited as an explanation for poverty and underdevelopment. But Banfield never explained what generated this kind of familism, beyond recourse to a kind of environmental determinism.

Doubtless, as noted above, this reflects weaknesses of contextualization.⁶ One wonders what lingering feudal deference or the damage caused by a recent World War might have contributed towards shaping local peasant attitudes? We are not told. Banfield also admits he had no time to explore the larger peasant holdings a greater distance from the centre of the community. Might their inclusion have modified his portrayal?

For these reasons, the study seems infuriatingly incomplete. For some, the book is best consigned to history, but for others, and not least some specialists in rural Italy, it remains a source of ‘never ending debate’ (Ferragina 2009). The source of that fascination, I would suggest, is not the book but the existence and meaning of the condition it first postulated – isolate social ordering. This remains a challenging blank on the map of social theory.

Escape agriculture: Haitian marronage

James Scott (2009) more recently reignited interest in isolate ordering as a form of moral economy by characterizing peasant worlds beyond the reach of states as being sustained by the development of what he terms ‘escape agriculture’. Peasants have an interest in covering their tracks (and hiding their farms) because isolation spells autonomy. This represents a major development in Scott’s thinking, since the peasant is no longer seen as being motivated by risk minimization considerations alone (Scott 1976). Autonomy is now seen as a virtue. It is a moral good to be free.

To probe isolate ordering, I might well have drawn on some of the ethnographic material Scott puts forward in his book on the autonomous communities of highland SE Asia. Specifically, hill rice agriculture in SE Asia was no relic from pre-history, but a newly elaborated type of agriculture designed in opposition to the lack of autonomy inherent in intensive wetland agriculture. It was a means to escape the state.

Instead, however, I choose to pay attention to the montangard Haitian maroon communities described in a recent book by Johnhenry Gonzalez (2019) where a Scott-like argument is advanced, but in which escape agricultures are demonstrably more closely linked to West African peasant experience. This will make a stronger connection with my main case study below, based on isolate ordering in the Upper West African forest under conditions of war.⁷

The Haitian revolution remains the only instance of a successful slave uprising leading to the formation of an independent nation-state. Saint-Domingue was the French half of the island of Hispaniola. In the eighteenth century, it was a prosperous colony based on sugar plantations dependent on slave labour. Toussaint L’Ouverture, a plantation slave turned militia commander who fought a brilliant guerrilla campaign against the slave owners, initially in defence of French republican interests, was a key figure in driving along the Haitian revolution (Hazareesingh 2020).

Intrigue in Paris by those seeking to restore slavery in Haiti stirred Napoleon to take steps against Toussaint and he died in captivity in France in 1802. The horse was out of the stable, however. Other commanders replaced Toussaint and continued his work of removing the planter aristocracy by abolishing the plantation system and adding a definitive break with France. Haiti became an independent nation.

Toussaint’s successors then wrestled with rebuilding the Haitian economy in the absence of slave labour. The problem was generic in the Caribbean with the spread of abolition in the aftermath of the Haitian revolution. As Gonzalez puts it,

the ‘British and Jamaican authorities can be compared to French and early Haitian authorities insomuch as they were all generally blind to the economic potential of peasant agriculture because of a wilful refusal to consider alternatives to the failing sugar industry’. (Gonzalez 2019, 161)

The insurrectionist leaders of independent Haiti sought to rebuild the plantation economy without slavery. Toussaint himself had positive memories of the sugar plantation where he was born a slave, perhaps because, skilled with horses, he escaped field labour to work as a trusted coachman (Hazareesingh 2020). He fervently wanted slavery abolished but at the same time tried to tie newly freed waged workers to their home plantations.

Those who were emancipated from field toil had other ideas. They wanted to work for no one but themselves, and certainly not under conditions that restricted where they could live. Many headed for the hills to seize their own land, joining others who had remained there as self-demobilized guerrillas. This cadre of independent-minded maroons turned unoccupied forest into a landscape of strongly self-reliant subsistence agriculture through merging local invention and African agricultural ideas.

Gonzalez notes that such maroon agrarian landscapes have a long history in Hispaniola (158–159). An uprising of Wolof slaves from Senegambia on a plantation belonging to Columbus’s son Diego Colon can be traced as early as 1522. Runaways then became roving cattle hunters and farmed small conucos (subsistence plots) in the hills. ‘Even as late as the early 20th century monteria [a lifestyle based on capture and slaughter of feral and semi-feral swine, goats and cattle] remained one of the main modes of subsistence in the rural interior’ (158). In regard to Haiti, specifically, Gonzalez concludes that ‘centuries of rough, rural freedom in el monte was the historical outgrowth of early colonial patterns of marronage and metropolitan neglect’ (159).

Marronage leaves few records, except in the form of techniques of subsistence designed to reduce reliance on others. Gonzalez reads these signs expertly.

One of [the] most enduring strategies has been a seemingly haphazard system of intercropping. Outsiders walking into a [Haitian] conuco might not even realize that they had left the jungle and entered a farmer’s field … By farming amid deep jungles and on high mountain peaks, and by avoiding the sort of seasonal mono-crop harvests that would make their labour legible and thereby taxable by the state, the Haitian farmers strove for freedom and domestic autonomy. (235)

Invisibility is the essence of ‘escape agriculture’.

More than a technical signature is needed, however, to fully substantiate a moral economy of isolate ordering. The evidence needs also to be social. Locating the social agency in marronage is a crucial test. We need evidence that social agents were actively creating a world of minimal social contact, protected by a distinct moral code.

Gonzalez provides some relevant evidence when he discusses the part played by sodality organization. The maroon communities of Haiti were extensively organized around ‘secret societies’, many of which offered in their names and practices evidence of West African origins. He then notes that

perhaps paradoxically, the most prominently militaristic of the African maroon organizations [the ones on which Toussaint and other revolutionaries often leaned for support] seem to have been the most quickly eradicated [when the Haitian revolutionaries tried to wrest back control of the interior], whereas the most evasive and secretive ones have endured. (Gonzalez 2019, 144)

One of the most enduring of these evasive secret organizations was Bizango, a sodality named after the Bissagos islands off the Upper Guinea coast. The islanders were enslaved by the Portuguese from the late fifteenth century and may well have developed their sodality organization as a technique of resistance. As Gonzalez puts it, ‘the small-scale farmers of the Bissagos islands had grown up in the original offshore university of marronage’ (145).

Citing the work of Haitian anthropologist Michel Laguerre, Gonzalez then suggests that these intensely secretive kinds of sodality offered protection against government officials or anyone else who might try to take away the land. ‘Societal songs warn members not to talk loosely about the society, its members, or its activities’ (149). When the threat is most extreme, it may be best to melt away, family by family, into the bush. Extreme stratagems, such as walking backwards through muddy sections of a track, to disguise the direction of travel, were adopted. Contact with strangers was discouraged. If talk is a kind of social grooming (as Dunbar 1993 has suggested) then even talk is an existential threat. Fading down the community to nothing becomes a means to protect the community.

Listening to silence: approaching African isolates

This silence fascinated Mary Douglas.⁸ She realized that those enculturated within specific modes of social ordering had their own distinctive ways of communicating their norms. For isolates, knowing when to remain silent is a first requirement of morality. Ritual can carry the load when talking stops. How does the social analyst approach rituals of silence? Douglas’s answer was ‘comparatively’.

She quickly deduced that supposed ‘inarticulacy’ often had a ritual form, but at first lacked tools to make much use of this intuition. She had been trained in colonial-period anthropology where divide-and-rule dominated administrative practice, and difference mattered more than interconnectivity. Ethnography tended to proceed one tribe at a time.

In her first academic job,⁹ however, she was exposed to the ideas of Daryll Forde, and his Belgian student, Jan Vansina. Both were interested in wider, comparative questions raised in neighbouring disciplines (history, geography, archaeology and comparative linguistics). This provided a supportive context in which to spread her net more broadly.

She tackled the question of economic comparison by assessing the Lele of her own African field experience with the neighbouring Bushong (Douglas 1962), a differently organized group with a different work ethic. Vansina had worked among the Bushong, and doubtless this spurred her to reflect comparatively.

With further fieldwork ruled out by war in the Congo Douglas subsequently developed a comparative ethnographic approach via work dominating her later years on competing and overlapping moral codes elaborated by two socially distinct groups of occupants of ancient Israel – those who had continued to live on the land during the Babylonian captivity and those who returned from this captivity (Douglas 1999b, [1992] 2001, 2004b).¹⁰

The comparative framework she felt she needed for this later work on the ethnography of the Hebrew Bible was first suggested by the work of Basil Bernstein (Douglas [1970] 1996; Perri 6 and Richards 2017, 52). Specifically, Bernstein’s notion of codes of communication suggested to her that traits such as deference and inarticulacy might not be deficiencies but manifestations of moral orders with their own internal logic.

All moral orders are exercises in restraint, but different orders impose different restraints. To capture this for ethnographic comparison, some kind of mapping of a range of moral styles of presentation would be necessary. The basic idea¹¹ (and diagrams) were first presented in the book Natural Symbols (Douglas [1970] 1996), but its reception was poor. The book was seen as no adequate successor to her widely praised book Purity and Danger, and one reviewer thought it ‘Catholic propaganda’ (Perri 6 and Richards 2017, 55). Douglas herself, admitted she had been less than clear in her aims, and it was not until some years later that she fully refined the simple, deductive account of the quadrants of social ordering summarized above (Douglas 1999a).

With the aid of her fourfold map Douglas was able more clearly to distinguish between different modes of moral ordering, and (as importantly) to separate out independent and dependent variables. Thereafter (and following Durkheim) ritual action was always the independent variable and moral ordering the result. Ritual action was what people collectively did; the culturally-expressed representations (totem, flag, or style of identity) were always what they made. We believe because we pray, and not the other way round.

At a stroke, she had cut herself free from much of the cultural determinism that dogged her discipline. But at the same time, she cut herself off from much of her anthropological audience, not helped, as Robbins (2012) remarks, by a general turn in the discipline in the 1980s away from social theory towards discursive modes of explanation more typical of the humanities.

The work on the Hebrew Bible was her solace in retirement. Through it she sought the evidence of moral conflicts in the editing of ancient Biblical texts. There was a problem, however, as pointed out by the Biblical scholar Lester Grabbe (Heald and Grabbe 2004); were these texts scholarly idealizations, or did they reflect different and distinct lived social realities? As with Gonzalez’s work on nineteenth-century Haiti, it is hard to know the truth from historical records alone, more especially when these records have been heavily shaped by representatives of some but not all the moral orderings present on the landscape.

Douglas’s late work is underwritten, therefore, by a sense of missed ethnographic opportunity. She returned briefly to the Lele in 1986 and discovered a society in the throes of working through major moral conflict involving Christian mission values and older concerns with witches (Douglas 1999c) but it was now too late for further African fieldwork. Others would have to take up the challenge.

Threaded through her final work (notably the collection of essays published as Jacob’s Tears [2004b]) is a concern for the clashes of social values apparent in the African ‘New Wars’ in the 1990s. At times, there is a palpable sense of post-colonial anger (Douglas 1999c, 2004a). In particular, she detected and called out a tone of moral self-righteousness in many Western accounts of these apparently ‘mindless’ African conflicts and was anxious for more attention to be paid to the entangled multiple moral economies sustaining them. Her final public words in 2007 via an interview in the Spectator magazine about how to address the crisis posed by militant Islam thus constituted a serious warning (6 and Richards: 9). We needed to probe very carefully for distinct and at times muted or silenced institutional voices beneath a cacophony of war-induced claim and counter claim, in order to detect separate and disjunctive strands of organization and world-making, without recognition of which long-term accommodation would be unattainable.¹²

An African civil war: Sierra Leone, 1991–2002

This brings us to a search for the silent world of the isolates on the war-devastated terrain of rural Sierra Leone. In stereotypical terms queried by Douglas, the civil war in Sierra Leone (1991-2002) was fought either for ‘greed’ or ‘grievance’ (Humphreys and Weinstein 2008). But its moral alignments, or lack of them, continued to puzzle analysts. Much effort was invested in trying to discover whether economic bans on ‘blood diamonds’ or militarily-enforced application of global human rights and (so-called) transitional justice made the misery and suffering of that war any the less (Smillie, Gberie, and Hazleton 2000; Kelsall 2009).

When the Sierra Leone war broke out, I was working in a village on the Liberian border, in the first area to be overrun by rebel forces. For a time, I had to pause village-level ethnographic work. Douglas, as a mentor, encouraged me to re-focus on my earlier material, to see if I could read (or re-read) in those findings some of the hidden moral alignments that might help explain behavioural facts of the civil war.

This earlier material derived from fieldwork in an off-road farming in the centre of the country (Richards 1986). Here, I had initially traced out two main organizational and moral strands. The first was the voice of hierarchy, expressed by the colonial administration and later by successor institutions such as the World Bank, requiring peasant farmers to align themselves, organizationally and in terms of values, with a global knowledge hierarchy – the seed-based revolution in rice farming techniques emanating from Asia, introduced initially via British colonial links with India and Burma, and latterly (under the Cold-War) by American private research foundations linked to Taiwan and the Philippines, sustaining an explicitly anti-communist (so-called) Green Revolution.¹³

The second voice, dominant in village society, was one of cooperation, forged through intermarriage among land-owning families (comprising about 70% of the population of the Mende-speaking rice-farming village where I worked). I painstakingly documented how seasonal labour bottlenecks were key social facts sustaining the village moral economy. Everyone was short of labour at busy periods. Whether a family was food secure or lived at risk of hunger the following year depended on how well labour bottlenecks were overcome.

Intermarriage was a key factor. Younger men worked for their fathers-in-law. For this purpose, they often ganged together into rotational labour groups. I joined one such group called Mbla (literally ‘father-in-law’). Morally (if I had been any good at the work), this would have provided me with an entitlement – bride service – to marry into the village. I would then have continued to work to support my wife givers for the rest of my life. Inter-marriage forms a tight circle of interdependence based on the exchange of labour and services.

This sustained what Douglas, referring to her map of moral quadrants, termed ‘enclave social ordering’ (Douglas [1992] 2001). A key feature of the enclave (in the case I described) was that the work-marriage circle was locked against outsiders unwilling or unable – e.g. through disability, weak land rights or time given up to other concerns, such as trade – to commit to the necessary society-forming actions. In turn, the circle of marriage was fiercely protective of those within it.

In the village, subsistence agriculture was equally resistant to the competing claims of trade (under which a moral order of individualism would have been cultivated) and the newly-introduced post-colonial system of Green Revolution wet-rice planting under which a moral economy grounded on principles of knowledge-based scientific hierarchy might otherwise have been fostered. Thus, local rice farmers held both individualism and hierarchy at bay.

Isolate ordering and the path to war

My book on this fieldwork (Richards 1986) documented various ways in which traders were levelled-down and also traced out local experimentation undertaken to attain independence from external knowledge sources. The fieldnotes on which it was based were full of discussions with villagers on why these responses were morally defensible, offering a view of village life which might, justifiably, have been criticized as functionalist. However, I had reckoned without the disruptive potential of isolate ordering, and the practical logic sustaining it.

The reality of what I had missed became apparent when the ‘rebel’ war broke out in 1991. Rebel attacks on village chiefs and rice traders required me to undertake an urgent re-assessment of a morally devastated landscape. Had the levelling down gone too far? The war was promoted by a poorly armed assailant (the Revolutionary United Front, a Gaddafi-supported sect) lacking the weapons and technique to mount conventional military attacks. The army pushed the insurgents back into the forest but lacked the training or motivation to properly end the conflict. As montagnard guerrilla commanders had found in revolutionary Haiti, the rebels fighters of Sierra Leone discovered ‘the forests and mountains have neither locks nor keys and we can enter [at will]’ (Gonzalez 2019, 154) and they began to wander where they chose, seizing subsistence and materiel along the way.

Local populations streamed out of the villages and headed for urban centres where protection was provided by international peace-keeping troops. Attention on the war in Sierra Leone was thus firmly focused on the huge number of internally displaced persons. For a time, Sierra Leone hosted the world’s largest humanitarian crisis.

Invitations to assess relief efforts took me into this world of urban camps, where I met some of my friends from the village. They told me something not apparent to the wider world. Not everyone had been driven out of the countryside. Some families adopted a two-pronged survival strategy. Some of the women and children were sent as refugees to town to gather relief supplies for the family, but other members stayed behind. In fact, these other family members went in an opposite direction, out from the village to secluded camps in the bush, reviving an old organizational modality from the days of the slave trade, so that they could better protect their land.

Meanwhile, the rebels had established their own fortified camps on riverine islands and escarpments, or deep within forest reserves, from where they raided more widely as conditions or lack of supplies dictated (Richards 1996). This meant that those villagers intent on protecting their land could survive in the bush for long periods unmolested, provided they took sufficient trouble to disguise their presence. The word in Mende used to describe this secretive lifestyle in the bush was sokuihun (corners). People now lived in corners.

These out-of-the-way nooks saw the return of a number of the historically attested practices characteristic of West African maroon sociogenesis (Mouser 2007; Rashid 2000; Sarro 2009). Specifically, corner dwellers lived well away for roads and took care to disguise the direction of travel when moving about (for example, walking along muddy tracks backwards). Fires were lit with great caution, not to give away positions. Careless talk was avoided. Essential to protection was a deep distrust of all but the immediate family, not least of fellow villagers who might be motivated by petty jealousies to betray them. Accordingly, traders, officials and other villagers were shunned. Freedom from harassment depended on closing down much of pre-war village social life; isolate moral ordering ruled.

Henceforth, I began more fully to understand the complexity of the moral landscape over which I had been working. Elements I had so far missed were essential to the mix. Isolate ordering was absent in my earlier account, but now I recognized it had been present all along but in abeyance. To disregard one of the orderings of Douglas’s institutional map was in effect to conspire in the erasure of a mode of existence that to some people was the difference between life and death. I began to comb previous fieldnotes for evidence of, and to ask new questions about, isolate social ordering and the moral economy it sustained.

I made a start by revisiting material on marriage. As noted, enclave ordering was woven together by marriage contracts. At one point, I surveyed all the marriages in the village. Seventeen of 98 marriages were with the family of the mother’s brother, and 15 of these were located in the oldest quarter (tawoveihun, ‘old town’ in Mende). Cousin marriage constituted the solid core of village society. This made it clear just how tightly the enclave was interwoven at its heart. It was hard for outsiders to join, since it required replicating a marriage pattern formed over successive generations.

It was important to ask what happened to those who could not or would not marry under village rules. The results of interviews by Humphreys and Weinstein (2004) with about 400 demobilized cadres of the rebel Revolutionary United Front provided a clue. Most of these recruits came from Mende-speaking villages in the south and east of the country. Many had been captured when rebels overran villages and forced to join the movement. When asked about benefits of the war few mentioned diamonds, the supposed driver of the conflict. However, a surprisingly large number – nearly a quarter – stated that they had been able to marry. The movement captured young women and redistributed them as partners to male fighters (Coulter 2009).

This raised the question of why previously, so many young people in rural Sierra Leone had found marriage so difficult to achieve. It turned out the problem had a long history in rural Sierra Leone, reflecting high rates of polygyny. The British missionary scholar, Kenneth Crosby undertook a survey of 20 small towns and villages in eastern Sierra Leone in the mid-1930s, finding that 51% of married men had two or more wives (Crosby 1937). In all, 842 men were married to 1,973 women. Only 82 women of marriageable age were unmarried, but there were 673 unmarried men of marriageable age. The upshot of this was that many men of marriageable age had no reproductive partner and found female ‘friends’ where they could, especially among a large number of wives of high-ranking chiefs.

Making a girlfriend (nyapoi) of a married woman was likely to lead to a court case in which the husband sued for ‘woman damages’. Analysing local court records in four chiefdoms in rural Sierra Leone, Chauveau and Richards (2008) report that such cases were the second most important class of local conflict after debt. These fines offered material redress to a man who had paid bride-service for a woman only to find she had gone off with another man who had paid her parents nothing. The possibility of suing for such damages strengthened the enclave, since it punished freeriding on exchange marriage, the basis of the circle of enclave solidarity.

But it was also a mode of redress open to abuse, especially in a context in which some men had (in Crosby’s term) ‘unsupervised’ partners. Crosby suggested that polygynously-married men short of labour for large family farms allowed neglected wives to encourage or entrap younger unmarried men and then jumped on them for damages. Generally, the accused would be unable to pay, and the sum would be commuted to farm labour. Villagers explained to me that money lenders sometimes attended court sittings, offering to buy the cases of those found guilty. They then sold on the labour of those forced into taking loans to pay their fines, constituting a kind of judicial servitude.

The chances of defending against such a case were slim. Chauveau and Richards (2008) note that a guilty verdict was returned in 94% of such cases in three sampled Mende-speaking chiefdoms in southern and eastern Sierra Leone. In one court in 2004, a chief had successfully sued five young men in succession for affairs with one or other of his several wives in a matter of months. Perhaps not coincidentally, the plaintiff was a large planter of cacao and was perhaps mainly looking for labour to brush his plantations.

Mokuwa et al. (2011) offer econometric evidence pointing to an association between the local frequency of these cases and the peak of labour demand on upland rice farms. They also note high numbers of cases co-vary with the amount of upland rice farming per chiefdom. This links woman damages to upland rice farming in chiefdoms where marriage-based exchange labour remains a factor in managing farm labour bottlenecks.

But how, then, does this connect to the issue of isolate ordering? A number of ex-combatants interviewed about the reasons for them to become associated with the RUF provided some clues. They frequently stated that excessive fines had driven them from their villages into the arms of the rebels (Richards 2005). These fines were often incurred in cases demanding woman damages.¹⁴

I then re-read some of my original fieldnotes from my first period of fieldwork in the 1980s. These reminded me that I had followed a case for woman damages over a period of several dramatic weeks, leading to the young man concerned fleeing the village to avoid his fine. I then lighted on a second set of notes taken with a young man who had explained to me how to avoid being trapped by any such accusation. He knew these cases were often manipulated; to avoid any ‘big person’ (numuwai) bent on turning a youth with weak family protection into an unpaid farm labourer it was necessary to retreat to the bush.

This young man became a friend. He and his wife and children had decamped to a remote farmstead on the edge of the village land. An item overlooked when writing my book speculated that he belonged to a fraction of the village population, perhaps descended from slaves and clients in earlier times. It is likely that there are such individuals belonging to all residential quarters in Mende villages, but slave origins are rarely discussed, and so how these people are related to the main land-owning families remains unclear. Some (like my friend) are clients of ‘big men’.¹⁵ They acquire a marriage partner and access to land with the help of a local land-owner known as a ‘stranger father’ [hota kee].

Those without patron and protector are liable to leave the village and migrate to towns, where they can earn daily wages and escape village predation on their labour. My friend, however, could remain. He had a protector, and access to land to feed his family. His concern, then, was to prevent himself from being framed for woman damages, so that his labour remained his own. He needed to avoid any open rift with his protector. One way was to retreat to his farm, and lie low, whenever possible, focusing solely on farming, and especially not paying visits to town during the high-risk season when heads of large family farms were hunting for extra farm help.

The rice farm was at some distance from the village, and off the beaten path, so my friend built a rudimentary homestead, with a sleeping platform, and remained there as much as possible. The couple used their time to experiment extensively with different rice seeds, thereby identifying hybrid seed types and gaining abundant harvests. In fact, they first helped direct my interest to this important but neglected topic, and I visited them on numerous occasions.

I asked why others did not follow this example. My friend said some did, in order to avoid becoming enmeshed in court disputes. But others took no such care. Once their crops were planted, they openly courted disaster by coming back to the main village and taking part in political intrigues as followers of rival taa gbakoi [‘important citizens’]. It was easy enough to cook up a case against these camp followers; cases were decided on testimony, and false testimony came cheap. Soon enough, a victim of entrapment might be found working for others.

Consciously taking the opposite course, my friend reduced his social and political agency to a minimum to protect his only tangible economic asset – the freedom to work for himself and his family and thus provide for his children without hindrance. In terms that would have seemed sensible to poor peasants in Montegrano, or in the montane Haitian interior, my friend explained to me that he and his wife consciously rejected forms of a community association in favour of a focus on their immediate family. The reason for isolate ordering was clear. Specifically, it was to gain protection from asymmetric societal obligations. Banfield was ill-advised to term this ‘amoral familism’, since to those who practice it, this family-protective social asceticism is moral to the core.

A moral landscape falls apart and is reintegrated

In 1997, with the collapse of the Abidjan peace process, the war resumed even more viciously than before, and I had to abandon fieldwork for a further period of about two years. The villagers who had earlier made me a welcome guest were certainly in no need of a visiting anthropologist to advertise their presence in sokuihun.

Eventual return to the field in 2000 was facilitated by a colleague working as a policy advisor for an international humanitarian agency. His organization had chosen to locate its forward base for serving hard-to-reach groups affected by the war in the area of my first fieldwork. He wanted me to attune agency activity to a fuller range of local expectations; the danger he foresaw was unwitting partisanship in relief distribution.

To distribute relief on this disassembled landscape without risk of re-inflaming war, the moral terrain first had to be carefully mapped and attention paid to disputes over relief distribution. We found that competing claims for humanitarian support readily mapped on to the four distinct varieties of local organization predicted by Douglas’s organizational map, and this became our guide on how to approach and solve dilemmas of relief (details are provided on Archibald and Richards 2002a, 2002b).

The agency’s forward base was now a market centre, populated by displaced people and migrants who returned from the city to access relief supplies, and characterized by competitive individualism. Relief supplies were openly traded. Over the wider region, army units and international peace-keeping forces exerted rudimentary hierarchical control on behalf of a distant government via checkpoints and infrequent patrols over the few motorable roads. Enclave mentalities remained strongly entrenched in villages and bush camps controlled by local civil defence and rebel militia, reflecting the local battle lines. Finally, we encountered scattered and disconnected clusters of corner dwellers living under isolate ordering.

We then asked groups to organize the distribution of relief supplies in accordance with what our map of the moral terrain told us. In villages organised by marriage, for example, the first names on the distribution lists were always the leading elders. It was explained that this was because they best understood the complex pattern of marital interdependencies around which food redistribution needed to be organized. Internally displaced people were excluded. They were strangers, outside the scope of the enclave system, and might (it was suggested) go home to seek relief. Enclave ordering trumped other social arrangements.

Youth militias found themselves excluded under arrangements that suited those under enclave ordering, and some threatened to decamp to the bush. Signs that some were already beginning to set up camps modelled along rebel lines told us this threat had to be quickly defused, and we convened community discussions on why the war had been fought, and what might now be done to prevent young people from being sucked into combatant factions again. Agreement was quickly reached to make distributional rules more inclusive.

A major problem was solved when rebel militia camps came under the scope of the national programme for disarmament, demobilization and reintegration (DDR). This was a responsibility for international peace-keepers, who mainly operated from urban centres. The benefit package was generous, and those combatants who still saw themselves under some kind of hierarchical command were keen to find a way into DDR. As a result, many fighters left the area and congregated in towns, where chances of admission to the DDR programme were better. This made it easier to handle remaining distributional conflicts, now mainly confined to local residents.

The agency itself was keen on the idea of reverting to market ordering and influenced by Amartya Sen’s theory of entitlements (Sen 1981) we worked out a way of registering family members separately so that every individual could take an entitlement and use it as she or he saw best. We started with seeds but recognized that as the local economy kicked into life cash distributions might be better. Retrospectively, however, this modality was a mistake, in that it placed too much trust in the agency’s own preferred moral ordering – competitive individualism. Market logic might complement family-based distribution but was powerless to reach the isolates, where some of the risks of immiseration were highest.

Those living beyond a certain line were deemed ‘rebel collaborators’, both by enclaved villagers and by those resuming life under individualistic ordering, irrespective of whether these groups were associated with the rebels or living in sokoihun under isolate ordering. Reintegration stalled. What seemed to be required was some public recognition of isolate ordering as a morally defensible local lifestyle, distinct from being organized as a rebel enclave, thus making its own distinctive contribution to the public good. Ways of talking about this would require attention to some of the most isolated localities, but we lacked intimate knowledge of their past.¹⁶ These were not issues ever discussed with outsiders.

In the event, local public authority came to our rescue, since many of chiefs driven out by the war were now returning, and anxious to prove their worth to sceptical and abandoned survivors of the war. A Mende chief is admired for being able to bring conflicted groups into some kind of reconciliation (Kamara, Mokuwa, and Richards 2022). Chiefs proved adept at accommodating different forms of social organization on this shattered terrain. Specifically, they were well aware that at times of particular difficulty it was an option for some of their subjects to retreat to distant farms, opting for isolation over community collaboration.¹⁷ Those with distant farmland were sometimes descended from former slave families, historically posted to the outskirts of the village. Chiefly mental maps of moral terrain thus already included isolate ordering, and it was enough to nudge this memory back into life. Allowances were made and aggravations began to abate.

In Douglasian terms, it is a task for public authority to hold open a space in which all four basic organizational and moral elements can achieve co-existence, both to provide for options, but also, eventually, to generate complementary support, conducive to what Durkheim termed ‘organic solidarity’. Durkheim’s own account of organic solidarity (Durkheim [1893] 1984) might be considered makeshift (Smith 2020). Perhaps we need to accept that solidarity is work for the people themselves, typically achieved by ritual means (Warfield Rawls 2005). An instance in Sierra Leone is when Christian and Muslim families prepare and share appropriate food for the religious festivals of neighbours of a different faith, thus enacting commitment to neighbourly tolerance. Public authorities demonstrate commitment to durable peace by encouraging ‘gifting’ across socially significant boundaries. Without huge amounts of ritual work of this sort, directed at all groups, peace will remain fragile.

An acid test of whether the process is working is how well isolate ordering is tolerated, since this is the fall-back position whereby the most impoverished or oppressed groups retain agency. The current pandemic of Covid-19 provided such a test. In 2020, government lockdown was imposed across Sierra Leone to try to halt the spread of the disease. A message from the chiefdom in which I worked told me that the stay-at-home order had been given in two versions by the chief. Either people stayed in the village, to subsist on what they had in store, or they left for their farms, to depend on the affordances of the bush. Some took this second option. Perhaps a willingness to offer options could be taken as a sign that a moral landscape disassembled by war was becoming a more complex space sustaining a greater range of social and moral codes conducive to peaceful co-existence.

Conclusion

Replacing the monoculture of competitive individualism by a more inclusive map of the moral landscape results in complexity, but it is futile to attempt to reduce this complexity by trying to determine which moral forms are superior and inferior. Each moral code is grounded in a material system of economic and social organization, and thus has its own reasons to exist. To assume (following Douglas) the potential co-existence of a limited plurality of elementary forms of moral ordering is a better framework, since this offers a range of options for oppressed or marginalized groups.

Argument is then less about the relative merits of moral codes and more about finding accommodation and tolerance. To ignore or seek to prohibit a mode of morality is to proscribe a material way of life. It denies human agency. For this reason, slavery can never form an acceptable or stable social system. As a human agent, the enslaved person cannot but seek to resist. It follows that it is equally unacceptable to take the next step and prohibit or ignore the moral economy of isolate ordering, since this closes off the maroon path of escape.

The later part of the paper has illustrated some of the ways in which war in Sierra Leone disassembled the moral landscape into its component parts. Post-war recovery then reassembled the parts. This process reveals some of the underlying dynamics of social formation among affected communities. The capacity of local public authorities to assist the process of reassembly is never exact. Effective intervention often grows out of tolerance of alternative organizational and moral pathways.

The tragedy of Haiti was that revolutionary leaders, caught between competing demands of royalism, republicanism and mercantilism, never found ways to build politically on the ‘escape agriculture’ nurtured by their ‘university of marronage’, the institution shaping the moral values of a significant percentage of the interior peasantry (Gonzalez 2019).

Sierra Leone was perhaps more fortunate. The reintegration of the rural areas after a decade of community destructive civil war in the 1990s was largely left to village factions to resolve, even if some humanitarian agencies sought to impose a moral agenda of their own. Isolate ordering played an important and positive role in this local recovery, by allowing those not yet confident in the motives of the post-war state or of the peaceful intentions of former combatant forces time and space for self-recuperation. It is a credit to chiefly public authority that it did not seek to control this space.

A non-judgmental approach thus held open an important opportunity for local agency to revive. Village leaders showed a surprising flexibility in terms of accommodating rival organizational forms – not least, a willingness to recognize the moral economy of isolate ordering, a mode of existence largely invisible to, or (if known) deplored by, humanitarian agencies. This offered a degree of respect for those whose only asset was their labour, in an environment where land intake and ‘escape agriculture’ remained important recovery options, as once had been the case in a post-slavery world of self-emancipation. This respect helped support a surprisingly rapid return to peace, even if much poverty and injustice remain.

Acknowledgments

Analysis, writing and Open Access publication costs for this paper were supported by CPAID grant ES/P008038/1 of the Economic and Social Research Council to the Firoz Lalji Centre for Africa at the London School of Economics, and this support is gratefully acknowledged. I would also like to acknowledge helpful advice from four anonymous reviewers (two for CPAID and two for the journal) and from four colleagues, Professors Perri 6, Tim Allen, Martha Prevezer and Ramon Sarro. The author bears the usual responsibility for errors and nonsense.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by CPAID grant ES/P008038/1 of the Economic and Social Research Council to the Firoz Lalji Centre for Africa at the London School of Economics.

Notes on contributors

Paul Richards

Paul Richards is Emeritus Professor of Technology and Agrarian Development at Wageningen University in the Netherlands, and Adjunct Professor of Environmental Studies at Njala University, Sierra Leone. He has written books on peasant experimentation, rural insurgency and epidemics. Currently, he is carrying out research on rural responses to Covid-19 in Sierra Leone.

Notes

1 Thompson (1971) gave renewed currency to moral economy, to be followed by Scott (1976); the phrase has a history traced in Götz (2015).

2 It would also have been the basis for scepticism about social capital (Putnam, Leonardi, and Nanetti 1993) since this kind of analysis offers isolate ordering a valuation of precisely zero in terms of social capital.

3 It should be added that the institution of inter-family cooperation in the face of sickness and death is so powerful that it translates to the city on migration, where there is not so severe a problem of labour cooperation to solve, since there are market options to keep hunger at bay. But institutional change should be governed by local need. Even in the modern African city the poor are also highly cooperative. Institutions change or die but should not be murdered before their time.

4 I thank one of my reviewers for suggesting the need to emphasise this point.

5 Consolidation of tentative peace settlements depends highly on informal public authority (Kirk and Allen 2022). Theros and Kaldor (2018) recognise in Afghanistan three distinct spheres of public authority they label ‘political markets, ‘civicness’ and ‘moral populism’, and explore the implications of working in a landscape of plural moral economies. My paper attempts to add isolate ordering to their list.

6 Banfield spent 9 months in the field with his family but reports that he never learnt Italian fluently; much of the ethnography depended on Laura Fasano Banfield, his wife, who was a native of the region, and acknowledged in the book as undertaking 70 interviews with peasants on which the analysis is based.

7 I am aware that many maroon societies developed long-term and more or less stable relations with surrounding and differently organised groups (Price 1983, 1996). Settled maroons perhaps ought to be located on Douglas’s institutional ‘map’ in the enclave quadrant. My own focus is on what might be termed ‘maroon sociogenesis’ where isolate ordering is of crucial concern (see Sarro 2009).

8 Today, anthropology has a literature on silence and its significance (see, for example, Basso 1970; and Dragojlovic and Samuels 2021) even if Douglas’s systematically comparative approach (Douglas [1970] 1996) has found relatively few followers.

9 In the Anthropology Department at University College London.

10 This took off from, but eventually turned on its head (Douglas 1999b), the celebrated chapter on the abominations of Leviticus in Purity and Danger.

11 First proposed by Durkheim (6 and Richards 2017, 214).

12 Douglas was well read in philosophy though she rarely engaged in philosophical debate. Perhaps the most influential philosopher in her last phase was Nelson Goodman (1978), in whose honour she and David Hull edited a set of essays, How classification works: Nelson Goodman among the social sciences (Douglas and Hull 1993).

13 The heart of the Asian Green Revolution in rice was, of course, the International Rice Research Institute in the Philippines, but the core technology (double cropping of non-photoperiodic rice) had first been worked out in Taiwan, using ‘temperate’ japonica rice types.

14 Young men most likely to challenge enclave social ordering were often migrants whose employment in mining had ceased. Additionally, their land rights were often insecure (Mokuwa et al. 2011). On migration and African wars more generally, see Sommers (2015).

15 One Mende village quarter surveyed by Little ([1951] 1967, 102) comprised 104 people, of whom 30 counted as patrilineal kin, 10 were members of the mother’s brother’s lineage, 32 were affines and 32 were ‘strangers’ (attached to the quarter through a senior man standing as a sponsor).

16 The Scottish missionary, William Cooper Thomson, encountered such a situation on route from Kukuna to Futa Jalon in 1843; a locust plague had driven local Susu landowners to lodge in the outpost villages of their slaves, where food remained available (Thomson 1846).

17 Alldridge (1901) offers a striking historical instance. In 1893 a meeting was proposed between the British governor and up-country Mende chiefs. The chiefs threatened to leave as soon as they had arrived due to an unexpectedly early rainstorm overnight. They explained that if rain continued their subjects would leave for their farms, upon which all community social activity would cease until after the harvest. The topic of seasonal isolate ordering was introduced into anthropology by Mauss and Beuchat ([1906] 1979).