The Method

We are a group of ten, calling ourselves the Acholi Land Lab. We are three women and seven men; one of us left formal academic education after one year of secondary school, while six have degrees and four have completed or are undertaking post-graduate qualifications; we range in age from twenty-seven to sixty-six. Six of us are, by profession, researchers, with the other four being an engineer, a full-time farmer, and two foresters, one of whom is also an artist and designer. Nine of us are active members of customary-land-holding groups in Acholi. Our landholdings range in size from 8 to 304 hectares. The smallest of the groups/families that we are part of, that own these holdings, consists of about 170 people, the largest, over 500. Luo is the mother tongue of nine of us and all of us are fluent in spoken and written English; by nationality we are nine Ugandans and one Briton. We represent different clans; our landholdings are all in western Acholi, in the districts of Amuru, Gulu, and Nwoya, though in some cases we have links to wider clans, including in the east, and in other regions. Apart from Author Four (the farmer, who lives about 30 kilometers to the southwest), our main homes are in Gulu, the largest town in Acholi and northern Uganda, which is where we meet.

In July 2022, four of us met to discuss how to move this project forward, identifying five others we would invite to join us. All of us knew some of the others before we started, but no one knew all. Some of us had worked together on research projects in the past. The key characteristics we shared were that we were active members of landholding groups and that we were critical and analytic thinkers. A tenth member, Author Two, was invited to join us at a later stage to make a particular contribution around language and hermeneutics, and to help develop our collective understanding of autoethnography. Author Five had been awarded funding for a research project exploring the governance of natural resources on customary land in Acholi. He had been given some ten hectares of customary land to farm in 2009 by a family to which he is closely connected, and this experience formed the basis of his doctoral research. Consciousness of the heavy and unsustainable use of trees on this landholding during the intervening thirteen years, albeit for mainly domestic and traditional purposes, led to the proposal for the project.

We did not approach the search for the ideal method to answer our research question by choosing from an academic basket of options. Rather, the selection of our approach was governed by the realization that we (and numerous others like us) were prime examples of possible key informants possessing the sought-for data. We were aware that we already knew as much about the topic as most of those we might interview or survey using conventional methodology, and any gaps in our knowledge could be remedied by conversations with our family members. Our research question, on how decisions are made by contemporary Acholi land-holding families, might better stand alone rather than being shaped and potentially distorted and prejudiced by questionnaires, surveys, and sub-questions. In this, as well as in rejecting a preliminary review of the literature, and in treating “all as data” we borrowed from Grounded Theory (O’Reilly Reference O’Reilly2009). However, our constitution as a group of researchers who were our own key informants allowed a very different approach to analysis, which proved to be inseparable from data collection. Our interest is in (and we represent) “people, who, historically speaking, would otherwise remain unknown, or whose stories are told by local elites or foreign humanitarians” (Ansoms et al. Reference Ansoms, Bisoka and Thomson2021:196). This, in conjunction with the fact that some of the researchers among us have viewed ourselves as ethnographers, led to us opting for a CAE identity, though there are, no doubt, alternative names for what we do. Our goal is less to develop theory than to create a narrative of contemporary customary ownership that can serve as a foundation for an empirically informed campaign to influence public attitudes in favor of the sustainable use of natural resources.

Our collaboration has consisted of six preliminary meetings in which we identified the skeleton of our process. This involved four exercises which we undertook individually: (1) a genealogy of our landholding group from its founder (in most cases a grandfather or great-grandfather, in one case a grandmother), including a census of those currently involved in and/or living on the landholding; (2) a map of the landholding using Google Earth, identifying its boundaries, size, and features visible from satellite photos; (3) a sketch map of the landholding identifying natural resources and other specific features; and (4) a written account of our place in our families and our experience of taking part in the research so far. Author Seven also serves as our IT lead, identifying suitable software for the exercises and coaching other members in their functions. Author Ten is our coordinator, scheduling our meetings and recording and transcribing them.

Collectively, we have participated in a series of ten all-day meetings in which we have taken turns presenting the materials we produced and giving an account of the governance processes in our families, often involving relationship and stakeholder mapping. These presentations have been interspersed with discussions. We identified key themes that had arisen earlier— for example “spiritual governance,” the authority of the ancestors or other spirits in a particular clan (kit me tekwaro) or Acholi-wide (cik Acholi); or “language of inclusion and exclusion” (such as calling an adopted child lutino aweno, a guineafowl, whose eggs, if found in the bush, are brought home and hatched by chickens); “language and action around boundaries” (there is a distinction between kigini, a physical boundary and wang coo, an abstract one)—and followed where this discussions led. This has been perhaps the most exciting and productive aspect of the project. As mentioned above, we have deliberately avoided the development of questions, beyond our core research question, which we see as an essential strategy in what Falola describes as “dismantling the colonial matrix of power.” Decoloniality, he writes, is:

The common language of academia and policy on Acholi land and property is heavily populated with words and ideas of colonial origin. Frank Girling, the anthropologist whose 1949 fieldwork led to the publication by the British government of his book, The Acholi of Uganda (Reference Girling1960), was a communist who had fought in the Spanish Civil War and was not at all in sympathy with the colonial project. In fact, he was expelled prior to the completion of his fieldwork by the Protectorate government, and his chapter on the British and South Asian communities in Acholi was edited out of the published version (Allen Reference Allen2019). Nonetheless, he was to a degree trapped within the academic theory of his time, structuralist-functionalist anthropology, although his writing suggests at times that he was uncomfortable with this paradigm. Girling gathered a remarkable amount of data in a very short time and his book remains relevant, but his was an era when anthropologists were confident that they were asking the right questions and rarely doubted whether they understood the answers they received. Girling’s descriptions of Acholi family and society are far from completely false and can offer useful insights when treated as a text to be interpreted; but they and what has come after, particularly the development sector literature produced during and immediately after the Lord’s Resistance Army (LRA) insurgency, distort, variously in the manner of a hall of mirrors or of crude ethnic stereotyping. Policy based on such accounts tends to magnify these distortions and can be deeply destructive.

Autoethnography is often viewed as a means of exploring subjects where deeply personal experiences can shine a light on matters that have been neglected or would otherwise be inaccessible (Chang et al. Reference Chang, Ngunjiri and Hernandez2013). Our focus does not valorize our inner experiences in this way. While our association was not random, we selected each other for this exercise based on being typical members of landholding groups in Acholi—our exceptionality is our collective research and study experience and our curiosity. We are doubtful of the usefulness in our case of a deep exploration of our subjectivity/objectivity, or insider/outsider qualities. While our perspectives on our families are certainly subjective and personal, much of the material we have been discussing is factual—what family meetings took place, who was there, what was discussed, how decisions were made, what were the concrete outcomes. We considered how land is distributed between family members, where our boundaries are, who lives on the land or farms there, what our natural resources are, and at what rate tree cover is being lost. Such matters, empirical and susceptible to external verification, are perhaps not the usual focus of autoethnographic projects. While we are not exploring introspective material accessible only to ourselves, we are exploring material that is only readily available to people in our situation. We have each of us undertaken decades of participant observation with a level of access far in excess of that available to a typical “outsider” ethnographer. In fact, matters of land are currently so sensitive in Acholi due to a perceived crisis of land grabbing that any outsider showing an interest in a specific landholding would encounter a high level of suspicion, and potentially violent resistance.

Our discussions shade into less verifiable matters of deduction and perception, for example, what are the factions in our family political structures? And then there are subjective, normative, and often emotional opinions, guesswork, or speculation: Are decisions made or is land distributed fairly? Who are the good actors and the bad? What are the motivations behind individuals’ behavior? Which ownership claims are strong, which weak, and which cynical? We are obviously insiders, but in the course of expanding and formalizing our knowledge through the initial exercises and then presenting our families to the group, and through identifying with the group as researchers, we have also adopted an outsider perspective.

As a number of scholars have pointed out, autoethnography eliminates the power imbalance between researcher and subject (Chang et al. Reference Chang, Ngunjiri and Hernandez2013). It might also have been possible for one of us to be an ethnographer and the rest to be a kind of elite focus group, but we doubt that this would have worked as well. For one thing, it would either have involved a massive commitment of voluntary time by the “subjects” or the problematic practice of paying research subjects. Our lack of ownership of something in which we were so clearly an essential element might have led to disenchantment. We would have been less motivated to be open and might have regarded the conceptually challenging aspects of the process as the responsibility of the sole ethnographer, not of the group as a whole. Our common identity as observers and observed, and as ethnographers who share a language and episteme with our subjects, has facilitated a process of hermeneutic examination of each other’s narratives, which might have been lost in a less equal process. The researchers among us have largely worked in the past as brokers and research assistants on the projects of outsider-academics and have been inspired to question these roles by our sometime colleagues in Democratic Republic of Congo, in Bisoka et al. (Reference Bisoka, Ansoms, Vlassenroot, Mudinga and Muzalia2020); our project offered us autonomy and agency as designers and leaders of a research project.

However, while individual autoethnography eliminates the power dynamic between researcher and researched, collaborative autoethnography can reinsert it; the group becomes the researcher, potentially in a power relationship with the individual members, and distribution of power within the group will be subject to its processes. However equitable a group may seek to be, societal norms around gender and age and other factors, such as level of education, may compromise its work to a more or less problematic degree. Power imbalances may also follow from the way the Global North academy esteems and rewards writing skills and academic qualifications (especially those awarded by Global North institutions) at the expense of all other capabilities and competencies involved in knowledge production.

To try to address these factors, this project has been designed to give weight and value to the contributions of those whose skills in language, direct access to its episteme, and critical thinking, allow effective analysis of their lived experience. These are fundamental to the generation of the knowledge sought for this project, and it is these factors which are most significant in respect to the group’s work. Author Five, the Briton, cannot contribute on such matters to anything close to the same degree as other members, even though he is part of a landholding group; but he makes other contributions to academic outputs through skills in writing and familiarity with relevant theory, though he will have less to add to our other planned outputs. His most significant contribution is access to British academic funding resources. Author Two also has a specialist role, having long experience on issues of Luo translation, though she is not an active member of a land-holding group. Author Seven is the only member of the group with advanced IT skills, which have proved vital. We are aware of each other’s levels of formal education, but this has very different implications in Uganda compared to the Global North. How far one progresses is driven much more by economic resources than by ability or motivation, nuancing assumptions about intellectual capacity. The different contributions of all the participants have been declared by the group to be of equal value, and our reflections on this suggest we have been successful in realizing this principle.

The process of writing this article involved two all-day meetings where we established the arguments we wanted to make and the various critical points, Author Five worked these up into a draft, which was then refined by the group over a further day and a half. This is not how most Global North-funded research in Africa works. As Author Eight put it:

Particularly because of this history, trust within the group has been an important factor enabling our functioning as a knowledge-producing collective, facilitated by a network of longstanding relationships, some going back more than a decade and being both professional and social in nature.

Co-ownership and shared responsibility for the whole project was also an essential factor in the generation of a virtuous feedback loop which turbocharged our process. Rather than research design, data collection, coding, data analysis, and reaching conclusions being discreet linear steps, we experienced these as occurring virtually simultaneously. A traditional chronologically linear research structure has certain advantages; it is easier to monitor, evaluate, and generally regulate, and the generation of specific outputs at various stages of the process can be used to validate claims of objectivity and scientific rigor. On the other hand, anyone who has undertaken a sustained solo qualitative research study will be aware of a parallel process in which each new seminal event, perhaps encountering a highly relevant article, a dramatic addition to one’s data, or a revelatory discussion with a colleague, involves an iterative revisiting of the entire exercise; the result is a circulating process that may be more or less conscious but is proportional to the depth of insight a project may produce. Our experience of foregrounding this process is that it has generated liminal space for inquiry where many of the usual constraints and formalities of social science research are suspended. Yet we defy anyone to demonstrate that our findings are less empirical than those produced by traditional approaches.

Ethical Issues

We agreed that members were at liberty to withhold any information that they felt might generate risks or that they were otherwise uncomfortable sharing. Any information that was shared was confidential to the group, and members were responsible for flagging up sensitive issues. Any information published or shared beyond the group should follow Chatham House Rules of non-attribution (unless otherwise agreed), with careful attention to ensuring that no identifying detail was included. All information published or shared beyond the group should first be seen and edited by all members of the group.

We decided this because most families have their secrets that they would not want shared externally, and there are practical considerations as well. In discussing our family politics, the factions and conflicts in play, we were exposing grievances or greed that could be exploited, for example, by predatory lawyers.

In addition, family land-holding groups belong to chiefdoms and clans, which have established public and private identities, including mwac, family or clan slogans or battle cries. These identities are sometimes liable to stereotyping by other clans, and may be marks of pride or shame, cruel or good-natured humor. Lamogi chiefdom members are alleged to be stubborn and aggressive, and eaters of bats, though they are also respected for their rebellion against the British in 1911; a traditional song celebrates but also mocks the occult powers of Paibona clan members. The rwot moo (chief) of Pabbo died after being bitten by a squirrel, and clan members can never mention being bitten themselves, as if someone laughs at them about it, they will die—any talk of squirrels with Pabbo members is likely to give offense. Some clans are immune to bullets, others cannot commit suicide. Members of Bwobo and Payera Paibwo chiefdoms should not be friends or intermarry due to legends of the battle in which the head of the Payera Paibwo rwot moo was cut off and sewn into the Bwobo royal drum. A particular branch of a clan may develop a local reputation as poisoners or thieves, or as any of a panoply of more or less dangerous spiritual beings, lajok: witches, sorcerers, shapeshifters, night dancers, night dreamers, and more. Such matters periodically arose in understanding relationships between our own and neighboring families or other groups. These are important considerations with respect to our research question, but they must also be handled with delicacy and discretion, ensuring that any of us who is impacted by such identities is comfortable with their representation.

Lastly, we agreed to full co-ownership of the research design, the process itself, and all outputs and publications. This included delegation/sharing of budget control and responsibility for deliverables under the grant funding by Author Five to the whole group. These were to produce journal article(s), and policy brief(s) identifying alternatives to unsustainable domestic natural resource consumption, in Acholi but ideally with implications for other African communities.

The Significance of Language

So far, our discussions and writings have been predominantly in English, with constant reference to and consideration of relevant Acholi terms. This has made sense in relation to the research funding; we are committed to writing academic papers which need to be in English, not least because there are no Luo language peer-reviewed journals, and because Author Five is not fluent in Luo. However, given our longer-term aim to influence community attitudes about sustainable natural resource use, it may be highly desirable that later phases of the project be conducted predominantly in Luo.

The Acholi language, a dialect of Luo, was first translated into English and written down by an Italian missionary, Fr. Crazzolara, in the 1920s and 30s (Reference Crazzolara1938). Since then, there have been a few further attempts to develop understanding of the language, mainly undertaken by foreigners. Recognition of the colonial nature of these exercises has been limited but highly articulate; Okot p’Bitek, the Acholi author of the Song of Lawino and Song of Okol (Reference p’Bitek1972), widely recognized as among the greatest works of twentieth-century African literature, was also an anthropologist and essayist: author of The Religion of the Central Luo (Reference p’Bitek1971). In this work, he observes that the name adopted by early missionaries for the Christian god was Jok Rubanga. The missionaries were sure that the Luo religion must conceive of a creator god. Quizzed about who this was, the Acholi respondents named the demon responsible for tuberculosis of the spine. Okot’s reverse biblical translations in which “God the Father” is rendered as “the Hunchback Spirit” are sharp illustrations of the problems and dangers of communication across epistemes. We do not question the truth of Ludwig Wittgenstein’s private language argument, that all language must in principle be translatable (Reference Wittgenstein1967). However, there is no reason to suppose that translation is straightforward or easy. A single word in a different language/episteme might in theory take a whole book to interpret; consider the sheer volume of words devoted to biblical hermeneutics over the past two millennia.

The Acholi word usually translated as “owner” is won. “Won” is also the translation for “father.” It is possible that the word “won” has two distinct meanings; on the other hand, it might be that it has a single meaning, a concept that accommodates its use in both translations, but which must hence be identical to neither. When one considers the vast difference between the English concepts of “owner” and “father,” it becomes apparent just how divergent from them a term that covers both would be from either. The word’s uses in Acholi include won cim, the owner of the phone; won Okot, the father of Okot (people are often known by such a title, referring to their firstborn); won tim, the owner(?)/father(?) of the bush (a traditional title for the spiritual guardian of lula, a hunting ground); and won ngom, a title given to traditional chiefs, usually but perhaps problematically translated as “owner of the land.” One possible translation for “won” offered by one of our group is “an authority over something as though you had fathered it.”

Reflections on such issues suggest how different the territory covered in a discussion or written description of Acholi customary ownership might be in Luo compared to the same discussion in English.

“Customary” Ownership, Family Politics and Decision-making

The policy and development sector debate on reform of customary land in Uganda and other parts of Africa revolves around an image of ownership of customary land as uncomplicated, where the supposed primary problem is bad people taking advantage of lack of formality and state involvement to grab land from others who unambiguously own it. This variously false or facile understanding continues to underpin national and international development agendas on land reform.

On the other hand, scholars, both foreign and local, have sometimes successfully illuminated certain truths about African customary land, primarily during the 1990s and 2000s. Hastings Okoth-Ogendo (Reference Okoth-Ogendo, Toulmin and Quan2000), Johan Pottier (Reference Pottier, Huggins and Clover2005), Celestine Nyamu-Musembi (Reference Nyamu-Musembi, Englert and Daley2008), Parker Shipton (Reference Shipton2009), and many others have written insightfully on African land. For example, on the subject of land rights:

And on notions of property:

As far as they go, these statements are not at all contradicted by our findings; however, the fact that they, and most other writings on customary land, are expressed negatively or comparatively, foregrounding Western capitalist norms of rights, rules and property, illustrates perfectly why there is much decolonizing work to be done (see Sabaratnam Reference Sabaratnam2017).

Our findings to date offer a de-mystification of “customary land.” Looking at our families, at some point in the past forty to one hundred years, a group of related people will have settled on a piece of land. If they were a small group, they may have negotiated to be attached to, or associated with, or become clients of a larger group. If they were a large group themselves, they just settled on empty land, and relied on their own capacity for occupation and defense. What happened thereafter will have been family life—no esoteric land customs, no exotic traditional systems of “customary” ownership. Rather, a typically complex (but perhaps unusually large, compared to some other parts of the world) family group, with all the attendant affections, rivalries, jealousies, intimacies, and inexorable mutability, as births and deaths occur; youth gives way to age; and marriage, adoption, and friendships generate new blood.

What we have seen in terms of governance is as varied as one might expect of families anywhere. One of us described an authoritarian family head, who made unilateral decisions and was largely obeyed, though on one occasion was beaten unconscious by his brothers when he went too far. Since his death the family has been leaderless, anarchic; no family meetings have been held, no collective decisions have been made, and there is no acknowledged head of the family. In all other cases, there are more or less accepted processes for collective decision-making and at least a titular head of the family. However, there are few patterns to observe. One group’s unwritten constitution recognizes eight decision-making households, representing each of the sons of the founder, though one of these who died is represented by his wife, and another very old brother is represented by his sons. In practice, the relative influence of the eight is far from equal, with one brother (who is not the formal head of the family, but who returned before the others in 2006, following the wartime displacement) controlling the bulk of the land. His authority is exercised obliquely through allowing or disallowing access to the land he controls by other family members.

In one of our families, the head and all influential members are women, while in another, women are not permitted to attend decision-making family meetings; in a third, women are allowed to attend but not to speak; while in a fourth, meetings cannot take place except in the presence of a key female elder. Reflecting Julian Quan’s (Reference Quan and Cotula2007) point above, one family is descended through the female line of the clan in which they are accepted as members, while in another, the founding father was called Ngeca, meaning “slave,” indicating that he was an abducted/adopted rather than an agnatic member of the clan. While the language is loaded with terms, many of them offensive, for members of the group who are not members of the clan by blood, we encountered normative standards of behavior around this issue; it is generally unacceptable to allude to such matters in public, and one of us remembered being beaten as a child for mentioning a respected elder’s non-clan origins. In some of our families, being a member of the clan by blood is an essential prerequisite for being a decision-maker. However, in one recent major intra-family dispute, a vote of twenty-nine elders was held, as seemingly the only way to break a deadlock of entrenched positions. Five of those who voted were from a different branch of the family, who were only present as mediators, and eight were not members of the clan—variously in-laws (husbands of daughters of the clan) and nephews (related through the maternal line).

It is a shibboleth of some traditionalists that it is a fundamental if often transgressed norm that customary land should not be sold. An examination of some of our family landholdings shows that land has long, perhaps always, been subject to acquisition and alienation, through exchange, gifting, and appropriation. All of our families have been involved in customary land sales of one form or another. In one instance, land had been gifted in repayment of funds loaned to the clan to defend themselves in a land-related court case; another case was reported where the clan decided to sell land to pay for the dangerously overdue last funeral rites of some increasingly vengeful elders. One of our families purchased additional collective land for the benefit of the whole group. In another, a large family with valuable but limited peri-urban land proposed to sell part of this holding to purchase a much larger area of rural farmland for the use of the whole group. Several of us have bought plots in order to enjoy the benefits of land where we exercise full control, a rare commodity (Meinert & Whyte Reference Meinert and Whyte2022), while also continuing to use communal family land. It is true, though, that land sales are often cause for conflict. Because families are organic and evolving, and the land-interested element of any family is constantly changing, it is often not possible to identify an exact number of specific individuals as the “owners” of a particular piece of customary land. Land has always been divided in many different ways and to different degrees; the exact terms of a division, for example whether a household is sufficiently autonomous to be entitled to give or sell any or all of “their” land, may be moot. This is also true of natural resources: are the trees on the piece of land you farm yours, or do they belong to all? How should the proceeds from an area of rock on a particular household’s land, sold to a road construction company, be divided? These are not, we think, issues of custom, but rather of family dynamics.