1. Introduction
Scholars like Clemmer (1958) and Sykes (1958) considered prisons as “small societies”, highlighting roles and hierarchies. Goffman’s (1961) work revealed that in “total institutions” isolation, rigid routines, total control and stigmatization help define the prison environment. Irwin (1970), through participant observation, described the values, attitudes and morals of the prison environment. Although inscribed in the subculture debate, prison life has been seen as something more complex than a mere exercise in control and punishment. Already Jacobs (1977) highlighted how prisons were porous, never really isolated from changes in the outside community (e.g., court decisions, social and political movements, or the emergence of organized gangs). Such exogenous changes manage to shape the internal structure of prison institutions.
In his examination of political violence in Northern Ireland through the narratives of ex-prisoners, Feldman (1991) uncovered the complexities of ethnographic research. The intertwined roles of “informant” and “informer” within the prison context reveal the nuances of surveillance culture (cf. Rhodes, 2001: p. 73). Feldman warns that in such a setting, participant observation may inadvertently align with the “others” who surveil (Feldman, 1991: p. 12).
Foucault’s influential work, Discipline and Punish: The Birth of the Prison (1977), is still crucial for understanding modern prison systems’ power dynamics and subjectivation processes. His analysis highlights Bentham’s Panopticon as a foundational model for contemporary disciplinary societies and biopolitics. Wacquant (2001) critiques the conventional “crime-and-punishment” paradigm, contending that the criminal justice system manages marginalized groups, influencing perceptions of race in addition to addressing criminal activities. Elsewhere, Wacquant (1999) notes the high presence of foreign inmates in European prisons, akin to racial disparities in U.S. jails. Ugelvik (2014) and others argue that detaining noncitizens implies “hyper-criminalization” of migrants, or a “crimmigration” (Stumpf, 2006). Fassin (2015) shows incarceration normalization accentuates socio-racial inequalities, requiring us to reassess the prison system and its pervasive societal impact.
This article explores the rhetorics and practices of global development interwoven with local power dynamics, utilizing the social and political functions of the prison institution in Tigray (Ethiopia).
2. Context and Methodology
As far as Ethiopia is concerned, particularly the Tigray region, where I carried out most of my research through medium- to long-term fieldwork between 2014 and 2019 (primarily for my PhD dissertation), conducting prison ethnography is nearly impossible. It cannot be conducted in the same way as one might imagine in Europe or in a clearly non-authoritarian democracy (see below). Only a small number of studies in this area focus on health matters (e.g., Abera & Adane, 2017) or political detentions (e.g., Guutama, 2003). A more recent study by O’Donnell (2019) looks at power relations in a southern Ethiopian prison.
Ethiopia, a federal republic formed from the struggle against the Derg1 dictatorship, transitioned from a dictatorial regime to a federal system in 1991. The 1994 Constitution introduced an “ethnic democracy” meant to promote decentralization and protect the rights of diverse ethnicities (Calchi Novati, 1994: p. 237). However, this framework primarily strengthened the Tigray People’s Liberation Front (TPLF), which dominated the Ethiopian People’s Revolutionary Democratic Front (EPRDF) coalition until 2019 (Calchi Novati, 2013: p. 328). Despite constitutional guarantees, the Ethiopian government frequently suppressed autonomist demands and violated human rights. The 2005 electoral protests exemplified this repression, as the government labeled demonstrators “dangerous vagrants” and “unemployed youth”, downplaying their legitimate grievances (Di Nunzio, 2014a), while others defined the regime as an “electoral authoritarianism” (Aalen & Tronvoll, 2009) and a “one-party state” (Tronvoll, 2011). These events further highlighted an emerging carceral state under a declared state of emergency (Toggia, 2008; see below: Section 4).
The crisis within the EPRDF deepened due to ongoing conflicts, including with Eritrea, and the death of long-time Prime Minister Meles Zenawi in 2012. In 2018, Abiy Ahmed, the first Oromo to become Prime Minister, dismantled the EPRDF from within to form the Prosperity Party, leading to an intensification of armed conflict with the TPLF, which still maintains control in Tigray (see Plaut & Vaughan, 2023).
My initial entry into the prison environment was during visiting hours, to meet with some male inmates who had previously been key ethnographic informants for my long-term research, particularly in the town of Mekelle. My research concentrated on the street fights between neighborhood youth gangs, as well as the local governmentality practices aimed at retraining youth considered “deviant” through microenterprise and microcredit programs (Marasco, 2024). In the subsequent phase of my research, I was able to interview prison officers and former inmates now on the outside. Eventually, I had the opportunity to engage directly with the management of the Mekelle Federal Prison, after being received by the Security Administration Bureau of the Tigray Regional State.
On numerous occasions, I found myself interviewing police officers at various district prisons (tabia - 
)2. Interviews with prisoners could not be recorded, as I was prohibited from entering with any electronic devices or even pen and paper. Consequently, the moments immediately following these ethnographic encounters were filled with intense diary writing, allowing me to capture the essence of the experience in the heat of the moment. The selection of interviewees in a prison setting always carries the risk of being influenced by prison authorities, who may filter the detainees available for interviews. However, by never requesting to speak with political prisoners but instead focusing on detainees accused of petty crimes (see below), I was able to overcome the initial difficulties and diffidence of the local authorities. In fact, they began to promote to me, as a Western researcher, the rhetoric of the complete success of rehabilitation and retraining processes.
For this reason, ethnographic information was also collected outside the prison environment using other methods, such as semi-structured interviews and focus groups. For example, this has been possible in those cooperatives implemented by the Mekelle city government for the “retraining” of unemployed youth with potential or previous judicial problems. In line with Feldman’s (1991) approach to social inquiries in the prison setting, participant observation takes a secondary role to the listening dimension. In fact, over the years, various returns to the research terrain allowed me to intercept the trajectories of my informants as they entered and left prison several times, even transitioning through retraining programs. This has allowed for a position of the researcher closer to the boundary between this “inside” and this “outside” and more sensitive to the degree of porosity between the prison institution and extramural society (cf. Cunha, 2020).
3. Esir Bet
In Mekelle, the large detention facility, located just outside the city, was built during the military regime (1974-1991) and at the time was called Esir Bet (
), where esir in Amharic means “detention” and bet means “house” (often used in place of “facility” in a general sense)—in short, “detention house”. From 1974 to 1991, this facility gained notoriety as a place of extreme oppression under the Red Terror of the Derg regime, when even small associations of political dissent led to arrest, torture or death (Young, 1997). Accounts in Jenny Hammond’s (1999) book unveiled the harrowing conditions inside Mekelle prison during this tumultuous period, detailing overcrowding, torture, executions, and bleak living standards that highlighted the dehumanizing treatment and dire lack of basic necessities that inmates faced (Hammond, 1999: pp. 35-100). The accounts cover the period from 1977 to 1986, a period also marked by a devastating famine in Ethiopia and Tigray, which compounded the difficulties of Mekelle, grappling with thousands of starving refugees and logistical difficulties due to military interventions, peaking at 50 - 60 deaths per day in March 1985 (Solberg, 1991: p. 119).
In the midst of widespread hunger, violence and despair, the TPLF successfully carried out the Gaazi operation, freeing over a thousand prisoners from Mekelle Prison. Led by the heroic General Hadush Araya, known as Hayelom (“Unstoppable”), on February 5, 1985, the assault on Esir Bet lasted only 15 minutes and resulted in the death of sixteen guards, with no TPLF casualties (Tareke, 2009: p. 105; Hammond, 1999: p. 11). This operation, named after a TPLF fighter killed in the second year of the armed struggle, represents a founding act for the new Tigray, which was fully liberated from the Derg only four years later with the conquest of Mekelle. The meticulously planned operation took the form of a myth of heroism: Hayelom spent weeks training his men and recreating the prison layout in the forest to prepare for the rapid assault (Hammond, 1999: pp. 166-167). After the triumph of the Gaazi Operation, the TPLF leveraged the symbolic capital it had acquired, creating a significant moral appeal for Tigrayan youth and inspiring new recruits to join the Front to lead societal reform and revolutionary change (Tareke, 2009: p. 105). The freed prisoners, echoing the concept of “individual regeneration” employed by De Baeque for the French Revolution, helped reshape society, like the transformative “new man” emerging from the fall of the Bastille, a symbol of liberation and rebirth (De Baecque, 1988: p. 196). By 1991, following the defeat of the old regime and the Revolutionary Front rise to power, political life in Mekelle, the region, and the country was being organized by the TPLF within the federal coalition (EPRDF). The old prison in Mekelle underwent a profound transition, a symbolic dismantling of the “Tigrayan Bastille”, repurposed as a center for metalworking, embodying rebirth and redemption for a previously coercive and alienating space. Hammond’s ethnography vividly captures the metamorphosis of these spaces amidst rapid transformations achieved through collective sacrifice, highlighting the recycling of military equipment for reconstruction, and the conversion of spaces of segregation and alienation enforced by the oppressive regime into sites of labor, rebirth, and renewal (Hammond, 1999: p. 315).
The transitional government in Ethiopia, led by Meles Zenawi until 1995, faced challenges related to separatist movements and electoral fraud allegations, leading to intense suppression of dissent (Kidane, 2001: p. 22). The Oromia Liberation Front (OLF) parted ways with the EPRDF due to fraud accusations, while subsequent elections saw EPRDF running unopposed following opposition boycotts. The Tigray region bore the brunt of the war with Eritrea from 1998 to 2001, suffering severe attacks on cities such as Mekelle and Adigrat. The early years of revolutionary democracy witnessed the EPRDF/TPLF adopting a stringent governing approach in response to emerging challenges. This period also coincided with an increase in “hunger crimes”, which in turn led to the expansion of prison facilities to house large numbers of new inmates. Ahead of the 2005 elections, a wave of arrests for political and “national security reasons” occurred, resulting in the detention of numerous journalists in 2003, showcasing the widespread judicial isolation of dissidents (Tronvoll & Hagmann, 2012). The EPRDF’s dominance was most evident during the 2010 elections, where the opposition suffered a major defeat. Toggia (2008) highlights a shift towards a carceral society post-Derg, with emergency measures and security policies being abused. The prison system, including Mekelle Prison, underwent structural and organizational growth to align with the evolving political climate in Ethiopia. Ethiopian administrations maintained repressive carceral systems to quell opposition, as documented by Human Rights Watch (2015). The Anti-Terrorism Proclamation (ATP) of 20093 introduced broad legal mechanisms, leading to numerous detentions under vague “terrorist acts” definitions, often resulting in prolonged pretrial incarcerations exceeding the legal four-month limit (Amnesty International, 2018).
In conclusion, the political landscape in Ethiopia following the transitional government era saw a tightening of control, a surge in oppression, and an expansion of prison infrastructure to stifle dissent.
4. Bet Hntset
Prior to my research in Mekelle, an ideological restructuring of the prison system was initiated, enabled by a significant project spearheaded by the Italian Cooperation for Development (IDC) in collaboration with the International Labour Organization (ILO). This project began in 2011, aiming to establish a sustainable development model through the implementation of retraining schemes focused on basic and vocational education for inmates, with a particular emphasis on forming labor cooperatives, especially for women (ILO-IDC, 2013).
One such cooperative, Pagume Construction Materials, is dedicated to the production and sale of building bricks. Inmates, both male and female, work here and also manage administrative tasks after receiving specific training. Additional cooperatives have emerged in agriculture, livestock, weaving, carpentry and metalworking, as well as in services like hairdressing and secretarial work. Simpler: Some of these cooperatives focus on producing cobblestones, modeled after successful initiatives developed by the German Agency for International Cooperation (GIZ)4. Like the GIZ model, the Mekelle Prison Project emphasizes the formation of productive citizens who are not just workers but also savers.
By establishing cooperative accounts, inmates can eventually access other loans, which they can repay through further work, thereby earning opportunities for new credit flows. This paradigm reflects micro-entrepreneurship and para-microcredit approaches aimed at empowering individuals to become self-entrepreneurs. Similar schemes have been seen in Mekelle, which were in that case targeted at youth groups labeled as “gangsters”, mainly because of their lack of “work ethic”, according to the narrative of government institutions (Marasco, 2024).
The program ties retraining to local microcredit variations, highlighted by the establishment of the Dedebit Credit and Saving Institution (DECSI) branch within the prison. It should be considered that the historical founder of DECSI is the Relief Society of Tigray (REST), which was established in 1978 as a TPLF humanitarian organization. Dedebit is the name of the locality where the first operational base of the TPLF against the Derg was located. This financial institution manages the deposit accounts for “retrained groups” in the city, including the inmates. Inside the prison, inmates can sell their products, while cooperatives producing construction materials benefit from social and political networks linked to the prison institution, city and regional government, ultimately aligning with the TPLF. The construction sector is supported at high levels, driven by government initiatives throughout Tigray and Ethiopia.
The final report of the Mekelle Prison Project highlighted its scalability and sustainability, aligning with the national growth plan and the United Nations Millennium Development Goals. Significant achievements in the project’s first two years included the training of more than 1200 people in economic activities, the vocational education of 610 people, and the creation of 31 cooperatives (ILO-IDC, 2013). The project has received international recognition for its success, with media coverage praising its transformative impact. An Italian (but U.S.-based) online magazine highlighted the Mekelle Regional Prison as a rehabilitation model to be preferred to the terrible conditions in Italian prisons (Bosco, 2013). In this report (4), the Head of the Security Administration Bureau of Tigray Regional State comments as follows:
One of the indelible scars left by the previous regime on our people is the prisons it set up in various places that served as centers to carry out torture and punishment […] Presently the correctional facilities in Tigray Regional State enable inmates to acquire knowledge and vocational skills that facilitate the conditions that allow them to turn into productive citizens. It has been some time now since prisoners began to appreciate that prisons are no longer the hell they used to be but are in fact centers of change where they endeavor to engage in development.5
In one fell swoop, international observers’ allegations of arbitrary detentions had been erased, and everything about a punitive society (cf. Foucault, 2016) was completely replaced by development narratives. The ILO report features diverse “witnesses”, including an unnamed boy incarcerated for theft. Other detainees’ crimes, documented anonymously, range from theft to misuse of office, but are often minor offenses. All express satisfaction with the new program, citing personal empowerment, repentance, and individual responsibility. The institution’s resources are seen as transformative, aiding in inmates’ self-improvement and societal reintegration. The evolution from a brutal Derg-era prison to a progressive model institution is highlighted as a testimony to personal growth and positive impact on society.
Di Nunzio (2014b: p. 446) aptly states that in a centralized and authoritarian state like Ethiopia, surveillance and discipline must reckon with the limits of their influence and capacity. In this context, only development programs that can prioritize the political concerns of the ruling party, primarily in producing citizens envisioned as agents of its control policies (ib.: 460), are empowered. Indeed, the involvement of international organizations, such as ILO in the transformation of Mekelle prison, ultimately indirectly supported the political programs of the TPLF, contributing to the consolidation of its power. This occurred because, both at the regional and federal levels (during the years of the TPLF-led EPRDF), the development project driven by the dominant political class, in the absence of a strong middle class, saw an overlap between government actors, party members, and development agents, with the TPLF using the rhetoric of the “fight against poverty” to consolidate its power (Vaughan, 2011: pp. 623-635). Despite criticism of its authoritarianism and human rights violations, post-Derg Ethiopia has been (and still is) a major recipient of international aid, described as a model of development thanks to its economic growth (which has even reached 10% of its GDP annually) and its progress first towards the Millennium Development Goals (Villanucci, 2014: p. 26) and then towards the Sustainable Development Goals.
Soon Mekelle prison changed its name to Bet Hntset (
), meaning “home for reform” in Tigrinya6. This is to create a clear break with the Derg era but also because reforming is at the center of institutional discourse inherent in prison. This is evidenced from the above project, even though for the Cooperation it is about putting empowerment at the center through micro-opportunities, that is, social reintegration through small economies that, it is hoped, will continue to help people and their families even outside of prison. The institutional discourses firmly focus on the contribution to the development of the nation and personal responsibility. The data, also found in the reports of the Mekelle Prison Project, of a prison population sentenced in over 80% of cases for small crimes (ILO-IDC, 2013: p. 5), is interpreted as a manifestation of “personal ignorance”, as a “wrong attitude”, a sort of personal resistance to wanting to emerge from poverty. This is in line with the policies implemented at the federal level by the EPRDF during its long rule and in Tigray by the still-dominant TPLF. These are initiatives against unemployment, aiming at creating jobs mainly through the activation of small entrepreneurship, where in government rhetoric becoming entrepreneurs and creating one’s own job is seen as part of a path of individual change oriented not only by liberal and individualistic principles, but also collectivist and developmentalist rhetoric (see Di Nunzio, 2015; Marasco, 2024). Regarding the above, I report the testimony of the manager of the Propaganda Office, a kind of political public relations office in Mekelle:
Those who are poor, we fix them, we organize them. They receive training and afterwards we provide them with a skills certification. Therefore, they will take part in a job and thus make money. The opportunity is given to them in this way [and not through forms of funding or economic incentive]. In this office we especially discuss the political aspects, how to implement in society a change of attitude. If there are some bad habits, we create awareness in society. For example, there are too many holidays in Ethiopia. All these holidays are not good for the purpose of eradicating poverty.7
However, these retraining and entrepreneurship start-up programs do not always succeed but turn into blocked life trajectories, where the only ones to gain in terms of consensus and economics are always and only the party (TPLF) through its various agencies and institutions (Marasco, 2024). A project inside Mekelle prison must therefore necessarily follow the government’s ideologies and avoid questioning the local blend of neoliberalism and collectivism. This, however, by shifting blame to individuals and neglecting the social context, depoliticizes what lies behind the “petty” crimes committed by the majority of Mekelle Prison inmates.
Consequently, for institutions, recidivism, if it is a matter of attitude, is like saying a “natural” disposition to crime. Not a condition of structural marginality in which it is difficult for an individual to experience viable and truly sustainable alternatives. If a person steals to survive, squeezed by an increasingly limited range, ending up in prison, he or she may consider themselves lucky in joining a cooperative project, for example, to produce bricks. This is what is officially communicated. However, the testimonies I have collected over the years, from former inmates who took part in these programs, mention the following possible scenarios that open up after release from prison:
a) Seeking employment in brick production (continuing the example), for which the skills acquired are not so exclusive as to offer an advantage over the enormous competition.
b) Rare success in obtaining a loan independently, or with the assistance of friends or family, to launch a business, only to discover that the market is rigged by political influences to favor para-government cooperatives (like the prison ones).
c) Receiving a loan from DECSI but failing to repay it, resulting in legal complications or harming guarantors (e.g., they risk losing a house mortgaged with DECSI).
These scenarios challenge the official narrative and provide context for the high rate of petty crimes (80 percent), revealing systemic barriers that perpetuate cycles of poverty and incarceration.
5. Stories from Inside
Since my first year of research in Mekelle, I have noted that the prospect of prison (whether pre-trial detention or after conviction) is a tangible and potential risk for many individuals, regardless of their social status, economic situation, or level of education. Not everyone fears being incarcerated for minor offenses like stealing a smartphone. I have met entrepreneurs who spent up to two months in a city district jail (tabia) following unfounded corruption accusations, exploited by competitors for their gain. While filing a false police report is technically a criminal offense, ambiguity in law enforcement allows for the indefinite detainment of the accused for “security reasons” while leaving the accuser unpunished, if this is in the interests of the authorities.
Further, I have met academics who fear they will be arrested for crimes of opinion and during the second year of research, my interpreter, a political science graduate, grew visibly tense when we encountered an old, abandoned public facility. It was a former neighborhood jail (tabia), where he was detained following a fight, unprovoked by him, at the age of 18, outside the school. On that occasion, because he did not call home and was denied the intervention of mediators (shmagiletat - 
)8, the judge sentenced him to four months in the large regional prison, later reduced to two, corresponding to the summer break from his studies, because he had as a mitigating factor the fact that he had been admitted to the university. Numerous are the stories of friends who found themselves caught up in street fights and public disturbances.
“Common detentions” (not for political reasons), entry into a “tabia”, where one can remain for months awaiting sentencing, is characterized by overcrowding, with all detainees usually crammed into a large room, often without access to an outside yard. It is interesting to hear from research participants, particularly gang members or those with a history of gang conflict, about their prison experiences. For Mogos9, a former street gang leader in Mekelle and now a respectable mechanic in the city, since his first detention, prison is seen as an extension of street life. Entering as a gang member, especially one who is findata (
)10 and hayal (
), meaning “muscular” and “powerful”, is different from entering as a thief.
When they opened the cell, it was completely dark and full of people. They were everywhere in the room, around thirty, but the space was really small. When you enter a cell, they immediately ask you for money “for a candle”. It happened to me that one of the prisoners asked me for this money. I said, “Oh, I have nothing,” so the other dared to say to me, “How dare you?” So I replied, “Go to hell!” Others intervened to calm us down. Shortly after, one of them lit a candle and recognized me. The guy said to the others, “Don’t you know him? This is Mogos.” The others also realized who I was, by word of mouth. At that point, though, I didn’t want to let it go, but just to beat the guy. Others intervened again, so I said, “Don’t talk to me about candles. I won’t give a penny.” They replied, “No problem.” Later, they gave me one of their mattresses. That way, I became the Abo Geza.11
Abo Geza (
) is an expression common to all prison testimonies. It means “father of the house,” and it stands for the head (inmate) of the cell. The demand for money is repeated in the testimonies as a rite of entry. It is found, in less epic tones than those used by Mogos, in Haftom’s account of the Enda Silus Gang:
Those inside the cell said nothing to me because they were afraid of me. They knew and feared me. So, I just had to lie down and sleep. But only because they knew me. If you go inside and they don’t know you, they put you through a lot. They ask you nay shm’a (literally “for a candle”). If you say you don’t have money, then you get beaten, they make you clean urine or sleep next to it and other things like that12.
The ritual always follows the same Tigrinya expression: nay shm’a (
), meaning “for a candle”. This phrase serves as a coded reference in case an overseer is listening, as extortion is theoretically punishable, while necessities like candles, bedding, and especially food can be brought in by relatives or purchased inside with money they provide. If Mogos feels ready to defy anyone who does not know him, choosing to oppose and literally acting “in the dark”, Haftom’s words make it very clear that his was, all in all, a lucky entry, in which he was able to spend the “street currency” of hayalnet (“power”)13, underscoring the value and usefulness of a gangster identity, in entering prison. It does not go as well for Joe, ex-member of another neighborhood youth gang, during his first stay in the tabia:
As soon as I entered, I paid “for a candle” and nothing more happened to me that was too serious. I, however, had to sleep on the floor next to the piss, because there was no one I knew there!14
For young gang members in Salsay Alem (“Third World,” nickname for Da Gabriel’s poor neighborhood), getting in and out of the tabia is commonplace. Their mella (
- “system”)15 is built around this reality. It is indeed about avoiding regional jail, and to do so is limited to petty theft, or types of crimes for which one can get out on bail. The system lies in the fact that the bail is paid by the stolen goods receiver, who will recover this money through subsequent thefts. Filmon, one of the Salsay Alem boys, adds:
When one of us enters the cell, he is immediately Abo Geza. We are always recognized, we are known as choma (
- “pieces of meat”, i.e. tough guys). And then it’s written all over our faces-see our scars? They are knife scars. At this point it’s up to us to ask for nay shm’a (“for a candle,” meaning to extort). If it is a deki sefer (“a son of the neighborhood”, even if not involved in the gang) we leave him alone, in fact we protect him. But when we get an outsider then he must give us what he has [...] If someone turns to the guards? No, they don’t. Because after that, sooner or later, we will all be on the street! And there the bills are paid.16
Becoming Abo Geza in the tabia legitimizes extortion. The material, symbolic, and moral economies of the street and prison mirror and amplify one another. In the Ethiopian prison system, at least as far as I have detected in Mekelle, the first difference between the tabia (the neighborhood prison) and the Bet Hntset (the regional prison) is that the former represents the social space closest to the street, and not only because these detention spaces are physically embedded in the grid of city streets. Proximity to the codes, rules, and illicit economies of the street mark distinctions among the inmates: those who are not hayal (“powerful”), or a choma (tough) like the boys from Salsay Alem, must quickly accept rules (foreign to them up to that point) and succumb.
Also interesting is the testimony of Hasos, a choma from Salsay Alem, when he was once challenged for command in the cell:
When they brought me to the jail, I entered and declared to everyone, “I am Abo Geza.” One of the prisoners stepped forward and said, “It’s me, be quiet.” We then fought. In prison, a fight needs to be instant; otherwise, the police catch you right away, and then they come after everyone, hitting with batons (or the thick cables of the high-tension lines). I immediately landed two blows to the head and subdued him. That’s how I became Abo Geza; from that moment on, it was my responsibility to handle the money, oversee everything, anything coming into the cell had to go through me, even the food was under my management, and I determined how much should be given and to whom.17
While waiting for release, Salsay Alem youths turn cell time into a small business. Internal extortion provides minimal compensation but builds symbolic capital, enhancing gangster reputation and street credibility.
6. An “Informal” Power Architecture of the Prison System
The question then arises as to what happens to this system, this economy of violence inside the city cell, when one enters the large regional prison. The dynamics undoubtedly change, becoming more complex as distinct discourses and competing power strategies intersect and intertwine. Hasos again recalls that:
We from Salsay Alem are respected, the most feared. Those who are not from around here, if they are transferred to Mekelle, they are warned by the other prisoners about who we are. Then there are leaders among us prisoners. The Abo Geza as in the tabia, one for each dormitory18, then there is the Merahi Hayli (
, “Head of the Force”) and the Abo Wenber (
, “Father of the Seat”). Of the latter two, the former deals with the blockade and the latter with all prisoners in the prison19.
However, the term Merahi Hayli could be translated as “second in command” in military jargon. Abo Wenber is akin to the “chairman”, signifying the person with authority, commonly used to refer to a head of a government office. It should be clarified what it means to “deal with” prisoners. It is not here a circuit of extortion and prevarication as in tabia. The three roles just described—roles that have been reported to me by all informants regarding their prison experiences and then confirmed by prison workers—are embodied by prisoners chosen by the inmates and then approved by the prison police. Hasos continues:
It works like this: if there is a problem, of any kind, the Abo Geza goes to the Merahi Hayli and he tries to solve it. If it’s a problem between inmates, the best thing is to solve it between us. If you have a problem with the guards, then it is the Merahi Hayli who talks to the police in your block. When he is not listened to, then the Merahi Hayli goes to his superior, to Abo Wenber, who solves problems directly with the prison police chief.
The three figures of representation and prestige among detainees behave as leaders, descending the pyramid in order to deal with simple issues among prisoners. For each cell block, there is an Abo Geza overseeing the group’s economies and privileges. Moving up the hierarchy, these representatives approach the coercive power of the prison administration, with the Abo Wenber (the apex) resolving disputes mainly with the highest-ranking officers, not directly engaged with the general detainee population. The Abo Wenber, symbolically titled “chairman”, serves as the closest mediation body to the judicial power. Many issues brought to the penitentiary police could involve concerns like food quality, living conditions, prolonged water shortages, etc. In essence, this process of mediation handles nearly all stressful prison situations.
These three figures are part of an intricate system of organizing and distributing tasks among prisoners based on participation. The system revolves around six committees: one for peace and order, one for health, one for food, one for educational programs, one for training and employment, and one for reconciliation (erqi - 
) processes. These six committees report to Abo Wenber, who is responsible for consulting them. To illustrate the operational structure of the committees, the Tiena Committee (“Health Committee”) serves as a pertinent example. It is made up of the various Abo Tiena (
), literally “Father of Health” in Tigrinya. There is one Abo Tenia for each dormitory. The prison consists of 6 blocks of 6 dormitories and 2 blocks of 3 dormitories. Thus, there are a total of 42 Abo Tiena who will go on to form the relevant committee. Each individual Abo Tiena is responsible for ascertaining the health status of inmates in their dormitory, proactively identifying and reporting issues, even if an ill individual does not seek assistance. The committee meets every morning to discuss health concerns, while also being able to consult with the Merahi Hayli.
This system works the same way for other committees. Very active are the committee for school education (block 1 is for juvenile detainees), the committee for training and work (which also organizes internal micro-enterprises), and above all the reconciliation (erqi) committee. For the latter, there are three supervising inmates per block, instead of one per cell, while other committee members are police officers and prison representatives, given its unique function aimed at early releases. To be released early, a detainee can request to leave through Yqrta (
)20, “forgiveness”; if the request is valid, a reconciliation process begins: the committee compiles reports, destined for the police and judicial authorities, regarding the applicant’s participation in prison activities, level of cooperation, performance of activities for the common good, etc. An essential element, for example, in the absence of a psychiatric section in the prison, is the care of detainees with mental disabilities, as evidenced by the testimony of my friend Beshir, a former leader of a gang from Quiha suburb:
I was the Mehari Hayli of Block 6 and I used to listen to the various Abo Geza, I used to compile reports which I would then hand over to the prison police. I had applied for Yqrta ( - “forgiveness”). I got out first thanks to the erqi committee, who wrote about me that I was a peacemaker, that I solved problems and above all that I knew how to take care of others, like mentally ill prisoners, even helping them to wash. Thanks to the erqi committee, the reconciliation process was started: the papers compiled were sent to the city police, who went to my accuser and asked him if he would forgive me—he said no, but fortunately the committee can request an internal prison process for forgiveness (yqrta). Thanks to this I was thus released early.21
More clarity on the internal prison pattern of power management and accountability is also given through the testimonies of some prison officers. Particularly valuable is the testimony of Gebre, 24 years old, who at the time had been working for three years in Mekelle’s Bet Hntset.
Clearly, we do not always keep our eyes on the prisoners. We do not know what they do alone in their cells, or how Abo Geza manages them. We are more concerned about the other levels, the Merahi Hayli, and the Abo Wenber. These three roles are chosen by the prisoners through a show of hands and they elect them... though it’s easier with Abo Geza since cellmates often already know each other and quickly choose their leader. It concerns us that they inform us. Generally, for us (police officers), it is important that the Merahi Hayli and Abo Geza have a certain type of attitude, respect the internal rules, know how to handle things... in fact, they are important to us, helping us in our work, especially when there are clashes between gangs like those of Salsay Alem22 and Adigrat. These “chosen prisoners” are the first to remove privileges from those who err. Indeed, these prisoner leaders also inform us who deserves the amokuro (
, i.e. “sentence reduction” in Tigrinya)23.
Unlike the yqrta (release for reconciliation and forgiveness), which can be requested and may not always be granted by the erqi committee, the amokuro is a sentence reduction available to all, a sort of already acquired pardon if one demonstrates impeccable behavior. When a prisoner is sent to the regional prison, the judge issues a sentence that indicates, for example, a three-year sentence, but also mentions the day on which the prisoner could be released for amokuro, which is equivalent to a one-third reduction in the sentence: thus, in this case, he or she would be released after two years. As demonstrated in Gebre’s testimony, reports provided by designated responsible prisoners (the “selected prisoners”) can result in the immediate revocation of eligibility for amokuro.
Hasos, member of Salsay Alem gang, spent a long period of detention in Block 1, designated for minors under the age of eighteen, and thus continues his account:
If you’re a detainee who doesn’t cause problems, you can also find ways to have your benefits, to access the entertainment areas more often, even to study... I took grade 11 (end of high school) there, now I know that inside you can also take grade 12 (preparatory to university)24.
7. From the Inside to the Outside: The Porosity of a Carceral Society
A complex internal power management framework emerges. Power dynamics and coercion are somewhat delegated to prisoners, with the most involved role being the Merahi Hayli (block leader). These leaders act not only as mediators and spokespersons for inmates but also as informants for the police. However, they are not viewed solely as spies, as the system (mella) of the wardens is the primary set of rules detainees learn upon entering prison.
It is evident how the rehabilitative function of the prison is strongly based on an action of “re-subjectivation”: everything is constructed in mirror image to the type of society—the “political dream” (Foucault, 1977)25—from which the prisoner comes, after having “betrayed” it, and in which they must be reintegrated. Thus, the mechanisms of widespread participation and techniques of mutual surveillance assume a highly significant and strategic function, linked to the extramural society in which those mechanisms and techniques are equally active, as will be seen shortly. Additionally, as seen, the prisoner leaders redistribute privileges bestowed from above: access to lighter work, the movie room, the tea and coffee area, soccer and basketball fields, as well as more water, food, tools, and objects (within internal security limits) that can materially improve incarceration. From a disciplinary perspective, they intervene by warning, even with severe threats, those engaging in violent conduct and causing disorder (
- naobi)26, such as in clashes between “corner groups” (as Hasos calls them). Furthermore, when the deterrent power of their redistributive authority is inadequate, these leaders, by reporting to the police, can identify which prisoners need to “learn a different attitude” (as reported by Gebre). They direct the “troublemakers” to work outside the prison, usually on city streets or remote farms, where they may be kept for a month, deprived of socialization or recreational areas. On the legality of the system, young prison officer Gebre expresses himself as follows:
There is no law for this system. The law says that a detainee must be guaranteed education, training, health, food, and so on. We do it with this system: it’s our internal (unwritten, informal) system. It is useful, it helps the inmates, and it helps us officers. We need to communicate to inmates that it is worth changing their attitude, to explain to them that there is always a chance to engage for themselves and for society. Prisoner leaders also serve to communicate this.27
In fact, this is the most entrenched reforming action taking place in the prison, and it is on this that all other interventions, such as the development programs themselves, educational training or experiments introduced from outside, are overwritten. Even for the selection of participants in work cooperatives, institutions resort to redistribution of privileges by means of leaders or representatives of inmates. A continuum should then be made explicit between the system of redistribution of functional control tasks inside the prison and that implemented outside, for example at the cooperatives of young workers, activated by the various Youth Affairs Offices (YAO) in Mekelle for the retraining of “unemployed potential criminals” because they lack “attitude to work” (Marasco, 2024). In this context, a contract is even signed between the group of retrained youth, municipal authorities, and the DECSI lending institution (see Section 4). The pact is signed by three youth representatives (a leader, a secretary, and a cashier), tracing both the model of distribution of authority within local youth gangs (Marasco, 2024), but also the very model of inmate representation within the regional prison. In addition, the pact signed outside the prison by the youth to be “rehabilitated” follows control mechanisms that act on the participants’ bodies. It is a contract that, in addition to containing an economic-financial part for the assignment of a job location (such as a parking space) and the opening of a savings account with DECSI, carries a section devoted to discipline in the strict sense. Youths promise the bank and municipal institutions to renounce alcohol, drugs and socially harmful behavior (11). Control over the actions of these boys undergoing rehabilitation is then exercised through visits by YAO employees, but mostly through any “complaints” from the neighborhood.
What remains to be explored is the question of the criteria by which leadership should be given to prisoners. Beyond the ability to mediate disputes and adhere to rules, it is unclear whether additional qualities are necessary for roles such as the Ago Geza, Merahi Hayli, and Abo Wenber. As Gebre explains:
Prisoners who have to serve too short sentences are not adequate as leaders. If they get out right away it’s a mess, because you have to let prisoners choose new leaders.... and then prisoners serving long sentences know the prison better than the warden (laughs) and are more reliable!
These considerations, while not immediately obvious, are strategically significant from a disciplinary perspective. Long-term inmates also hold substantial symbolic capital (Bourdieu, 1998: p. 47) within the inmate hierarchy, reinforcing their authority. It is understood that surveillance and coercion are the predominant components of rehabilitative factors, even before the introduction of the recent innovative programs, which in any case are not universally accessible. This dynamic underscores the complex interplay between institutional control, inmate leadership, and the broader social structure within the prison environment.
The patterns of power and control, which are reintroduced into inmates’ behaviors in order to be internalized, also reflect the core of extramural surveillance practice. To clarify this last point, I will simply recount that during the years of my research, whenever I was conducting an interview at the various district prisons (tabia) attached to police stations, there was a constant coming and going of very young boys and even underage street workers, such as shoe shiners (
- listro)28 or street vendors. On the walls of each tabia, there was posted a kind of work schedule and street corner assignment for each of these young vendors. “An organization for their safety”, was usually the comment provided by the police. It is not coincidental that there is a proverb in Mekelle: “Never discuss politics in the presence of a listro.” Shoeshines are in fact commonly perceived as spies, known in Tigrinya as joro tabia (
), meaning “ear of the tabia”.
Over the years, I partly delved into the criminal and judicial aspects of the lives of my informants, approaching lawyers and city prosecutors. After a while, an atmosphere of relative trust and confidence developed with some of these, such that I was able to touch on sensitive issues like security and information management for judicial purposes. A city prosecutor once told me:
In the city of Mekelle alone, there are about 3000 spies. Among them, a portion have official training (degree) in espionage. The others are simply people from the community who do it as a second job because the pay is very low. It is usually a small reimbursement for following someone in a cafeteria or other places. “High-ranking” spies, on the other hand, are internal to the administration and are trained and equipped with modern devices. In general, it is best to be wary of everyone.29
On the use of spies, as a tool of the judicial system, the prosecutor adds:
Spies typically do not appear in court but provide reports to government offices or law enforcement, like the police. The “Order keeping” office is a primary contact. Occasionally, they testify in rare cases when no other witnesses can prove the facts.
From my informant’s account, then, a minimum of one in 100 people in Mekelle alone would be potential spies, although it is a common saying-even among my interpreters that “one in three people” in the city are. Once, at a downtown police station, I was fleetingly shown, posted on the wall but hidden behind a celebratory party poster (TPLF), a list of at least four pages: according to the police officer, authorized to speak to me, these would be the spies active only in that urban area, put on the “payroll” by his station. Not only the practice, but especially the flaunting, in society, and not only in front of the researcher then, of the idea that the police have spies everywhere, is a well-tested government technique.
Sentence discounts (amokuro) or pardons for release (yqrta) are not the only ways to reduce stress within the prison institution30. In Ethiopia, amnesty is granted at each legislative change, thus following the election of a new parliament. Working on the imagery of forgiveness is a very effective mechanism that the EPRDF put in place after each new election cycle, especially considering that arrests and convictions have always increased exponentially before voting to dampen political dissent (see above). It is clear that ordinary prisoners are the preferred recipients of such measures, not those accused of high treason or special laws implemented during the state of emergency. This is a crucial point for understanding the rest of the discussion.
8. Discussion
The fact that in a punitive society (cf. Foucault, 2016) like Ethiopia, especially in the Tigray region, I was able to conduct ethnographic research in prison thanks to permission from the Tigray police and prison authorities may seem paradoxical. However, considering the work done by Italian Cooperation, now fully in the hands of the local government, through cooperative work and microfinance schemes and the ideological cleansing experienced by the institution, the research permit finds rationale: as an Italian in that context, I would witness the productive, rehabilitative, and organizational capacity of the large prison. In the logic of the local TPLF administrators, I would likely contribute to promoting the narrative of the model prison externally and unwittingly be co-opted into the local ideological circuit. My research, however, was not overshadowed by institutional narratives, a risk that is always present unless a reflective approach is adopted in anthropological analysis (Clifford & Marcus, 1986). Ethnographers must critically examine their role, presence, and the textual representation of their work. My presence in the field undoubtedly elicited specific reactions and narratives, which must be carefully analyzed to construct a comprehensive understanding. Consequently, this last part of the article will also make use of the ethnographic material contained in my field diary.
I am absolutely convinced, especially in retrospect, that the local government and the institutions responsible for the prison structure truly believed, before and during my investigation, that they were managing a panopticon, as originally conceived by Bentham (2009 [1791]), meaning the best possible prison, according to the idea of control and security that the overseer can have.
Informal power structures are also observed by O’Donnell (2019) in a prison in the Southern Nations, Nationalities, and Peoples’ Regional State of Ethiopia (SNNPR), and they also impact the society at large. In that prison, inmates have developed an internal governance code that regulates daily life, dispute resolution, and maintenance of order. This self-governance system reduces the need for constant intervention by prison staff, who are outnumbered by inmates. However, in both the Mekelle context and O’Donnell’s study, the “attitude to work” plays a crucial role in shaping inmates’ experiences. They are highly motivated to work to generate income, as they must provide for themselves in terms of clothing, hygiene items, and other necessities. While this reinforces a sense of purpose and autonomy, as seen in our ethnographic section, it also strengthens dependence on authority (both informal and formal), which has the power to revoke even the smallest privileges. Work appears to function more as a tool for instilling discipline rather than a path to redemption. This reiterates and confirms what I have observed over the years outside the prison walls in Mekelle: the prevailing political narrative asserts that, thanks to the TPLF, people are able to access and maintain employment. This is attributed both to the peace and development it ensures and to the job opportunities it directly provides. Those who remain unemployed are often portrayed as having a poor “attitude” toward work (cf. Marasco, 2024).
O’Donnell (2019) also notes that the material conditions in the Ethiopian prison he observed are extremely poor, but inmates have access to goods and services they can purchase with income generated from work. According to the author, this differs from many Western prisons, where inmates have access to basic services but often lack opportunities to generate income. Furthermore, O’Donnell (2019: pp. 274-275) highlights that work within prison creates inequalities, including economic disparities, among inmates. Some, thanks to their entrepreneurial skills or specific competencies (such as weaving or painting), manage to earn significantly more than others. However, this does not lead to social tensions, because the prison’s internal governance code typically defuses any signs of protest through mechanisms of mutual aid, keeping the promise of rewards alive, and access to “forgiveness”. I say “typically” because in 2018, a breakdown in the system occurred, leading my interlocutors to question their belief that their prison model was functioning perfectly and was immune to disorder.
In July 2018, I was authorized to conduct research at Mekelle Regional Prison, accompanied by a friend who was a local radio journalist. My research was abruptly interrupted on the 25th of that month by an event that, as I later discovered, had repercussions not only for Mekelle and its prison. On that day, I was in the administrative and management office area, preparing to begin interviews with inmates. I was in conversation with the deputy warden, who was explaining how the facility had achieved a safer and more peaceful environment compared to past periods, though he did not specify which periods he was referring to. Below are the events as recorded in my field diary:
Suddenly, gunfire erupts, causing us to jump from our seats, with the deputy director reacting first. AK47 shots are fired from the guard towers, creating a chaotic atmosphere. The deputy director orders my colleague and me to stay in the room, keep low, and not leave for any reason, before hastily departing. We hear screams and chaos outside the room, with administrative staff, including women, in tears. Despite the unfolding crisis, we obediently adhere to the directive. After about twenty minutes, the deputy director returns in a bulletproof vest and armed with a rifle, accompanied by three equally armed police officers, and we are escorted outside the facility amidst ongoing gunfire. As we exit, we witness an army truck unloading fifteen soldiers before departing, with two more trucks arriving. Gunfire persists everywhere. The deputy director leaves us on the street shouting “For another day!”, as he returns inside surrounded by soldiers, with gunshots intensifying. Seeking refuge from the escalating situation, we join a group of people who find shelter in makeshift small cafes by the building entrance, actually private houses where owners offer refreshments to visitors outside the gates. After listening to gunshots and military vehicle noises for about an hour, my interpreter and I decide to leave. The prison is ablaze, with at least a quarter of the structure burning, generating a dense black tower of smoke rising into the sky. Choosing to walk towards the city center, we face a two-hour journey to reach transportation. Along the way, we encounter groups of women praying in teff fields, singing and pleading for mercy for those inside the prison. Distraught individuals, worried about their incarcerated family members, travel in the opposite direction, alarmed by the billowing smoke. One mother, desperate for information about her son, tearfully recounts her situation and pleads for help. Back in the city, evening internet and national TV broadcasts reveal that the prison rebellion extends beyond Mekelle, impacting major detention facilities across the country.
Right in the Mekelle prison (the “model prison”), the most intense clashes with the worst consequences took place. Some grievances among the inmate population actually began two days before the reported events. In different areas of the country, detainees had started to demand the application of a special amnesty law issued by the Parliament a few weeks prior. In fact, the controversial law proposed by the then newly elected Prime Minister, Abiy Ahmed Ali, had granted amnesty (and thus release) only to political dissidents accused of treason and detained by the previous government, leaving all other detainees waiting for a general amnesty disappointed. The news of the prison uprising also leaked abroad. On a US-based online news site, for example, the Minister of Communications of the Government declared that numerous detention centers had been set on fire by detainees over two days, with many attempting to escape, primarily in the city of Mekelle, in the Tigray Regional State and then in Debre Markos, Woldiya, and Finote Selam in the Amhara Regional State. Deaths and injuries inevitably occurred. The outbreak was spreading, and there were concerns about disturbances even in Addis Ababa’s prisons. The government declared the prisoners’ complaints illegitimate because they misunderstood the new amnesty law and urged the prisoners to refrain from using force.31
So, the model prison, much written about as a place from which inmates “would not want to escape”, was the scene of the worst clashes that occurred in the context of a national prison protest. The amnesty issued by Prime Minister Abiy Ahmed sparked a fierce conflict, turning the field of ideologies into a field of violent political struggle over rights perceived to have been denied. The government justified its stance by consistently asserting the illegitimacy of the prisoners’ demands. The release measure was “extraordinary” since it was only directed at political prisoners, not extended to all common offenses as it would have been in the case of a change of legislature, that is, the entire parliament. The EPRDF and the subsequent government had indeed only changed leadership, with Abiy, to avoid early elections. For the detainees, however, such explanations were perceived as pretextual. Moreover, the purely formal question of the lack of a change in legislation was relegated to the background by the emphasis on “renewal” adopted by the new Prime Minister during his first speeches in the squares and on television. Presenting himself as a transformative figure, he rallied the public, intellectuals, the military and all other national institutions with calls for “renewal”, “revolution”, “democracy” and “openness”, while placing significant emphasis on the idea of “rectifying past mistakes”.32
Abiy’s rhetorical spirit was like an electric shock along the country’s social spine. At the time of his proclamation and immediately after, more than half of the people I knew were enchanted by the new leader. Many friends and collaborators saw him as the bringer, finally, of true parliamentary democracy, even if they watched and judged him from Tigray, that is, from the region whose politicians (of the TPLF) had just had to relinquish the leadership of the great national coalition (EPRDF). While the Prime Minister’s rallies in squares and on television energized the public, they fostered a deep sense of betrayal within prisons. Common prisoners simply could not justify, beyond official and legalistic explanations, why the very government that professed itself radically “different” and innovative, in the post-Derg era, was addressing only a small part of prisoners, the old “traitors” of the State, who had moreover (until then) never received any favors or sentence reductions. Abiy had then secured a peace deal with Eritrea, just 16 days before the Mekelle Prison’s riot, by sealing a strategic alliance with the Eritrean dictatorship, a move that would soon after win him the Nobel Prize and greater international prestige. However, the release of political prisoners had represented a significant moment of symbolic and ideological capitalization for his leadership. This move did not take into account, however, that, in the Ethiopian penal state, prisons were and still are already places of high symbolic intensity, as well as centers of specific discursive practices, and that post-election amnesties represent the main regulation valve of political and judicial stress and a mechanism of “restitution” (to the streets) of lives that are continually torn from it.
In the conflict that was unleashed in November 2020 and ended two years later between Tigray (still led by the TPLF) and the federal government led by Abiy Ahmed, the self-aggrandizing strategies and control policies of both sides operated at full capacity.
Abiy gained support from foreign nations through the supply of weapons and equipment (from Turkey, the United Arab Emirates, and Iran)33, but also through the intervention of troops from Eritrea, with whom peace had just been agreed, and also in diplomatic terms (especially from Russia)34. The TPLF, on the other hand, opened the prison doors to call for the new battle for freedom. During my last period in Tigray, just before the civil war, the number of regional special forces had been greatly increased (cf. Plaut & Vaughan, 2023: p. 223). At a checkpoint, while traveling on public transport, one of the “new” soldiers of Tigray Special Police Force is a familiar face. It was my friend Kibrom from the old Quiha gang, whom I had visited several times in Mekelle prison itself. His life was taking a new course. He could redeem himself perhaps from the fraud charges (from his point of view undeserved) that had stopped his career and put his family in crisis, due to incarcerations and the difficulty of finding a stable job. Then again, as he told me some time later,
Weyane35 is still the same, it has its flaws, but in the end it rewards those who fought for it. Maybe this is my chance.36
I recall the earlier words of my informant, disillusioned about retraining programs and critical of their inability to permit real life change. An earlier study had highlighted how such programs work, becoming a concrete form of gain and social mobility for its adherents, solely when they are implemented in sectors and activities that are already well established, especially in local informal markets (Marasco, 2024).
9. Conclusions
Yet despite this, the system of redistribution of prestige and power, internal to the large regional prison, managed to reiterate a process of subjectification aimed at not eroding the consensus implicit in the heirs of the Revolution in Tigray. In Kibrom’s life, as in that of many others, there is a continuum between the prison and the street involving consent and acquiescence to power, grafted into the deep nerve of a carceral society. It is a governmental porosity facilitated by the ideological management of international development logics, starting with the “attitude to work” conceived as an impulse to an individualistic readiness to lift the whole country out of poverty, but which actually functions as a process of accumulating symbolic capital that power can redeem in times of need.
After the Tigray War (2020-2022), the Ethiopian prison system became a crucial tool to consolidate political power and manage the aftermath of the conflict, with direct implications for the treatment of political prisoners. Prisons were used to detain opponents of the federal government, particularly members of the Tigray People’s Liberation Front (TPLF) and supporters of Tigrayan resistance. During the conflict, thousands of people were arrested on vague charges, such as “collaboration with rebel forces” or “violation of regional bans”. Following the Pretoria peace agreement (2022), the federal government released some detainees but excluded those accused of “serious crimes”, maintaining a selective logic that penalizes dissidents37.
A report from the New Lines Institute (2024) documented acts of genocide against Tigrayans, including systematic rape, induced famine, and destruction of vital infrastructure. Prisons became theaters of torture, arbitrary detentions, and sexual violence against political prisoners, often justified as “counterterrorism”. The war accelerated the transition to a more centralized political system, with Abiy Ahmed’s Prosperity Party replacing the previous ethnic-based federal model. Prisons became spaces to neutralize inconvenient opponents or perceived ones38.
The European Union, through resolutions of its Parliament (2022), condemned human rights violations in Ethiopian prisons39. Despite facing criticisms for failing to honor commitments to judicial transparency, Ethiopia remains one of the largest beneficiaries of international aid. The international community, divided between support for Abiy (China, Russia) and Western condemnations, has no ability to influence prison reforms.
The partial release of prisoners in 2023, framed as a gesture of pacification, served more as a propaganda tool than a genuine step toward reconciliation. Several years after the conflict, the exact number of individuals who disappeared or were presumably imprisoned during the war, by opposing factions, remains unknown and has become a further cause of constant tension. The indiscriminate use of arbitrary detention still persists across Ethiopia. For instance, Amnesty International (2024) documents thousands arbitrarily detained in the Amhara region, a federal ally and central to the Tigray war. Activists, journalists, academics, and civilians were held without charges, often in inhumane conditions, denied medical and legal aid. These detentions occurred during a government operation against a regional “subversive” armed group, escalating into widespread repression. The report highlights violations of international law, including the right to a fair trial and prohibitions on arbitrary detention, alongside allegations of physical and psychological abuse.
Many political prisoners remain incarcerated, while post-war reconstruction has been used to bolster government narratives of a “return to order”, ignoring demands for transitional justice. Without a change in direction, prisons will continue to be pillars of a repressive system, fueling ethnic tensions and instability in the Horn of Africa.
Acknowledgements
This research was carried out as part of the Italian Ethnological Mission in Tigray-Ethiopia (IEMET), directed by Professor Pino Schirripa, whom I thank for his valuable advice. I would also like to thank the Italian Ministry of Foreign Affairs and International Cooperation (MAECI) for its support of the Mission.
NOTES
1The Derg (or Dergue, lit. “committee”) was a military junta that ruled Ethiopia from 1974 to 1991, following a coup that removed Emperor Haile Selassie from power (although it had already formally ceased to exist in 1987).
2Tigrinya.
3Proclamation No. 652/2009.
4See: https://www.giz.de/en/downloads/giz2012-making-good-governance-tangible-en.pdf (accessed 09-01-2025).
5Italics are my own.
6Tigrinya is the language spoken by Eritreans and Tigrayans. Most of the expressions collected in this study are therefore in Tigrinya. There is obviously a coexistence and contamination with Amharic (which is the language of the bureaucracy and the federal administration). In the text and in the notes, the language each expression belongs to will be indicated.
7Interview conducted on November 9, 2015 in Mekelle.
8In Tigrinya, the term shmagle (
), singular form, refers to both the elder and the mediator in conflicts resolved through social courts, before escalating to civil or criminal court systems. This traditional reconciliation practice is becoming less common, particularly in cases of street violence and disruptions (see Marasco, 2024).
9The real names of my informants have been changed to protect their privacy.
10Amharic word widely used in Tigrinya nowadays.
11Interview conducted on October 21, 2015 in Mekelle.
12Interview conducted on October 23, 2015 in Mekelle.
13The Tigrinya word for “power” is hayli (
). So “powerful” is hayal (
). My interlocutors use hayalnet (
), a term widespread among the urban youth population. It is a Tigrinya neologism, like most terms with the suffix-net (typical in Amharic).
14Interview conducted on October 28, 2015 in Mekelle.
15Both Amharic and Tigrinya (the word reported in the ge’ez alphabet is in Tigrinya).
16Interview conducted on August 16, 2018 in Mekelle. The expressions are all in Tigrinya.
17Interview conducted on August 9, 2017 in Mekelle.
18Smaller cells are dormitories of 70 inmates; others house 120.
19The expressions are all in Tigrinya.
20Both Amharic and Tigrinya.
21Interview conducted on September 18, 2019 in Mekelle.
22See Section 5.
23Interview conducted on August 07, 2017 in Mekelle.
24Interview conducted on August 09, 2017 in Mekelle.
25Foucault (1977: p. 198) writes: “The exile of the leper and the arrest of the plague do not bring with them the same political dream. The first is that of a pure community, the second that of a disciplined society. Two ways of exercising power over men, of controlling their relations, of separating out their dangerous mixtures”.
26Tigrinya.
27Interview conducted on August 07, 2017 in Mekelle.
28Tigrinya word derived from the Italian lustrare, meaning “to polish to a shine”.
29Interview conducted on October 11, 2015 in Mekelle.
30The request for Yqrta (pardon) can be made at three specific times: to be released for the new year, on Martyrs’ Day, and on the day of liberation from the Derg. The request must be submitted at least two months prior to allow for analysis, reconciliation, and mediation processes.
31See: http://archive.fo/ulKdI [accessed 28/12/2024]. Many correspondents from Ethiopia, in order to continue their work and publish in foreign-based newspapers, continue to write anonymously.
32As a Pentecostal, the “born-again” narrative (see Meyer, 1998) is in Abiy’s wheelhouse. On the topic of Pentecostal regeneration related to “late capitalism”, see: Schirripa (2008).
33See: ISPI, 2023.
34See: Plaut & Vaughan (2023: p. 333).
35Weyane (
) means “revolt” in Tigrinya, but it is also commonly used as a synonym for the TPLF, which spearheaded the revolt against the Derg from 1975 onward. By the same process of symbolic appropriation, the peasants’ insurrection in 1943 against the emperor was renamed Kedamay Weyane (
), i.e., “first revolt”, drawing a historical continuity with the ruling party (cf. Tareke, 1996).
36Interview conducted on October 1, 2019 in Mekelle.
37https://adf-magazine.com/2023/07/information-scarce-about-missing-ethiopian-prisoners-of-war/ (accessed 15/02/2025).
38https://newlinesinstitute.org/rules-based-international-order/genocide-in-tigray-serious-breaches-of-international-law-in-the-tigray-conflict-ethiopia-and-paths-to-accountability-2/ (accessed 15/02/2025).
39https://www.europarl.europa.eu/doceo/document/TA-9-2022-0351_EN.html (accessed 15/02/2025).