INTRODUCTION
When a visitor enters the office of the Criminal Investigations Department of the Ghanaian police, it is at first impossible to distinguish plain-clothes police officers from civilians. Every visitor tries to do so, as the distinction between state officials and civilians is the expected and implicitly stipulated interactional setting of everyday police work. In the course of criminal investigations in northern Ghana, however, the actors' actions put this distinction into question. Civilians arrest suspects on their own, taking over what are tasks for the police. Police officers are heavily influenced by civilians, often forming paid alliances with them. Criminal investigations are an on-going negotiation between complainants, suspects and police officers, each trying to ensure a favourable outcome for themselves. This can be described as brokerage in an administrative setting (Blundo Reference Blundo and Lund2007: 133–5; Blundo & Olivier de Sardan Reference Blundo, Olivier de Sardan, Blundo and Olivier de Sardan2006: 87), yet brokerage is the main activity of all actors involved, not only of a group of professional mediators. Through an ethnography of the specific challenges of police work, this article describes the ambiguity of such mostly informal practices of state officials in West Africa. However, it also argues that these should be understood as a necessary adaptation of police practices to a context of low legitimacy and resources. Placed in relation to practices described by researchers on the police in European countries, they are local solutions to comparable challenges of police work, and not fundamentally different.
The boundary between state officials and civilians in West Africa seems blurred because of these informal practices, but just as important are practices of officials which limit and frame these negotiations. Ghanaian police officers chose to maintain the boundary at specific times, retaining their role as state bureaucrats. They even explicitly expressed their wish to construe a clear separation between themselves and civilians: a sergeant put this concisely, saying that ‘there should be no open doors in the police’. State officials seek to maintain this boundary because it provides their organisation and themselves with the qualities of distance, sole competence and impartiality, in other words with aspects of stateness. Instead of understanding the police and its officers as instruments of the abstract ‘state’, this paper explores the practices by which actors attempt to affirm the stateness of their organisation. Exactly because police officers in northern Ghana so manifestly struggle to achieve a clear separation, this research on boundary work can contribute to broader understanding of the state.
Blurred boundaries is an often used concept in the anthropology of the state, and is also the title of one of the most influential articles (Gupta Reference Gupta1995). Attesting that public servants collapse the distinction between state and society, public and private, many anthropologists have called into question the adequacy of these categorisations in the everyday realities of non-Western countries (Ferguson Reference Ferguson2006: 112; Gupta Reference Gupta1995: 384; Mbembe Reference Mbembe2001: 51; Poole Reference Poole, Das and Poole2004: 43; Roitman Reference Roitman2005: 157). Blurred boundaries is a key term to describe informal everyday practices of officials, a spatial concept that makes it possible to link everyday practices of officials with theories of the state.Footnote 1 The sparse research on the police in Africa mirrors the notion of an unclear separation between state and society. Hills (Reference Hills2000: 10) first defines one of the core attributes of a police force as establishing an ‘identifiable boundary’, and then shows how police organisations in Africa are unable to provide such a boundary. Hornberger (Reference Hornberger2004: 213) understands policing as a ‘negotiation of this boundary between state and society’, but concludes that the police in South Africa have been transformed into a private source of security, similar to vigilante organisations and private security companies.
The way the boundary concept has been employed has two main shortcomings. First, many researchers seem to regard the blurred boundary between state and society as a phenomenon particular to non-Western, especially African, states, and regard it as an aspect of state failure. In doing so, scholars implicitly compare an idealised image of Western states with the realities of non-Western states. They ignore sociological studies of Western bureaucracies, which describe a ‘relative autonomy’ (Lipsky Reference Lipsky1980: 16) of public officials or a complex web of partly illegal relationships between bureaucrats and clients (Luhmann Reference Luhmann2000: 261).Footnote 2 Researchers on legal pluralism question the reality of a state monopoly on coercive force (Benda-Beckmann et al. Reference Benda-Beckmann, Benda-Beckmann, Eckert, Benda-Beckmann, Benda-Beckmann and Eckert2009: 15). As early as the 1920s, Weber (Reference Weber, Roth and Wittich1978: 216), for instance, saw bureaucratic authority as an ‘ideal type’, always modified in reality. Research on the police is currently interested in similar topics, analysing the pluralisation of policing (Newburn & Reiner Reference Newburn, Reiner, Maguire, Morgan and Reiner2007: 932), and the necessary reliance by the police on informal practices (Reichertz Reference Reichertz, Reichertz and Schröer2003: 40), as well as researching intense police–private interest cooperation, leading to ‘boundaries’ becoming ‘diffuse’ (Hoogenboom Reference Hoogenboom1991: 28; see also O'Neill Reference O'Neill2005: 130). Although all of these phenomena are understood as a blurring of boundaries, they are not seen as totally collapsing the distinction between state and society. Rather, in the context of Western states, they are conceptualised as necessary for the functioning of state organisations, not as a sign of state failure. Hansen and Stepputat (Reference Hansen, Stepputat, Hansen and Stepputat2001: 16) argue that the abandonment of distinct boundaries is an elaboration process of the state in Western Europe, and Santos (Reference Santos2006: 44) remarks that even the withdrawal of boundaries is organised and enacted by state officials. There may be differences in the scope, frequency and degree of the legal framework applied to regulate the blurring of state–non-state boundaries by public officials, but the blurring itself is inherent in all states and not specific to ‘African’ or ‘failed’ states.
A second problem concerns the inclination of many researchers to disregard the ambiguity of the boundary and its maintenance by public officials through their everyday practices. The ethnographies of African bureaucracies already cited nonetheless reveal distancing practices by public officials (Gupta Reference Gupta1995: 380), criminal prosecution of illegal practices (Roitman Reference Roitman2005: 157), and even public officials explicitly criticising this ‘blurring’ (ibid.: 162); yet this information is not taken into account in the later analysis. Lund (Reference Lund and Lund2007: 1), in contrast, acknowledges this paradox: ‘On the one hand, actors and institutions in this field are intensely preoccupied with the state and with the distinction between state and society, but on the other hand, their practices constantly befuddle these distinctions’ (see also Fuller & Harris Reference Fuller, Harris and Fuller2000: 24). Researchers on the police, especially, have begun to comprehend the blurring of boundaries in West Africa not as a failure of the state, but as a means of extending state control. In Baker's view (2007: 138; see also Göpfert Reference Göpfert2012: 54), non-governmental security organisations are still in the orbit of the state, and Kyed (Reference Kyed, Jefferson and Jensen2009: 57–8) sees marking and breaching the boundaries between state and non-state jurisdictions as practices of the police to reclaim legitimacy. Blurring is an enigmatic term that problematically conflates the recession and the extension of control by state organisations.
To tackle such shortcomings, this article develops theoretical observations out of and closely linked to ethnographic data. This approach makes it possible to grasp the complexities of police work, and to link everyday practices with the abstract functioning of state bureaucracies. Interactions in the course of criminal investigations involve complex communications, choreographies, and spaces. The specifics of these interactions are the important data, because these determine how the state is realised. My own experiences were part of the data, as police officers asked me to assume the role of a police officer, that is, to support them in raids and other tasks. My position in relation to the state organisation was not fixed, and attempts to incorporate or exclude me revealed both the fluidity of the positioning and the struggle to remain a distant state organisation. Instead of assuming the boundary to be an interactional setting with assigned roles that is either blurred or fixed, the mundane practices of state officials can be re-examined as the result of forms of boundary work and boundary-shifting by actors.Footnote 3 Through a close analysis of the everyday practices of state officials, the drawing of the boundary behind the blurring can be analysed in order to grasp the ambiguities of the process and to further develop the original concept.
This paper is based on fifteen months of ethnographic fieldwork on the police in Ghana. The author conducted participant-observation and interviews as an ‘attachment’ in several police units (on access, see Beek & Göpfert Reference Beek and Göpfert2011). By this emphasis on everyday police practices, the article inevitably loses ethnographic depth regarding local communities, and historical and political aspects.Footnote 4 The paper focuses on the Criminal Investigations Department (CID) of the regional headquarters of the police in Wa, the capital of the Upper West Region, the poorest and least developed region in Ghana, with a population of 230,000. Most of the population is Muslim, and the main ethnic groups are the Wala and the Dagara. About 200 police officers serve at police headquarters, which is situated in the centre of the city.
SUBMITTING A COMPLAINT: MAINTAINING THE BOUNDARY
The starting point for nearly every criminal investigation by the police is a complaint by a civilian. Civilians dealing with an occurrence that could be categorised as criminal are supposed to submit a complaint at the police station. But in everyday life, civilians prefer other solutions. They approach religious authorities, chiefs (who are perceived as an official judicial authority for minor cases), the head of the family, or the district assembly, trying to settle their disputes without the involvement of legally relevant public officials, the police and the courts. While some local providers are perceived as state officials or are publically funded, none have jurisdiction over criminal cases from a legal standpoint. These multiple forms of dispute resolution have been researched under the heading of ‘legal pluralism’ since the 1970s (see Benda-Beckmann & Benda-Beckmann Reference Benda-Beckmann and Benda-Beckmann2006). They are not directly part of this research but provide an important context for criminal investigations. This avoidance of the police has external and internal causes. The Wala often refer to the ideology of tijaabunyeni (meaning roughly ‘we are all the same’ in Wala), which encourages the settlement of disputes within family and ethnic networks. Even before submitting a complaint, relatives of the suspect pressure the victim of the crime to settle the case without involving the police and the courts. The selection of the organisation to deal with the criminal occurrence depends on availability, perceived chance of success, and costs (Baker Reference Baker2004: 206). Only a fraction of potentially criminal occurrences are reported to the police, yet this phenomenon is not exclusive to northern Ghana, but is recognised worldwide in legal sociology (Blankenburg Reference Blankenburg1995: 13). The reporting of a case marks the beginning of a series of interactions over the course of the criminal investigation:
In the late evening, Counter NCO Boateng and the station orderly sat behind the counter in the charge office. The counter, one metre wide, made from massive darkened wood, separates the police officers from visitors. At the back of the room was the cell door. The prisoners stood directly behind the iron bars in their underwear, looking apathetic and exhausted.
Two young men and a woman entered the charge office. The woman cried, appeared traumatised, and had fresh wounds in her face. Boateng let them wait, taking his time and first finishing his meal. Always looking at the ground, the woman then told the story of an unprovoked attack by another woman. The NCO began writing it down in the diary of action, but complained, grumbling about the lack of medical forms. Yet, he explained, it would not be a problem, as he had copied the forms on his own and could provide them for 4 cedis (approx. $US3). One of the men accompanying the woman complained about this need for payment, but the woman gave the sergeant 3 cedis, concealed in her hand. Then Sgt Boateng sent the woman to the hospital to fill out the medical form.
Half an hour later, another woman arrived and greeted Boateng like a long-time friend. She told the same story and presumably was the other woman involved in the struggle. Boateng was friendly and sat with her on a bench in front of the charge office, recording a second diary of action.Footnote 5
According to formal law, the police must always begin an investigation once a civilian files a complaint, opening an investigation and an accompanying case docket. In practice, police officers at the counter refuse many complaints by ignoring people when they try to file a grievance. Similar forms of discretion are an underlying and necessary principle of police work everywhere, as is the underenforcement of the law concerning certain offences (Waddington Reference Waddington1999: 38, 61). This specific form of duty avoidance is a reasonable practice for Ghanaian police officers: the counter NCO has a heavy workload and twelve-hour shifts. Many minor cases, mostly assaults and familial disputes, are not pursued by the victim and are thus perceived as pointless paperwork. Other cases, in which the suspect is well connected, could be risky for the police officer involved. Lastly, because it is the detectives, not the policemen in the charge office, who later carry out the criminal investigations and receive any possible tips, there is no long-term incentive. The counter itself creates a visible physical division between civilians and the uniformed police officers. There is also a social distance relating to ethnicity and language. The police in Wa are mainly southerners, who are quartered in barracks and often do not speak the local languages.Footnote 6 Yet by speaking English, as instructed in official regulations, the officer emphasises the formal character of the interaction, de-emphasising the ethnic dimension.
The police officer maintains a boundary between the state organisation and society by his practices, embedded in organisational procedure and architecture. From an ethnographic perspective, stateness emerges as a result of mundane practices by actors (see Pates Reference Pates2009), rather than by conceiving the state as an entity that works through state officials and represents itself in artefacts. With these everyday practices, in other words with boundary work, actors ascribe certain qualities to an organisation like the police: the organisation is distant, solely competent regarding criminal cases (ignoring local forms of dispute resolution) and impartial (employing bureaucratic procedures). These qualities distinguish the police from competing local forms of dispute resolution, and are all aspects of stateness of an organisation which emerges if actors successfully endow their organisation with them. Nevertheless, such qualities are based on a preconception of the state–society divide, as the processes of imagining and producing the state are intertwined (Midgal & Schlichte Reference Migdal, Schlichte and Schlichte2005: 14–15; Mitchell Reference Mitchell and Steinmetz1999: 77). Everyday interactions are not the sole source: boundary work is about maintaining and not about newly setting this boundary. Other sources include the making of boundaries at an inter-organisational level (especially towards the political system), and specific historic configurations (see Bierschenk Reference Bierschenk2008).
The police's distance from society and its sole competence are the main reasons why people prefer the police to local dispute resolution in certain cases. The visible cells make every civilian aware that the police enforce the law in an uncompromising and punitive way, outside the normal social order. Human rights organisations may harshly criticise the conditions in police cells, but do not have a serious impact. One officer remarked: ‘I pity the prisoners. When I beat them, I only beat them with a cane.’ While most disputes are dealt with by alternative means, it is exactly this uncompromising punishment of wrong-doers that many civilians seem to expect and desire from the police when submitting a complaint. Both reporting women above, for example, were infuriated because of their recent fight. When threatening others with police action, civilians use the term ‘I put you in the counterback.’ By this they explicitly refer to the cells and the punitive aspect of criminal investigations. Filing a complaint often serves as an escalation at the end of a lengthy dispute, to jump-start a dispute that has not been resolved in attempted settlements. Police officers maintain the boundary as much as possible in the publicly accessible charge office, because this boundary provides their competitive advantage. They establish the police as a distant organisation that does not depend on compromise, in comparison with local forms of dispute resolution. They also present the police as the sole competent organisation that can take on all criminal cases, regardless of which local provider previously claimed jurisdiction.
Another incentive for the two women above to go directly to the police is prompted by police procedure. Police refer to the civilian who is the first to report a case as the complainant. Police officers in northern Ghana grant the complainant several advantages. One complainant, who tried to be the first to report a case in a year-old dispute, remarked that ‘the first complainant is the best person’. As complainants, individuals try to establish their perception of the case first, while the other party will be arrested and become the target of the criminal investigation.
From the complainant's perspective, the boundary is primarily a hindrance when trying to submit a complaint in the charge office. Civilians have to circumvent it just to be able to submit their case. For Schaffer and Lamb (Reference Schaffer and Lamb1974: 79), such circumvention of the formal system is inevitable in certain access situations, and Bierschenk (Reference Bierschenk2008: 129–30) identifies money, connections and status as currencies used to influence administrative procedures.
The first woman depicted the attack as a brutal assault, trying to reach the officer on an emotional level. She also brought two friends as intermediaries, which is a common practice and often involves local authorities. But bringing supporters and trying to convince the officer of the seriousness of the crime did not work in her case. An assault is not considered a serious offence by the police, and her supporters did not have any real influence. NCO Boateng offered her a different approach, providing, on compensation, a copy of a required document and framing this as mutual help. By virtue of money being handed over the counter, the counter becomes a space for social exchange instead of one of separation, convincing the police officer to accept the woman's complaint.
The second woman utilised her personal relationship with the NCO to report her case. Boateng left the counter and talked to her outside the police station, treating her like an acquaintance. Blundo (Reference Blundo and Lund2007: 132) describes how clients always seek to personalise an administrative relationship. Like money, personal connections allow civilians to submit the complaint outside the logic of the boundary, framing the interaction as a social one. Connections to detectives, often with local authorities as mediators, can enable civilians to approach investigating detectives directly, circumventing the charge office completely. Civilians try to approach the highest-ranking police officer with whom they or a close acquaintance has a connection.
ARRESTING A SUSPECT: SHIFTING THE BOUNDARY
Criminal investigations are carried out by a single detective of the CID. The aim of every investigation is to make an arrest and obtain a cautioned statement. Because of the lack of forensic science, physical evidence is rarely collected. Investigating a criminal case is mainly a communicative task. Plainclothes detectives work in their offices and in the street. They are given far more leeway than uniformed police. In theory, every action has to be ordered by the unit commander, but Inspector Barima let detectives act independently and often post-dated orders.
Criminal investigations in northern Ghana are not an automatic process, propelled by the organisation, but require the use of time, money and connections by the complainant. Investigations are often delayed and officers simply say kO bra, meaning ‘go and come later’ in Twi. Detectives often expect incentives for their pursuit of the case, mostly small amounts of money called a ‘dash’ or ‘tip’. Police officers and their superiors perceive the acceptance of this money not as corruption, but as part of establishing a connection, a social exchange (Blundo & Olivier de Sardan Reference Blundo, Olivier de Sardan, Blundo and Olivier de Sardan2006: 5; Hasty Reference Hasty2005: 274). In portraying the civilian as a ‘friend’, which in Ghana entails solidary gift-giving, police officers make the payments indistinguishable from clearly instrumental ones.Footnote 7 Complainants do not necessarily share their view, but may also try to present their interaction with the detective as a friendship, to encourage further action and influence the detective's perspective on the case. The partially changed allegiance of the officers is reflected in their choice of words, as they sometimes call the tips ‘sitrep’, an abbreviation of ‘situation report’. A sitrep is actually a summary of a case which enables senior officers to control the work of their subordinate detectives. The same word is used for the handed-over money which connects detective and complainant. The influence of the civilian is apparent in the vocabulary – the important sitreps do not circulate between the detectives and their senior officers, but between the detective and complainant. By describing these actions as a shifting of the boundary, rather than blurring, they can be analysed more accurately. Both police officer and complainant shift the boundary; the complainant expands his influence and police officers are in effect transformed into his supporters. Actions circumventing the boundary, as described at the counter, ultimately shift it.
On rare occasions, a complainant's influence on a detective becomes so strong that the complainant, to some extent, becomes the de facto employer of the detective. The detective does not register the case, although he still arrests suspects and threatens them with legal proceedings. Detectives do not have to share the tip (an otherwise expected practice) with their colleagues and close superiors in such cases. This can lead to two CID units investigating the same case on behalf of different complainants. The two women mentioned above started two cases, so that each had to try to focus the investigations on the other party. Hornberger (Reference Hornberger2004: 228) describes a similar practice in South Africa, arguing that the police are ultimately transformed into a privatised source of security. This is also part of the popular imaginary in Ghana: police officers are envisaged as being bought by the highest bidder, with envelopes of money circulating to the highest ranks.
But shifting the boundary not only transgresses the impartiality of law enforcement; the limits of this shifting and its ambiguous direction are equally important. Ghanaian police officers have a strong sense of limits, even though they enable boundary-shifting to an extent – but the exact degree of shifting is the result of continuous argument and negotiation. Shifting is limited by constant boundary work. The Ghanaian saying ‘You can never be a friend to a policeman’ reflects this: it means that despite a good relationship, a police officer may still enforce the law against you. Detectives mostly ignore their monetary interests in first-degree felony cases threatening social order. Detectives also sometimes defy influence on moral grounds. Notably, there is a system of mutual control by all officers that makes a categorical privatisation nearly impossible. The ambiguity of boundary-shifting becomes apparent in the vocabulary of police officers, who mostly characterise their connection to the complainant as friendship, suggesting reciprocal help and influence. Detectives actually depend on such help by complainants: because of their low-level legitimacy and ignorance of local languages, police expect complainants to pursue the investigation partly on their own.Footnote 8 Officers give complainants advice on how to find stolen property or identify suspects. They expect them to help in any arrests that may take place, as in the following encounter:
Detective Arday and Detective Jawula went into the inner city to make an arrest. Both were unarmed, but carried their handcuffs visibly. They followed the directions given by two complainants. From an outsider's perspective, these two came across as the real leaders of the group, guiding the detectives. When the group reached the house of the suspect, the complainants entered first, without knocking or speaking. The detectives followed. Det. Arday explained curtly that we were there because of a stolen mobile phone. The two male suspects looked dumbfounded and irritated. The detectives did not even touch the suspects. Going back, the suspects were in front, followed by the complainants. The detectives followed them, appearing uninvolved in the whole process. During our walk, an aggressive argument broke out between suspects and complainants. The detectives did not interfere, just urging everyone forward. Immediately after arrival at the station, the suspects were processed and put into the cells.
(revised fieldwork notes)When arresting a suspect, the support of complainants is a necessity, arising from the lack of administrative infrastructure (such as street names, house numbers and identity cards). Complainants also serve as translators, if there is no common language between detectives and suspects. Due to the police's lack of full legitimacy, the population is generally unwilling to point out suspects or help the police in any other meaningful way. Tips play a practical role, too, as the CID is not provided with any cars, and a tip enables police officers to pay for a taxi. Additionally, involving complainants serves as a tactic by Ghanaian detectives to cope with the problem of bystanders. In northern Ghana, bystanders often attack police officers during arrests. Because the field of interaction in public also includes these potentially threatening bystanders (see Paperman Reference Paperman2003: 410), the detectives stage the arrest as a private scene between two disputing parties – the complainants lead the way, the actual arrest takes place inside the house, and the detectives refrain from involving themselves too much. As a private dispute, the scene requires the discretion of uninvolved parties, reducing the likelihood of interfering bystanders. Police procedure advises the deployment of armed uniformed police during arrests, but detectives mostly refrain from this, calling their arrests ‘invitations’. For complainants, this tactic is problematic; they appear as auxiliaries of the police and are often pressured and threatened.
If detectives are really interested in a case due to promised sitreps, a perceived moral obligation, or a threat to the social order, they utilise their social networks for information-gathering. Constable Mensah, for example, found the thief of a police officer's motorcycle in the course of four hours, with no initial leads, just by relying on his contacts. Detectives establish these networks through their contacts with complainants and even with suspects whom they have treated decently. These contacts are called ‘informants’ or ‘friends of the police’, the latter term signifying a reciprocal claim on support and security. Such ‘friends’ sometimes even work informally for the police on a regular basis (like the interpreter Darko), and play a role similar to the civilian intermediaries of the customs authority described by Blundo (Reference Blundo and Lund2007: 135). Yet the designation ‘friend of the police’ is quite open, and each complainant is one, to an extent. Mediating, respectively shifting the boundary, is less a fixed social role and more a necessary action in police–civilian interactions.
Police officers shift the boundary in the other direction by incorporating complainants and ‘friends of the police’. Thereby they enable criminal investigations in the context of low legitimacy and a lack of knowledge, extending the control of the state organisation and facilitating the social integration of a bureaucracy. Not so much ‘distorted’ forms of official policing (Hornberger Reference Hornberger2004: 226) as an adaptation of police procedures to the local context, these practices simultaneously threaten the impartiality of the investigation.Footnote 9
TAKING A STATEMENT – THE DYNAMICS OF BOUNDARY-SHIFTING
Statements obtained from suspects are called ‘cautioned statements' and are the cornerstone in solving a case, due to the reliance on such statements in a possible trial. Because of the right to refuse to give evidence, the suspect has a structurally strong position while giving a cautioned statement. Schröer (Reference Schröer, Reichertz and Schröer2003: 66) conceives interrogation procedures as an attempt by detectives to disguise this fact, and to trick suspects into not seeing or using their advantage. The CID detectives in Wa utilise similar tactics, simultaneously threatening suspects and giving them a way out through cooperation:
The cautioned statement took place in the second office of the CID, a small and dimly lit room. The suspect, thirty-years old, sat in the middle of the room. The two desks and several stools were facing in his direction. Detective Dwamena sat on one of the desks, while Darko the interpreter sat positioned between him and the suspect. The statement concerned a theft case. The interrogated suspect was a trader charged with ‘dishonest receiving’, meaning buying stolen goods, the offence was the byproduct of the real case against the suspected thief. Det. Dwamena spoke English and Darko translated into Wala, mixing in English phrases. Dwamena was in charge of taking the statement, but in reality Darko did most of the interrogation. He talked at the suspect, urging him to ‘help the police’, his tone and demeanour friendly, but serious.
The suspect looked confused and denied all accusations. As the interrogation progressed ten minutes without results, Dwamena became slightly agitated, pacing up and down behind the suspect. He loudly remarked: ‘This man is a criminal.’ Becoming frustrated, the detective hit the man lightly from behind with his open hand. Later he brought the actual thief into the room, confronting the trader with the thief's statement and heavily incriminating him. The trader still denied all charges. All three leaned on the trader, urging him to testify for his own good, on one occasion hitting him again. Darko plainly told the trader: ‘Give the police officer something to write down.’
Since the trader was insisting on his statement, Dwamena changed his tactic. He remarked aloud: ‘Then we bring him [the trader] to court, the main suspect will be free and he will be sentenced’! All involved started to get up. But as the trader did not react, they shortly sat down again. Another detective entered the room. Dwamena explained the situation, to which he pragmatically responded: ‘Then we beat him’, picking up a bike lock. Following Darko, I left the room at this point.
(revised fieldwork notes)The interior of the interrogation room is arranged to confuse the suspect. All desks point in his direction. Surrounded by police officers in a dimly lit room, he has no control over the space and cannot even see all the officers. The interrogation procedure constantly escalates, basically offering him two counterparts at every turn, that is, Dwamena and Darko. The translator Darko is also an ethnic Wala, speaking the same language and wearing the same clothes as the suspect. He is not a police officer and officially does not even receive payment. In court, he will serve as an ‘independent witness’, vouching for the proper taking of the statement by the police. Darko made use of his unclear location, offering the suspect a friendly connection to the police officers, similar to Darko's own relationship. Describing a possible statement as ‘help’ for all involved, Darko tried to emphasise the personal level in the interrogation.Footnote 10 The interaction between the interrogating officer and the suspect is framed as an impersonal relationship. Confronted by two accusing witnesses, the suspect learns that it is not the police officers themselves but a third party who has got him into trouble. Furthermore, the words ‘give the police officer something to write down’ portray the procedure as an anonymous writing process. Neither the accusation nor the procedure seems to originate from the questioning officer, Dwamena. The translator offers to shift the boundary, while the police officer maintains it.
Offering the police the desired statement can be more than a way out – it sometimes marks the beginning of a long-term friendly connection to the police officers. The actual thief in the case, brought in here as an accusing witness, was in a similar interrogation some days before. But he became an ally of the police, offering the names of the traders involved and supporting the police in their arrests.
Denying the charges leads to further escalation of the interrogation: first the appeal that the suspect would help himself by confessing, then the categorisation as criminal, then the threat of a trial, and finally violence. The categorisation as ‘criminal’ and the threat of a trial underline the potential social implications of a criminal investigation.Footnote 11 The detectives thereby demonstrate their power to exclude people from the community (see Waddington Reference Waddington1999: 41). ‘Criminal’ or ‘hardened criminal’ is a category used by police officers and civilians alike for foreign armed robbers, drug addicts, and thieves of the Zongo quarter. Based on moral judgements more than judicial definitions, the categorisation of criminals places them in a clear dichotomy in which they are perceived as the enemies of the community (Buur & Jensen Reference Buur and Jensen2004: 147; Kyed Reference Kyed, Buur, Jensen and Stepputat2007: 148; Pratten Reference Pratten2008: 75–6). Civilians normally resist and obstruct police interventions, but demand and support investigations against perceived criminals.
The use of violence is an exception in the CID of Wa and is only used when police officers are at a loss. In most cases, the police do not necessarily need a ‘cautioned statement’, because the statements of the witnesses are sufficient. Even when a statement is needed, however, police mostly use subtler forms of interrogation. Beating the suspect from behind is a common practice that uses the fear of pain as a psychological tactic while providing anonymity for the officer. Despite their harsh criticism of corruption, the public mostly seem to expect and accept violence in police interrogation, as it supports their image of the police as a strong and uncompromising organisation. Violence also quite plainly shows that the suspect, as a criminal, has lost all protection and is now treated as an enemy. The term criminal is used by police officers as a legitimisation of violence, but the violence is also a physical performance of their power to exclude.
In most instances, this power is not demonstrated by blatant violence, but by putting a suspect into the cells after the cautioned statement. Bringing suspects in handcuffs to the charge office, stripping them of their clothes and belongings, and putting them in this hazardous space, are inherently violent procedures. The suspect is visibly staged as a ‘constitutive other’ who no longer belongs to the community (Kyed Reference Kyed, Buur, Jensen and Stepputat2007: 151). Thus, police officers perform their power physically, while simultaneously performing it in their writing. Taking statements is primarily a process of transforming a morally charged story into a fact-based text that is in accordance with judicial and police internal standards.Footnote 12 The police exclude the self-portrayal and moral reasoning of the suspect from their texts. They show that they can use the law as a resource (Waddington Reference Waddington1999: 39). Threatened with social exclusion and an unintelligible judicial procedure, suspects have to try to evade these threats immediately after their imprisonment. The bars of the cells seemingly mark the unshiftable boundary.
CLOSING A CASE: RE-APPROPRIATION AND THE LIMITS OF SHIFTING
Imprisonment is not so clear-cut in reality. Women and older people, for instance, are allowed to sit on a bench outside the cells. Even inmates inside the cells are not cut off from contact, as their relatives have to supply them with food. There is a continual coming and going, providing channels of communication between the inmates, their relatives and the detectives. After forty-eight hours, according to regulations, every suspect has to be released (except for serious offences). A trusted contact of the inmate can post bail, often accompanied by a ‘bail bribe’ for the responsible police officers. The cells not only symbolise the power to exclude by the police, but ultimately also show the openness of the category ‘criminal’ and the flexibility of the boundary.
After being threatened with a trial and realising the consequences, every suspect tries to evade having going to court by mobilising relatives and other networks. Via an intermediary, they approach the complainant to settle the case, invoking the appropriate local dispute resolution. At the same time, they also approach the investigating detectives, urging them to stop further investigations and arrests, a practice locally called ‘begging’. According to judicial regulations, police officers are not allowed to close cases, but can refuse them by declaring them settled, arguing they were civil cases from the beginning, or claiming that the statements of the complainant could not be verified. Thus police headquarters are a place of constant negotiation. Waiting lines, for suspects and their intermediaries who want to bring forward their requests, are set up in the parking lot, the corridors, and the offices. These negotiations are performed as private conversations, whispered or behind doors. Civilians try to approach the highest-ranking police officer they know to make an agreement. Senior Officer Afriyie explained in an interview the difficulties of this practice for the police:
Ahh, you know they come so much … you see, I have changed. I realised that they overdo it, so I changed gradually. So, what I do, I don't like to know details about it. This guy just came up. Then he said, ‘I have a little bit of a problem. You know I bought a moto, I have two motos, I registered one, I didn't register the other one. So my brother took it without wearing helmet.’ And I said: ‘Look, it's too common, this excuse, and I don't want to be part of it.’ That is the wrong thing to do. But at the same time, I will not dismiss him right out. I called one of my officers, to go wherever the moto is, and find out: if he was not rude and if his behaviour was apologetic, then he could be considered and given a warning letter. And they went and didn't come back to tell me, so I really don't know… In fact, if it doesn't work, he doesn't need to come again, because I don't want to overdo things, otherwise the people at MTTU [the Motor Traffic and Transport Unit] will think that I am frustrating their efforts …
So on typical cases, let's say there is an issue, a problem, and they come. They will dare not to give me money, because they know they will get problems. But here too the risk is: they will now use the investigator. They will use the investigator in terms of trying to influence him. So if you are not careful, and you get the docket, and you allow the investigator to talk you into his side, you have a problem. But we are aware, the other senior officers and me, we are aware … Whatever they bring to us, we try to be very careful. But people use them to pass gifts or something to us. Sometimes, even we don't know …
(interview)The senior officers serve as contact points for many civilians, who cultivate this valuable connection with small presents and invitations to social events. Afriyie describes his possible latitude as limited because he tries not to hinder his subordinates and demands respect from suspects. Because Afriyie himself refuses money, he mistrusts his detectives and tries to control them by their dockets, but denies that any money given to them would ever reach him. The detectives have a different view. Constable Amoah told of a case in which a mobile phone was stolen from an acquaintance of his:
So those guys, sitting near the mobile phone, were spectators. So definitely they were the suspects in the case. So I invited them. I wrote a letter to the school, I detained them, I questioned them, they were given bail, and they left. And one of the suspects was from Accra, she knew one of the commanders in Accra. He called Osei-Bonsu [a senior officer in Wa] that there is a case in Regional CID, and that the policeman is demanding an amount of 450 cedis [about US$300] from them and they fear it will be sent to court. And so she doesn't understand why the policeman should … So I was called to explain and I said: ‘I've never said that. I investigated the case. I said that the others should pay the cost of the mobile, but I never mentioned any amount that they should pay to me or to the complainant.’ Osei-Bonsu said, why should I do that. Why should I demand surety for mobile phones. He said, it was bad [repeated two times]. I said: ‘No, they opted to pay for the cost of the phone, because the parents, they wouldn't like the children to go to prison and leaving education behind. Because if they should go to prison, the school would dismiss them for stealing. They thought it wise to pay for the cost of the phone to discontinue the case.’ So the man called me and said: ‘Why should you investigate such a case? Why should someone come to you and explain she was on a motorbike and the mobile got lost? She made a case, and you didn't refuse the case, and picked up investigations of the case, why? Are you stupid, are you mad?’ [repeated two times]. He abused me in front of men, in front of civilians and suspects in his office.
(interview)Constable Ammoah saw the youths as culprits and urged them to find an agreement with the complainant. Whether he really conveyed the complaint's claim for money directly to the youths, as the high ranking senior officer alleges, remains unclear. Using their connections to high-ranking officers in Accra, the suspects convinced the senior officer to pressure Constable Ammoah, demanding that he drop the case because of formal errors. In Ammoah's view, the senior officers acted like allies of the suspects, obstructing a lawful criminal investigation.
Police officers and senior officers accuse each other of befriending civilians and being influenced by status, connections and money. Civilians call a contact person in the police a ‘friend in need’. Such a police officer is expected to help when civilians or their relatives are the target of a criminal investigation. The networks of the ‘friends in need’ are congruent with the networks of the ‘friends of the police’ – a police officer gaining support from a civilian is expected to help him or her in return. The networks enable criminal investigations despite a low-level legitimacy and other issues, but ultimately threaten the impartiality of the law and the solidarity of the police. Police officers see the shifting of the boundary with complainants as necessary during criminal investigations. But they harshly criticise the same shifting if suspects use similar networks to elude their grasp. For civilians, the shifting of the boundary allows a case to be closed, which in most instances is intended by complainant and suspect alike. A civilian explained:
If you take my family to court, you are only accusing the one who has offended you, but the whole family will hate you for that. They will come to you: if this is what our brother has done, then we settle it. But if you ignore and go to the police, they won't have anything to do with you … This is why they just mobilise themselves, tijaabunyeni, master we beg you. They settle it so that the relationships stay intact.
(interview)Because going to court leads to social tensions, both complainant and suspect try to avoid bringing a case to trial. A settlement allows a solution that incorporates multiple normative orders and moral beliefs outside state law. The shifting boundary of the police allows civilians to take back possession of a dispute, and thus limits the police power to exclude such orders and beliefs. A statistical analysis of dockets suggests that only about 5% of all cases go to court. The other cases are closed, settled, or – the majority – still pending.Footnote 13 These statistics are not specific to Ghana; Blankenburg (Reference Blankenburg1995: 9) gives a similar figure of 10% for the German legal system. Though similar in scope, the way suspects avoid being brought to court differs (Bierschenk Reference Bierschenk2008: 119). Even when, despite their low-level legitimacy and other problems, police officers succeed in arresting someone and obtaining a statement, the case can still be forcibly stopped or settled by the suspect. Going to court is unusual, signifying either a lack of support by the suspect's social networks or weak connections on the part of his network – the former often indicating that his relatives and friends consider him a criminal as well.
During criminal investigations, police officers work on the assumption of a fixed social order, based on state laws that are seen as unambiguous. Settling a case is, by contrast, an ambiguous process, incorporating and negotiating multiple views on social order, determined not by the police but by the negotiation process of the involved civilians. This is reflected in the choice of words of police officers, who legitimise settlements with the oft-repeated sentence: ‘We want to forgive them’. Instead of referring to the law, the officers invoke Christian morality. They allow the disputing parties to find a solution appropriate for them and enable the social integration of law enforcement (see Kyed Reference Kyed, Jefferson and Jensen2009: 55). But abandoning these state laws challenges the legitimacy of the police, in a way fundamentally challenging their sole competence and impartiality, qualities that bestow stateness on the organisation. Santos (Reference Santos2006: 70) suggests that the ‘porosity of the boundaries between the different existing legal orders’ can be framed either as ‘cross-fertilisations’ or as ‘cross-contaminations’.
However, police officers limit and shape the context of settlements. Detectives never take an active part in local dispute resolution processes. For the most part, as mentioned, detectives do not allow settlements in first-degree felonies or other serious breaches of social order. As the state is the formal complainant, it is the police who finally agree or disagree with a settlement by closing or continuing the case, which often involves a final payment of money to the police officers responsible. The ‘moral community’ may be constituted by the everyday practices of non-state actors, as Buur and Jensen argue (Reference Buur and Jensen2004: 146; see also Comaroff & Comaroff Reference Comaroff, Comaroff and Tyler2008), but in Ghana its limits and final legitimisation are set by the police. In effect, police officers mediate between state law and local normative orders – allowing out-of-court settlements by boundary-shifting, but maintaining the boundary by representing themselves as the final legitimising authority, limiting and framing the local negotiation.
This process of dropping a case takes months and often ends in pending cases, in which there is no final resolution by the police. When dispute resolution by local providers fails, the complainant can go to the police again – and the whole process of criminal investigation begins anew.
The more one learns about paying complainants, problematic investigations and dubious informants, the more the practices of the Ghanaian police seem to fit the concept of blurred boundaries. This article argues that the boundary is still worked on; it just becomes difficult to see it because actors shift it constantly. Contrary to the notion of a non-functioning police, the shifting of the boundary enables criminal investigations by extending the police and permitting the negotiation of multiple normative orders. The entanglements of boundary-shifting also threaten the impartiality of law enforcement – the opened doors cannot be closed again. But the question of who uses whom, and whether the police officers incorporate a civilian into their ranks or rather act for private interests, cannot be determined in one situation, as the direction in which the boundary is being shifted is perceived differently by complainant, police officer and suspect. The direction of the shift can change with every interaction, and is only conclusively set at the end of a criminal investigation. These constant oscillating movements and the multiperspectivity of the boundary lead to the persuasiveness of the metaphor ‘blurring’.
Waddington (Reference Waddington1999: 125) argues convincingly that deviant police behaviour mostly occurs in ‘contexts in which the normal distance between officers and suspects is diminished’. The more visible negotiations of this distance in Ghana reveal that this normality has to be produced by everyday boundary work – in African as in Western countries, though the boundary may be more worn out and weakened in the former. The distinctions between police officer and civilian, state and non-state organisation, and the implicit qualities of stateness like distance, impartiality and sole competence, remain meaningful for all actors involved.
