The Evidence of What Cannot Be Heard: Reading Trauma into and Testimony against the Witness Stand at the International Criminal Tribunal for Rwanda

This paper explores the silences and the gaps that cut through witness testimonies at the International Criminal Tribunal for Rwanda (ICTR) by applying a trauma lens to the narratives that emerge on the witness stand and by contrasting those with a survivor testimony. It compares the recollection of a traumatic experience with the production of legal meaning. To do so, it focuses specifically on a survivor testimony shared with the author at the Rwandan Nyange memorial in 2014 where the crimes in question happened, and the ICTR The Prosecutor vs Athanase Seromba trial that relates to the events at that particular site. This paper shows that the experience of trauma not only challenges the language of law but also blurs the legal narratives and functions of tribunals like the ICTR.


Introduction
This paper contrasts the recollection of a traumatic experience with the production of legal meaning. To do so, it analyses and compares a survivor testimony that was conducted at the Rwandan Nyange memorial in 2014 and the The Prosecutor vs Athanase Seromba case at the International Criminal Tribunal for Rwanda (ICTR). 1 Between 10 and 16 April 1994 around 2,500 people were murdered or bulldozed to death in the Nyange parish. Jacques is a survivor of these massacres at Nyange church; he lost his wife and his two children during the genocide against the Tutsi. 2 His recollection of this traumatic experience will be contrasted with the legal account of what happened in Nyange parish. Father Seromba was indicted at the ICTR for crimes against humanity and genocide; in 2006 the ICTR Trial Chamber III (hereafter the Chamber) found Seromba guilty of counts of genocide, conspiracy to commit genocide and crimes against humanity.
The article applies a trauma studies lens to a legal phenomenon, and uses the insights of narrative studies to illuminate legal obscurities. Applying this approach is important because what must be heard in court cannot be articulated by legal language (Felman 2002). This paper shows that the experience of trauma not only challenges the language of law but also blurs the legal narrative and functions of tribunals like the ICTR.
However, in pursuing such an inquiry, it should be acknowledged from the outset that the dialogue between the discipline of law, (poststructuralist) trauma and narrative studies is characterised by epistemological boundaries. Although a new awareness of trauma in legal practice has recently emerged, this scholarship has been limited to the lawyer-client relationship and the impact greater trauma awareness might have on a victim's mental health (see, for example, Katz and Halder 2016). What is often forgotten is that law is inherently related to an injury, which was the central insight of Felman's analysis of the OJ Simpson case. She writes 'the trial has attempted to articulate the trauma so as to control its damage', but 'the trial has become itself a vehicle of trauma, a vehicle of aggravation of traumatic consequences rather than a means of their containment and their legal resolution' (Felman 1997: 743). In a similar vein, philosopher Francois Lyotard, reflecting on the Holocaust, warns us that 'the perfect crime does not consist in killing the victim or the witnesses … but rather in obtaining the silence of the witnesses, the deafness of the judges and inconsistency of the testimony' (cited in Hirsh 2001: 536). Such insights connect to central aspects of this article: the silences and the traumatic experiences that cannot be heard on the witness stand.
The article is developed across three sections. The first briefly outlines the function of ad hoc tribunals and contrasts those with the functions of the testimonial process. In this 'testimonial process' the person who has experienced violence comes into existence as a survivor rather than just as evidence as at a tribunal. I argue it is the mode of testimony in this process that is crucial in understanding how the individual is positioned as a survivor in the testimonial process. The modes of testimony construct the survivor as a person who has experienced an uncanny event and remembers it whereas a tribunal constructs the witness as simple evidence to make the legal case. This highlights the different structural preconditions in which traumatic narratives are formulated. Therefore, the analysis in this section also engages with the critical work on the extent to which international criminal trials really can, and do, write history (see Wilson 2011: 1, also Gaynor 2012Simpson 2007). The analysis contributes to these critiques by revealing the ways in which fragments of the past are at best side-lined, and at worst silenced or eradicated, from the historical account the trial establishes.
The second section introduced narrative study and the particularities of traumatic narratives and then sets out to read trauma into and the survivor testimony against the witness testimonies heard in trial. 3 I borrow the approach of reading into and against from Shoshana Felman (1997, 2002 who applies this particular methodology to a comparative analysis of the OJ Simpson and IJCJ&SD 53 Online version via www.crimejusticejournal.com © 2017 6(1) the Eichmann trials. Felman reads trauma into the trials themselves and in the analysis thereof. In addition, she reads the OJ Simpson trial against Tolstoy's novel The Kreutzer Sonata. Here, I read the survivor testimony produced in the testimonial process against The Prosecutor vs Seromba trial (hereafter the Seromba case) and read trauma into both narrative accounts. This particular methodological approach produces new knowledge on and insights into a legal phenomenon that cannot be explained by mere legal analysis and makes visible the excluded, the marginal and the invisible. To do so, this empirical section draws on excerpts from the survivor testimony, trial witness statements, 4 cross-examination and the Seromba judgment to illustrate the argument. I show that the traumatic experience is eradicated from the witness stand and that it is instead replaced by a linear, chronological, factual and precise narrative account in order to support legal meaning-making. The third section, the conclusion, draws together these findings and relates them to broader discussions of trauma and law, of victims' positions in international criminal trials, and of how history is written by international criminal trials, as well as some ethical reflections on writing about violence and trauma.
It should be noted at the outset that the survivor testimony and the legal trial do not aim to produce the same kind of conclusion, nor do they strive towards the same kind of effect or impact. The trial wants and needs to establish the truth about what happened and its judgment suggests a (illusionary) finality (Agamben 2008: 17-19). The survivor testimony, in contrast, is a search for meaning-making, a search for expression and a symbolic understanding of an uncanny and so often unspeakable past (Felman 1997: 738;Laub 1992Laub , 1995. The analysis of the Seromba case is not meant to be a critique of the trial process because there are some things a trial cannot do by its very nature. The aim is to read narrative and trauma into a legal trial, alongside a survivor testimony, in order to propose a new methodology that allows a different and, in fact, deeper understanding of the trial process at international courts. Although the article illustrates its arguments by using and comparing one testimony and only a limited number of documents related to one single case before the ICTR, I submit that such an approach can fruitfully be applied to further cases within the ICTR, as well as other international and domestic trials, in order to expose the quite different narratives and understandings of the same horrific events. This article seeks to contribute to the project of southern criminology by applying trauma, narrative studies and testimony to an international legal phenomenon. The inquiries of mainstream criminology largely focus on the global North, often neglecting patterns of violence, and the experiences and perceptions of violence, in the global South. This trend is perhaps surprising given that most intra-state conflicts happen in the so-called 'least developed' countries. As Carrington and colleagues acknowledge, criminology is mainly a 'peace-time endeavour' with much of the research focusing on 'justice as a domestic and national project' (Carrington, Hogg and Sozzo 2016: 3). At present, ciminology remains largely blind to the cultural and transformative aspects of violence and post-violence contexts; it tends to ground its analysis in metropolitan readings of delinquency and patterns of criminal behaviour in industrialised nations (in the global North).
But collective violence not only alters social relationships and communal trust but also the ways its legacy is understood and addressed, legally. Crimes on a large scale, mainly categorised as crimes against humanity and genocide, do not only concern the individual (legal) responsibility of an 'other' but, as Edgar Faure argues, concern 'a criminal enterprise against the human condition' (cited in Hirsh 2001: 533). Similarly, Hannah Arendt has famously argued in the context of the historical Eichmann trial that 'genocide is an attack upon human diversity as such, that is, upon characteristics of the "human status" without which the words "mankind" or "humanity" would be devoid meaning' (Arendt 1994: 268-69). This is to say, writing, addressing and acting upon such crimes requires us to look beyond metropolitan readings of delinquency and criminal behaviour and to develop novel epistemological and legal tools of knowledge production and inquiry. This paper therefore wants to promote thinking about mass violence beyond disciplinary boundaries.

Functions of the ICTR and of the testimonial process
Between April and July 1994 approximately one million people, predominantly Tutsi, but also moderate Hutu, were killed in the genocide against the Tutsi. The evidentiary foundation of the Tribunal and its function to produce a historical record of the atrocities are important considerations of this article. As with regard to the latter, there is an expectation that international trials produce a historical record of what happened in the countries concerned (Gaynor 2012;Osiel 2012;Wilson 2011). Israeli Prime Minister Ben-Gurion, for example, stressed before the start of the historical Eichmann trial that 'it is necessary that our youth remember what happened to the Jewish people. We want them to know the most important facts of our history' (Arendt 1994: 10). Further still, war crime trials shape political life and collective consciousness through the 'juridical memory' they produce and reproduce (Simpson 2007: 79).
However, the desire to simultaneously establish the criminal liability of the defendant can be at odds with the task of producing a historical account of the atrocities that took place. This holds particularly true for crimes against humanity and counts of genocide, such as incitement to or conspiracy in genocide, where it is paramount for the prosecution and the Trial Chamber, respectively, to establish the broader historical context the crimes are embedded in and that made those heinous acts possible in the first place (Dembour and Haslam 2004;Osiel 2012;Wilson 2011). Simpson concludes on this point, 'there is in particular a tension between law in its history-making mode and law in is judicial mode'. This history-making function of international trials is supported precisely through human witnesses (Gaynor 2012;Simpson 2007).
Since the Nuremberg trials, the evidentiary foundation in international criminal law has fundamentally changed (May and Wierda 1998Wierda /1999. Whereas the Nuremberg trial almost exclusively relied on documentary evidence, the more recent international tribunals heavily rely on witnesses to build the legal case. The historical Eichmann trial was the first trial that introduced the human witness as part of the truth finding endeavour. It sought not only to establish facts but also to use law to transmit the uncanny history as an experience, as a tool of unimaginable facts and as a tool of communication (Felman 2002: 133).
Even though the human witness gives rise to the vulnerability of establishing a truth 'beyond reasonable doubt' (Felman 2002: 134), the recent international trials heavily rely on human witnesses (May and Wierda 1998Wierda /1999; see also Combs 2016). In fact, witnesses are the pillar of international criminal trials. This development is partly due to the complexity of the crimes they prosecute. Without human witnesses those trials would not be able to produce probative evidence of 'what happened' (Wilson 2011 Crucial for the purpose of this article is that witnesses are treated as mere evidence in the case of the ICTR. The Tribunal's Rules of Procedure and Evidence (adopted in 1995 and subsequently updated) (hereafter, the Rules) do not define the witness other than in the context of the admissibility of evidence. The same holds true for the Seromba judgment that only refers to witness statements as exculpatory or probative evidence. In other words, the witness is desubjectified. In a similar context, Jonneke Koomen (2013) found that witness statements that were collected in the field by Tribunal investigators were already then the product of encounters and hierarchies in the field. In addition, by couching witness statements, collected in the field, into a chronological timeframe in order to give them legal meaning, the human experience of the atrocity has already been eradicated in that early stage of the trial process. 8 In what I call the 'testimonial process' however, the narrativization of the traumatic experience can unfold. In and through this process, the person who has experienced violence comes into existence as a survivor rather than just as evidence. 9 I argue it is the mode of testimony in this particular process that is crucial in understanding how the individual is positioned as a survivor.
In explaining how this is so, it is helpful to draw on Holocaust studies, particularly on those studies that examine survivor testimonies. One characteristic of a traumatic experience such as the genocide in Rwanda or detention in concentration camps is that the self is annihilated and the story is neither told nor remembered. The experience becomes 'absent', as something that has yet to be experienced (Laub 2005: 257), or remains an 'unclaimed experience' (Caruth 1996). In the process of giving testimony this becomes part of the social world. In their pioneering work, Dori Laub and Shoshana Felman speak of giving testimony as a social process that requires a great deal of attention and careful listening (1992: 70). In addition this process requires a listener who recognises and reaffirms the realness of the story told (Laub and Felman 1992: 68). Testimony can therefore be understood as storytelling to an emphatic other. Moreover, Laub (2005) develops different characteristics such as the dialogical relationship between survivor and listener, the reaffirmation of the story and the secret 'password' that together form the necessary holding space to set in motion the narrativization of the traumatic experience. 10 All together, these particularities form what I define as the mode of testimony that is crucial for the testimonial process.
This article draws upon my wider research on memorialisation and transitional justice in postgenocide Rwanda. I sought to speak to people who had experienced the genocide, with the aim of collecting 'survivor testimonies' from particular memorial sites and comparing individual narratives of what happened with official narratives and 'architectonical' narratives inscribed in the materiality of the memorial sites (see, for example, Viebach 2014). I did not select interviewees in a structured way but asked 'care-takers' (Viebach 2014) if they wanted to share their memories with me. I defined 'care-takers' as survivors who work pro bono cleaning and preserving the human remains and dead bodies that are displayed in so many of the memorials. 11 I also differentiated between those 'care-takers' and official memorial staff who would usually work for the Rwandan Commission for the Fight Against Genocide (CNLG) and would conduct, at least at the national and bigger memorials, the tours for visitors. 12 Moreover, I also spoke to survivors who did not work as 'care-takers' at the memorials, though they had experienced the massacres at the particular sites. I chose national memorial sites as well as very remote memorials mainly unknown to outsiders (for example, at Kanduha, Kinazi, Cyahinda or Cyanika). In addition, I covered the different provinces so as to have a geographical balance, something which is particularly important given that the genocide played out differently across the provinces. In total, I interviewed around 40 individuals, including 'care-takers', memorial staff, staff of survivor organisations such as Ibuka and Avega, staff of the CNLG and individuals remotely involved in commemoration and the memorials. I also conducted two additional focus group discussions with female survivors. For the purpose of this article it is important to note that I only conducted five survivor testimonies that are related to massacres in Nyange, Gisenyi, Kanduha, Bisesero and Nyanza. 13 The Nyange church survivor testimony and the IJCJ&SD 56 Online version via www.crimejusticejournal.com © 2017 6(1) associated Seromba case was chosen because it is illustrative of the overarching argument of this article. 14 In summary, this section has discussed the different functions of trials and the testimonial process. This highlights the dissimilar structural preconditions in which traumatic narratives are formulated. As explored in the next section, this deeply affects the ways in which such narratives are (mis)understood and interpreted.

Narratives beyond time and space: Reading traumatic testimony into and against the witness stand
In order to highlight the ways in which the events of 10-16 April 1994 in the Nyange parish are narrated in the two settings, I first turn to general ideas on narrative and the shape of traumatic testimony.
The core of any narrative and storytelling is the plot. It forms the causal link between events in a story, functions as the structure of the story, and the means by which otherwise mere occurrences are made into moments of the unfolding of the story (Polletta et al 2011: 111). At the same time, narratives are a form of discourse and can be identified by their structural and formal features (Polletta et al 2011: 112). One of these formal features is what is termed the 'Aristotelian configuration' (Ricoeur 1984). Ricoeur argues that when stories are told there is a certain pressure to deliver them within an Aristotelian conventional narrative configuration, one in which concordance looms large, where there is a sense of the connection between events, and where the conclusion is 'congruent with the episodes brought together by the story' (Ricoeur 1984: 67). Thus, stories are narratives told according to the conventions of linearity, continuity, closure and omniscience that are often taken as a quasi-natural condition of narrative (Brockmeier 2008: 28, in Andrews 2014. There is, Ricoeur suggests, always a pressure to transform a chain of events into a meaningful whole. Moreover, only when stories are emplotted can our lives, our experiences, become meaningful. He writes, 'we tell our stories because in the last analysis human lives need and merit being narrated. This remark takes on its full force when we refer to the necessity to save the history of the defeated and the lost. The whole history of suffering cries out for vengeance and a call for narrative' (Ricoeur 1984: 75).
Part of this 'wholeness' is time: time lies at the heart of narrative. On this point, Jenny Edkins describes the linear or narrative time: … is a notion that exists because we all work, in and through our everyday practices, to bring it into being … the production and reproduction of linear time take place by people assuming that such a form of times does exist, and specifically that it exists as an empty, homogenous medium in which events take place. (Edkins 2003: xiv-xv) Further, she explains that language, and thus narrative, is inevitably linked to social structure and power: language is part of the social order, so that when this order falls apart during genocide or mass atrocity, so does language (Edkins 2003: 8; see also Scarry 1985). In that sense, the traumatic experience keeps one from forming a meaningful narrative. Narrative time and the 'wholeness' of the story delivered by its conventional configuration is precisely what makes a traumatic narrative so difficult to convey. The trauma does not fit into this conventional narrative form.
Now, juxtaposing trauma and narrative, I suggest that it is the traumatic occurrence that the narrator fails to integrate into a plot, into a conventional narrative form. Developing this further, it is helpful to return to trauma and Holocaust studies again. Cathy Caruth explains, for instance, that trauma is always a story of a wound that cries out rather than a mere pathology. It is a story IJCJ&SD 57 Online version via www.crimejusticejournal.com © 2017 6(1) 'that addresses us in the attempt to tell us of a reality or truth that is not otherwise available' (Caruth 1996: 4). She further argues that the truth in its belated character cannot be linked only to what is known, but also to what remains unknown in our very actions and our language. To compound the point further, Caruth maintains that 'the story of trauma … far from telling of an escape from reality -the escape from a death, or from its referential force -rather attests to its endless impact on a life' (Caruth 1996: 6).
The core of understanding traumatic narratives, therefore, is that neither the narrator nor the listener knows what is yet to be said or cannot be said. Despite a vast amount of documents or other evidence, the listener to a traumatic story faces a situation where he or she comes to look for something that is in fact non-existent since it is a record that has yet to be made (Laub 1992: 57).
Against this backdrop, we can now explore and read the traumatic narratives on the witness stand and in the testimonial process against each other.

Form, structure, facts and an experience
Let us start with some ideas around the form and structure of narrative accounts as they are produced in court. David Hirsch (2001: 530), drawing on the trial of Andrei Sawoniuk, 15 observes that the memoir is acted upon by the rules and norms of the legal processes, particularly by the process of cross-examination and by that of the sifting out of evidence which is deemed to be inadmissible. In the courtroom, the narrative takes a specific form: here, a 'law-type' statement invites negotiation of meaning with falsification and verification such as by asking witnesses on the witness stand questions such as 'are you sure' or 'I have heard you saying this differently before', or 'evidence suggest'. This excerpt from a cross-examination of a defence witness by the prosecutor is very illustrative of this 'law-type' form. The witness CBR has participated in the killings at Nyange church (TRA03317/1): Q. I wish to refer to the transcripts of the last session and I shall also refer to the written statements, and we are looking at page 24 and page 25 of the transcripts … Mr Moss put a question to the witness in regard to the day of Friday, the 15th of April … 'The name of the priest who was with the gendarmes, was that name mentioned by your leader?' Now Mr Moses's question concerning that episode: 'Was the name of the priest who was with the gendarmes, was that name mentioned by your leader? Answer: Yes. They mentioned the name of the priest Ndugutse. They didn't even tell us the name of the priest, whether it was a father or priest, they were saying Seromba and they said that Seromba did not allow us to get into the courtyard of the presbytery before we removed this filth'. Now, Mr. Witness, you made that statement on Friday. Do you admit to having made this statement, which was recorded by the stenographers? A.
Yes, I can confirm having made that statement concerning that Friday, as you have stated, because the person who was asking me questions wondered whether the name of the father had been mentioned, and I said that that was the case. Q.
In other words, you confirm that the words you spoke regarding your leader, the words were [pause] they said that, 'Seromba did not even allow us to enter the courtyard of the presbytery before we removed the filth'. I've read it again. A.
Yes, that was what was said. It was, in fact, the time when the priest had not allowed or prevented the people to enter the courtyard of the presbytery, they were asking for the killings to stop … Yes. Q.
Thank you. I want to refer now to page 39, and here you are being crossexamined by Defence counsel regarding the same circumstances … In relation to that incident, where the priest sent back your leaders at the gates to the presbytery, this time around you said …: 'He stopped them from doing so. He asked them what they wanted, and the authorities told him that they wanted to kill the refugees who were in that part. He told them, "Listen, look around, first of all, clear this filth". There were dead bodies and bricks.' The same circumstance, the same incident that you are referring to now, when you were being examined by the Prosecutor, it was your leaders who said that Seromba did not want to let you kill the refugees before you clear the filth. Now, the same incident, you place in the mouth of the priest the following, he said, 'Listen, look around and clear this filth'. First of all, what is the true situation? A.
Yes, but the two versions are not different. … It becomes apparent through the cross-examination of witness CBR, the prosecution attempts to transform what Hirsch (2001) labelled the 'memoir' of the witness into evidence. However, here the prosecution attempts to undermine the credibility of the witness by referring to statements made in the direct examination of CBR by the defence council (DC) the party which had called him as a witness (a stage of the process known as the examination-in-chief). The narrative of the witness is put under scrutiny and the meaning of the content is heavily debated between him and the prosecutor. In that process the 'evidence' or the account of the witness is rendered inadmissible.
In the testimonial process, however, narratives take an open form encouraged by the intimate relationship between listener and survivor. For instance, as an interviewer, I would never ask closed questions or direct the testimony in a certain direction or press it in a conventional form. Turning back to the court, Hirsch (2001) further states that, in the process of forming memoir into evidence, criminal trials give extraordinary events a routine form: they abstract, shape and civilise them. For Jacques the genocide is clearly an extraordinary event that has fundamentally changed his life and altered his being in the world. In his testimony to the author, he says: In addition, law understands facts restrictively, which means that the collection of legal evidence privileges positive or objective facts. Facts are only considered as such if they are precise, pedantic and quantifiable, and structured within a true/false dichotomy (Dembour and Haslam 2004: 163 On that day, I did see Father Seromba. Q.
Witness SE13, you told us during your testimony this morning, that the bourgmestre had requisitioned gendarmes from the préfecture to bring them to Nyange; is that correct? A.
Yes, you are right. Q.
They were supposed to provide for the safety of the refugees at the parish; is that right? A.
Yes, that was why they were at the church. Q.
Did you see them on the 16th of April while the church was being demolished? A.
They were present and there were four of them.
The defence counsel, Mr Monthé, attempts to establish positive and objective facts by frequently asking witness SE13, 'is that correct?' or 'is that right?' He thereby reduces the events to singularities of objective facts and creates a binary and very simplistic narrative that fits the conventional narrative feature. The witness testimony continues: Q. Witness SE13, can you explain to the Chamber, the circumstances and the atmosphere in which the church was demolished? You state that Nkinamubanzi was there, and that he was driving the bulldozer. What happened thereafter? Did you remain there to observe what happened? A.
Assailants had come from practically everywhere. The doors to the church were locked, and the refugees inside had barricaded the entrance of the doors with benches, such that they could not be opened. The people outside the church could not go inside to find [pause] or to get at the refugees. That is why they were throwing stones at the window panes above the doors. So, above the doors were window panes. It is true they had difficulty throwing stones. There was total chaos outside. People were trying to get into the church, and they were waiting for people to come out so that they could kill them. Q.
Do A. When insecurity threatened our commune, there were certain people who engaged in those activities. When I say that people were bloodthirsty, I am referring to those who carried out killings. Usually, there were hoodlums on the hills who heeded no advice. When they started killing people, they didn't stop; they continued. That is why I say that those hooligans had killed, those on the hills, and they knew that only the refugees in the church had not been killed. So they came to kill them. They were present and no one could stop them from executing the orders they had received.
In this extract, witness SE13 is directed to establish precise and objective facts. When he talks about people being 'bloodthirsty', the DC pushes him to define what he means by it. Questions are asked in a linear structure, where the single events and circumstances are 'emplotted' in a chain of events that suggest an allegedly wholeness of the narrative emerging on the witness stand. The DC obviously wants to establish the 'fact' that the massacre on 15 April and the bulldozing of the church the following day could not have been prevented by the defendant, priest Seromba.
In contrast, Jacques' testimony neither refers to facts nor to a precise or quantifiable true/false dichotomy. Consider, for example, this longer excerpt from this testimony that describes the early days of the 'events' at the Kivumu commune: When the plane crashed [in] this district of Nyange they started to kill in the commune Kivumu. The church was called Nyange. On 10 April they killed the achronomyst of the commune. The same day the préfet called the bourgemestres of different communes to make plans how to kill the Tutsi and to encourage them. The next day on 11 April the bourgemestre of Kivumu assembled authorities to pass instructions from the préfet. On the same day after the meeting they sent a letter to me because I was a business man at that time. The letter said that the authorities would need my car. The letter reads 'according to the security meeting you have to bring your car on 11 April today at 4.15pm for the security of the commune' [Jacques actually showed me the letter that he kept ever since he received it.] Despite of so many Hutus having cars, I was the only one who was asked. I brought the car to the Komini and they asked me to drive them to the border of the commune Kivumu/Kibirira but I asked why they made me do this. So I gave them my driver, because I didn't believe them ... I was glad I didn't go because they could have killed me. After giving his car away on the 12 April the killings officially started. They started killing in the villages, so people sought refuge in the church. They used my car to take the people from the villages to the church. They told Tutsi they were safe in the church. Even those who were hiding believed them and went to seek refuge in the church.
In this account of the events between 10 and 11 April, Jacques refers to information that he cannot possibly know or have witnessed himself ('hearsay' evidence, in legal terms). He was hiding with neighbours and only came to the Nyange church on 13 April. Reading his testimony from the standpoint of facts, precision and true/false dichotomy, he cannot know what they did with this truck because he gave it to his driver. He also was not at the meetings he describes. We also never learn how he convinced 'them' to let him go after he gave 'them' his truck or if his driver actually delivered the truck. 20 We also do not get to know who 'they' in his account are. He refers to both Hutu more generally and to authorities and 'they' who started killing Tutsi. Yet, he refers to common knowledge and broader social narratives of what happened on those days that fed into his own narrative account and how he remembers what happened.
Giving testimony is always more than simply narrate or to report a fact or an event, even to relate an experience. We can understand Jacques' testimony more generally as what Felman refers to as a 'responsibility to truth' (Felman 2014: 322): 'to speak from within the legal pledge of the IJCJ&SD 61 Online version via www.crimejusticejournal.com © 2017 6(1) witness oath, whether one is actually on the witness stand or not, whether one is in a trial or one is in the court of history'. In this sense, to testify is always metaphorically to take the witness stand, and the narrative account of the witness is engaged in both an oath and a commitment to one's own narrative that addresses the other. Jacques neither simply reports facts or events, nor is he referring to a factual truth. His testimony is a responsibility to a truth that is defined by how he remembers the events.
The courts in contrast may produce a vast number of facts, but the traumatic experience remains silent. In the Seromba case, for instance, it was difficult for the Chamber to establish the exact number of bulldozers that were used on 16 April to destroy the church and to kill Tutsi hiding in the church. On that point the judgment compounds, '13 witnesses testified to having seen a bulldozer at Nyange church, while 7 others mentioned the presence of two bulldozers'. The Chamber finds that the discrepancy between the accounts is due to the difficulty they had in 'identifying the type of vehicles present at Against this backdrop, how does Jacques describe the events between 13 and 15 April?
On 13 April I came to the church, leaving my hiding place with neighbours. They said things were getting worse. But in the church I kept myself in the corner of the entrance and my children were gathering near the altar together with other children. They didn't know I had entered the church. I kept in the corner all the time … My fellow businessmen bet on my life. If they found me they said to other perpetrators they would give them 100000 RWF. On 14 April many Tutsi were in the church and the Hutus were gathering around the church … I managed to flee the priest compound. They were killing the whole day of 15 April from 10am through the night … I left the roof at 9pm in the night but killings still went on. In the night I left I was running around the bush until I reached the river Nyaberongo; I knew how to swim but I wanted the river to kill me. I chose to be drowned instead of being killed by machetes. After what I saw here, dead bodies everywhere, people hacked into pieces, I really wanted to die. But when I went in, I managed to reach the other side. On the other side was Gitarama; I was wounded all over, especially on my legs because of glass, but I could walk slowly until I reached my younger brother, who was living in Gitarama. There, it was somehow peaceful because they hadn't started to kill there. But the next day they started killing, so I and my family went to seek refuge in the marchlands (Kabgayi) where so many Tutsi were.' In his account of the events between 13 and 15 April at Nyange church, Jacques does not testify according to a descriptive account of what he observed, rather he offers the very personal and private effect of survival and of resistance to extermination, the very crime priest Seromba was found guilty of committing. His narrative account shows that the act of testifying, in and of itself, is vital not so much because of the historical or legal information that can be extracted from it but because of the depth of darkness that it begins to make visible to those who were not there.
IJCJ&SD 62 Online version via www.crimejusticejournal.com © 2017 6(1) Such stories testify to a struggle of survival then and now (Hartman 1996: 142). Jacques wanted to die, being drowned in the river after seeing how people were hacked into pieces by machetes. He testifies to the struggle of survival and in that moment also to the cruelty of survival. His account makes visible the despair and betrayal in saying how Tutsi believed the authorities that they were safe in the church. Further, he recalled how his fellow businessmen who knew Jacques for years bet on his life as if he wasn't a human being, but an object. His ordeal doesn't end at Nyange church or in the river. He has hope when he arrives at his brother's home, a treachery 'peace' as he describes it. But the next day he must learn that his struggle for survival continues. His testimony breaks with the marchlands. At this point during the testimony, he becomes very upset and we need to take a break. Tutsi were hunted in the marchlands. Accounts of these survivals are indeed 'unspeakably' horrific (see also Hatzfeld 2007).
Later in the testimony he explained to me that his father and brother are not buried at the Nyange memorial. Indirectly, he here suggests they both died in the marchlands. The darkness of his testimony also becomes visible by the way he talks about his children. His children were in the church, but he stayed in the corner of the entrance, where they could not see him. And, here, we must ask what does it mean to know that we cannot save our own children? It is precisely this silence that cuts through his testimony, the recollection of losing and witnessing the killing of his children, wife, brother and father ('I remember how they were killed'). This mode of testimony breaks the frames of law and testifies, on the one hand, to its very rupture in the wake of 'crimes against humanity' and, on the other, to the very human nature of survival. The testimonial process, in a relationship between listener, the 'other' and narrator, is the place and the process in which the traumatic experience comes into existence and is therefore knowable. We become more familiar with Jacques's trauma of survival, although we will never truly understand the deeper meaning of the silences that cut through his testimony.
In contrast, legal knowledge may produce vast numbers of 'facts', but the traumatic experience is silenced because the trauma does not speak from the language of facts. Consider this short excerpt from the examination-in-chief of witness CBN by the prosecutor to further illuminate this point (TRA003309/1): The prosecutor does not follow up on the death of Miriam, the fact that she was humiliated or her legs spread apart. Indeed, the prosecutor continues the line of questioning, by simply asking 'You said that Seromba sent some people out. Can you estimate the number of people sent?' Again, the trial is only interested in establishing meticulous and detailed facts. Only minutes later CBJ must take a break. He says, 'I have a problem. Before I came into the courtroom, I talked to the prosecutor regarding the wound I had which I sustained as a result of the Interahamwe. My head is heated up, and I'm feeling confused. This is the situation I am facing now. I don't think I can understand the questions that are put to me now.' The traumatic experience of witnessing how Miriam died and his physical injury, is not allowed to come through in the trial. It also seems arbitrary to establish the fact of one particular person, given that around 2,500 people died in the parish between 11 and 16 April 1994. At the end of the trial, the judgment establishes the death of Miriam 'beyond reasonable doubt'.

In-between time and timelessness: Reading trauma time against chronological time
Let us revisit the chronological aspects of narratives in order to read the testimonial accounts on the witness stand. Generally, we need narratives or stories to make sense which they obtain by giving them a chronological sequencing and instilling in them a wholeness they might not have.
IJCJ&SD 64 Online version via www.crimejusticejournal.com © 2017 6(1) White has argued that events are expected 'to display the coherence, integrity, fullness, and closure of an image of life that is and can only be imaginary' (White 1987: 24).
I would argue that this is even more the case in the context of a trial. Events must be told and configured in a way that can be given legal meaning. In order to produce a narrative that meets the requirements of 'beyond reasonable doubt', the testimony in court must follow the traditional Aristotelian narrative structure. For example, the examination-in-chief of witness CBN is pressed into this conventional Aristotelian structure by the prosecutor's way of questioning (TRA003309/1): Q. Witness CBN, there is a photograph being displayed; Prosecutor's Exhibit P. 3-21. It's on the screen before you. What is in that photograph? A.
It is a photograph of the church. Q.
Which church? A.
It is the Nyange church. Q.
In April 1994 did you go to Nyange church? A. Yes.

Q.
On what day did you arrive at Nyange church? A.
It was on Tuesday, the 12th of April. Q.
Can you give us an approximate time of day: morning, afternoon, night? A.
I got there at about 1 a.m., if I reckon it … Q.
On what day did you depart or leave Nyange church? A.
I left the parish on Friday, the 15th April. Q.
In April 1994 who was the parish priest at Nyange? A.
It was Seromba. Q.
What is that person's other name? A.
On the 13th of April, did you see Seromba; yes or no? A.
Did you hear Seromba say anything? A.
Yes, I heard him say something … A.
In the morning we asked him to say mass and he told us that he couldn't waste his time saying mass for Tutsi, that our brothers, the Inyenzi, 21 the Tutsi Inyenzi, had attacked the country and they had even killed President Habyarimana … Q.
At which location was Seromba when he said those words? A.
He was standing in front of the church. Q.
Did you see any weapons inside the church? A.
Yes, there were; there were traditional weapons.
This examination-in-chief illustrates how the prosecutor attempts to establish a coherent sequence of the events that happened. In the line of questioning the prosecutor continues to lead the witness through each day, until 15 April when the witness fled the parish thereby focusing on little details of the events so as to make sense of what happened. The examination-in-chief continues, now establishing what happened on 15 April as follows: Q. I'd like to turn your attention to the 15th April 1994, and if you could tell us, step by step, what happened that day. A.
Yes, I can tell you. In the morning we faced the attacks that were launched against us until we were beaten following the grenades that were thrown by the assailants at us, up until the time when I, myself, fled; that was in the afternoon. In the meantime this is what I saw: on the upper floor of the building, Father Seromba sought refuge there, but he did not come down.

Q.
Before the grenades were used, what weapons or items were used in this attack? A.
The assailants were throwing stones at us and we did the same, and at one time they threw grenades at us. The assailants had traditional weapons, spears and others. Q.
Was anyone affected by the grenades? A.
Some people were affected by the grenades. I was standing near a tree in the rear courtyard of the presbytery. Girls were wounded. I was throwing stones while taking cover behind a eucalyptus tree. Q.
Was anyone killed? A.
Those people moved towards the church. I don't know whether they died, but they seemed to have died. Q.
Who was in the church during this attack? A.
I haven't quite understood your question, Prosecutor. Q.
Was anyone inside the church at the time of the attack? A.
There were people inside the church. Some people were inside the church while others were defending themselves outside. Q.
Did the attackers ever come inside the church? A.
The assailants were around the church. They came into the courtyard when we were repelled. They came in a vehicle and they would hit us with sharp objects. That was when I fled … Q.
Can you estimate the time of day you ran away from Nyange church on 15 April? A.
I fled in the afternoon at about 1 p.m.
By the way of questioning, the witness testimony is given a wholeness and a legal meaning that can be used by the Chamber to make sense of what happened and to establish whether Father Seromba is guilty or not. The witness testimony has a clear ending, which is at 1:00pm on 15 April, when witness CBN fled the parish. We do not get to know where his journey of survival led him to and how it ended. The legal wholeness of the testimony is established with the witness leaving the 'crime scene'.
Moreover, the trial is concerned with a chronological sequence of events and with the facts that establish the event as truthful. For instance, the Seromba judgment is structured around a chronology of events that was introduced by the indictment and maintained in the judgment. The Chamber begins the 'fact' presentation with ruling that it has been established during the trial that it is beyond reasonable doubt that father Seromba 'in light of those testimonies … acted in a number of ways which show that he was responsible for the daily management of Nyange parish during the Arpil 1994 events' (Seromba Judgment, 01-66-0276/2: para 15, p. 38). The judgment continues detailing the events in chronological order from 6 April to 16 April. This shows that, in the language of law, there must be an exact chronicle ordering of temporality. However, such precise temporal order can pose difficulties for traumatic testimony.
When looking at narratives on the witness stand, it is important to consider how time is experienced by a survivor of atrocity as opposed to how legal procedures integrate time aspects in constructing legal meaning of an event. When listening to a testimony the listener usually ascribes meaning to what has been said using a chronological timeframe (beginning, middle and end) (Edkins 2003: xiv-xv). As aforementioned, the emplotment brings single occurrences in a chronological order. Trauma, however, changes the conceptions of time in the body of the survivor (LaCapra 2001: 90). Lawrence Langer compounds further on that point that, in the body of a survivor, there exist two times (Langer 1997: 57-58): a chronological and a durational time.
He explains further that the traumatic experience 'is not part of ... historical past, but of [the] durational present' (Langer 1997: 59). Jacques for instance moves between time and events in his testimony: IJCJ&SD 66 Online version via www.crimejusticejournal.com © 2017 6(1) On 14 during the day the Tutsi won the battle; the Hutu did not manage to come in. In the evening the bourgemestre and other leaders and these two businessmen had a meeting at the priest compound. When Tutsi saw the authorities meeting there they thought they might discuss ways to save them. But this was not true, because next day, 15 April, many cars, trucks and buses came full of Hutus with the purpose to kill everybody in the church. But after the meeting (30min) two of the guys, a judge and a policeman were also in the meeting. They came to the church to collect Hutu women who were married to Tutsi men. They had a list of those women. They were calling them to come out.
In this part of the testimony, chronological temporality and coherence are lacking. Jacques fails to integrate the events in the wider context he refers to and fails to establish an emplotment, which would bring the events (meetings, arrival of trucks, the collection of Hutu women) into a coherent order. The meeting, he talks about, in his account happens on 14 April and then again on 15 April at the same time when the buses, carrying Interahamwe and probably villagers, arrive. Clearly his point of reference is not the chronology or the facts of how the events unfolded during these days, but the knowledge that, with the arrival of new forces, the Tutsi in the church and the priest compound would not be able to defend themselves. In other words, they knew they would die. This account also talks of despair and hope that is betrayed when he recalls the meeting saying that they, the Tutsi in the church, thought the leaders would discuss ways to save them.
What comes alongside the chronological timeframe of narratives is a suggested closure. As Andrews (2014) explains, the narrative configuration of linear time seemingly gives the trauma narrative what it does not have: an experience which is contained in time; an experience that happened in the past and indeed has ended, is finished now. In the examples above the witnesses must be able to narrate a story in the past through the past tense, which suggest an ending. Furthermore, when considering the trial judgment, it becomes clear that it forecloses the future since the judgment is 'final'; what happened is past and this is where the legal story ends.

Conclusion: Law's past and trauma's present
Reading the survivor testimony against the witness testimony on the witness stand has revealed the difficulties of the trial to hear and address the trauma of genocide and crimes against humanity. Felman reminds us that there are limitations in the possibility of seeing and hearing trauma. She argues that there exists a structural exclusion from our factual frame of reference that is determined by a built-in cultural failure to see, and I would add, hear, trauma (Felman 1997). Law is a language of abbreviations, of limitation and totalisation that rules out what cannot be disclosed in language and words. And this is exactly what Lyotard (cited in Hirsh 2001: 536) prompts us to understand, when he asserts that the perfect crime is not the death of the witnesses, but obtaining the silence of the witnesses and the deafness of the judges. To carry this thought further, law's story focuses on ascertaining the totality of facts. The facts of when, where, what and how many, is what the trial is concerned with. This is the evidentiary foundation that the Chamber uses to rationalise its judgment and its legal quest for justice in the Seromba case. The court needs to establish the facts in its totality as 'beyond reasonable doubt' in its judgment. But as Arendt has argued in respect of the Eichmann trial, 'reality is different from, and more than, the totality of facts and events which anyhow is unascertainable' (Arendt 1993: 261).
Law is, by definition, a 'discipline of limits' (Felman 1997) that also applies to the way time is understood and interpreted in the trial process. The analysis has shown how the crossexamination, the law-type questioning, produce legal narratives that not only distance and totalise the singularities of 'events' but also form a legal temporality that fails to include, admit or acknowledge trauma time, except as a rupture of the legal frame. Temporality is indeed problematic with regard to trauma. The judgment makes past out of the massacres that happened at Nyange church and brought a legal closure whereas the traumatic experience for Jacques lives IJCJ&SD 67 Online version via www.crimejusticejournal.com © 2017 6(1) in the present. Claude Lanzmann, with regard to the Holocaust, even emphasised that the Holocaust is not a memory but reveals itself only in a hallucinated timelessness (Lanzmann cited in Felman 2002: 153), something the trial can neither think of, nor include.
Moreover, reading the survivor testimony against the witness testimony has highlighted how the witness evidence given at trial reveals the crime, while the survivor testimony speaks to a traumatic experience that cannot be translated into legal idiom. Alongside the crime comes the position of the 'victim' as a witness who in the court procedure automatically falls under law's scrutiny. That is to say, the witness is judged by the credibility in the narrative account her or she testifies to in court. But as the analysis has revealed, the experience produces traumatic narratives that lack coherency, closure and consistency.
Reading trauma into the witness stand might illuminate why the Chamber found so many inconsistencies in witness testimonies in the Seromba case. Indeed, as Nancy Combs has observed, inconsistency in witness testimonies in international criminal tribunals is not only a problem before the ICTR, but a 'serious problem in international criminal law' (Combs 2016: 5).
Combs further argues that inconsistencies arise when witness testimony diverges from pre-trial witness statements and that those are often related to details of dates, distances, duration and numbers (Combs 2016: 6). Combs argues that around 50 per cent of inconsistent testimonies are given at the ICTR. She admits, though, that such can occur due to false testimony, interpretation problems and cultural misunderstandings, and claims that these cases of inconsistencies relate to facts that 'can't be forgotten or confused'.
The first section of this paper showed that, in the early stage of the trial process, the witness statement is already a text that is shaped by hierarchical encounters in the field and that the traumatic experience is cleared out of the 'legal text'. This obviously poses a difficulty when, years later, the accuracy of the witness testimony is assessed against this first statement that is so often used in cross-examination to undermine the witness's credibility. The analysis has revealed, in contrast to Combs findings, that a more sympathetic understanding of traumatic testimony appreciates that, while there is always an oath to tell the truth, everything in this truth can be forgotten and confused. A pure legal perspective neglects these fragments of both traumatic memory and experience. I therefore suggest there is a need to consider reading trauma into inconsistencies in witness testimonies before international criminal trials in order to illuminate different perspectives as to why those 'inconsistent' testimonies occur.
Reading trauma into and testimony against the witness stand importantly speaks to humanistic claims that storytelling by victims in juridical proceedings is a desirable and necessary act to build collective peace. Often there have been claims that victims are healed by giving testimonies in court or before Truth and Reconciliation Commissions. Osiel argues, for instance, that the 'victimwitness' appeals not only to the judges but implicitly to the community at large and that, therefore, the legal arena provides the victim with a superior platform to articulate their stories as influencing the collective memory of the events its judges (Osiel 2012: 2, 30).
It seems, though, that historically the Eichman trial has been rather an exception in the ways 'victim-witnesses' were allowed to recall their stories in court. The foregoing analysis of how the witness is de-subjectified and how the traumatic experience is not heard because it is silenced in the legal structure of the trial procedure has clearly shown that the 'real' story remains mute and the legal audience deaf to it. This dominant view that testifying in court leads to a catharsis for 'victims' and an acknowledgment of their suffering seems much more debatable in trials such as the ICTR. Rather, the findings of this paper seem to complement the work of those who argue that today's international trials leave little room for witnesses to tell their stories (Dembour and Haslam 2004: 153). In the case of Rwanda, local victim organisations such as Avega (the organisation of genocide widows) or Ibuka (the umbrella survivor organisation) even refused to IJCJ&SD 68 Online version via www.crimejusticejournal.com © 2017 6(1) work with the Tribunal because of repeated mistreatment and the lack of protection of witnesses in the court room. 22 As the analysis of the court transcripts revealed, the only way to tell the story is in the form of giving legal evidence that at the same time clears the human experience out of the legal forum in order to establish a legally authoritative account of the crime and of 'what happened'. There is an inherent tension between the survivor who speaks on the witness stand and whose story cannot be heard. The survivor is only positioned as evidence with probative value that fundamentally neglects the fact that he or she has experienced inhumane suffering. The court, therefore, cannot provide a testimonial holding space that would be needed in order to form and convey a traumatic narrative. The assumption of so many scholars that victims' pain and suffering is acknowledged in legal proceedings should be reconsidered and challenged by more empirical research. The results presented here however, build a fruitful starting point to initiate a more nuanced debate about the position of victims in juridical proceedings, nationally and internationally.
The analysis in this articles also engages with the critical work on the extent to which international criminal trials really can, and do, write history (see Wilson 2011: 1;also Gaynor 2012;Simpson 2007). It was shown how the trial marks what history remembers and what history forgets, or in the words of Felman, 'what is pragmatically included in and what is programmatically excluded from collective memory' (Felman 1997: 766). The collective memory created in court programmatically excludes the traumatic testimony, but pragmatically includes the evidentiary linear narrative that supports legal meaning-making.
This leads to the human experience of the mass crime being cleared from the legal records. Therefore, we need further outlets for a traumatic past to emerge, to be heard and understood if we are to generate a fuller and more sensitive picture of mass atrocities and genocide. Importantly, though, we must be aware that law relates to history through trauma, through the social function of the trial as a structural, procedural and institutional remedy to trauma (Felman 1997: 766). Yet, as trauma studies teach us, a trauma cannot simply be remembered when, in the first place, it cannot be grasped, transgressed into language (and legal idiom) or heard by its (legal) audience.
In addition, trauma is constantly re-enacted in the present because it cannot be transgressed. But law closes the possibility of reinterpretation and negotiation of the traumatic past, and the judgment closes all meanings by presenting a totality of facts. However, the archive of the tribunal(s) preserves the trauma, its incoherence and inconsistency; the challenge here is to see, understand and to read trauma into the archival records. The methodology presented here enables us to assess this invisible traumatic material in a way that lets us hear the silences and see the gaps that cut through the testimonies on the witness stand.
In conclusion, it is important to note that trauma is unique, both individually and historically, although some aspects of it share some characteristics (Andrews 2014: 148). That also means that the representation of trauma faces epistemological and ethical challenges (Feldman 1991; Robben and Nordstrom 1995;Young 1988). As researchers we should be cautious when arguing to 'speak for the subaltern' or 'give voice' to victims, particularly when the research is characterised by hierarchical encounters between global North and global South. Robben and Nordstrom importantly note that 'one can count the dead and measure the destruction of property, but victims can never convey their pain and suffering to us, other than through the distortion of word, image and sound. And rendition of the contradictory realities of violence imposes order and reason on what has been experienced as chaotic' (Robben and Nordstrom 1995: 12). We have come across this in the witness testimonies and in Jacques' story. Yet this paper has rendered this contradictory reality meaningful by providing it the linear (beginning, middle, end) narrative structure that is needed to understand and contextualise what is said, written and heard, and what is ultimately needed to write an academic piece of work. When IJCJ&SD 69 Online version via www.crimejusticejournal.com © 2017 6(1) writing about narratives of violence, we must, therefore, reflect on our own subject position in the broader epistemological community and how this impacts on the way we present and make use of traumatic experiences in our work. Nevertheless, this paper has hopefully contributed to the project of southern criminology and its democratisation of knowledge production by making visible the invisible, the marginal and the excluded.  17 For a recollection of the killings from a survivor perspective, see also Jean Hatzfeld (2007Hatzfeld ( , 2010. 18 Witness BZ1 states in his witness testimony that he witnessed how Anicet Gatare asked a gendarme, in exchange of some money, to be killed in order avoid an atrocious death [by a machete or club]. Gatare was killed as a matter of fact. However, the Chamber could not establish whether this happened through a machete or through shooting (Seromba Judgment, 01-66-0276/2: para 200, p. 56). 19 Anasthase Nkinamubanzi appeared before court as well and was a defence witness. On the stand he claimed that he was coerced to drive the bulldozer. Witness FE31 testified that another driver who was asked to bulldoze down Nyange church, refused to follow the order and was killed. The Chamber regarded this statement as not credible. 20 Indeed, in the examination-in-chief of prosecution witness CDL, he refers to Jacques' vehicle and states, 'he was a Tutsi and his vehicle was taken away from him. He had gone into hiding. During that period [name redacted] was the driver of that car' (TRA003315/2). 21 Inyenzi is the Kynarwanda term for cockroach and was used to refer to Tutsi during the genocide. 22 A report by the International Federation for Human Rights into the treatment of witnesses before the Tribunal concluded that the complaints raised by those survivor organisations were justified (October 2002, No 329/2).