Multiple representations of a human rights violation : Competing discourses in TRC narratives and related media texts *

The Truth and Reconciliation Commission of South Af rica (TRC) was established on the principle that uncovering the truth of a divided an d unjust past was a prerequisite for reconciliation and the building of a united nation for the future. Both "truth" and "reconciliation" are contested notions. The recogni tio by the TRC Commissioners of the difficulty of establishing a single uncontested ver sion of past events is demonstrated in their acknowledgement that they worked with different kin ds of "truth", most notably "factual/forensic" and "personal or narrative" trut h (TRC Report , I, 110-114). "Reconciliation" is also a polysemous and somewhat ambiguous word, w hose meaning for different people at different times is shaped by the context in which i t is used. In the context of the work of the TRC, reconciliation, as the postscript to the Inter im Constitution quoted above suggests, was


Introduction
The pursuit of national unity, the well-being of all South African citizens and peace require reconciliation between the people of South Africa and the reconstruction of society.
… there is a need for understanding but not revenge, for reparation but not for retaliation, a need for ubuntu but not for victimization. 1   The Truth and Reconciliation Commission of South Africa (TRC) was established on the principle that uncovering the truth of a divided and unjust past was a prerequisite for reconciliation and the building of a united nation for the future.Both "truth" and "reconciliation" are contested notions.The recognition by the TRC Commissioners of the difficulty of establishing a single uncontested version of past events is demonstrated in their acknowledgement that they worked with different kinds of "truth", most notably "factual/forensic" and "personal or narrative" truth (TRC Report,I,(110)(111)(112)(113)(114)."Reconciliation" is also a polysemous and somewhat ambiguous word, whose meaning for different people at different times is shaped by the context in which it is used.In the context of the work of the TRC, reconciliation, as the postscript to the Interim Constitution quoted above suggests, was doi: 10.5842/36-0-38 broadly understood as being about restoring relations between a divided people in order to create a healthy and united society.Seen as a pre-requisite for the achievement of this aim was the need for "understanding", "reparation" and "ubuntu".The concept of 'ubuntu' emphasises the interdependence and obligations of mutual respect, value and acceptance of members of a community. 2  The TRC, Posel (2006: 89) reminds us, played a critical role in suggesting that one element of ubuntu ("one form of our human mutuality") is "a mutuality of speech"."Speaking out", before members of the TRC either in camera or at the public Human Rights Violations (HRV) and Amnesty hearings, about wrongs and abuses suffered or perpetrated was seen as essential to the project of recognizing the common humanity of all citizens and of creating the ground for reconciliation.The degree of its success in fulfilling this reconciliatory role has been widely debated. 3Concerning truth, although there is general agreement that the TRC broke a long silence, revealing truths about South Africa's violent past that could never again be denied, many testifiers were left with unanswered questions (see Villa-Vicencio and du Toit 2006).If reconciliation is viewed as "the formal attainment of the political and constitutional unity of the country", then the TRC is seen as "an important moment" in the transitional process (Gerwel 2000: 280).But reconciliation on the national level cannot be divorced from reconciliation on other levels, community and personal, since all are interdependent and all were affected by the apartheid structures. 4On the latter two levels, while the TRC marked up some notable successes, 5 there is also evidence that for many testifying before the TRC failed to bring about the better life that they had hoped for.
The focus of this paper is a set of narratives offering different perspectives on a single event, the death of Ashley Kriel, a young Western Cape activist, in July 1987, which illustrates Details of it were rehearsed in the TRC Report on the findings of the Amnesty Committee, first made public in 1999 and published in full on the TRC web-site.Each telling of the story is a representation, offering a different angle on the event and the people involved, and as such each is, in its own way, relative, partial, and, at times, in conflict with other accounts.
The failure to find a version that could bring closure on the event for all the participants is evident in comments made by Kriel's sister, Michelle Assure, in an interview published in the Cape Times (18 February 1999) after the announcement of Benzien's amnesty.
This paper offers a close linguistic analysis of these narratives.The study is undertaken in the belief that Discourse Analysis (DA) can complement work done in other fields, for example, politics, sociology or history, which have examined material primarily for content, for what has been said rather than how it has been said and written. 6People's ways of speaking, their linguistic and structural choices, are frequently indicative of their attitudes, worldviews and social relationships.DA offers a methodology that allows a close look at linguistic choices of speakers and the ways in which these choices encode possibly opposing views on events and participants.
DA also enables at least a partial rediscovery of meanings that have been overlaid by other meanings in frequent retellings.In the words of Bakhtin (1986: 94) every utterance is a "link in the chain of speech communication": texts feed off and into other texts, and in the representation of or response to an earlier text, meanings are often highlighted in new ways.In addition, when a narrative is retold over a period of time, changes in the social climate will be reflected in both its telling and its reception.Close examination of the texts included in this paper shows what shifts and re-accentuations may be wrought in the telling of an event when different perspectives come into direct contact in a textual chain or network which extends across a period of radical social and political change.I shall argue that close scrutiny of the linguistic choices of the different narrators of the Ashley Kriel story reveals traces both of the political and historical context of the TRC and of deeply entrenched attitudes and worldviews of the narrators, and may thus indicate how these factors have operated in the complex matter of reconciliation that was the TRC's mandate.
Working with the Bakhtinian notion that utterances (texts) are ideological and interactive both inter-and intra-textually, I use the analytical tools and concepts of Critical Discourse Analysis doi: 10.5842/36-0-38 (CDA), as developed by Fairclough (1992Fairclough ( , 2001)), and of stylistics, as explored in the work of Fowler (1986), Toolan (2001) and Simpson (1993Simpson ( , 2004).In the following section of the paper I comment on the distinctive character of TRC discourse, and briefly outline the theoretical concepts and approaches that have guided my analysis.In section 3, through a detailed analysis of the texts, I seek both to show how opposing attitudes and perspectives are encoded in the language of the texts, and to examine how these narratives resonate with each other and against the changing historical context.In my conclusion I comment on the nature of the attention given to the various perspectives and what this implies about the possibilities for reconciliation.

2.
The theoretical approach

The discourse situation
The discourse situation of the TRC was unusual in a number of ways.Between the time that the events described in the narratives examined in this paper took place and their retelling at the TRC hearings, South African society had undergone radical changes.With the ending of the Afrikaner nationalist hegemony came a shift in the ownership of the "discourses" of power.By 1996, when the first TRC hearings took place, voices which had at one time been dominant had been subdued.Conditions on "sayability" -who could speak, what could be said in public, how it would be evaluated and by whom -were changing.At the HRV hearings of the TRC, formerly voiceless people were given a public platform and a sympathetic audience as they told stories of past wrongs, some of which had never been spoken of even in private.At the Amnesty hearings, testimonies of former apartheid functionaries resonated against a new political context.The testimonies of most public hearings were relayed to a wider audience through media channels freed from the blanket of restrictions under which they had been placed in the 1980s by a government wishing to control any perceived threat to its power.
Other unusual features are apparent within the immediate context of the Commission's hearings.Firstly, there was no ready-made format or register for the Commission to draw on, and traces of multiple discourses -for example, legal, counselling, theological -appear in the language of both commissioners and testifiers.In the recorded testimonies participants can be doi: 10.5842/36-0-38 seen finding their way into appropriate modes of speaking and interacting with each other.In addition, testifiers varied in their success in adapting to the formal context in which they spoke, in the narrative skills that they brought to the hearings, and in their understanding of what was required of them.
Secondly, the purposes of the HRV and Amnesty hearings were different, as were the kinds of interventions of the commissioners, whose attitudes and affect influenced the ways in which stories were told.It has been argued that the HRV narratives were substantially co-authored by the panels of commissioners and facilitators hearing them (Blommaert et al. 2001;Bock et al. 2000;McCormick et al. 2006;Verdoolaege 2006).It is also clear from some hearings that tellers had their own understanding of what was being sought or of what they wanted to tell, and that these agendas could be missed or misunderstood even by sympathetic commissioners.The mode of conducting the Amnesty hearings was more adversarial.Since applicants for amnesty were required to make full disclosure of abuses committed by them, the panel of commissioners was naturally concerned to probe the accuracy of what they were saying.
Finally, 'truth', as I have suggested, is a complex notion, opening up the question "Whose truth?"Recognising this complexity, the Commission adopted four "notions of truth": (i) factual and forensic, (ii) personal and narrative, (iii) social, and (iv) healing and restorative truths (TRC Report,I,.Factual and forensic truth, that is verifiable details about events and trends, was a primary object of the Amnesty hearings and also of the HRV hearings.At the HRV hearings, however, testifiers were invited to tell their own stories, "his or her truth as he or she sees it", a personal, experiential truth.Social truth was seen by the TRC as emanating out of a dialogue between different viewpoints, and restorative truth as being part of the healing process.Admitting different kinds of truth made it inevitable that the Commissioners would have to weigh up accounts given by former opponents which differed in both emphasis and selection of detail.

Critical Discourse Analysis
For the analysis of texts grounded in a painful recent history of the society and country where they were produced, and where the past continues to affect the lives of many people, CDA offers a productive analytical approach and procedure.It emphasises the links between doi: 10.5842/36-0-38 discourse and the social practice within which the discourse is produced, and views discourse analysis as a means both of revealing these sometimes hidden links and of contributing to or effecting social change (see Wodak and Meyer 2001;Fairclough 1992Fairclough , 2001)).Because Fairclough's approach, which involves the close linguistic analysis of texts, suits my own aims, I have drawn on the analytical terms and categories he uses.The texts selected for analysis represent two "orders of discourse", that is, two sets of discursive practices associated with a particular social institution and comprising a network of associated genres, discourses and styles (Fairclough 1992).The two orders under examination here are the emergent order of discourse of the TRC, with its genres of various kinds of public hearings, report writing and so on, and that of the South African press.Central to Fairclough's approach is the recognition that the boundaries between the different elements, styles, genres and discourses within any order of discourse are fluid, and that texts often display a shifting across boundaries which indicates changing patterns of ascendancy in the social structures from which they derive.The crossing of boundaries also takes place within genres or discourses, as prior texts and their ideological overtones are either absorbed, ironically highlighted or contradicted within new textual contexts.Through intertextual inter-action old texts are invested with new resonances, and these movements see "texts historically as transforming the past … into the present" (Fairclough 1992: 85).Such shifting both across and within textual boundaries is well illustrated in the trajectory of texts analysed in this paper, which are framed in contexts representing both the old and the new political order.
For the analysis of other linguistic aspects of the texts, I have followed Fairclough, Fowler, Toolan and Simpson, all of whom draw on Halliday's (1995) Systemic Functional Linguistics (SFL).SFL is a semantic theory, which argues that linguistic choices made in specific contexts reflect the purpose(s) which language is intended to serve in those contexts.Thus speakers' and writers' lexical, syntactic and textual choices are indicators of their interpersonal relationships and how they perceive their own and others' social identities, and of their ideational meanings, that is, how they experience the world around them.Clues to both interpersonal and ideational meanings are found in the speakers' choices of expressions of modality, 7 of transitivity patterns (process-participant roles), 8 in the selection of whose voices to quote, in their lexical choices, and in rhetorical features such as the structures of opposition created within and between testimonies.These choices will be examined in the textual analyses that follow.

Textual analysis
The following texts constitute a story about the telling and reception of narratives relating to a deeply traumatic event.The set of narratives resembles a network rather than a chain, since the intertextual links reach both synchronically and diachronically (horizontally and vertically): the stories resonate against each other both within the contexts of their original telling and hearing, and across the changed social and political environments of the year in which Ashley Kriel died and the time of the TRC hearings.
The extracts from the newspaper reports from 1987 and 1996-7 introduce the story, and their respective reporting styles provide indicators of the discursive context, that is, of some of the dominant discourses and of what was "sayable" in each period (see Blommaert 2005).My analysis focuses principally on extracts from the testimonies of Ashley Kriel's sister, Melanie Adams, and the former member of the Security Branch, Jeffrey Benzien.The TRC was required to find answers to questions which are paraphrasable as "What happened, to whom and by whose agency?" and "How were people's lives affected by these events?"The perspectives on the event given by each of these testifiers highlight one of the problems arising from the TRC's acceptance of the notion of multiple truths, in this case, its inability to establish definitively what were the circumstances of Ashley Kriel's death and thus to open the way to a better understanding between these protagonists.Later texts illustrate the reception of Adams' and Benzien's testimonies, and suggest further impediments to the achievement of reconciliation in this case.

3.1
The discursive context: media texts "Speech genres are the drive belts from the history of society to the history of language."Bakhtin's observation (1986: 65) is substantiated in examples from the South African press reporting on political events in 1986-7 and in 1996-7, respectively. Bennett and Verbist-Serekonyane (2000: 260) note that whereas in the 1980s the predominant vocabulary in media reporting of protests and unrest was that of "conflict and confrontation", by 1996 vocabulary choices tended to reflect the media's "aim to highlight the desire for reconciliation", thus helping to prepare the ground for the TRC.The shift is illustrated in the following short extracts from articles from two Cape Town newspapers -The Argus, an afternoon paper, and doi: 10.5842/36-0-38 the Cape Times, the morning daily.In addition, the extracts interestingly reflect whose voices and whose discourses could be heard.
Extracts (i) to (iv) come from an article, headed "Man killed as police raid house in Athlone", which appeared in The Argus, 10 July 1987.(Italics highlight phrases commented on in the analysis.)(i) A MAN was shot dead and four people, including a woman, were detained for questioning during a police swoop on a house in Athlone.
(ii) Details of the incident have not been disclosed as investigations have not yet been completed.
(iii) However it is understood that a large number of policemen were involved in the raid, which led to the recovery of a "heavy calibre firearm" and a hand-grenade of Russian origin.

(iv)
Police said that during the swoop, there was a "skirmish" and a 22-year-old man was shot dead.
In December 1986, the apartheid government had added to the already draconian restrictions on press reporting, 9 rules making it illegal for reporters to participate in political protests or report on most forms of unrest or about the arrest or detention of government opponents.As a result the primary official sources of information for journalists were the daily police reports, as indicated in (iv).In these extracts in the almost complete absence of any other voice or opposing account, and of mediating or distancing phrases (note the bland inquit verb "said"), there is little to differentiate the reporter's perspective from that of the police.
Given the source, the lexical choices are unsurprising."Swoop" (in extract (i)) and "skirmish" (iv) (the quotation marks perhaps representing the reporter's only moment of self-distancing) are euphemistic.The end-focus in (iii) falls on "heavy calibre firearm" and "hand-grenade of Russian origin", phrases which portray the police action as a pre-emptive strike against a perceived communist threat, a strategy which the government of the day frequently used to justify its security measures.
doi: 10.5842/36-0-38 Passive forms -"was shot dead", "were detained" (i) -allow deletion of the agent and foreground the recipients of the action, the "man" and "four people, including a woman".The "swoop" has been attributed to the police, but other nominalisations, such as "questioning" (i), "investigations" (ii) and "recovery" (iii), like the passives, function to obscure agency, to abstract the particulars of the actions, and to dodge the question of responsibility.
This was the only report carried by The Argus on this event.The Cape Times, rather less submissive, covered Ashley Kriel's death and funeral extensively.The following extracts from a report in the Cape Times of 15 July 1987, headed "Kriel 'product of '85 upheaval'", show other perspectives and emphases, as the reporter records the views of Ashley's family and his fellow activists (my italics): (v) Mr Ashley Kriel, 20, shot dead by police in a "skirmish" in Athlone last Thursday, was an African National Congress guerilla who joined the armed struggle after the 1985 schools upheaval in Bonteheuwel, anti-apartheid organisations said yesterday.

(vi)
The family at first would not speak to the Cape Times, but yesterday Mr Kriel's elder sister, Michelle, told journalists police had continually visited the family's Vlamboom Road home since his disappearance two years ago.
(vii) She said police arrived at the family home on Friday and, after a thorough search of bedrooms, cupboards and bedding, had "bluntly" informed them of her brother's death.
She had then been asked to find someone to accompany her to identify the body.
(viii) A spokesman for the Cape Youth Congress (Cayco) said Mr Kriel joined the former Bonteheuwel Youth Movement (now Cayco) in 1982 when he was 14. "His qualities as a leader and a person who took the struggle for freedom and justice seriously made him an automatic choice as Cayco Co-ordinator on stage," she said.
(ix) Late in 1985, because of "continual harassment by the police", he stopped working in Cayco and decided to leave the country and join the armed struggle.
Intertextual echoes highlight the contrast with The Argus report.Ashley Kriel's name (released on 11 July) is foregrounded both in the headline and in the initial sentence; the police are named as agents of the shooting; and the scare quotes around the word "skirmish" imply the writer's scepticism about its aptness.In this report, the voices and point of view of the family and anti-apartheid organizations are given a good deal of space, as are details of doi: 10.5842/36-0-38 Ashley's career as an activist.The family are shown as active participants having the choice of speaking -or not speaking -to the press.His sister, Michelle, tells how they had been harassed from the time of Ashley's disappearance, and of the insensitivity of a further search of their house before police "bluntly" (her word, presumably) informed them of their brother's death (vii).
A substantial part of the report as a whole consists of quotations from spokespersons from four anti-apartheid organizations, exemplifying the oppositional discourse of the South African struggle.Phrases italicized in the extracts shown above ("the struggle for freedom and justice", "the armed struggle") are followed by others in the remainder of the article, such as, "bearing the brunt of particularly repressive state action", and by descriptions of Ashley Kriel as "a leader in the struggle" and "an example of commitment".This reporting of competing voices gives an additional layering to the discursive context of the time, as well as testifying to the courage of the Cape Times editorial staff.
The shift towards emphasis on reconciliation and commemoration of struggle activists, noted by Bennet and Verbist-Serekonyane (2002), is observable in the Western Cape English language newspapers before and during the TRC's HRV hearings (my italics): (x) "It was wonderful the way people listened.I think a few months ago there might have been a kind of booing."(Archbishop Tutu, quoted in The Argus, 7 August 1996).
(xi) Today it will be Bonteheuwel's turn to relive its role in the struggle, with several wellknown cases to be heard.(Cape Times, 5 August 1996).supported the aims and work of the TRC.On the one hand, some families of victims opposed the notion of amnesty for perpetrators; on the other, many felt that the past should not be reopened, the argument being that healing could best be served by "forgetting" past wrongs.

Testimony of a human rights violation
The testimony of Ashley Kriel's sisters, Melanie Adams and Michelle Assure, at the HRV hearings held at the University of the Western Cape, 5-7 August 1996, received a good deal of attention in the news media.At the TRC public hearings, assisted by a facilitator with whom they had previously gone over the story, testifiers were invited to tell their personal narratives in their own words and were thus given a chance to try to make sense of what had happened to them or to family members.Ashley Kriel's sisters' declared purpose in testifying was to "get to the truth" of the manner of their brother's death.It becomes apparent from their testimonies that "truth", for them, meant acknowledgement that Ashley's death had not been accidental and had been much more prolonged and brutal than the policemen involved had admitted at the inquest.
Melanie Adams' testimony includes a brief biography of Ashley, the manner in which the family learnt of his death, the funeral, and the inquest into this death (held in 1987).Michelle Assure tells of identifying his body and seeing evidence of an extended and bloody struggle at the house where he was killed.Questions from the commissioners elicit further details relating to police methods, conflicting evidence about Ashley's death and their late mother's probable wishes with regard to their appearance at the hearing.
A pattern of evaluation emerges in Melanie Adams' description of Ashley's character and early involvement in the struggle.Her statement that Ashley was not a "terrorist", as he had been "branded" by some members of the community, but a "freedom fighter" suggests political tensions within the community.Later, through a series of oppositions, she contrasts ways in which he was viewed, on the one hand, by activists and community members and, on the other, by the South African Police (SAP): as he developed into "an outstanding leader in the community" and a "very good speaker", so he "became a very great threat for the SAP".In contrast to this portrayal of her brother as a heroic community leader, she consistently describes the police as insensitive, rude and aggressive.
Against this background Ms Adams tells the story of the inquest, part of which is given in the following passage.(The text is punctuated as it is in the official transcript.)The ways in which the testifiers, Melanie Adams (at the HRV hearing) and Jeffrey Benzien (at the inquest and later at his Amnesty hearing), viewed the manner of Ashley Kriel's death are evident in the linguistic expression of roles of the participants in the various processes recorded, what is known within the Hallidayan SFL framework as "transitivity".Transitivity is the grammatical system by which we construe our experience of the world.Choices of process and participant types express the ways speakers perceive the "goings on", events and happenings in the world: who does what, to whom and how.Transitivity patterns encode speakers' experiences of reality, both the outer experiences of events (what happened and by whose agency, who verbalized these events) and their inner experiences of perceiving, understanding, feeling and so on.Benzien is described as the "guy who actually killed Ashley" and, by implication, likely to be falsifying the evidence (2).The way she turns the process of verbalization into an event (4, 7) heightens the significance of the contradictory information.These structures affect the modality of her testimony and underline the cause of her distress: all is hearsay or allegation, nothing is certain.
Details of the action are contained in the reported clauses.Material processes of action all relate to the struggle between Ashley and the policemen, as it was reported by the police: Ashley "drew a gun" (13) and "they tried to disarm him" (14).Ms Adams' summary of Benzien's account echoes features of the 1980s' police discourse.Intertextual links are significant here, as she repeats the core of the inquest report.The shooting is represented as an event, in which agency is conferred on inanimate things: "the trigger hooked onto his jersey and the shot went off".The fact that she is translating the Afrikaans of Benzien's original report and maintaining its process-participant structure (as will be seen in the analysis of Text B) suggests that his version is seared into her memory.
Ms Adams evaluates her experience overtly with the phrases "most terrible" (1) and "very disturbing" (8). 11The intensity of her need to know what happened is conveyed by the cleft structure: "[t]hat's why I should like to get to the truth of this matter" (9).Implicit evaluation is present in her repetition of Benzien's version of the event, in the parallel structure of "I don't know if it was him but she thought it was him" (24, 25), and in the contradictory details of the accounts: Benzien asserts that it was all over in a minute (18); other people "heard screaming for about an hour" (20).Another dissonance sounds in her representation of Benzien's speech in Afrikaans as opposed to the English of her testimony -the switch in language highlighting the antagonism expressed previously between "us" (the family) and "them" (the police).The account is pervaded by her suspicion, articulated earlier in the testimony, that this was "a pre-meditated murder" and that the truth had not yet been told.

Jeffrey Benzien's amnesty application
Another side of the story was told by Jeffrey Benzien at his Amnesty hearing (14 July 1997), which also attracted a great deal of media attention.The first part of Benzien's hearing, the prepared statements and initial interrogation, are conducted in Afrikaans.At the start Benzien departs from the expected format with a spontaneous apology to the victims and their families.This is followed by his account of the acts for which he is applying for amnesty and his declaration of his affiliation with the police force, his political objectives, motivations and beliefs.He then reads the statement that he had delivered at the inquest held after Ashley I grabbed my right hand with the firearm out of his grip.He turned to his left, whilst he was still in a sitting position in order to free himself and get up.Sergeant Abels, meanwhile tried to restrain Kriel I however, realised that Kriel was getting into an upright position and from my position at that stage, which was behind him, because he was turned away from me, I jumped on his back in order to pin him down to the ground once again.
With me on his back, he thrashed in all directions and tried to enter the house.At some stages we were on the ground and other stages we were kneeling or -it was during this stage that I heard a shot.I realised that it was his firearm which was still in my right hand which had gone off.
I realised that Kriel had been wounded and I noticed blood at his mouth and nose.I immediately instructed Sergeant Abels to cuff the deceased's left hand as well, and to guard him whilst I immediately went to my vehicle to get help on the radio.
The struggle couldn't have lasted for more than a minute.At no stage did I cock the weapon and in the struggle, I didn't notice whether it had been cocked.
However, I am of the opinion that the deceased must have cocked the weapon before opening the back door of the house and had concealed the weapon underneath the towel and the jersey.
(Amnesty  The curious discursive dissonance that marks entire testimony before the TRC is explained in part by the fact that as part of his application for amnesty in 1997 he read (and insisted upon the "truth" of) the 1987 inquest statement.The statement was written in the terse style of a police report of the time, with a view to justifying Benzien's conduct in the incident and to satisfying his superiors.The details in the extract (Text B) emphasise the physical actions of the participants and the mental actions of Benzien.
Benzien constructs the event as a series of swift responses on the part of the police to the actions of an armed terrorist.The transitivity patterns (in the numbered independent clauses in Table 2) show Benzien portraying himself and Abels as acting in self-defence against Kriel, the aggressor who was resisting arrest.While material processes of action are attributed almost equally to each party, the policemen's actions of taking out handcuffs, handing them to Abels, trying to restrain Kriel, jumping on his back (2, 3, 7, 9, 12) all appear to be responses doi: 10.5842/36-0-38 to Kriel's attempts to jump up, grab the gun back and "thrash" around in order to free himself (5,6,8,13,14).
Transitivity patterns in this passage also reflect the interests of the speaker, who not only gives his version of the story, but represents himself as the main observer and protagonist, who perceives, assesses, responds quickly to what is happening, gives orders, and takes immediate steps to deal with the "accident".Benzien portrays himself in the role of senser of the mental processes of perception, "heard" (17) and "noticed" ( 22), and cognition (10,18,20,27).
The most interesting linguistic feature is the convoluted structure of the clauses describing the central event.Not only is the actual shooting represented as an event, that is, as an accident, but it is embedded in layers of clauses subordinated to a verb of perception: "I realised (my italics) that it was his firearm which was still in my right hand which had gone off" (18-19); and the details appear to be highlighted by the two cleft structures, "it was during this stage that I heard a shot" (17) and "it was his firearm … which had gone off" (19).These features, together with the deletion of the agent in ( 21), implicitly draw attention to the event, but emphasise Benzien's role as perceiver, not actor.The three negative statements (25,26,27) lend further emphasis to the speed of the action, the absence of agency and, by implication, the correctness of Benzien's responses.The final statement affirms his opinion that Kriel was the agent of his own death, both because he was a "terrorist" and because he was in possession of a gun.
The jarring effect of this statement is a consequence of its "re-entextualisation" (see Blommaert 2005) in the context of the TRC hearings ten years after the event.Tracing the prevailing modes of institutional police discourse in South Africa from Union (1910) to the present, Arend (2002) notes a shift in the years immediately preceding the first democratic election in 1994 from the discourse of "counter-insurgency" and the embattled (Afrikaner) "volk" towards a "human rights" discourse.He points out that many policemen, whose professional identity was tied up with the old counter-insurgency discourse, found it difficult to make the shift to the new discourse and its ideological implications.That Benzien has partially adapted to a human rights discourse becomes apparent in other parts of his testimony.But the following passage where he is cross-examined by a TRC lawyer, Ms doi: 10.5842/36-0-38 Inthanga, shows him shifting footing (in the sense used by Goffman 1981) in a way that suggests that he is still trapped between the two discourses.(The turns are numbered for ease of reference.)

Text C
(1) MS INTHANGA: I want you to explain to the Committee as to whether throughout the struggle that you had with Mr Kriel, did this take place outside or inside the house, where exactly did this take place?
(2) MR BENZIEN: In the proximity of the -the whole incident took place in the proximity of the back door and at the furthest just through the threshold, sir.
(3) MS INTHANGA: When the family visited the house on the following day, blood was found in the kitchen on the kitchen walls and in the bathroom.Could you explain how this could have occurred?
(4) MR BENZIEN: I have no idea where this alleged blood was found if any was found sir.If I may, but it would be a speculation.Maybe the defence would know better, wasn't that house still under police guard for a number of days after the time for investigation purposes?It is speculation Mr Chairman.Maybe the parents of family could say if they had been in that house within a day of the incident.I am not sure.
( Turn (6) categorically affirms his version of the event; and (8) places him within the human rights discourse, as does his use of the lexical form "freedom fighter" and his affirmation that the "death of a human being is always to be regretted" in exchanges that follow this passage.
Benzien was also applying for amnesty for the torture of a number of other former activists, who were present at the hearing and who were allowed to interrogate him.Geschier and Lubbe (2002: 284) point out that the kind of inconsistencies noted above continue through his answers to their cross-examination, as he alternates between professing "to understand the concepts of 'forgiveness' and 'reconciliation'" and attempting to re-establish his control of the narrative by questioning their version of events and claiming to have forgotten the details of torture that they remember.Whether his forgetfulness is deliberate or a sub-conscious defence mechanism against having to face the enormity and responsibility of his deeds, 13 his version of events was unacceptable to many members of his audience.For those who, like Ashley's Kriel's sisters, did not believe that he had told the whole truth, Benzien's testimony was damaging to the possibility of reconciliation on a personal and community level.For others, among them Afrikaner poet and journalist Antjie Krog (1998), the hearing awakened complex and ambivalent emotions, of distaste, pity and shame.

3.4
The Amnesty Report Despite Benzien's admission that he did on occasion lie in his official reports, and despite finding some inconsistencies in aspects of his testimony, the Amnesty Committee granted Benzien amnesty on the grounds that "the possibility exists that he did not intend to kill Kriel" (Amnesty Decision).The version of events given by the Committee in its Decision bears a strong resemblance to Benzien's testimony, in both the factual details and the linguistic structures, as may be seen from the extract from the Decision presented as Text D below.
(Independent process clauses are numbered.) Text D (1) Benzien suspected that Kriel might be armed with a pistol or hand grenade, (2) so he moved quickly, (3) put his arms around Kriel's arms and chest trying to pin his arms to his body.(4) Benzien identified himself as a policeman (5) and told Kriel that doi: 10.5842/36-0-38 essential to dialogic engagement, an "actively responsive understanding" of the expectations and point of view of the other. 15  Ashley Kriel's sisters felt that they had been disappointed on both counts: they felt deprived of the whole truth about their brother's death, and they failed to get from the perpetrator acknowledgement of full responsibility for that death.In his application for amnesty, Benzien accepted responsibility for Ashley Kriel's death in so far as he admitted to holding the gun.
But he denied both his own agency in the shooting and the prolonged violence that was alleged by other witnesses.For Kriel's sisters, therefore, appearing before the TRC had Another instance of what Tutu (1999) calls "the long reach of apartheid" is visible in the language of this trajectory of texts and says something instructive about the social context in South Africa during the period in which the TRC was working.Bakhtin's notion of 'heteroglossia' is useful here: inherent in this concept is the idea that all instances of language use are sites of tension between the interests and discourses of more powerful groups and the constantly challenging voices of the less powerful.Press in association with the Dictionary Unit for South African English, 1996).The concept, as Posel (2006: 88-89) points out, "has been widely commodified, even caricatured", yet it was a founding principle of the Constitution and the TRC.

4.
De Gruchy (2002) suggests four levels on which reconciliation can be considered: theological, interpersonal (for instance, between victim and perpetrator), social (between divided communities), and national (the political settlement).
6. See Foster's (2006) account of various studies that have attempted to measure both the meaning of reconciliation and the degree to which it has been achieved.

7
. Fowler (1986: 131) defines modality as "the grammar of explicit comment, the means by which people express their degree of commitment to the truth of the propositions they utter, and their views on the desirability or otherwise of the states of affairs referred to".In addition to the modal verbs, modal adverbs and sentence adverbs, he includes evaluative adjectives and adverbs and verbs of knowledge prediction and evaluation as markers of modality.All of these features will be considered in the analysis.

8.
In the following abbreviated list of process types I have followed Eggins (1994) and Simpson (1993Simpson ( , 2004): (i) Material processes are processes of "doing", of action.10.This aspect of the passage has been described in Bock et al. 2000. 11.My use of "evaluative" derives from Labov's (1972)  12.
The form of address "sir" can probably be explained by the fact that Benzien felt that his remarks should be addressed not to the interrogator but to the presiding judge-a mark of Benzien's adherence to traditional forms of courtroom discourse and perhaps of police respect for authority.

13.
Bar On (1999) makes the point that perpetrators may try to preserve some shreds of their integrity by such forms of denial.

14.
Foster quotes a survey conducted by the Institute for Justice and Reconciliation in which 23% of the respondents' equated reconciliation with forgiveness (2006: 78).
15. Dialogical interaction does not necessarily have to take place face-to-face; as Bakhtin (1986: 91) writes, "every utterance is a link in the chain of speech communication of a particular sphere", and filled with the echoes and resonances of other utterances in that sphere.
16. Rousseau and Fullard (2003) argue that certain academic critiques of the TRC have continued this closing down of heteroglossia.Their contention is that some critiques that see the TRC as a failed historical project or as failing to create a consensual national memory tend to construct a monolithic image of the TRC, and overlook the many varied discourses and ambivalences emanating from it.die opdrag on Kriel se hande vas to boei.Net nadat Sersant Abels die een handboei om Kriel se regterpols geplaas het, 4 het Kriel opgevlieg, in 'n sittende posisie, terwyl 5 hy my regterhand waarin die vuurwapen steeds was, vasgegryp het. 6Ek het my regterhand met die vuurwapen [inaudible] met geweld uit [-] sy greep geruk.. 7 Hy het na links [-] omgedraai terwyl steeds in 'n sittende posisie in 'n poging om sy hande onder [-] hom te kry ten einde op te staan. 8Sersant Abels het intussen die handboede handboei wat slegs aan Kriel se regterpols was [-] um beet to kry en hom probeer vashou. 9Ek het egter besef [dat Kriel besig was om orent to kom] en van my posisie waar ek op daardie stadium agter hom was [--] um op [--] aangesien hy weggedraai het van my af [-] op sy rug gespring en ten einde hom weer op die grond vas to pen.

(
xii) The controversial death of one of the Cape's best-known uMkhonto weSizwe (MK) soldiers, Ashley Kriel, will be recalled at a hearing of the Truth Commission's human rights violations committee at the University of the Western Cape today.(The Argus, 5 August 1996) While these latter examples suggest a more receptive climate and audience at large for testifiers telling their narratives of human rights violations, by no means all South Africans Kriel's death in 1987.Text B is an extract from this inquest statement.Although Benzien was doi: 10.5842/36-0-38 speaking in Afrikaans at this stage, the text that appears on the TRC web-site (from which this passage is taken) is the English translation.(See Appendix for the Afrikaans version.)Text B I still held this firearm which I took from Kriel in my right hand and with my left hand, I took my handcuffs from my pocket and I handed them to Sergeant Abels, with the instruction that he should handcuff Kriel's hands.Just after Sergeant Abels had placed the one cuff around Kriel's right wrist, Kriel jumped up into a sitting position and grabbed my right hand in which this firearm still was.
brought neither truth nor the forgiveness necessary for reconciliation.As Melanie Adams said at the HRV hearing, any chance of forgiveness depended upon full knowledge "because what can I forgive if I don't know what happened".In the end they were left with the unresolved questions and contradictions expressed in Ms Adams' testimony, unable to achieve closure.The linguistic choices, highlighted by the analyses of the extracts from the testimonies of Melanie Adams and Jeffrey Benzien, reveal not only different presuppositions on the part of each speaker, but deeply opposing attitudes and unreconciled evaluations and discrepancies in their modes of viewing and re-constructing the incident.In addition the possibility of dialogic co-operation was undermined by, on the one hand, what might be called Benzien's desire for self-preservation, and, on the other, by the sisters' suspicion and mistrust, heightened by years of apartheid rule and their harassment by the security forces.
Directed or transitive material processes have 2 key participants -actor and goal; nondirected (intransitive) processes have 1 obligatory participant -actor.Simpson also distinguishes material event processes (performed by an inanimate actor).(ii) Mental processes include those of perception, reaction and cognition (key participants: senser and phenomenon).(iii) Verbal processes are processes of saying (key participants: sayer, verbiage).(iv) Behavioural processes are part action and part mental (psychologically driven) processes (key participant: behaver).(v) Relational processes are processes of being: in one of two modes -identifying or attributive; each mode can be in one of 3 types -intensive, possessive or circumstantial.(vi) Existential processes involve the words "there is/are".doi: 10.5842/36-0-389.These included prohibiting any journalist from being "within sight" of any "unrest", restricted gathering or "security action".See Anthonissen 2001 for a detailed account of the restrictions on press freedom in South Africa in the 1980s.
account of the elements of personal narrative.Evaluation is found in intonation and syntactic and lexical phrases or clauses which interrupt the sequential flow of narrative action clauses and indicate what the teller sees as the point of her or his narrative.
That was the most terrible experience because we as the family were treated as the perpetrators.Benzien, this is the guy who actually killed Ashley, according to his statements it was an accidental death.But what came about in the inquest was that Ashley's jersey had hooked on the trigger and the shot went off accidentally.But at the same time while I was Saturday at this briefing workshop of the TRC, something new came to light, which was very disturbing.That's why I would like to get to the truth of this matter.Someone said that, according, in fact according to Benzien and Abels they had come, posed as water works, council people to get into the house, Ashley recognized them, drew a gun, they tried to disarm him and in that process the gun, the trigger hooked onto his jersey and the shot went off.Now according to doi: 10.5842/36-0-38TextABenzien it happened as he said in Afrikaans, in 'n omesientjie, in a split second, alles op die drumpel van die agterdeur [trans: all on the threshold of the back door].But according to witnesses again, they've heard screaming for about an hour as if someone is being tortured, the next door neighbours maid could inform us the particular Friday night after we heard about his death that he was screaming and they had him surrounded, I don't know if it was him, but she thought it was him.(Helderberg/Tygerberg,Monday 5 August 1996, UWC Hearing, Case No. CT/0061 Victim: Ashley Kriel.Official transcript, pp. 8 (l.24) -9 (l.13))

Table 1 .
In Table1, I analyse the process types of the independent and the reported speech clauses in Text A (see note 8).The processes (material, mental, verbal, behavioural and relational) are highlighted in bold.I have interpreted phrases such as "according to his statement"(2, 17, 19)as verbal processes.The processes in clauses (4) and (7) are ambivalent: they are in fact verbal processes that Ms Adams represents as events.Analysis of process types of independent and reported speech clauses in Text A doi: 10.5842/36-0-38

Table 2 .
Analysis of process types of independent and reported speech clauses in Text B answer echoes the simple narrative past tense verbs of the inquest report.His second reply (4) is confusing, as he draws on features of legal discourse, "alleged" and "speculation", and admits uncertainty, "maybe" and "I am not sure". 12It is not clear whether doi: 10.5842/36-0-38 these hedges imply that the other "witnesses" are lying or whether he is avoiding answering the question.In either case, what is implied contradicts what Kriel's sisters have testified.
The TRC was established at a time when the transition to a culture and social practices more reflective of human rights in South Africa was in progress but by no means complete.CDA reminds us that social practices inform the discourse ready-made for it; it drew upon a multiplicity of discourses, genres and styles (with their associated ideologies and world views) as it developed its methodology and practices.Unresolved tensions between old and new discourses are apparent in this set of testimonies and contribute to the lack of resolution experienced by Ashley Kriel's sisters.One of the achievements of the TRC was to provide a forum for heteroglossia, giving the right of speech to many to whom it had previously been denied, and thus encouraging the new national discourse which, asGobodo-Madikizela (2006: 74)suggests, would be fundamental to "meaningful social transformation" and the rebuilding of relationships.Newly empowered voices are reflected in the many extracts from testimonies of human rights violations quoted in the TRC Report and in the media, as has been shown above.However, in the final volumes of the TRC Report, heteroglossia closes down.16Thediscourse of the amnesty decision handed to Jeffrey Benzien bears traces of the police discourse of the apartheid regime, and the voices of other testifiers, Melanie Adams and Michelle Assure among them, are silenced.The final volume of the TRC Report (2003) contains summaries of the stories of some 19,000 "victims".Ashley Kriel's is summarized as follows: Many of the texts I have used in this paper have been discussed in a postgraduate course on "Discourse Analysis and the TRC" which Kay McCormick and I taught for three years at UCT.I am indebted to the students who contributed to these discussions and to the insights offered by my colleagues, especially Kay McCormick and Claudine Raffray.
ways in which language is used, setting boundaries to what can be said and how it can be said; on the other hand, speakers may challenge the discursive limits and open the way to changing of social attitudes and values.The TRC, as has been said, did not start with an doi: 10.5842/36-0-38 institutional Ubuntu is defined as "human-heartedness; compassion; the qualities embodying the values and virtues of essential humanity, or of Africanness" -see A Dictionary of South African English based on Historical Principles (Oxford: Oxford University doi: 10.5842/36-0-38