Iberica 13 Ibérica 39 (2020): 319-344 ISSN: 1139-7241 / e-ISSN: 2340-2784 Abstract This study critically scrutinizes the discursive practice of state killing in capital trials. The quantitative and qualitative analysis compares the experientially- constitutive roles of lexico-grammatical choices in attributing agency to the actions of the defendant and victims in lawyers’ opening and closing speech in the penalty phase of a high-profile capital trial. Integrating Van Leeuwen’s social actor representation framework (2008) into Halliday’s transitivity analysis (1994), the study identifies the prosecution’s and defense’s polarized referential practice for the defendant and victims as well as descriptions of their actions. It is argued that, rather than existing as facts outside the trial discourse, aggravating and mitigating factors are morally and normatively constructed in real-time discursive interaction through these extreme, and at times disturbing, reference and transitivity choices, potentially leading to a death recommendation. Keywords: capital trial, closing, opening, penalty phase, reference term, transitivity. Resumen Acciones, actores y agentividad en la fase de pena de los juicios capitales: comparación de dos géneros Este estudio examina de manera crítica la práctica discursiva asociada a la sentencia de muerte durante el juicio. Mediante un análisis cuantitativo y cualitativo se comparan los papeles que desempeñan las elecciones léxico- gramaticales en la atribución de agentividad a las acciones del acusado y de las víctimas en el discurso inicial y final de los abogados en la fase de pena de un juicio capital muy mediático. Con la integración del marco de representación de actores sociales de Van Leeuwen (2008) dentro del análisis de la transitividad de Actions, actors, and agency in the penalty phase of capital trials: A comparison of two genres Krisda Chaemsaithong & Yoonjeong Kim Hanyang University (South Korea) krisda@hanyang.ac.kr, goyun@hanyang.ac.kr 319 Ibérica 39 (2020): 319-344 KRISDA CHAEMSAITHong & YoonjEong KIM Halliday (1994), este trabajo identifica las polarizadas prácticas de la acusación y de la defensa en la referencia al acusado y a las víctimas, así como a las descripciones de sus acciones. Se señala que, más que existir como hechos más allá del discurso que tiene lugar en el juicio, los diferentes factores agravantes y atenuantes se construyen moral y normativamente en la propia interacción discursiva en tiempo real mediante estas marcadas elecciones referenciales y de transitividad, en ocasiones perturbadoras, que potencialmente pueden conducir a recomendar la pena de muerte. Palabras clave: juicio capital, discurso inicial, discurso final, fase de pena, término referencial, transitividad. 1. Introduction Perhaps nowhere else are language practices more consequential than in the penalty phase of capital trials. Unlike the guilt phase, where jurors assess whether the facts and information presented by opposing lawyers are proved beyond reasonable doubt, the penalty phase entrusts jurors with weighing aggravating circumstances against mitigating1 ones before a recommendation of death or life imprisonment is rendered. In this phase, words can lead to the ultimate taking of someone’s life. Previous studies, although few in number, provide some evidence that language serves as a resource for mediating jurors’ perceptions of the individual on trial and their decisions in the penalty phase. For instance, Conley (2016) observes how the prosecutor makes little reference to the defendant in his closing address, but takes the effort to present the crime and its gruesome details (135-136). References to the defendant, if any, are a generalized type, with some modifier (e.g. “the person that killed like this”). In contrast, proper names appear to be used to refer to the victims, thereby “bringing them, as individuals, into the jurors’ focus” (138). Attending to the narrative structure of closing statements, socio-legal scholars Costanzo and Peterson (1994) find that defenders tell complex and textured stories, presenting a tragically flawed character, emotionally and socially deformed by years of neglect and abuse, which cause the defendant to commit the crime. In contrast, the prosecutor’s story features a simple plot and character, emphasizing an evil character, motivated by greed and sadism. The crime is thus unmistakably the defendant’s own choice rather than the product of external forces beyond his control. Despite such differences between the two sides, Kaplan (2012) argues in contrast that both sides’ narrative strategies 320 are essentially similar because defendants in the defense’s narrative also “choose” to commit the murder. The only distinction is that in the defense’s narrative, childhood abuse or a different cause is used as an explanation for the exercise of free will. what all these studies share is that at the heart of the guilt phase are the different perspectives in which the defendant and her actions are viewed as relating to other social actors in the represented world, and are positioned in terms of more versus less agency, dynamism, and affectedness. Simply put, agency is central to determining the defendant’s blameworthiness and, ultimately, the “appropriate” form of punishment. The present research goes beyond global, overarching narrative structures to explore the issue of agency in lawyers’ accounts and how it is constructed and negotiated through local lexico-grammatical choices. This study quantitatively and qualitatively compares the ways in which opposing lawyers attribute (or deflect) agency to the actions of social actors, focusing on the following experience-shaping resources: reference terms and process choices. This study attends to the opening and closing speech—the first and final opportunity, respectively—for lawyers to persuade the jury that the defendant should be spared or killed. In these genres, lawyers become sole storytellers, who have complete control of their linguistic resources. Most scholars agree, and research demonstrates, that these monologic speech genres influence jury decision-making (Matlon, 1993; Rieke & Stutman, 1990). As the opening statement aids in framing a dispute and creating strong mental images that will endure throughout the trial (Powell, 2001) and can influence the jurors to draw tentative conclusions early on (Lind & Ke, 1985; Pennington & Hastie, 1991; Spieker & worthington, 2003), the closing argument constitutes “the dramatic highpoint of the capital trial” (Burt, 2008: 903), and “[r] egardless of the evidence, whichever attorney can send a trier off to deliberate with the most ringing and eloquent soliloquy will emerge victorious” (Bergman, 1989: 222). As such, these discursive events go beyond giving a preview or summarizing the facts and evidence to persuade how the recommended sentence will serve the interests of society and to fit the choice of death or life in prison with widely held moral norms. Three questions guide this research: 1) what are the forms, functions and frequencies of reference terms employed by the prosecution and defense in the opening and closing speech? 2) what are the forms, functions, and frequencies of process choices in these genres? 3) To what extent do the two sides differ in the use of the two resources? As will be shown shortly, despite ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 321 individual differences, the prosecution and defense exhibit consistent polarized patterns of representational choices in the two genres. It is argued that, rather than existing as facts outside the trial discourse, aggravating and mitigating factors are morally and normatively constructed in real-time discursive interaction through these extreme, and at times disturbing, reference terms and process choices. This study thus contributes to not only revealing how referential and process choices are instrumental in making moral appeals to the jury in this phase of capital trials but also to placing under critical light the practice of state killing. This article begins by discussing the concept of agency and how it can be attributed or withdrawn lexico-grammatically, goes on to explain the data and methodology before presenting the quantitative and qualitative findings. It concludes with reflections on how representational resources function to negotiate the appropriate sentencing of the individuals on trial. 2. Actors, actions, and agency Following Duranti (2004: 453), this study understands agency as the property of those social actors that have some degree of control of their behavior, whose actions in the world affect other social actors’ (and sometimes their own), and whose actions are the object of evaluation in terms of their responsibility for a given outcome. Along the same lines, Taylor (1985) notes that “we think of the agent not only as partly responsible for what he does, for the degree to which he acts in line with his evaluations, but also responsible in some sense for these evaluations” (25). Agency is encoded in such features as reference terms and verb choices (o’Connor, 2000: 39). The former includes labeling options for referring to (or excluding) social actors, which have been shown to manage interpersonal distance, involvement, and solidarity for the referent and topic being discussed (Murphy, 1988), ascribe identities (Chaemsaithong, 2018) and signal evaluation and reality claims (Tracy, 2011; Rosulek, 2015). Verb choices deal with the way in which a specific kind of action or event is construed (as doing, saying, thinking, being, or simply existing), known as transitivity system. Both of these resources illustrate the experiential (or ideational) metafunction of language (Halliday, 1994). According to Van Leeuwen (2008), in referring to another individual, the speaker first faces two value-laden choices: whether to exclude or include KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344322 another social actor in the discourse. when excluded, that social actor’s existence is obscured and her agency eliminated. If included, she can be represented either in terms of her unique identity through nomination, or in terms of identities and functions she shares with others through categorization. nominations range from very formal (last name only), semi- formal (first and last), to the least formal (first name or nickname). To bestow authority and power to the referent, nominations can be titulated, as in Doctor X, while a nickname creates intimacy for her. Rosulek (2008: 538) reports that the prosecution in a sexual abuse case refers to the victim with a nickname to make her more childlike and personal to the jury, while the defense uses her last name. Categorization can be divided further into functionalization and identification. The former includes labels that describe what the referent does, that is, her role in a context (e.g. victim, or defendant). Identification represents the referent in terms of something she more or less permanently or inexorably is through labels that differentiate people into classes, which further include physical descriptions (size or weight), social relations descriptors (kinship terms), or other classification labels (e.g. age, character, race, religion, or sexual orientation). All categorizations are inference-rich and carry social connotations. Bennett (1978), for example, notes that when cross-examining a witness, the lawyer employs the kinship term “your son” to refer to the defendant so as to support the argument that, as a mother, the witness is likely to protect the defendant. Finally, a referent can be impersonalized, which de-emphasizes her human characteristics and instead foregrounds a specific non-human identity aspect. This strategy may create in-group solidarity, as in The defendant wants to punish America, where the name of the country metonymically represents all American people and positions the defendant as an outsider of that group. Alternatively, impersonalization can also estrange the referent. Luchjenbroers and Aldridge (2007) show that prosecutors emphasize the defendant’s violence through animal metaphors, such as “predator”. All in all, different reference options enable the presenter to highlight or eclipse specific aspects of the referent, including power, authority, ability, strength and character. Co-occurring with the referent in a clause is a description of her involvement in an event through the verb phrase. According to Halliday (1994: 106), an event can be represented in a manageable set of process types: doing ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 323 (material process), sensing (mental process), being, having or locating (relational process), saying (verbal process) and existing (existential process).2 These processes are shown in Table 1, with examples from my data. Like referential choices, process choices are ideologically significant. Selecting one process over another results in a completely different experience and reality being construed. For instance, the material clause the defendant killed the child can be expressed as a behavioral process the child died or as an existential process there was killing. These alternatives obscure the actor and sidestep her responsibility, so that it appears that the incident simply happens. A mental process such as the defendant wanted to kill, which reveals what otherwise is an internal feeling, portrays the actor as having the desire to commit the illegal act. In addition, an actor’s involvement in a process may be allocated varying degrees of agency. As Table 2 illustrates, the first possibility is when the actor appears in the grammatical subject slot of an effective process or in the by- agent phrase of a passive process. In this case, the agent directs her action towards a goal and is, thus, attributed full agency, power and responsibility (Dreyfus, 2017). At the other extreme, the referent’s agency may be deflected when she is cast as the traditional in/direct object of an effective active process or the subject of a passive verb. That she plays a recipient or beneficiary role can signal vulnerability (Van Leeuwen, 2008). Also lacking in agency and responsibility are events that are construed as if they just happen by themselves, as in the middle voice or nominalization, because the focus is taken away from who carries out the process (Dreyfus, 2017). In between these two possibilities are cases where agency is mitigated, notably in an agentless passive clause or an active clause without a goal (i.e. ineffective processes). In such cases, agency is implied or can be inferred, but the processes do not bring out an effect on others. KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344324 Process type Example Material The defendant killed Leo. They marched resolutely to the crowd. The family was broken apart by Dzhokhar. Mental Dzhokhar adored his brother. Relational qualificational Around his friends, Jahar is somebody who seemed independent and cool. possessive The defendant had plenty of time to reflect. The defendant’s lack of remorse locative The defendant was in front of the crowd. Verbal It’s Lingzi Lu screaming in pain. Existential There are others who found their lives irrevocably changed after the crime. Table 1. Process types and examples. I Analyzing clausal patterns in eight Chinese cases, guang (2010) finds that material processes occur most frequently, followed by verbal, relational and mental processes. Behavioral and existential processes are infrequent. A useful finding is that the defendant appears as the agent in material processes more frequently in the prosecutor’s discourse than in the defense’s. This is argued to serve to portray the defendant as active and aggressive. However, the ways in which the victims are represented are not discussed in guang (2010). Focusing on closing statements of the guilt phase, Rosulek (2015) finds that the prosecution foregrounds the defendant and their agentive roles in the crime more than the defense does and that the defense includes less personal information about the victims. All in all, previous research provides evidence that referential practice and process choices construct different versions of reality in the guilt phase. It remains to be seen how such resources function in the penalty phase of capital trials. 3. Data and methodology Drawing from the official transcript of the penalty phase of United States v Tsarnaev (no. 13-10200-gAo), the data consists of the opening and closing speech of the prosecution and defense team (26,017 words in total). Each team assigns a different lawyer to deliver the speech, hence four presenters in total. This sentencing trial took place in 2015, after the jury had convicted Dzhokhar Tsarnaev and found him responsible for having killed three people and wounded about 300 people in the 2013 annual Boston Marathon bombing, murdering a university police officer Sean Collier three days later in an attempt to take his gun, and carjacking a driver in Cambridge while trying to escape from Boston. ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 325 Agency deflected Agency mitigated Agency attributed Tamarlan influenced Dzhokhar. There was killing. The innocent people were killed and hurt [by ?]. Dzhokhar acted under the influence of his brother. (Def) The defendant killed multiple people. (Pro) Her daughter was blown apart by this defendant. (Pro) Table 2. Degress of agency. 3 T In this guilt phase, the prosecution argued that death is the only appropriate punishment due to such aggravating factors as the death of an 8-year-old child, the choice of the Boston Marathon as a target because of the opportunity for maximum bloodshed. The defense, in contrast, argued that the 19-year-old defendant participated in the crime only after falling under the influence of his older, radicalized 26-year-old brother, Tamerlan Tsarnaev. This psychological influence is claimed to be a major mitigating factor, so that family members, friends and witnesses were summoned to testify that the defendant had a turbulent childhood but had done well in school and was liked by many people. From a linguistic perspective this case constitutes a unique opportunity to witness how representational resources are utilized in the negotiation of agency, free will, and creation of “us/them” for at least two reasons. First, this case features not only the most destructive act of terrorism in the US since the 1995 oklahoma City bombing but also the likelihood of mental influence, which is a very controversial issue in the scientific community. one common argument propelled by supporters is that a powerful individual can exercise control over another “weaker” individual (Fournier, 1992: 612). However, this claim is difficult to prove as it involves the evaluation of the intangible notion of free will and in the courtroom is discursively managed through expert testimony, which is not only inconclusive but also conflicting (warburton, 2003). Second, the case also brings to the fore the issues of human nature, justice and revenge. Many witnesses are likely to seek revenge in such a situation, while the purpose of a civil society is to rise above the instinctive desire for revenge and determine what is just and humane, in which case power struggle between the two sides can be expected. Methodologically, as we are interested in how individuals are made into certain kinds of people and their roles in carrying out actions and events, we first divided social actors in the lawyers’ stories into three main groups, consisting of the defendant Dzhokhar Tsarnaev, the defendant’s brother Tamerlan Tsarnaev, and the victims. Then, we identified and performed a frequency count of the reference terms made for each group, using Van Leeuwen’s model discussed above. In this step, all the reference terms were included and counted separately. Pronouns, which are low in semantic content, were excluded. with regard to the analysis of the process types, we first identified the processes with which the three groups above were associated. The processes KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344326 can occur in a finite or non-finite, and in a main clause or subordinate clause. For the purposes of this paper, a social actor is considered to hold an agent role when she is the doer of the process and, so, is given some degree of agency (i.e. in the grammatical subject position of an active verb, or in the by-agent phrase of a passive). In contrast, a social actor holds a recipient role when she is acted upon or is the beneficiary of the process, and therefore is lacking in agency (i.e. the traditional direct and indirect object of a non- passive verb or the subject of a passive voice verb).3 Possessives (as in Dzhokhar’s computer) were treated as agents, as they can be paraphrased as Dzhokhar had a computer. we excluded those social actors that are peripheral to the actual process (as in The defendant placed the bomb near the crowd), as these instances are rare and are part of a prepositional phrase rather than processes. All the processes were then coded based on Halliday’s framework above. In cases where a process is realized periphrastically (as in he wanted to kill innocent people), we recognized it as consisting of two processes (in this case, a mental and a material process). However, in the case of periphrastic auxiliaries, which function more grammatically as aspectual markers, only the main verb is considered (as in He’s going to shoot the officer or He kept sending stuff to his brother) To illustrate our analytical process, consider the following excerpt: Dzhokhar Tsarnaev [agent] executed [mat] a police officer [recipient] targeted [mat] simply because he [agent] was [rel: qual] an officer, and Dzhokhar Tsarnaev [agent] wanted [men] his gun [agent; rel: poss], an officer [agent] who sat [mat] in his cruiser [agent; rel: poss] and died [mat] after he [recip] was shot [mat]. In the above example, references to the major social actors are underlined, showing nominations (Dzhokhar Tsarnaev) and identification terms (a police officer and an officer). The processes are italicized and coded in brackets, together with the roles of the social actors (either agent or recipient). To allow a comparison of the results for different text lengths, all the frequency counts were normalized to a common basis of 10,000 words. To ensure the consistency of the analysis, each researcher first analyzed the transcripts separately and resolved coding discrepancies. Finally, our discourse-oriented qualitative analysis was performed to identify salient patterns displayed by each side and also to consider how a specific reference form and process may be pragmatically motivated in relation to the speaker’s ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 327 goals. Authorial evaluation, which indicates either a positive or negative stance of the presenter, was also taken into account in this step. 4. Findings 4.1. Referential practice Finding 1: As the overall frequency counts in Table 3 indicate, in both the opening and closing, the prosecution focuses much more on the victims than on the defendant and his brother. In particular, nomination, identification, and impersonalization constitute this side’s major strategies to reference the victims. on the other hand, the defendant is mostly represented either by nomination (as we saw in the opening statement), or by functionalization (as we saw in the closing statement). In comparison to the defendant, the defendant’s brother is not referenced as frequently in either genre. This suggests that the prosecution’s discourse is about assigning agency and responsibility to the defendant, thereby limiting the roles of his brother to a partner in crime, as opposed to a leader and an influencer. The defense, in contrast, consistently makes the case more about the defendant and his brother, mostly through nomination and identification. note that references to Tamerlan occur about two or three times more than they do in the prosecution’s speech. In addition, the defense rarely, if at all, mentions the victims in both genres. In effect, the issues of victims and damages are silenced, and attention is drawn to the defendant in relation to his brother. Each reference strategy will be broken down and discussed in more detail below. KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344328 Opening Prosecution Defense Defendant Brother Victim Defendant Brother Victim Nom 123.93 16.52 110.16 83.48 85.22 - Func 19.28 5.51 35.80 1.74 1.74 3.48 Iden 2.75 16.52 148.71 43.48 12.18 3.48 Imper 2.75 8.26 49.57 5.22 1.74 - Total 148.71 46.81 344.24 133.92 100.88 6.96 Closing Nom - 10.12 191.91 107.93 81.70 5.05 Func 111.57 2.98 29.75 6.05 1.01 4.03 Iden 7.44 25.29 144.30 36.32 28.24 3.03 Imper - - 63.97 3.03 1.01 - Total 119.01 38.69 429.93 153.33 111.96 12.11 Table 3. Lawyers’ references to social actors in two genres. ( Finding 2: In both genres, the two sides display starkly distinct patterns of nominations for the defendant and his brother, as illustrated in Table 4. To begin with the opening, the prosecution exclusively uses the semi-formal label Dzhokhar Tsarnaev. nominated references to defendant’s brother occur much less frequently and are almost always in the semi-formal format. The reason is likely that using the last name alone would not distinguish the two individuals, and first names would create solidarity for the defendant, which in turn would work against the prosecutor’s intention. The semi-formal label, therefore, solves this dilemma for the prosecution. with this semi-formal label identifying the defendant as the actor of violent actions, the lawyer can pragmatically place the defendant at a distance from the jurors: (1) why is the death penalty the appropriate and just sentence? Because Dzhokhar Tsarnaev planned and he plotted to kill. Because when Dzhokhar Tsarnaev sauntered down Boylston Street and took a pressure cooker bomb into the crowds, he created a grave risk of death. (Pro opening) The defense, in contrast, almost exclusively uses an informal label for both the defendant and his brother in the opening. Doing so creates solidarity for them, so that although they committed a crime, they are just “like us”. what is most interesting is that the defendant is represented by his nickname Jahar, known among his friends. Indicating intimacy between interlocutors, the nickname shapes the social relations between the lawyer and the defendant by adding a sense of engagement with the defendant (Brown & gilman, 1960; Brown & Ford, 1961). ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 329 Opening Prosecution Defense Defendant Brother Defendant Brother Informal - Tamerlan 2.75 Jahar 80 Tamerlan 81.74 Semi- formal Dzhokhar Tsarnaev 123.93 Tamerlan Tsarnaev 13.77 Jahar Tsarnaev 3.48 Tamerlan Tsarnaev 3.48 Formal - - - - Titulated - - - - Total 123.93 16.52 83.48 85.22 Closing Informal - Tamerlan 1.49 Dzhokhar 100.87 Tamerlan 80.69 Semi- formal - Tamerlan Tsarnaev 8.63 Dzhokhar Tsarnaev 7.06 Tamerlan Tsarnaev 1.01 Formal - - - Titulated - - - Total - 10.12 107.93 81.70 Table 4. Nominations for the defendant and his brother in two genres. I F ( O In the closing, the defense’s referential practice is consistent with what is witnessed in the opening—placing almost equal emphasis on the defendant and his brother in an attempt to explain how their relationship contributed to the crime and creating familiarity with these actors through first names. note, however, that in the prosecution’s closing, nominations do not occur (but, as will be discussed below, the functional term defendant is employed instead). Finding 3: The defense’s opening and closing speech shows much more variation in the use of other reference terms for the defendant and his brother than does the prosecution’s. As exhibited in Table 5, classifications predominate, introducing into the courtroom various identity aspects of the defendant and, to a lesser extent, of his brother. For example, the terms aider, abetter, participant mitigate the defendant’s agency and his roles in the attack, while influence and bomber attribute full agency to the defendant’s brother. Labels describing their age, size, and character also construct polarized identities for the two brothers: (2) jahar really was what he appeared to be: a lost teenager with very little motivation to do anything much on his own, who had been raised all his life to take direction from the most powerful adult, by 2013, the only powerful adult in his world. (Def opening) of note in the prosecution’s discourse is the functional term defendant (particularly in the closing), which roughly occurs at the exclusion of other referential terms. This specific term situates Dzhokhar solely according to his legal status, eclipsing other personal aspects. The emphasis on this legal identity appears to be pragmatically motivated. As evidenced in (3), the term is repeated in two successive clauses even when a pronoun would suffice: (3) The defendant took all of that away from loving, loving, caring, positive people. This defendant blinded the mother, maimed their six-year-old daughter, ripping off her leg... (Pro Closing) In the prosecution’s opening and closing, Tamerlan is simply referred to as the defendant’s brother and is described as a partner, conspirator, easy target [of assigning blame], and substitute for the defendant. These terms clearly lessen his full responsibility for the crime. KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344330 Finding 4: In both genres, the prosecution’s references to the victims show a greater variety than the defense’s. As regards nominations, labels of all formality levels are employed. As Table 6 illustrates, informal and semi-formal nominations occur most frequently in the prosecution’s discourse. The use of first-names brings the jurors pragmatically closer to the victims, while listing out first and last names shows that the victims are of diverse backgrounds: (4) Because of his actions and role in this conspiracy, he [defendant] maimed Jeff Bauman, Erika Brannock, Celeste Corcoran, Mery Daniel, Rebekah Gregory, Patrick Downes, Jessica Kensky, Martin... (Closing Pro) In contrast, the defense does not use nominations at all in the opening. Although they do acknowledge the death of a police officer through titulated ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 331 Opening Prosecution Defense Defendant Brother Defendant Brother Func defendant 19.28 partner, conspirator 5.51 the person who is being sentenced for 1.74 the man who conceived, planned, led 1.74 Iden class phy rel - - 17-year-old high school senior, a lost teenager, quiet kid, good kid, coolest kid 27.83 a young Chenchen overseas, that sort of person, powerful adult 3.48 - - little brother 5.22 - brother 2.75 brother 16.52 (younger) brother, 10.43 older brother last adult family member, oldest son 8.70 Imp his hands 2.75 easy target, substitute 8.26 his face, his closed left eye 5.22 easy target 1.74 Closing Func the/that defendant, terrorist, partner 111.57 partner in crime 2.98 participant, aider, abetter 6.05 bomber 1.01 Iden class child, adult 5.95 - that kid, child, teenager, 8-year-old Dzhokhar, quiet kid, good friend, young man 32.28 skilled boxer, 15 or 16 year old Tamerlan 7.06 phy - - little boy, small boy 2.02 bigger brother 1.01 rel brother 1.49 brother 25.29 younger brother 2.02 (older) brother, 20.17 Imp - - puppy 3.03 major influence 1.01 Table 5. Use of other reference terms for the defendant and his brother in two genres. nominations in the closing, the defendant’s agency is deflected and instead assigned to Tamerlan: (5) Dzhokhar didn’t pull the trigger. He may be responsible for the death of Officer Collier, but in a sense of weighing that for punishment, consider who killed Officer Collier, who pulled the trigger. (Closing Def). Finding 5: As with nominations, a variety of categorization terms for the victims are found in the prosecution’s opening and closing. In particular, as presented in Table 7, classifications and relationship descriptors predominate, serving to make salient the fact that the victims are vulnerable and innocent human beings and that the defendant caused their families to break apart, which sensationalize the issues of damage and victimization. In addition, impersonalization, which is not found in the defense’s discourse, constitutes another prime means for the prosecution to assign responsibility. The victims’ body parts are singled out to emphasize the damages sustained and highlight the destructive effects the defendant caused, as in (6). Example (7) illustrates how metaphors and other identification labels are used for the same victim in a short stretch of discourse. (6) You saw the autopsy photos of Martin Richard, Krystle Campbell, and Lingzi Lu. The bombs burned their skin, shattered their bones and ripped their flesh. It disfigured their bodies, twisted their limbs and punched gaping holes in their legs and torsos. (7) Lingzi was their only child, their future. That future ended on April 15th, 2013. She was her father’s jolly elf. She was the beautiful nerd. (Pro Closing) KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344332 ( ( Opening Prosecution Defense Informal 46.82 - Semi- formal 41.31 - Formal 2.75 - Titulated 11.02 - Total 101.90 - Closing Informal 74.38 - Semi- formal 108.60 - Formal 2.98 1.01 Titulated 5.95 4.04 Total 191.91 5.05 Table 6. Nominations for the victims in two genres. F These terms are scant in the defense’s speech, and when present in the defense’s speech, these victimized actors are not related to the defendant in any way: (8) now, you’ve all probably realized by now that no punishment could ever be equal to the terrible effects of these crimes on the innocent people who were killed and hurt. (Def opening): 4.2. Event description Finding 6: In both genres, the prosecution significantly represents the defendant as the agent of effective material processes, as Table 8 indicates. This means that the defendant’s actions are always directed to a goal, usually a victim. The verb kill constitutes about 35 per cent of these processes, and another 30 per cent include such processes as murder, mutilate, injure, terrorize, and maim: ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 333 1 ( Opening Prosecution Defense Func survivors, victims 35.80 survivors, victims 3.48 Iden class innocent children, jolly girl, eight-year-old boy, spectators, young beautiful people 82.62 innocent people, those people so young and full of promise 3.48 phy small boy 2.75 - rel son, daughter, grandchildren, best friend, brother, sister, family, colleague 63.34 - Imper those beautiful faces, limbs, the talented fingers, symbols of loss 49.57 - Closing Func victims, perceived enemies, the wounded, amputees, the dead 29.75 victims, the injured 4.03 Iden class innocents, 12-year-old body, children, beautiful nerd, lovely, caring people, dad’s princess 65.46 people whose lives were ripped apart 1.01 phy little Leo 10.41 - rel father, mommy, daddy, family members, only child, loving son 68.43 family members, their families 2.02 Imp mangled and ruined limbs, the light of his life, moral compass, whole body, severed femora, skin, internal organ, jolly elf, future 63.97 - Table 7. Other reference terms for the victim in two genres. (9) Dzhokhar Tsarnaev was willing to cross every line for personal glory. Kill innocents with a bomb: Done. Kill a police officer: Done. Kill a child: Done. (Pro opening) Ineffective material processes, through which the prosecution mitigates the defendant’s agency by relating what he did after the attack, further construct Dzhokhar as cold-hearted and remorseless: (10) while his victims lay dead and dying and bleeding...there’s the defendant, strolled into whole Foods like it was an ordinary day and shopped for milk. That same evening, at 8 p.m., he got on the Internet and tweeted to his friends, “Ain’t no love in the heart of the city. (Pro Closing) In contrast, the defense represents the defendant in effective material processes at much lower frequencies, and in such cases, the defendant is portrayed as a benevolent actor: (11) He made her [a friend] laugh and feel good. He was respectful to the other girls. He treated them with respect. (Def Closing) The defense’s ineffective processes downplay the defendant’s role in planning and committing the crime, making him a vulnerable subject of his brother: (12) The idea the younger brother follows and supports the older brother is part of who they both were...when Tamerlan decided that it was time, his little brother went with him. (Def opening) (13) He followed Tamerlan around like a puppy. (Def Closing) of note in the defense’s speech is deflected agency (14) or, at the other extreme end, his recipient role (15), signaling his lack of power: (14) He didn’t beat anyone up. He didn’t take advantage of people. (Def opening) (15) Tamerlan set himself and his brother on this terrible course...he would pull the younger brother with him. (Def opening). KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344334 Finding 7: As Table 9 indicates, the two sides attribute starkly different degrees of agency to the defendant’s brother in performing material processes. In the prosecution’s discourse, Tamerlan’s role is almost inconsequential. Interestingly, Tamerlan’s material processes, whether effective or ineffective, almost always appear in a co-ordinated structure, with the defendant as the other agent: (16) Also consider how the defendant and his brother killed officer Sean Collier...You saw how they deliberately walked together across the campus, and they went straight to the door of his car....The fact that the defendant and his brother targeted him... (Pro Closing) In contrast, the defense consistently assigns Tamerlan much more agency in executing effective material processes, especially those having to do with planning, leading, and committing the attack (17) and negatively impacting other people (18): (17) who had the gun? who was shooting at the police? who shot Collier with the gun? Tamerlan. (Def Closing) (18) Tamerlan was a very tough guy...who was suspended from high school for assaulting another student and who was arrested for assaulting his own girlfriend. Tamerlan turned a lot of people off. (Def opening) His negative persona is further enhanced by ineffective material processes: ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 335 Opening Prosecution Defense As agent As recipient As agent As recipient Agency attributed 165.24 - 48.70 - Agency mitigated 79.87 - 40.0 - Agency deflected - - 8.70 36.52 Total 245.11 - 97.40 36.52 Closing Agency attributed 226.82 - 42.36 - Agency mitigated 69.34. - 38.33 - Agency deflected - 8.11 3.03 34.29 Total 296.16 8.11 83.72 34.29 Table 8. Material processes executed by the defendant in two genres. F ( ( ( ( (19) Tamerlan, who had changed dramatically, becoming very radical, left for Russia on a trip we now know was to wage jihad, to take up the fight in the mountains. (Def Closing) Finding 8: As Table 10 shows, other notable processes in the prosecution’s speech are relational and mental processes. About 80 per cent of the prosecution’s qualificational processes clearly indicate a negative stance towards the defendant, as in (20). Also, the defendant is described as possessing crime-related materials and violent character as well as having premeditated the crime through the mental process knew (21): (20) This is Dzhokhar Tsarnaev unconcerned, unrepentant, and unchanged. without remorse, he remains untouched by the grief and the loss he caused. And without assistance, he remains the unrepentant killer that he is. (Pro opening) (21) He knew that the marathon attracted families and that people go there with their friends, so he knew that his bomb was likely to kill and mutilate parents in front of their children. He also knew that the last stretch down Boyston Street...drew huge crows. He knew that by placing his bomb there, he had a good chance of killing and injuring hundreds of people which is exactly what happened. (Pro Closing) In contrast, the defense constructs a different persona for the same individual (22): KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344336 Opening Prosecution Defense As agent As recipient As agent As recipient Agency attributed - - 73.04 - Agency mitigated 11.01 - 43.47 - Agency deflected - 2.75 3.48 17.39 Total 11.01 2.75 119.99 17.39 Closing Agency attributed 21.67 - 38.34 - Agency mitigated 11.36 - 26.23 - Agency deflected - 8.67 - 5.04 Total 33.03 8.67 64.57 5.04 Table 9. Material processes executed by the defendant’s brother in two genres. F ( ( (22) Dzhokhar respected and loved his older brother...He was the shy, quiet, respectful, hard-working kid that teachers and friends came in here and told you about. (Def Closing). Finding 9: In both genres, the prosecution predominantly represents the victims as agents of ineffective material processes or recipients of the defendant’s effective material processes, as Table 11 shows. By mitigating or deflecting their agency, the defendant’s actions appear ruthless and unjust: (23) when she finally had enough breath to breathe, she screamed in agony. She was left to try to recover in the same hospital room as her daughter, another family blown apart by this defendant and his brother. (Pro Closing). ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 337 Opening Prosecution Defense Defendant Brother Defendant Brother Relational qual 44.06 16.52 40 43.48 pos 68.85 2.75 62.61 36.52 loc 11.02 - 1.74 1.74 Mental 16.52 - 26.08 15.65 Total 140.45 19.27 130.43 97.39 Closing Relational qual 20.23 1.44 59.51 19.16 poss 47.67 2.90 78.68 31.27 loc 4.33 2.90 7.06 7.06 Mental 86.68 10.11 47.40 24.21 Total 158.91 17.35 192.65 81.70 Table 10. Relational and mental processes involving the defendant and brother in two genres. F Opening Prosecution Defense As agent As recipient As agent As recipient Agency attributed - - - - Agency mitigated 79.86 - - 1.74 Agency deflected - 71.61 - - Total 79.86 71.61 - 1.74 Closing Agency attributed - - - - Agency mitigated 84.79 - 2.02 - Agency deflected - 123.47 - 13.11 Total 84.79 123.47 2.02 13.11 Table 11. Material processes executed by the victims in two genres. In those rare cases where the defense acknowledges victimization, agency is grammatically obscured through agentless passives and nominalization of processes: (24) The story of the Boston Marathon bombing is about resilience and the strength of the spirit of those so deeply affected by these senseless and catastrophic acts. Finding 10: Also substantial in the prosecution’s speech are the victims’ relational (25) and mental processes (26), by which the victims’ personal character, subjective pain, and suffering are described in detail. As Table 12 indicates, this strategy is even more pronounced in the closing speech. (25) Before he murdered them in some of the cruelest ways imaginable, they were sons, they were daughters, they were grandchildren, they were brothers and they were sisters. (Pro opening) (26) She [victim] remembered every detail of the blast. She suffered excruciating pain…She just wanted to die because the pain was too much. (Pro Closing) In the defense’s discourse, the victims’ relational and processes are sparse, and when present, it is primarily for the purpose of defending an argumentative position. Example (27) illustrates how the defense advances a question in anticipation of the jurors’ doubt, where the victims’ sufferings are acknowledged just to pre-empt other alternatives. (27) The question could be….why shouldn’t he [defendant] suffer as his victims did?...There’s nothing wrong with those questions…but there is something wrong with thinking that the answer will be found in imposing the sentence of death… (Def Closing). KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344338 5. Conclusion This study has examined representational choices that negotiate different degrees of agency and responsibility for the main social actors in the opening and closing of the penalty phase of a capital trial. while individual differences between lawyers exist, this study has been able to identify the prosecution’s and the defense’s polarized linguistic patterns in both genes. As regards referential practice, the prosecution individualizes and personalizes the victims mainly through informal or semi-formal nominations, categorizations, and impersonalizations in both genres. Such emphasis on the victims positions them as having no control over what happened to them and aggrandizes the magnitude of the impacts they sustained. when it comes to the defendant, the prosecution references him either by a semi-formal label or by his legal status, thereby not only drawing attention to the agent of the criminal acts but also distancing him from the jurors. The defendant’s brother is mostly excluded. In contrast, the defense, consistently in both genres, avoids mentioning the victims and equally represents both the defendant and his brother, mainly through informal nomination and categorization. This emphasizes their human attributes. However, it is observed that references to the defendant’s brother convey negative evaluation, thereby increasing his share and blameworthiness. with respect to event description, the two sides also fare differently. The main strategies are summarized in Table 13. ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 339 ( ( Opening Prosecution Defense Rel qual 46.82 1.74 poss 41.31 3.48 loc 5.51 - Ment 35.81 - Total 129.45 5.22 Closing Rel qual 50.58 - poss 81.82 3.03 loc 7.44 - Ment 41.66 1.01 Total 181.50 4.04 Table 12. Relational and mental processes involving the victims in two genres. In a more critical vein, this study reveals how referential practice and process description comes to be instrumental in struggles of power. However, the contest in the penalty phase does not seem to be as much about constructing an alternative version of reality as about attributing different degrees of agency and constructing polarized identities for them, thereby invoking moral, normative values for the same individuals. This indeed stands in stark contrast with the fact-finding, evidence-based setting, where the guilt phase takes place. This particular case has resulted in a death sentence.4 whether or not such recommendation testifies to the perlocutionary effects of the prosecution’s representational strategies is difficult to say precisely at this point. nor does this study attempt to make a claim about the effectiveness of such strategies used by either side. Further research that compares the use of these strategies in other cases may shed more light on this issue. This study does, however, make explicit how the heinousness of the crime and dangerousness of the defendant are discursively constructed in the minute details of representational choices. The justification of state killing, which is thought to be objective, impartial and neutral under the rule of law, is in fact imbued with value-laden representational practice and strategic assignment of agency and responsibility (or concealing these issues). At the risk of venturing into subjective territories, we are concerned that one negative consequence of this dueling representational practice may result in a process that lacks fairness and integrity such that the individual to be executed is distinguished not by her past records, her culpability, or the circumstances of KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344340 A Prosecution Defense Defendant - is attributed full agency and responsibility mainly by positioning him as a) agent of effective material processes with victims as recipients, b) agent who shows awareness of actions, and c) agent with negative qualifications crime-related materials - is represented as having no or weak agency and responsibility by positioning him as a) agent of ineffective material processes, b) as recipient of processes executed by others, and c) agent of positive qualificational processes Brother - is infrequently represented as involved in processes, hence deflecting agency and responsibility - is attributed agency as a co-participant of effective material processes - is attributed high agency and responsibility and represented as a) agent of effective material processes which negatively affect others, and b) agent of negative qualifications and possessor of terrorist materials Victims - are attributed little agency by being represented as a) recipients of effective material processes carried out by the defendant, b) sensors of mental processes to emphasize pain and suffering, and c) carriers and possessors of innocent and vulnerable traits - are mostly not represented Table 13. Summary of characteristic process patterns in each side’s discourse. I the crime, but by the quality of representation and presenter’s labelling skills. with a view towards a more humanistic judicial system, the findings are hoped to inform legal professionals and the public at large, some of whom may be summoned for duty jury, to be more aware of the power of linguistic choices in shaping their sentencing decisions. Acknowledgements This research is supported by funding from Hanyang University (HY-2020). Article history: Received 4 February 2019 Received in revised form 10 June 2019 Accepted 21 April 2020 References ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 341 Bennett, W. (1978). “Storytelling in criminal trials: A model of social judgment”. Quarterly Journal of Speech 64: 1-22. Bergman, P. (1989). Trial Advocacy in a Nutshell. St. Paul, MN: West Publishing. Brown, R. & M. Ford (1961). “Address in American English”. Journal of Abnormal and Social Psychology 62: 375-385. Brown, R. & A. Gilman (1960). “The pronouns of power and solidarity” in T. Sebeok (ed.), Style in Language, 253-276. Cambridge: Cambridge University Press. Burt, M. (2008). “The importance of storytelling at all stages of a capital case”. UMKC Law Review 77: 877-910. Chaemsaithong, K. (2018). “Referential practice and contested identities in legal narratives”. Lingua 212: 44-59. Conley, R. (2016). Confronting the Death Penalty: How Language Influences Jurors in Capital Cases. Oxford: Oxford University Press. Costanzo, M. & J. Peterson (1994). “Attorney persuasion in the capital penalty phase: A content analysis of closing arguments”. Journal of Social Issues 50: 125-147. Dreyfus, S. (2017). “‘Mum, the pot broke’: Taking responsibility (or not) in language”. Discourse & Society 28: 374-391. Duranti, A. (2004). “Agency in language” in A. Duranti (ed.), A Companion to Linguistic Anthropology, 451-473. Malden: Blackwell Fournier, V (1992). “The admissibility of expert testimony in brainwashing-related cases: Should witnesses be fryed?” Santa Clara Law Review 32: 607-647. Guang, S. (2010). “An analysis of the transitivity system in courtroom discourse”. Chinese Semiotic Studies 4: 245-257. Halliday, M.A.K. (1994). An Introduction to Functional Grammar. London: Arnold. Kaplan, P. (2012). Murder Stories: Ideological Narratives in Capital Punishment. Lanham: Lexington Books. Lind, A. & G. Ke (1985). “Opening and closing statements” in S.M. Kassin & L.S. Wrightsman (eds.), The Psychology of Evidence and Trial Procedure, 229-252. London: Sage. Luchjenbroers, J. & M. Aldridge (2007). “Conceptual manipulation with metaphors and frames: Dealing with rape victims in legal discourse”. Text & Talk 27: 339-359. Matlon, R. (1993). Opening Statements/Closing Arguments. San Anselma, CA: Stuart Allen. Mayr, A. & D. Machin (2012). The Language of Crime and Deviance: An Introduction to Krisda Chaemsaithong is Professor of English Linguistics at Hanyang University, where he teaches and supervises research on courtroom discourse from a socio-functional linguistic perspective. His research topics include person reference, speech reporting, politeness, and narrative structure. Yoonjeong Kim is a doctoral candidate at the same institute. Her research interests include stylistic positioning of characters in narratives and visual rhetoric. NotEs 1 Aggravating factors are those circumstances that increase the severity or culpability of a criminal act, thereby warranting a death sentence. Mitigating factors are those circumstances that provide reasons as to why a life sentence, as opposed to death, should be granted. 2 Halliday (1994) also proposes behavioral processes, those processes that are partly material and partly mental in nature, such as watch, dream, breathe, die, etc. In this paper, behavioral processes will be categorized under material processes, where the agent has lesser agency than those (effective) material processes with a goal. 3 Following Mayr and Machin (2012) and Rosulek (2015), we implement a broad classification of agent and recipient, as opposed to more fine-grained labels Halliday (1994) proposes (such as actor for the agent of material processes, senser for the agent of mental processes, etc). KRISDA CHAEMSAITHong & YoonjEong KIM Ibérica 39 (2020): 319-344342 Critical Linguistic Analysis in Media and Popular Culture. London: Continuum. Murphy, G. (1988). “Personal reference in English”. Language in Society 17: 317-349. O’Connor, P. (2000). Speaking of Crime: Narratives of Prisoners. Lincoln: University of Nebraska Press. Pennington, N. & R. Hastie (1991). “A cognitive theory of jury decision making: The story model”. Cordoza Law Review 13: 519-557. Powell, G. (2001). “Opening statements: The art of storytelling”. Stetson Law Review: 89-104. Rieke, R. & R. Stutman (1990). Communication in Legal Advocacy. Columbia, SC: University of South Carolina Press. Rosulek, L. (2008). “Manipulative silence and social representation in the closing arguments of a child sexual abuse case”. Text & Talk 28: 529-550. Rosulek, L. (2015). Dueling Discourses: The Construction of Reality in Closing Arguments. Oxford: Oxford University Press. Spieker, S. & D. Worthington (2003). “The influence of opening statement/closing argument organizational strategy on juror verdict and damage awards”. Law and Human Behavior 27: 437-456. Taylor, C. (1985). Human Agency and Language (A. Kozulin, Trans). Cambridge, MA: MIT Press. Tracy, K. (2011). “What’s in a name? Stance markers in oral argument about marriage law”. Discourse & Communication 5: 65-88. Van Leeuwen, T. (2008). Discourse and Practice: New Tools for Critical Discourse Analysis. Oxford: Oxford University Press. Warburton, I. (2003). “The commandeering of free will: Brainwashing as a legitimate defense”. Capital Defense Journal 16: 73-97. 4 Interestingly, to the best of our knowledge, the cases studied so far in an on-going project (over 10 cases) show similar lexico-grammatical choices and they all resulted in the defendant’s execution. However, there is one exception, where the defendant received multiple life sentences without parole (instead of execution), but the defendant in this case was tried as a juvenile (under 18 at the time of criminal violations), and the law in that jurisdiction prohibits execution. Thus, although the prosecution employs similar strategies in this case, it is the age factor that determined the final sentence choice. ACTIonS, ACToRS, AnD AgEnCY In THE PEnALTY PHASE oF CAPITAL TRIALS: A CoMPARISon oF Two gEnRES Ibérica 39 (2020): 319-344 343