Introduction

Increasingly, nations around the world have abolished capital punishment and adopted lifelong incarceration as the most severe means of punishment. However, in a small group of countries (including the US and Thailand, among others), the government still maintains a legitimate monology over the killing of an individual it has convicted, and the public’s view toward death penalty tends to be supportive. Thailand is a good case in point. Previous studies in the last four decades (from 1980 to 2020) that surveyed attitudes toward death penalty all find that the majority of Thai citizens are pro-death penalty, with “in favor” results ranging from 69% to 99% (Sritiravisarn 1980; Satjanit 1990; Kangsanthia 2000; Khwandee 2007; Sai-ngam 2020). Most of the informants believe that executions can prevent future crimes and protect against more crimes by repeat offenders. This is despite scholarly findings that executions are sometimes imposed on the innocent, carry high administrative costs, and do not deter crime (Nagin and Pepper 2012). To date, scholars have attempted to explain why popular support in certain countries remains high. One strand of research posits that certain characteristics, including regime type (democratic vs autocratic), religion, colonial heritage, and demographic composition, constitute major determinants of the adoption or abolition of capital punishment (e.g., Greenberg and West 2008; Anckar 2014). The other strand adopts a more socio-cultural view, proposing that support for legitimate killing is gained through the process of constructing shared public beliefs in everyday experience (e.g., Lesser Wendy (1998); Sarat et al. 2014; Yap and Tan 2020). To that end, executions are made to appear different from violence outside the law and, thus, are purified to create uninteresting, nonsadistic, administrative deaths (Sarat 2001). Part of this naturalized process involves the removal of executions from the public eye in the mid-20th century, so that citizens cannot raise questions about what they see and forget that they are killing people (Foucault 1977; Richards and Easter 1992; Linders 2002). However, despite this shift in visibility, “state killing lives in culture as a set of images and imaginings, a sight unseen except in the most mediated way,” and thus “we should attend to those images and imaginings, asking…what messages they convey…” (Sarat 2001, p. 188).

The present study aligns with the latter strand and proposes to examine an important mediated cultural source where state killing thrives, namely, newspaper reports on executions, which purport to cater to the public information about state killing. However, this study is distinct in that it investigates the issue from the lens of critical discourse analysis, which regards the centrality of language as a means of social construction, with an aim to unpack implicit linguistic strategies that position the reader to favor or disfavor the event and people depicted therein. The main argument of this study is that the press collaborates with the state in constructing shared beliefs about capital punishment by sanitizing state killing and making it appear different from the crimes committed by executed individuals. This is achieved through lexico-grammatical choices, which we term “representational strategies.” These choices deal with the selection of words to describe people, events, and speech that is congenial to the author’s ideological position.

The data set consists of a specially complied corpus of execution news reports, collected from quality Thai newspapers over a period of 20 years (1997 to 2017). These news outlets are not state-owned or funded and together enjoy a countrywide circulation, attracting domestic readership. One of the unique features of this corpus is that it spans the time when the country switched from firing squad to lethal injection, the latter being claimed to be a more “humane” method of killing. Another characteristic feature is that these texts do not openly engage in debate about capital punishment, and because of this, they constitute a fertile ground to scrutinize how subtle discursive choices are manipulated to influence the reader to take a particular view toward the executions. It is also interesting to find out whether (and if so, how) there are discursive changes which may be indicative of different ideological shifts with respect to crime and punishment in the two periods.

The study is guided by the following questions: 1) What are the forms and pragmatic effects (i.e., considering the linguistic choices and their meaning potentials in the context of use) of representational strategies with respect to the people, events and speech descriptions in these news reports?, 2) What, if any, are some shifts in the representational patterns when transitioning to the use of lethal injection?, and 3) What messages do these choices convey about capital punishment?

This study offers a number of contributions. First, it examines a group of news coverage that has received scarce scholarly attention (in comparison to crime reports, for instance), but likely plays an important role in supporting dominant ideologies about crime and punishment (i.e., that criminals are viewed as rational actors who bear individual responsibility for the crime committed). Second, while previous studies on the discourse of death penalty remain limited, and those that exist focus their attention on what public attitudes towards death penalty are, the current study’s focus on the detailed lexico-grammatical choices adds to the existing literature by bringing to light how these attitudes and the reality of crime are discursively shaped in otherwise objective reports. The demystification of linguistic choices helps emancipate offenders from alienating discursive practices that are cloaked “in the language or responsibility, censure and blame” (Loader and Sparks 2011, p. 104) and constitutes the first step toward promoting a socially inclusive and mutually respecful community. Third, the findings from the Thai context allow for inquiry into other contexts with respect to the roles of language in constructing deviant subjects on the one hand and legitimizing state power on the other.

This article begins by synthesizing lexico-grammatical concepts that constitute representational strategies and reviewing relevant literature, goes on to introduce the data under study and the analytical method employed, and presents the findings in detail.

Representational strategies: lexico-grammatical resources

The present study adopts a functional view of language, conceptualizing language as made up of “networks of interlocking options” (Halliday 1994, p. xiv). These options do not simply have syntagmatic and paradigmatic relations, but they are chosen according to the communicative purposes, ideological positions, and interests. They are thus value-laden and play a role in mediating reality (Fowler 1991). Three of these options, which form the focus of this paper, include how language users describe people (social actors), events and happenings, as well as the words of others.

Social actor representation deals with reference terms, mainly expressed through noun phrases. A comprehensive inventory is proposed by Van Leeuwen (2008) that distinguishes three major reference strategies (each with sub-categories): nomination, categorization, or impersonalization. Nomination is accomplished through naming, which varies in terms of formality, ranging from very formal (last name only), semi-formal (first and last), informal (first name only) to least formal (nickname). Titles may also be used to authorize the referent (such as “Correctional Supervisor”), while the removal of a title from a socially accredited person may de-authorize him or her (Hart 2014). Nomination highly personalizes the referent, as a person’s unique identity is presented (van Leeuwen 2008, p. 40).

Categorization, on the other hand, includes labels that describe various socio-cultural characteristics of the referent, which can be functionalization or identification. The former consists of referencing people in terms of what they do, that is, their role in the context, such as “police officer,” “executioner,” or “victim.” The latter, identification, describes social actors by way of what they more or less inherently are (van Leeuwen 2008, p. 42). This includes classification (e.g. age, ethnicity, gender, race), physical identification (e.g., size, color, weight) and relational identification (interpersonal relationships with others). Any of these choices singles out and foregrounds one identity aspect of the referent.

At the opposite end, impersonalization involves de-emphasizing the human characteristics of the referent. This option can conceal agency and evade responsibility. For example, when the object used stands for the user as in “The drug put the inmate to sleep,” attention is shifted away from the actor of the killing. Relatedly, a social actor may not be included in the text at all, treated as if non-existent. The syntax of pro-drop languages, including Thai, facilitates this option. However, in functional terms, whether the referent is included or not is treated as a strategic choice in communication. Palakornkul (1975), for example, shows that speakers opt for this strategy when they cannot identify the status or age of the interlocutor.

Descriptions of events and happenings are executed through verb phrases, also known as processes. In Halliday’s model (1994), six process types are distinguished, as shown in Table 1. An event can be represented as an action (material process); physiological and psychological response (behavioral process); cognitive state (mental process); expressive action (verbal process); relation between two entities in terms of possession, location or attribute (relational processes); or existing state (existential processes)

Table 1 Process types based on Halliday (1994).
Full size table

These choices have implications on agency and responsibility. Some processes highlight an agent and his or her power to affect others and bring about change in the world (e.g., material processes in the active voice), some convey less agentive power (e.g., behavioral processes), while others introduce a recipient of others’ action (e.g., as the object of an active material process clause or the subject of a passive voice clause). Previous studies (e.g., Trew 1979; Clark 1992; Teo 2000) show how news reporters position readers to blame certain groups by presenting them as agents of material processes. At the same time, agency can be concealed through existential processes or passivization.

Just as ideologically significant are verbal and mental processes, which allow language users to represent the words and ideas of others, also known as intertextuality. For instance, the choice of which words to quote and whose voices are included in the text can mediate the perception of the reader. It is argued that voices of powerless groups (such as ethnic minorities) are often excluded from the text (Teo 2000). Similarly, in the case of mental processes, because it is impossible to access others’ thoughts directly, a specific choice is always an act of interpretation on the part of the reporter to put those projected thoughts in the text. Stenvall’s analysis of “fear” in news reports on terrorism (2007) convincingly shows how news agencies often represent fears as powerful, free-floating entities, as in “British Airways cancelled a London-Washington flight on Friday as security fears grounded a US-bound plane for the seventh time in just over a week” (p. 215). Here not only is the agent who fears concealed, but the emotion is also constructed as the agent that has the power to ground the plane (instead of a human agent). This kind of representation undermines the factuality of the report and possibly allows the authorities to win the citizens’ acceptance for strict security measures. Linguistic choices such as the ones discussed above, together with their ideological functions, work subliminally and are not readily transparent, and it is these linguistic features that this work is concerned with (as opposed to explicit emotive and sensational use of language for which many newspapers are known for).

These theoretical concepts have been successfully applied to crime reporting in the press, and the current study finds the following insights particularly helpful. For instance, Clark (1992) reports “fiend naming” in the press, which refers to extreme negative labeling attributed to criminals, such as “monster.” Analyzing Australian news reports on drug dealing activities of a local minority gang, Teo (2000) finds that reporters overwhelmingly choose identification terms (including age and race terms) to color the perception of what offenders do. For example, the drug dealers in his study are homogenized through such identificational labels as “Vietnamese,” “Southeast Asian,” or “Asian kids” (Teo 2000, p. 17), while quotations are sought mostly from police officers or experts in criminology, thereby denying the ethnic community a chance to be heard. Along the same lines, Jewkes (2015) shows that in popular journalism, there is an oversimplified representation of criminals, emphasizing negative qualities such as selfishness, unconventional lifestyles, different ethnic characteristics, thereby reducing the causes of crime to mere individual pathology.

Examining a factual television crime show, Machin and Mayr (2013) find an emphasis on negative personalities of the offender through such evaluative expressions as “dangerous predator” or “strange individual.” In contrast, law enforcement officers are referenced in occupational terms (“detectives”). In addition, relational and mental processes also work to create an abnormal figure and a criminal mind, such as “is a strange person,” and “thought he was untouchable.” Comparing how the British and German press construct offenders and victims, Tabbert (2015) observes that both sources label the offenders through functional terms, emphasizing what they do (“attacker,” “gang,” or “rapist”) or their social roles (“husband”), while victims are recognized by identificational features, such as age, gender, or family relations. All in all, previous studies provide ample evidence that representational strategies can create textual persons for offenders and victims and popularize the current policies of criminalization: people commit crimes because “they” are not like “us,” and the society will be safer when “they” are punished (see Simon 2007 for the discussion on governing through crime).

The present study is also informed by a few previous studies on execution reports. Halmari (1999) examines 36 articles collected from a local newspaper in Huntsville (Texas, USA). The findings reveal “emotionally loaded” expressions that reference the executed in the headlines, such as “rapist” and “killer,” while the process of death is linguistically minimized to suggest that the execution is not painful, such as “closed his eyes,” or “gasped three times.” In a subsequent study, focusing on execution reports on one executed individual, Halmari and Ostman (2001) reveal recurring discursive choices in execution stories. Particularly relevant to this study is juxtaposition of the execution information (with the agent suppressed) and the crime committed, as in “Tucker, 38, is condemned to die by lethal injection for the pickax deaths of two people” (p. 815). Headlines also feature pro-death penalty by way of quotes from local people, as in “Survey says Karla Faye should die” (p. 810). In this way, this newspaper overtly reports objective facts, but implicitly manipulates the reader to take a pro-death view. Suriya (2011) examines a group of Thai execution news from 1998 to 2009. Particularly relevant to this study are lnexical selection (noun phrases and action verbs) and intertextuality. It is revealed that these devices represent the death row inmates as law breakers, ruthless and incompetent people, while the executions are represented as legal and humane. However, the study does not consider other possible forms of person reference (such as agent omission or suppression), along with co-occurring lexical strategies (e.g., titles, nicknames, or various manifestations of impersonalization), or transitivity choices and the way in which they are intertwined with agency and responsibility. Nor does it take stock of the voices that are included and discuss the functions of those voices. While these previous studies have provided a number of invaluable insights, the current study departs from its predecessors in that it takes a comparative, diachronic approach to examine a larger pool of data and, by way of linguistically-grounded typologies, attends to not only all possible forms of the linguistic choices in question (the “what”) but also the way in which they position the reader to support death penalty (the “how”).

Data and methodology

The corpus for this study comprises two groups of reports, collected from major daily quality newspapers archived at the National Library of Thailand. The reports are in Thai and English, and, in cases of the Thai data, direct translation was carried out and presented here to preserve the lexico-grammatical choices of the source language and cross-checked by two language experts. For the purposes of this study, this process should not present any problems, as the study’s main concern lies in the ideational metafunction of language (i.e., relating experiences and ideas), which is by and large universal. Any significant lexico-semantic departures from English will be discussed and the original provided, where relevant.

The reports can be classified into two groups. One group features 69 articles written between 1997 and 2002 (Period 1), during which firing squad was the method of execution. The other group features 39 articles written between 2003 and 2017 (Period 2), after lethal injection was adopted as the sole means of execution. The rationale for these two groups of reports is that we wish to compare the representational choices over time and see whether changes in the execution method are intertwined with discursive choices. Despite the difference in data size between the two periods, we do not view this issue as necessarily affecting the findings. Quantitative findings (i.e., frequency counts) will be discussed in percentage terms, relative to the total frequency counts in each period. At this point, some contextualization of the data is in order.

The firing squad method in Thailand involved setting death row inmates behind a screen, tying them to a pole, with their eyes blindfolded and a bunch of flowers, a candle, and an incense stick in their hands as part of the final rite. A target was then painted on the screen, so as to allow the executioner to pull the trigger and hit the heart without having to see the person. Some of the bullets were reported to have missed the heart, prompting a second round of firing and causing great pain and suffering to the executed. Lethal injection, on the other hand, is believed to be a more “humane” and less “painful” practice. The authorities reason that the body is not damaged, the drugs (consisting of a slee** medicine, muscle relaxant, and heart stopper) work quickly and effectively as intended, and that the execution takes place in an air-conditioned room, kept away from the public (Chulachain et al. 2005, p. 35). However, studies reveal that problems abound with this method, ranging from poorly qualified or unconcerned execution personnel and inadequate amount of drugs administered to lack of proper monitoring once the process starts (Koniaris et al. 2005; Heath 2007). Some medical scholars go so far as to argue that there is no evidence to support the use of lethal injection, be it in terms of the order in which the drugs are administered or the dosage (Zimmers et al. 2007). Suffice it to say that although touted as the most humane method, lethal injection is fraught with questionable practical and moral issues (see Sarat (2002) for a thorough discussion of lethal injection).

For the purposes of this project, we excluded those articles that are not execution reports proper (e.g., editorials). Each article tells of the execution performed on a previous day. By and large, they feature one executed individual, although in some cases, up to five executions were performed on the same day and subsequently reported. As can be seen in the breakdown of the data in Table 2, there are different reports in the corpus on the same executed individuals. We consider this a distinct advantage, as we can compare different news outlets with respect to their linguistic choices.

Table 2 Overview of the corpus.
Full size table

With respect to the analytical steps, the researchers first identified referential expressions for the major social actors in the reports (i.e., the executed individuals, executioners, and victims), both in headlines and content, using van Leeuwen’s inventory discussed above. Every representational choice was taken into account. For example, “a six-grade student aged 5 years old” consists of 3 descriptions: six-grade (functional term), student (functional term), and 5 years old (identification term). Next, based on Halliday (1994), the process choices involving these social actors were considered. For mental and verbal processes, we took stock of whose voices were included (and excluded), and the functions of these voices in context. Relevant grammatical operations were also marked (e.g., nominalization or passivization). These steps were carried out independently by each researcher, and any coding discrepancies were discussed and resolved subsequently. Finally, we also paid attention to authorial evaluation (i.e., explicit signals of attitudes and stances) and observe how they interact with representation choices.

Findings

We organize the findings in two sections: representational choices in the headlines and content. The focus is on the similarities and differences between the two periods under study.

Headlines

Finding 1

In both periods, the executed constitute the main element, predominantly represented in functional terms (93% and 95% of the headlines in Period 1 and Period 2, respectively), while the executioners are either excluded or impersonalized (98% and 100%, respectively). The victims only appear in sensational crimes (6% and 2%, respectively).

As observed by Suriya (2011), the headlines foreground the executed individuals through legal categories that convey their guilt (e.g., “prisoner,” “convicted inmate”) or their roles in the crime (e.g., “killer,” “drug dealer”). In contrast, the executioner is backgrounded and, in my cases, excluded. This suppression is what Huckin (2002) refers to as “manipulative textual silence.” For example, in (1), the executioner is simply omitted from the subject position, leaving only the verb in the initial position, while the executed and the victim are left intact. In (2), the executioner is represented metonymically in the name of the institution for which he works, while the executed is represented through a title and a functional term. An important effect of this is that the state’s desire to kill a person is purified and linguistically transformed into a routine procedure.

  1. (1)

    [agent omitted] Executes the inmate charged with murdering-burning little kid (06/11/1998, Thairath)

  2. (2)

    Bangkwang Prison fires Sergeant Major Gunman. (30/10/1998, Khaosod)

    The contrast between the offender and executioner continues in Period 2. Note the agent is mystified by way of nominalization (e.g., lethal injection) or omission:

  3. (3)

    4 inmates executed by lethal injection for the first time (13/12/2003, Dailynews)

  4. (4)

    [agent omitted] Administer drugs to execute 2 inmates on meth charges (25/08/2009, Neawna)

  5. (5)

    Lethal injection for two ya ba dealers (25/08/2009, The Nation)

It is worth pointing out that in all of these examples, the crimes that the executed commit are also explicitly or implicitly conveyed. In (2), for example, the functional term “gunman” implies that the executed is a contract killer (which is unexpected for a “sergeant major”), while in (5), the functional term “ya ba dealers” suggests that the individual is executed because he is a drug dealer (methamphetamine is commonly known as “ya ba” in Thai culture).

As for the victims, it appears that those who do not fit the “ideal” victim stereotype (Christie 1986) are not represented in the headlines, including cases where there are no direct victims, such as drug trafficking. People who are vulnerable (both young and old) or publicly known tend to be highlighted in the headlines. In these cases, nicknames and/or identifications (size and age) are often supplied that not only generate sympathy but also highlight the executed’s transgression of social norms. For instance, we see earlier the use of “little” in (1) and “eleventh-grader” in (4), which foreground the helplessness of the victims. Below are more examples:

  1. (5)

    Child-killer put to death. (19/06/2018, Bangkok Post)

  2. (6)

    Executes Pun Saithong, lusty murderer who rapes and kills Sister Oom (22/06/1999, Dailynews)

In (5), the victim and the offender are lexically compounded in one functional term: “child-killer,” emphasizing that the offender is not just any killer, but a child killer. In (6), note the stark contrast between the victim and the offender. The victim is referenced by way of a nickname, prefaced by the kinship term “sister,” which establishes an interpersonal tie with the reader, whereas the offender’s name is untitled and evaluated negatively in the subsequent appositive “lusty murderer.” Note that, in Thai society, nicknames are common among intimate group members, and it is precisely the fact that nicknames are never observed as reference terms for the offenders that pragmatically frames them as outsiders.

Finding 2

In Period 1, reporters’ attitudes toward the executed are expressed explicitly. However, explicitly evaluative expressions are on the decline in Period 2.

With explicit negative evaluation, the executed are treated as societal outsiders and, in many cases, disturbingly dehumanized. For instance, the evaluative expression “brutal” is common (as in 7), and so is the distal deixis “that” before a nomination (as in 8 and 9), which is pejorative and contemptuous (comparable to Latin iste). What is more, the executed’s human quality is dismissed, so that what gets executed is a legal case (“second case”), animal (“savage”), or a weapon/organ (“spatha,” “the hand”). By pragmatically positioning the executed at a distance from the reader and disqualifying them from empathy, executions are legitimized.

  1. (7)

    Fires a brutal child murderer; Bangkwang checkmates the second case (06/11/1998, Thai Post)

  2. (8)

    Execute that (scroundrel) savage ra** and killing 2-year-old kid inhumanely (24/08/2001, Dailynews)

  3. (9)

    Bangkwang executes 5 male inmates, that (scoundrel) spatha, the hand that killed Tung Hua to pay for his karma (09/10/99, Khaosod)

In Period 2, only 2% of the headlines register explicit evaluative expressions, illustrated in (9). The rest of the headlines employ impersonalization, as in (10):

  1. (9)

    Department of Corrections performs lethal injection on brutal eleventh-grader killing case for the first time in 9 years (19/06/2018, Matichon)

  2. (10)

    Administers drugs to execute 2 male inmates of meth charges, 1 million dollars seized (25/08/2009, Khaosod)

However, the absence of explicit evaluation does not mean that the press’s attitude toward capital punishment becomes less supportive. Interestingly, in 85% of the headlines of Period 2, we find adverbials that qualify the frequency of the execution:

  1. (11)

    Lethal injection performed for the first time in 9 years (19/06/2018, Kom Chud Luk)

  2. (12)

    Give a shot to execute 2 male inmates selling drugs for the 2nd time in 6 years (25/08/2009, Kom Chud Luk)

  3. (13)

    Give a shot to execute 2 drug-selling prisoners, 5th and 6th cases after abortion of squad firing [in 2003] (25/07/2009, Ban Meung)

  4. (14)

    Execution (for real) in almost 9 years (20/06/2018, Neawna)

Note how the prepositional phrases in the above examples, such as “for xth time in y years,” or “after abortion of squad firing,” set up a scope for the frequency of the events in question. Something that happens once or twice in six or nine years, or the 5th and 6th executions within a six-year period (from 2003), is not likely to be perceived as a common event. What these adverbials do is portray executions as infrequent and selectively imposed on a few people, thereby not only creating newsworthiness for the events but also implicitly authorizing the executions.

Finding 3

In both periods, non-violent material processes are predominantly used to represent state killing (with a suppressed or institutionalized agent).

The material processes chosen for state killing belong to a small group of lexical items, as shown in Table 3. In Period 1, the formal lexical choice “to execute” comes to the fore. Other frequent expressions include “to target-shoot,” which conveys precision, and the onomatopoeic verb “bang-bang,” which imitates the sound of a mechanical gun. As a result, the intent and action of killing are eliminated. In addition, “for display,” “live show,” or “settle accounts with” treats the condemned as objects to be eliminated from society and portrays death as a way to deter future crime. In Period 2, the headlines almost invariably make use of “give a shot to execute X,” thereby turning the execution into a medical procedure. While we do find a few instances of death-related expressions (such as “lethal injection for two ya ba dealers,” “child killer put to death,” and “inject poison into the body”), the sense of killing a human is missing because the verb is nominalized and/or the recipient is not a person but a legal case.

Table 3 Material processes that represent execution in the two periods.
Full size table

All in all, in both periods, most of the headlines feature lexical choices that suggest that state killing is a bureaucratic procedure. Compassion for human life or any tint of violence involved in the process of firing (or lethal injection) is removed from these verbal choices.

Finding 4

In both periods, behavioral, mental processes, and relational processes are used to represent the executed’s reactions before and during the execution.

In addition to the offenses, the reader is also allowed a glimpse of how the executed react in the headlines. However, the descriptions are quite different in the two periods. In Period 1, negative reactions appear to be the news value:

  1. (15)

    Fires a round of bullets, but does not die and screams loudly (7/11/97, Thairath)

  2. (16)

    Executes female inmate. Needs to fire 2 rounds [for her] to pass (24/11/99, Siamrath)

The underlined processes in (15–16) construct the executed persons as recalcitrant subjects. In (15), the contrastive “but” cancels the implicature in the previous clause, while the deontic modal “need to” in (16) implies extra effort to kill. As a result, the headlines in these periods emphasize the unexpectedness of the inmates’ survival, when in fact, the barbaric nature and the botched procedure could have been pointed out to the public. Contrast this example with headlines from Period 2:

  1. (17)

    Injects poison to execute 4 inmates. It only takes 25 minutes! to take their last breath and die peacefully (13/12/2003, Thairath)

  2. (18)

    Administer drugs to execute 4 prisoners for the first time. All goes well—pass away in 35 minutes (13/12/2003, Matichon)

Together with other contextual cues (e.g., “only take 25 minutes,” “all goes well,” and “in 35 minutes”), these behavioral processes imply that the execution is a quick and painless one, and there is no problem with this method of killing.

Sometimes we find the executed’s reactions as part of a reported clause in Period 2, as signaled by the verb “revealed” in (19–20):

  1. (19)

    Final moments of the death row inmates revealed: asks to touch a monk’s robe to send his soul to heaven! (26/08/2009, Manager)

  2. (20)

    Execution moments revealed: the inmate is sad and sheds tears (26/08/2009, Dailynews)

In these instances, the reports prioritize the privileged access to the death chamber, granted to a few witnesses. Like (17–18) above, the procedure is represented as uneventful with the executed seemingly accepting the impending death. All in all, the reactions or emotions reported in both periods convey that, despite some unexpected technical issues, the executions are carried out mostly successfully, and they are effective in causing fear for the offenders and, by extension, future offenders.

Representational practice in the content

Similar representational patterns are observed in the content. That is to say, in both periods, the emphasis is on the condemned with respect to their identity, past actions, and reactions in the death chamber. However, these elements are backgrounded for the executioners. The following are characteristics unique to the content.

Finding 5

In both periods, there is overlexicalization and repetition for the condemned and victims.

In the content, the executed are repeatedly represented through the following formula: “(convicted) (fe/male) inmate + first name + (last name).” In fact, this choice by and large constitutes the sole reference choice in both periods (95% in the first period and 97% in the second). While it is an objective fact that the referents are convicted inmates, we take issue with its use for a number of reasons.

  1. (21)

    After the injection, convicted prisoner Teerasak recited Arabic prayer because convicted prisoner Teerasak was a Muslim. He mumbled the prayers, without saying out loudly up until the drugs took effect. Convicted prisoner Teerasak passed away peacefully. (19/06/2018, Khaosod)

Here the title “convicted prisoner” is used to preface the name three times in a short stretch of text, when other choices, such as “Mr. + name” would pragmatically suffice for subsequent mentions.

Like the condemned, the victims tend to be overlexicalized to create victimhood and damage magnitude:

  1. (22)

    The said convicted prisoner attacked and extorted personal property from the victim and stabbed the dead person 24 times altogether, causing the victim to die (19/06/2018, Manager)

Here three referents are in place successively. However, the second referent “the dead person” is pragmatically infelicitous: when the attack takes place, the victim is not alive. This is evident in the subsequent clause: “causing the victim to die.”

Finding 6

In Period 2, referents to the executioners become scarce. In particular, nominations are absent.

While a great deal of information is supplied about the executed or the victims, little is known about the executioner. As a result, the executioners remain aloof from the reader, and the relationship between the executioner and the executed emerges as a professional one, when in fact, it is a violent one and requires close contact, particularly when the substances are injected. In Period 1, the identity of the executioners is explicitly exposed in the reports. Their first and last names, as well as professional ranks, are spelled out:

  1. (23)

    In the death chamber, the executing officers are Mr. Chaowares Jaruboon, First Executioner, and Mr. Sanan Boonloi, Second Executioner, were getting ready at the scaffold, with high-rank officers, consisting of [Mr. + name], Prison Commander… (08/11/1997, Matichon)

However, in Period 2, this personalized representation does not occur in any reports. Only generalized functional terms are used, as exemplified in (24):

  1. (24)

    The person undertaking this new execution method is a prison officer from the Department of Corrections who was well trained by officers from the Department of Health and who will replace Mr. Chaowares Jaruboon, the last firing executioner… (13/12/2003, Dailynews)

Note in (24) the contrast between the representation of the two executioners. One is anonymized and referenced through a non-specific noun phrase “a prison officer from the Dept. of Corrections,” while the other is personalized with the first and last name.

While it is not possible to state precisely the motivation behind this shift, a possible reason is that the change is in line with the international standards. For example, in the US, the executioners’ identity is confidential (Roko 2007). Regardless of the motivation (which can be multi-factorial), what this concealment or anonymization of the executioners means in practice is that public scrutiny of the executioner’s past execution records and qualifications (i.e., whether they are “well trained” as claimed) is not possible. This is particularly important in cases when lethal injection is ineptly administered, the revelation of the identity of the executioners would allow the public to assess the qualifications of the executioners and the implementation of the procedure, if need be. The press could have served as a responsible agent and performed this essential check and balance function in a democracy, at least with respect to the representational practice of the executioners.

Finding 7

In both periods, the executed’s actions are predominantly conveyed through (effectual or unlawful) material processes (89% in the first period and 91% in the second). Other process types are also observed (to be discussed in Finding 8). In contrast, the executioner’s processes are either ineffectual or lacking in agency, and no other process types are observed.

A staple in all the reports is that the condemned’s past deeds are described in detail through material processes as acting upon a victim or committing unlawful deeds (e.g., selling narcotics). In (25), the underlined material processes produce impact on the victims and display malicious intent; the rest are behavioral and mental processes that construct him as deviant. In (26), the material processes “commit” and “sell” (as well as the relational process “have”) are attributed to the executed individuals.

  1. (25)

    The last inmate was convicted male inmate Tapoiho or Tongyoon or Boy, aged 51 years old, Burmese, was given a death sentence for killing his wife and 2 nephews by chop** with a hatchet, as he was angered after his wife kicked him out of the house because he did not behave well. This defendant also crossed the border illegally and killed a border-patrol officer. He just got out of jail in 1977 but committed a crime again. (24/11/1999, Siamrath)

  2. (26)

    At Bangkwang Prison (Nonthaburi), on 24 August, Mr. Natti Jitsawag, Director of the Corrections Department, reveals that he has received the order from the Prime Minister to execute two male inmates, namely, convicted male inmate Jirawat Pumpruk, aged 45 years old, and convicted male inmate Bundit Jaroenwanit, aged 52 years old. Both committed the unlawful deed of having and selling methamphetamine pills. (25/08/2009, Dailynews)

In contrast, the descriptions of the executioner’s actions consist of details of the weapon used and the number of bullets. The executioners simply go on duty and undertook the killing task in a precisely trained manner, with the bullets simply working their way, as exemplified in (27). However, in Period 2, in line with Finding 6, rarely do we see an actor involved in the process of the execution. In (28), that the agent of the material processes is the “prison,” while in (29), the verb is omitted, and the passive voice and nominalization are used, all of which grammatically remove the agent:

  1. (27)

    When the officers took male inmate Supoj to the scaffold, and Mr. Chaowares Jaruboon, First Executioner of Bangkwang Prison, assumed position and took an aim, male inmate Supoj shouted, “It’s time to shoot.” A round of continuous gunshots broke out, amounting to 10 shots. After the end of the gunshots, male inmate Supoj tilted his neck toward the scaffold and died instantly (30/11/1998, Matichon)

  2. (28)

    As for the process in the death chamber, the prison kept the inmates shackled and had them lie on the execution table, on which a white cloth for covering the body is placed, and secured the inmates; arms to the table in a cross position, and stuck the needles into both writs and gave three shots…(25/08/2009, Khaosod)

  3. (29)

    Both convicts remained shackled inside the injection chamber when they lay down on the beds. Their arms were stretched out, with three syringes [verb omitted] beside each man. The first contained a soporific drug… The injection process took about 25 minutes, after which the doctor…checked whether the convicts were dead. (25/08/2009, The Nation)

In these examples, the executioner (or the institution) behaves strategically and professionally, which is in contrast to the unlawful deeds by the executed.

Finding 8

In both periods, voices of the powerful are overwhelmingly incorporated, while voices of the condemned and death penalty opponents are marginalized.

The distribution of the voices quoted in the texts is shown in Table 4. In both periods, state personnel constitutes the majority. Perhaps a unique characteristic of the reports from the Thai context is the inclusion of the voices of monks, who are invited to give a last sermon to the condemned prior to the execution. Next are execution witnesses, all of whom are reporters and whose voices frame the narrative of the execution, for example, “Reporters reveal that prior to the execution….” Least frequent are the voices of death penalty opponents.

Table 4 Distribution of voices.
Full size table

Interestingly, these voices perform distinct functions. When state personnel and monks are quoted, their messages are clearly in support of the system in use, including its utility for future crime deterrence (30) or unproblematic and successful implementation (31). Likewise, the views of the monks (32), with whom the executed spend their last moments, show approval of the execution as an appropriate punishment. Essentially, the content all empowers the execution as well as capital punishment and confirm its validity.

  1. (30)

    Director of the Corrections Department added that…while many countries have abolished capital punishment, there are other countries that retain the system like Thailand, such as the US and China, which prioritize the protection of society and the majority of the citizens from becoming crime victims over the protection of the rights of the offender. Therefore, the Corrections Department hopes that this execution will be a cautionary lesson for those who think of committing a serious crime or violate the law and remind them of this punishment. (19/06/2018, Dailynews)

  2. (31)

    Mr. Pittaya Sungnakin, Director of Bangkwang Prison, said that the prisoner is in the process of constructing a new execution chamber with 0.09 acre, featuring 2 execution tables and…With respect to today’s first lethal execution, it went well. There was no problem, whether it is the drugs or witnesses; everything was prepared. It can be held that this method of execution is not painful, involves no blood shedding, which is way better than firing squad. I believe that we will be completely prepared in the near future. (14/12/2003, Khaosod)

  3. (32)

    As for Phra Khru [monk’s rank comparable to a bishop] Srinonthawat, head of Bangpraektai temple, who was the only monk that gave the final sermon to the two inmates before the execution, revealed that during the lethal injection, convicted male inmate Bundit was sad and down, with tears rolling down the cheeks all the time, whereas convicted male inmate Jirawat remained still in silence, before asking to touch his robe so that the soul is blessed to go to heaven. “I myself also feel pity and sympathize with the fate of the two individuals. But it is a matter of karma. One must reap what one sows,” Phra Khru Srinonthawat said. (26/08/2009, Manager)

When the executed is given a voice, it is only for the purpose of dramatizing the execution narrative. The quotes in (33) and (34) illustrate this well. It is as if the reports are meant to tell the readers not to commit crimes, so they do not have to have their last words with their loved ones.

  1. (33)

    Male inmate Panom Thongchanglek, aged 32 years old, a contract kill from Chumporn, was waiting, as if accepting his fate. He was the only convicted death row inmate in this room. “Bring it on, Sir, I’m ready,” he said. “Sir, will I get to say goodbye to my parents?,”

    Panom asked. (15/12/2003, Thairath)

  2. (34)

    [Before the executions], they were allowed about an hour to write a will or call their loved ones. Jirawat called his mother, who had met him earlier in the day during a visit. He told her that he was going to be executed and asked her not to cry. He then spoke to his daughter, asking her to be a good girl. “Love you” was his last words to his daughter. (25/08/ 2009, The Nation)

Only one report features the voices of capital punishment opponents. In this instance, the quotations criticize the system in place and arguably provide a balanced view of capital punishment:

  1. (35)

    Amnesty International rejected execution. Srirak Plipat, its Thailand director, said there was no evidence that convicts felt less pain and that execution reduced crime. “It’s just an illusion that society becomes more peaceful with execution,” he said, adding that countries which returned to execution such as Canada witnessed rising crime rates. (13/12/2003, Bangkok Post)

Unlike some reference terms, none of these quotations is derogatory or disturbing. However, considering the distribution and the functional pattern of the voices, the issue is that the reports appear to be more interested in seeking the perspectives and opinions of certain groups, as though it is only this group that can offer insightful or valuable information. The other groups, including the executed and opponents of capital punishment are marginalized. These voices are denied the chance to be heard and understood from “their” perspectives, when in fact they may invalidate the premises that the execution is reliable, effective, and humane. Some executed individuals may insist on their innocence, some may oppose death as an unjust form of punishment, while others may state that they deserve to be punished (Schuck and Ward 2008).

Conclusion

This study seeks to explore how execution news in the Thai press discursively purifies state killing and implicitly promote death. Firstly, the study examines the forms of the representational choices for the major social actors. The findings show that the reference terms and action choices are starkly different for these social actors. The executioners are represented professionally, while the executed individuals are de-humanized and de-personalized. The selection of spoken words and speakers is also significantly contrastive. The reports embrace the points of view of those in the legal and judicial system and rarely incorporate the voices of the death row inmates. Secondly, some shifts in the representational practice patterns are detected. For example, the identity of the executioner is backgrounded in Period 2, and evaluation for the executed individuals appears to be less explicit in Period 2. Despite these differences, and this leads to the third major point, the study shows that person reference, process choices, and quotation patterns aid in the process of purifying capital punishment. The critical message put forth by this study is that: in the periods examined, the press mediates the perceptions of offenders, executioners, and victims by manipulating linguistic resources to bar the reader from recognizing the violence of the execution while aggrandizing the violence committed by the condemned. Most likely, this practice would yield no positive outcomes, be it prevention of future crimes, serving justice, or creating an informed public. Note that “manipulation” here does not necessarily imply intentionality, but it is more about the meaning potential constructed in the discourse of these news texts.

Overall, the findings based on this study appear to be consistent with the patterns identified in previous studies, discussed in the literature review. Just as in general crime reporting, social actors are represented unambiguously in execution reports as well. Offenders and victims are labeled in ways that reinforce respective stereotypes, making them “ideal” offenders and victims. In many cases, explicit evaluative expressions are chosen to describe these individuals, revealing authorial attitudes that are extremely hostile and void of empathy. At the same time, reporters also stress the violence and cruelty of the crime by singling out aspects of identity of the victims through representational practice (such as youth, vulnerability, or social status). In contrast, law enforcement and judicial officers are portrayed as acting professionally and dispassionately to uphold the law, and the executions as a non-violent, routine procedure successfully implemented.

This study also offers insights from a comparative perspective. We have observed some changes in the representational practice over time. After the adoption of lethal injection, execution reports become less explicitly evaluative with respect to the naming and process choices for the executed and the victims. In addition, the identity of the executioners in the second period is withheld from the reports, shifting from personalization to anonymization (or impersonal institutionalization). This, however, does not suggest an ideological change. The executed are still viewed as incorrigible offenders who deserve to be punished for what they do and for the safety of society. Quotation patterns, which remain constant over time, also testify to this pro-death stance: certain groups whose views are uncritically in support of capital punishment are quoted, while dissenting opinions are virtually excluded. The perspectives of the executed are useful only in terms of creating a vivid reconstruction of the executions. Compared with previous studies on execution reports (Halmari 1999; Halmari and Ostman 2001), this stereotypical representation of offenders and victims appears not to be a culture-specific discourse pattern, but a cross-cultural one. That is, offenders are represented as outsiders from the mainstream society, without any information that would helpfully contextualize the offense or humanize them, while victims, especially vulnerable victims of sex crimes, receive attention and are represented through linguistic choices that call for sympathy and compassion.

Sontag (2004) convincingly argues that the media encourage the consumer to think about the world in terms of memorable events rather than lengthy complex processes. This appears to be true of the execution reports under study, as evidenced in the representational choices. Reporters create narratives in which an offender commits legal violations (with victimization in some cases), and commensurate punishment is implemented as a result. The reports analyzed in this study do not leave room for the possibility, for example, that offenders are often victims too (e.g., unequal access to resources, illness, or discriminatory practice), or that executions can be botched.

It is expected capital punishment will remain in place for years to come. However, for a fundamental social change to occur, this discourse-based study takes one small step in that direction. By deconstructing linguistic choices in execution news, it raises important questions about the representational practice in these reports and makes transparent the processes that enter into the construction of the narratives of executions, thereby laying the groundwork for more informed views of executed individuals and the justice system in place.