Evaluative stancetaking in courtroom opening statements

1 Introduction

Recent research has embraced a view of legal discourse as evaluative and interpersonal rather than merely objective and impersonal, unveiling how legal professionals intervene in their texts not only to present information but also to compare, contrast, and evaluate their claims against their opponents’, and build solidarity with the targeted audience (e. g., Hobbs 2003, 2008; Tracy 2011; Cotterill 2010; Chaemsaithong 2012). According to Breeze (2011), legal discourse may be characterized by its reliance on evaluative expressions, notably adjective/adverbs such as correct(ly), reasonable/-ly, and appropriate(ly). These expressions reflect the epistemological premises and disciplinary values that operate within this discourse community.

Indeed, evaluation has been found to be prominent in various genres of legal discourse: expert witness testimony (Chaemsaithong 2012); direct and cross examination (Hobbs 2003; Cotterill 2003; Heffer 2007); lawyer–client meetings (Kozin 2008), judicial summation (Henning 1999), and written opinions of district and appellate courts (Finegan 2010; Hinkle et al. 2012). It could be argued that what legal professionals mainly do is evaluate, and that legal communication, like other social interactions, crucially depends on evaluation as a semiotic resource for interlocutors to negotiate and manage the frame for interpreting the ongoing discourse. The resulting texts reflect this concern and are pervaded by evaluative language.

One particular genre of legal discourse where evaluation is likely to be present is the opening statement, that is, an introductory speech made by the attorneys for each side at the start of a trial. In early opening statements (1759–1789), attitude markers are used to interact with the jurors (Chaemsaithong 2014). Studying legal narratives, Harris (2005: 222) also notices that, in the opening statement, lawyers make extensive use of “those very resources of belief, opinion, intent, and subjective evaluation which the rules of evidence prohibit”. It is these stance resources that transform a referential account into an “evaluated point-laden narrative” (Toolan 2001: 238). These observations warrant further investigation, as they stand in stark contrast to the legal requirement that lawyers may “do no more than to inform the jury in a general way of the nature of the action and defense” (Best v. District of Columbia 291 U.S. 411, 54 S. Ct. 487, 78 L. Ed. 882 [1934]; my emphasis).

Until now, the genre of the opening statement has been understudied in the literature; past studies have focused on specific stance markers, which in turn limits the strength of the claim that legal discourse is evaluative and interpersonal. Thus, while the idea that lawyers do not simply present facts is not entirely new (see, for example, Rieke and Stutman 1990), it is the aim of this study to examine the discourse moves that convey evaluation in the context of opening statements. To that end, the study will explicate the process of stancetaking by qualitatively and quantitatively analyzing five high-profile Anglo-American trials. Two research questions that particularly guide this investigation are: (i) How do lawyers enact stance in the opening statement? and (ii) How do these stance resources contribute to lawyers’ storytelling and the construction and maintenance of their relationships with the jurors? As will be shown shortly, expressions displaying authorial stance permeate the opening statement of the defense as well as the prosecution, with five key features showing similar frequency patterns. Of these features, linguistic devices that signal commitment or lack thereof (i. e., boosters and hedges) appear to occur frequently and outnumber explicit attitude markers, which are low-frequent. These devices thus significantly create what Amsterdam and Brunner (2000) term “implications” in lawyers’ stories, which in turn function to shield the speaking lawyer from “the risk of refutation by stopping millimeters short of assertion, and from the risk of backfire by embroidering with equivocations at the margins of commitment” (Amsterdam and Brunner 2000: 14). It is in large part these value-laden linguistic choices that enable a lawyer to construct an “interpretative frame” for the silent audience (Amsterdam and Brunner 2000: 172). Ultimately, these resources may influence jurors to draw at least tentative conclusions at this initial stage (Lind and Ke 1985; Pennington and Hastie 1991; Spiecker and Worthington 2003).

This article begins with a description of the discursive characteristics of the opening statement, and proceeds to discuss the concept of stancetaking and the corpus. It then presents the findings in details, and concludes with some observations on the use of authorial stance in this genre.

2 The opening statement: Discursive characteristics

The opening statement is the first opportunity for the trier of fact to hear a comprehensive statement of each party’s factual claims; hence it is counsel’s first interaction with the jury. This phase of the trial starts with the party that has the burden of proof (i. e., the plaintiff’s attorney in a civil trial or the prosecution in a criminal trial), and is then followed by the defense’s presentation. In the adversarial system, when the defense presents their case, they will try to construct facts that suggest either a counter-narrative or the weakness of the prosecution narrative. The narrative construction in this initial phase of the trial will later influence the type of facts that counsel will try to construct from the witness examination and that the defense will try to deconstruct in their cross-examinations.

Generally speaking, the opening statement allows attorneys from either side to introduce themselves and the parties involved in the lawsuit, outline the important facts of the case in the form of a narrative, explain the applicable law, and make a request for a verdict. However, what makes the discourse of the opening statement particularly peculiar is its dual nature. The official website of the US federal courts states that “although opening statements should be as persuasive as possible, they should not include arguments” (Administrative Office of the US Courts 2014). This part of the trial is, in principle, limited to outlining facts and intended to be informative, rather than argumentative. What this means in practice is that, to create a successful opening statement, lawyers need to stimulate the interest of jurors, build rapport with them, and, at the same time, come close to being argumentative (Tanford 2009: 147). They can, for instance, offer a discussion of the anticipated evidence and “facts” they intend to prove, but they may not assert personal opinions, comment on the evidence, explain how a law may be applied to the facts, or arouse the emotions of the jurors by making negative judgments about the other party in scurrilous terms. In reality, however, such limits are hard to enforce straightforwardly, and are usually left to the discretion of the trial judge. Cotterill (2003: 68–83) demonstrates this well in a corpus analysis of lexical choices made in the opening statements of the OJ Simpson criminal trial. Strategic word choices are found to help prime the jury into viewing the events and participants in a given way. Taking this as a point of departure, the current study seeks to investigate the ways in which stance resources allow lawyers to be persuasive while appearing non-argumentative.

The discursive characteristics and constraints of the opening statement are different from those of witness examination in a number of ways. First, instead of being a dyadic interactional situation, the opening statement is a monologue delivered to a silent, overhearing audience with no interruptions by other participants (except when opposing counsel raises objections for possible misconduct). Second, while witness examination is a jointly negotiated discourse created by the lawyer and the witnesses, the opening statement can be argued to manifest the lawyers’ complete control over the linguistic choices, thereby testifying to lawyers’ pragmatic awareness of the audience and their needs. Third, the opening statement is directed specifically to jurors, instead of just being a presentation in their presence. Finally, the narrative in the opening statement is in large part woven into one unified speech, as opposed to fragmented stories (Harris 2005). These characteristics reveal that the opening address is a phase where the lawyers of either side are highly motivated to use stance resources allowing them to align and negotiate their positions with the jurors for a favorable verdict. It is this stancetaking process and its pragmatic effects that will be explored in subsequent sections.

3 Evaluative stancetaking and legal discourse

From a discourse-pragmatic perspective, the process of stancetaking has been investigated under several labels. To a large extent, those labels refer to the same phenomenon. Two of the most widely used terms are evaluation (Hunston and Thompson 2000; Camiciotti and Bonelli 2004; Bednarek 2006; Holmgreen and Vestergaard 2009) and stance (Biber and Finegan 1988, 1989; Hyland 2005; Englebretson 2007a; Baratta 2009; Johnstone 2009). Other terms that have also been used include appraisal (Martin 2000; Martin and White 2005), voice (Ivanic and Camps 2001; Hyland 2008), and point of view (Simpson 1993).

Among the first to systematically investigate stance expressions were Biber and Finegan (1988: 12), who write: “[b]y stance we mean the lexical and grammatical expression of attitudes, feelings, judgments, or commitment concerning the propositional content of a message”. In a later study, epistemic commitment is more broadly subsumed under assessments: “personal feelings, attitudes, value judgments, or assessments” (Biber et al. 1999: 966). Thus, their framework covers epistemic expressions (i. e., expressions of certainty/doubt, actuality, precision, and limitation), attitudinal expressions (i. e., personal attitudes or feelings), and style-related expressions (i. e., a speaker’s comments on the communication itself). The most overt expressions of stance, it is argued, are those including first-person subjects, as they attribute stance directly to the speaker (Biber et al. 1999: 976–978). However, studies using corpus methodology are mainly concerned with identifying stance styles by grouping texts into clusters that show a similar exploitation of stance expressions. As a result, these studies do not focus on any specific genre.

In a much similar way, Thompson and Hunston (2000) conceptualize evaluation as “the broad cover term for the expression of the speaker or writer’s attitude or stance towards, viewpoint on, or feelings about the entities or propositions that he or she is talking about” (2000: 5). They identify four parameters of evaluation: certainty/likelihood, desirability/goodness, obviousness/expectedness, and importance/relevance. Thus, modality expressing certainty or lack thereof is treated as a sub-category of evaluation because it offers ways for speakers to signal stance on a proposition or entity. An important insight is that, depending on the context, any expression can be used evaluatively, and that evaluative practice can be recognized when there are signals of subjectivity, comparison, and social value.

From a systemic-functional linguistic perspective, Martin (2000) and Martin and White (2005) approach the concept of stance under appraisal. Their view builds upon the traditional concepts of affect and epistemic modality, distinguishing between attitude, engagement, and graduation. Attitudinal stance indicates affect (emotional evaluation), judgment (moral evaluation), and appreciation (esthetic evaluation). Engagement, on the other hand, concerns “the ways in which resources such as projection, modality, polarity, concession and various comment adverbials position the speaker/writer with respect to the value position being advanced and with respect to potential responses to that value position” (Martin and White 2005: 36). Graduation deals with the ways in which the other categories are graded for their intensity. While this broad categorization is interesting, it is not clear why certain features, such as modality or comment adverbials, should be classified under engagement rather than attitude, when in fact they primarily indicate the speaker’s decision to withhold or express complete commitment to a proposition.

In contrast to Appraisal theory, Hyland (2005) sees stance as constituting a category distinct from engagement, although acknowledging that they both are evaluative resources and that there are overlaps between the two categories. According to him, stance

can be seen as an attitudinal dimension and includes features which refer to the ways writers present themselves and convey their judgments, opinions, and commitments. It is the ways that writers intrude to stamp their personal authority onto their arguments or step back and disguise their involvement. (2005: 176)

This concept of stance is further redefined in terms of voice in his more recent publication: “stance refers to the writer’s textual ‘voice’ or community recognized personality” (Hyland 2008).

Hyland (2005, 2008) identifies four components of stance:

1. hedges (devices which withhold complete commitment to a proposition), e. g., possible, may, could, tendency

2. boosters (devices which allow writers to express their certainty in what they say and to mark involvement with the topic and solidarity with their audience), e. g., should, definitely, of course

3. attitude markers (devices which indicate the writer’s affective, rather than epistemic, attitude to propositions, conveying surprise, agreement, importance, frustration, rather than commitment), e. g., believe, remarkable, extraordinary, interesting

4. self-mention (the use of first person pronouns and possessive adjectives to present information), e. g., I, we, our

The first two categories above seem to correspond to Thompson and Hunston’s (2000) “certainty/likelihood” parameter, and the third has some overlap with the parameters “desirability/goodness”, “obviousness/expectedness”, and “importance/relevance”. Similar to Biber and colleagues’ views, Hyland’s stance features reinforce a basic distinction between meanings that comment on the epistemic status of an entity or a proposition and those that indicate a speaker’s personal attitudes, emotions and assessments, along with ways to indicate the degree of those personal positions. Hyland’s concept highlights the element of authorial presence, which concerns the extent to which speakers choose to project themselves into the discourse.

In the present study, stancetaking is understood as a process in which speakers evaluate and position themselves in relation to their objects of talk and interlocutors. For the purposes of this paper, Hyland’s (2005, 2008) system of classification is adopted because his conceptualization of stance is comprehensive, ranging from authorial presence to the signaling of attitudes, feelings, judgments about a proposition or entity in the text. In addition, Hyland’s categories are function-based; in the absence of previous in-depth studies that record markers of stance in legal discourse or in the opening statement, this function-based model is particularly suitable, as it allows linguistic expressions to be matched to specific functions, thereby making possible quantitive comparison of the data across trials.

In legal discourse studies, stance functions have received a considerable amount of attention. Matoesian (1999) demonstrates how a defendant in a rape trial, who is also a physician, uses modal perfects to encode an epistemic stance of possibility and proffer alternative medical diagnoses of the victim’s injury. In so doing, he can background agency and involvement in the injury, thereby undermining the prosecution’s case. This is in stark contrast to lay witnesses who can testify only to their personal observations. More recently, examining a focus group session in which police officers discuss the effectiveness of the community police training to which they have been subjected, Matoesian (2005) illustrates how stance is lodged not only in verbal expressions but also in non-verbal signaling practices. In this particular institutional discourse, stance helps to mediate a conflict over professional jurisdiction between the police professionals interviewed and the social scientists responsible for the community police training.

Finegan (2010) chooses to examine adverbials of attitude and emphasis in written opinions or decisions of the Supreme Court of the United States. His findings reveal that, contrary to public belief and warnings against the use of attitudinal and emphatic expressions in legal-drafting textbooks, such expressions abound in the data examined (e. g., properly, correctly, simply, indeed, and clearly). The study offers a glimpse of the degree to which the language of emotion finds its way into legal opinions.

Mazzi (2010) offers a linguistic analysis of judicial evaluation strategies in US Supreme Court judgments. Focusing on evaluative lexis, Mazzi finds that judges take a stance via a range of strategies, including straightforwardly evaluative verbal and adjectival items (e. g., disagree and incorrect) and the more finely-grained pattern “this/these/that/those + noun” (e. g., omission, problem, argument, statement, and conclusion). Such labeling nouns signal positive or negative semantic prosody depending on contextual elements (e. g., supported by or refutes itself). This study raises awareness of how judges craft their reasoning through subtle stancetaking expressions, enabling them to manipulate readers and persuade them to see things in a particular way.

In a much similar way, Gozdz-Roszkowski and Pontrandolfo (2013) explore the role of evaluative phraseology in the discourse of Supreme Court judges in the United States and Italy, focusing in particular on the pattern: “the + noun + that” (e. g., the view that, the notion that, the conclusion that, etc). The researchers identify different negative and positive evaluative patterns that collocate with the head nouns and their syntactic positions. For example, the phrase the fact that in subject position co-occurs with negative particles, while this collocational pattern does not occur when the fact that is the object of a clause. The study demonstrates that apart from overtly evaluative lexical items such as problem, wrongly, or disagree, evaluative stance may not always be readily recognized when considered in isolation, but the evaluative force emerges in negative or positive co-texts. This, in turn, reveals that evaluation in this institutional discourse spreads across phraseologies.

Focusing on the relationship between naming practices and speaker stance in appellate marriage cases, Tracy (2011) finds that plaintiffs are referred to not only by their court role, but by reference terms, including sexual orientation (gay, lesbian), sex (same-sex couples, men who want to marry men), and citizen-linked references (these unequal citizens, second class citizens). Each choice of expression reveals the stance that legal professionals adopt at the outset. For example, a judge in the author’s data juxtaposed the term homosexual couple with married couple in “extend benefits that would otherwise be only available to married persons to homosexual couples” (2011: 77). This reflects the judge’s ideological position of heterosexuality as the norm, and may point to her support of a one-man/one-woman definition of marriage. The study shows the roles of reference terms in influencing how the cases should be perceived and decided.

All in all, the studies above do much to inform the present study. However, they tend to focus on particular features and constructions (such as evaluative reference terms, adverbials, or epistemic modality), thereby only partially mapping stancetaking in legal discourse. In addition, no studies have focused on the genre of the opening statement. Thus, to complement the picture of stancetaking in legal discourse, this study will explore a range of stance features in a group of high-profile trials.

4 Data and methodology

The current study draws on a corpus consisting of written transcripts from five American trials that were selected because of their very high visibility. High profile cases are worth investigating because they often tap into bigger issues in society, and the jurors in charge may harbor stronger opinions about the culpability of some defendants because they have learned about the case from the media. For example, terrorist trials bring to the fore issues including loss of lives, death penalty, and justice; then again, trials of idolized celebrities and well-respected individuals raise a valid question as to whether the aim of the trial is to ensure that justice is to be done to all, or just to cater entertainment to the public (Ewing and McCann 2006). Also, the outcome of these trials is difficult to predict, even for the attorneys involved. These cases, therefore, constitute rich resources for analyzing how the presenting lawyers’ goals, understandings and biases are reflected in the discourse strategies they create. A short description for each trial is provided below.

Case 1: The State of Alabama v. Betty Woods Wilson (1993) (7,822 words). As one of Alabama’s most sensational murder cases, this trial involves Betty Wilson who was tried for allegedly hiring handyman James White to murder her husband Jack Wilson, an ophthalmologist, at their house in Alabama. The motive was said to be the inheritance of Dr. Wilson’s $6.3 million estate. Since there was no physical evidence connecting Mrs. Wilson to the murder, the only evidence the prosecutor had to convict her was the testimony of White, a known alcoholic, drug addict and a diagnosed delusional schizophrenic. Both Wilson and White were convicted and sentenced to life without parole.

Case 2: United States v. Timothy James McVeigh (36,016 words). Deemed the most destructive act of terrorism on American soil until the September 11, 2001 attacks, the so-called Oklahoma Bombing incident was committed by Timothy McVeigh, who, motivated by his hatred of the federal government and angered by how the government was perceived to mishandle the Waco Siege in 1993, detonated an explosives-filled truck parked in front of a federal building in downtown Oklahoma. McVeigh was tried and convicted in 1997, and executed by lethal injection.

Case 3: Commonwealth of Virginia v. Lee Boyd Malvo (2003) (18,843 words). This case involves a 17-year-old defendant who stood trial for his role in the 2002 sniper attacks around Washington, D.C. Together with 42-year-old Muhammad, he was involved in killing ten people and wounded three others. The defense team claimed that Malvo was brainwashed by Muhammad into committing the crimes, which would not have happened if he had not been under Muhammad’s control. Nevertheless, Malvo was found guilty and sentenced to life in prison without parole (while Muhammad faced the death sentence in a separate trial).

Case 4: The State of California v. Michael Jackson (2005) (36,257 words). Investigations in this case started after a British documentary had been broadcast in March 2003, presenting an interview of Michael Jackson and a (then) 13-year-old boy named Gavin Arvizo. In one scene of the show, Jackson was seen holding hands with Arvizo and, at one point, Arvizo rested his head on Jackson’s shoulder. Moreover, the film also showed the guests’ discussing sleeping arrangements. Jackson was charged with four counts of molesting a minor, four counts of intoxicating a minor in order to molest him, one count of attempted child molestation, and one count of conspiring to hold the boy and his family captive at his Neverland Ranch.

Case 5: United States v. Zacarias Moussaoui, an alleged 9/11 co-conspirator (2006) (14,806 words). Moussaoui, a French citizen, was tried in a United States court on charges of involvement in the 9/11 attacks. Moussaoui’s path to arrest started when the flight instructor assigned to him had suspicions as to why Moussaoui would seek simulator lessons when in fact he lacked basic plane knowledge. After some convincing, the flight instructor’s supervisors contacted the FBI, who eventually arrested him. Moussaoui admitted having planned to hijack a fifth plane to crash into the White House. Several pieces of evidence were found that showed his connections with Al Qaeda. As a result of his conviction, he is serving a life sentence without parole in Colorado, USA.

The data for my corpus was collected using both quantitative and qualitative techniques. As the purpose of the paper is to investigate how stance is realized linguistically, the transcripts were scanned for expressions of authorial stance. To collect the stance markers in opening statements, I did not first run automatic searches for a number of predetermined forms, but scanned the transcripts manually. The software AntConc 3.4.3 m was then used to help quantify stance expressions, and all frequency counts were normalized to a common basis, per 10,000 words, to allow a direct comparison of results for the different functional features. Finally, the search results were double-checked to verify that they pragmatically function as evaluative devices.

Some of the features could readily be identified (for instance, self-mention pronouns, epistemic-based boosters, and hedges); attitude markers, on the other hand, were not as easily recognizable, mainly because attitudinal expressions are much more context-dependent and based on the issue being argued. In identifying attitude markers, I relied on the following points of consensus among researchers (see Hunston 2011: 12–19). First, evaluation is primarily subjective, expressing a personal opinion and endorsed only by the speaker (Thompson and Hunston 2000). This helps distinguish between stance and descriptive information. Second, considering an item in question along the dimensions of desirable/undesirable, true/false, important/unimportant, and expected/unexpected (i. e., such as surprising or intriguing) can, to a large extent, ensure agreement on what counts as a stance marker among researchers and even ordinary readers (Thompson and Hunston 2000: 13–19; Escandell-Vidal et al. 2014: 159–160).

On the basis of these criteria, the verb agree in “Mr. Moussaoui … proclaiming that … And on this point, and it may be the only one, I wholeheartedly agree with him” (Zacarias Def ^[1]) is an attitudinal expression, as it is suggestive of the lawyer’s position. In contrast, the same lexical item in “she [a witness] had no choice but to agree for herself and the children to participate …” (Jackson Pros) does not count as an authorial stance expression, because it does not signal explicit speaker’s attitude towards a particular issue or statement: it is descriptive, but not evaluative. Similarly, the reporting verb show is considered a booster in “the logs show they were brought back to the ranch” (Jackson Pros) and in “the phone records show they were hijackers” (Zacarias Pros) because it is strategically chosen to highlight what follows as a fact. However, this is not so in “we can show the cancelled ticket” (Jackson Pros), and thus this instance does not count as a booster. All in all, my approach is function-based, rather than form-based.

5 Findings

The overall frequencies in Table 1 reveal that stance expressions are an integral part of the opening statement. Boosters and hedges predominate, occurring at about the same rate (74.90 and 74.82 instances per 10,000 words, respectively). These two linguistic devices work together to vary the strength of the lawyers’ statements, thereby contributing to crafting a well-balanced blend of assurance and circumspection. Within the group of self-mention devices, first-person singular pronouns show a slightly higher frequency than first-person plural pronouns (58.64 and 53.72 instances per 10,000 words, respectively). The least frequently occurring device is attitude markers (35.96 instances per 10,000 words). This suggests that in their opening speech, lawyers rely more on personal pronouns to project their stance than on explicit affective stance markers.

A more revealing picture emerges at the level of the presenter’s side, shown in Table 2. In comparison to the overall frequency counts in Table 1, it is clear that, with the exception of first-person singular pronouns, each side employs a similar pattern of stance presentation: the frequency counts of each specific feature appear to fall in the same range. However, there are two notable exceptions, thereby slightly affecting the overall frequency counts, namely, first-person singular pronouns and boosters. This may be attributed to individual preference, rather than to any systematic linguistic variation. For example, the use of first-person singular pronouns by Wilson’s prosecutor alone constitutes about half of the occurrences (i. e., about 40/10,000 words), and McVeigh’s prosecutor appears to use the booster show in presenting his evidence twice as frequently as the other lawyers do. As a result, these two factors somewhat obscure what would otherwise be very consistent patterns.

Quantitative results do not adequately explicate the common ways in which lawyers from both sides orchestrate these features to achieve certain pragmatic purposes. Thus, a detailed qualitative analysis of these features is in order below.

6 Conclusion

By way of conclusion, this section attempts to answer the two research questions posed in the introductory section, and discuss a number of issues of theoretical import with respect to evaluative stancetaking.

The first research question that guided this investigation is the following: How do lawyers enact stances in the opening statement? That is, what linguistic resources do lawyers employ to evaluate social actors and events in their opening narratives?

As my analysis shows, opening statements can be characterized in terms of the use of evaluative stance markers, ranging from lexical stance expressions, such as attitude markers, hedges and boosters, to grammatical stance markers, such as personal references. The study finds that the lawyers appear in the text with a considerable degree of visibility, and the discourse of the opening statement in effect embodies the lawyer’s view of “factual” events leading to the trial, displaying different degrees of certainty, and involvement, and affect.

Viewed from the perspective that language is a system of choices (Halliday 1994; 2013), these evaluative stance markers are among the many resources the lawyers could have selected (for example, increasing the degree of impersonality of their messages). Accordingly, the presence of these stance markers in both the defense’s and prosecution’s speech suggests that such overt self-indexes and attitudinal displays are meaningful and motivated in the context of opening statements. In conjunction with other interpersonal resources such as politeness (Obeng 1997), humor (Hobbs 2007; Forabosco 2011), stance devices, I argue, are persuasive “weapons” in the courtroom (Forabosco 2011: 353). In contrast to the claim that some linguistic items, for example, she and I, only mark reference, that is, carry no evaluation and are “reference simpliciter” (Schegloff 1996), this study offers further support to Tracy’s argument that “it is questionable that even ‘simple’ pronominal reference forms are entirely devoid of evaluative significance” (2011: 68). As shown above, self-reference pronouns are loaded with evaluative significance, although such reference forms vary in the degree of moral load they carry.

Interestingly, of the set of evaluative stance markers in the corpus (first-person singular pronouns, first-person plural pronouns, hedges, boosters, and attitude markers), attitude markers are least used; further, within this group of attitude markers expressions displaying (dis)agreement and significance of information occur at high rates. A possible explanation for the relative low frequency of attitude markers might be that the other four categories can be neatly specified, thereby facilitating counting, whereas attitude markers pose a challenge for quantification because they depend on the context and the issue argued for or against. However, this may also well be one of the characteristics of the genre of opening statements. Indeed, there has been some evidence that stance markers show generic and disciplinary differences. For example, in academic writing, hedges occur most frequently, and in the domain of biology, they were found to be about 6 times higher than attitude markers, whereas in Philosophy, they were only 2.5 times higher (Hyland 2005: 187). In the case of opening statements, where argumentation is not allowed, attitude markers are likely to be avoided by experienced lawyers, as they explicitly signal affective stance. While analysis of a larger data set of opening statements is necessary to confirm this hypothesis, it is felt that our findings offer a sufficiently comprehensive sketch of lawyers’ practices in the discourse genre under study; indeed, the function-based, text-driven approach applied to a manageable set of courtroom data yielded a collection of stance expressions sufficiently large for a consistent treatment of all stance categories

In an earlier diachronic study (Chaemsaithong 2014), I also found high occurrences of first-person pronouns (99 per 10,000 words for first-person singular pronouns; 8 per 10,000 words for first-person plural pronouns). The present study differs from that earlier study in that in contemporary opening statements, first-person plural pronouns significantly outnumber first-person singular pronouns. This may be attributed to changes in the legal system. In the early eighteenth century, defense lawyers were not used, and the defendants had to prove their innocence in face of the prosecuting lawyer’s (or the judge’s) examination. As a result, the use of first-person plural pronouns was not as frequent. In the late eighteenth century, defense lawyers were introduced, who sought to expose the weaknesses of the prosecution’s case (Hostettler 2006: 40). With this new adversarial system in place, the prosecution has come to represent the people at large and has at its disposal the police department with its investigators and laboratories, while the defense must find its own resources. This in turn has led to the needs for each side to establish authority and solidarity with the jury panel, which can in large part be afforded by the use of stance markers.

From a discourse-cognitive perspective, the frequent use of stance markers not only constitutes an important element of the lawyer’s argumentative moves but also points to the lawyers’ conceptualization of this monologic genre as having an underlying dialogic structure (Bakhtin 1981), independent of the overt communicative structure (i. e., monologue). The opening statement is an instance of fictive interaction (Pascual 2006a), as it involves “interactants who do not always correspond to the addresser and addressee in the actual situation of communication, or even to referents in the current discourse space” (Pascual 2006b: 248). As this study has shown, lawyers take an active role in reaching out to the audience in the courtroom and, through stance markers, turn the jurors into co-constructors of the discourse. When directly referencing themselves, the lawyers presuppose the existence of real-time conversation partners. At the same time, they have to adjust the strength of their claims to negotiate the intended level of claim and the level of audience acceptance. All of these evaluative practices are done in the service of opening a discursive space where the lawyers display involvement with the issue they are presenting and where they strive to create a more convincing representation of their narratives.

The second research question was the following: How do these stance resources contribute to the lawyers’ roles in storytelling and to the construction and maintenance of their relationships with the jurors?

It has been argued that to be successful in their presentation, “lawyers must … convey power, authority, and credibility to the opposing sides, to their clients, and where necessary, to juries as well” (Bogoch 1999: 333). Indeed, as this study shows, to achieve such a goal necessarily involves a good command of stance features. Lawyers have to know when to signal their presence, and when to step back and lessen their involvement in their discourse.

However, to reveal the lawyer’s involvement, state of knowledge, viewpoint and judgment is not the sole function of stance markers. There is also a relational side to stancetaking. When a lawyer takes a particular stand, it is via a process of dealing with various points of view and value positions. By using stance resources, lawyers can produce a voice that is able to acknowledge, engage with or disalign itself with respect to other positions which are in some way alternatives to that being advanced in their speeches. In this way, stance resources map out lawyers’ relationships with various participants (including, among others, opposing lawyers, the jurors, and the general public) and other value positions brought into communicative play by the text.

For instance, the use of evaluative verbs such as demonstrate or assume can help create a voice that indicates, respectively, the lawyers’ endorsement and rejection of the current viewpoint, signaling to the jurors that they stand with or against the source in advancing that particular position. Similarly, via the use of epistemic expressions such as clearly and maybe, the voice actively presents itself as aligning with the construed reader or opening up the space to alternatives, so that it appears as entertaining or at least sharing the same belief or concern. Thus, the significance of stance expressions is not simply indicative of the lawyer’s position, but also constitutive of his response, support, and challenge to some external source.

In a more critical vein, it is worth considering that even though lawyers have been found to use many instances of inclusive we or boosters such as evidently or clearly to index shared stances, the truth of the matter is that in the actual courtroom practice many of these instances do not necessarily include the jurors. Because these resources give an impression of inclusiveness to the interlocutor, it is not likely that the jurors could take their time to consider whether they were actually included in this group. It is through these linguistic devices that the lawyers subtly make the discourse essentially argumentative and lead the audience to make certain inferences. This in turn stands in stark contrast with the primary purpose of the opening statement (at least theoretically), which is to inform the jurors about the case in a general way.

It is hoped that this research been able to explicate and demystify how lawyers communicate their stances in the initial phase of trials. The findings of this study may have practical implications for training jurors and the public at large to become aware of, and be less easily swayed by, persuasive interactive techniques. An interesting question to be pursued in future research is to look more closely at cases where lawyers may mark their stances in other ways, such as via prosody or bodily gestures. We also need to examine opening statements in other legal cultures as well in order to present a more complete picture of the kinds of engagement that are acceptable as well as expected cross-culturally, as the linguistic resources lawyers select are always relative to a particular audience and the socio-cultural contexts in which they are used (Chang 2004).