Play-by-Play Justice: Tweeting Criminal Trials in the Digital Age

Abstract

Journalists routinely live-tweet high-profile criminal trials, a practice that raises questions about access to justice and the principle of open court. Does social media open up the justice system? There is a normative debate in the literature about the use of Twitter and social media in the courtroom. This paper takes on this debate by exploring the relationship between digital technologies and criminal justice. Through a systematic examination of journalists’ tweets during two key trials (Ghomeshi and Saretzky), we ask to what extent can the live-tweeting of court proceedings achieve greater access to justice in Canada? We argue that while the live-tweeting does provide more access to court, potentially furthering the principle of open court, the nature of this access provides little in the way of increased engagement with the public and its understanding of the legal system. This paper makes contributions to both the legal studies and digital politics literatures.

Résumé

Lors de procès criminels très médiatisés, les journalistes font régulièrement de la diffusion en direct par le biais de publications sur Twitter, une pratique qui soulève des questions quant à l’accès à la justice et à l’égard du principe de la publicité des débats des tribunaux. Les médias sociaux ouvrent-ils les portes du système de justice? Il existe un débat normatif dans la littérature sur l’utilisation de Twitter et des médias sociaux dans la salle d’audience. Cet article aborde ce débat en explorant la relation entre les technologies numériques et la justice pénale. En examinant systématiquement les tweets des journalistes au cours de deux procès d’envergure (Ghomeshi et Saretzky), nous nous demandons dans quelle mesure la retransmission en direct des procédures judiciaires, par le biais de la plateforme Twitter, arrive à permettre un meilleur accès à la justice au Canada. Nous soutenons que si la diffusion en direct via Twitter offre un meilleur accès aux tribunaux, ce qui pourrait éventuellement favoriser le principe de la publicité des débats, la nature de cet accès ne permet pas d’accroître la participation du public et sa compréhension du système juridique. Cet article offre une contribution à l’étude du droit et à l’étude de la politique numérique.

Introduction

In 2014, Jian Ghomeshi was charged with four counts of sexual assault and one count of overcoming resistance by choking originating from three complainants. At the time of his arrest, Ghomeshi, a former pop musician, was the host of the popular CBC program Q. Given the high profile of the accused, the resulting trial in 2016 ignited a firestorm in the traditional media, and also on social media. Journalists, interest groups and individuals used social media, especially Twitter, to discuss the machinations of trial and issues related to rape culture and sexual assault in Canada. While this was not the first Canadian trial to be discussed in real-time on social media, several media outlets noted the overwhelming attention this trial received online. Women’s magazine Chatelaine suggested there is “value in live-tweeting Jian Ghomeshi’s trial” by revealing “in real time—what sexual-assault complainants face on the stand” (Giese 2016). The Canadian Journalism Project concurred, suggesting that live-tweeting “demystifies” the court process (Watson 2016). Others were less charitable. In a Globe and Mail editorial, Nathan Gorham (2016) argued that the Ghomeshi trial demonstrates that Twitter is a poor medium to provide access to the court system and that television would be preferable. In the law and society literature, a debate has developed on the extent to which social media helps or hinders court processes (for instance, Goehler, Dias, and Bralow 2010; Winnick 2014; Packer 2012; Sossin and Bacal 2013; Hall-Coates 2015).

In this article, we take up this debate. More specifically, we analyze the role of live-tweeting in criminal trials. Live-tweeting is the communication of events in real time on Twitter. Twitter (a micro-blogging platform) allows users to post short messages known as “tweets,” which other users (followers) can interact with by replying or sharing the message with their own followers (retweet). This analysis is guided by two research questions: What is the nature of live-tweeting in criminal trials? Does live-tweeting enhance the principle of open court? The objective of the first question is to provide an empirical analysis of what live-tweeting in a criminal trial looks like. As will be discussed, there is limited empirical research on this in either the digital politics or legal studies literatures. There is, however, considerable normative discussion in the popular and academic literature about the role social media might play within the court system and the extent to which social media might help or hinder court processes. The objective of the second question is to reflect on live-tweeting within the context of this broader normative debate.

In order to answer these questions, we conduct a content analysis of live-tweeting by journalists in two criminal trials that occurred in Canada, the aforementioned Ghomeshi trial and the murder trial of Derek Saretzky that took place in Alberta in 2017. We focus on live-tweeting by journalists, rather than members of the public, because only journalists are permitted to live-tweet in the courts where the Ghomeshi and Derek Saretzky trials were held (Ontario and Alberta). In these jurisdictions, only journalists have the presumptive right to live-tweet in courtrooms, that is, the ability to live-tweet without permission of a judge or judicial officer. In some Canadian courts, journalists and the public have the presumptive right to live-tweet (Puddister and Small 2019). Moreover, news organizations have begun to use live-tweeting as a way to reach out to audiences (Broersma and Graham 2012). The comparative research design of exploring two different cases allows this analysis to go beyond any peculiarities of one trial in understanding live-tweeting.

This article is organized as follows. First, we review the literature on the principle of open court and the potentials of social media. Following this, the two trials, Ghomeshi and Saretzky, are briefly described. The methodology and selection of data are outlined, followed by the results. We find that live-tweeting criminal trials is overwhelmingly play-by-play in nature, rather than providing context and analysis of trial proceedings. In light of this finding, we argue that live-tweeting does provide more access to the court, potentially furthering the principle of open court. That said, the nature of this access provides little in the way of increased engagement with the public and arguably does little to improve the public’s understanding of the legal system.

This article makes a number of contributions to both the digital politics and legal studies literatures. While there is extensive analysis on the impact of Twitter and other social media in governmental, electoral, and advocacy contexts in Canada, courts are an understudied political institution in this regard. Our study contributes to the study of the political and legal uses of Twitter in the courts through an empirical analysis. We also make a methodological contribution by developing an original technique for analyzing live-tweeting in court proceedings, by using a coding scheme that was derived inductively and deductively. Finally, we make an empirical contribution to the normative debate developing in the law and society literature regarding the role and potential value of social media in the legal and court system.

Literature Review: Open Court & the Potentials of Social Media

The aphorism “not only must justice be done; it must also be seen to be done,” is one of the most fundamental principles of our justice system.¹ This principle holds that public access to the courts and their decisions is essential for ensuring legitimacy in court proceedings and transparency of the legal system, with the ultimate goal of promoting public confidence in the system. The public must have confidence in the legal system to promote and respect the rule of law, ensuring the protection of a necessary element of a democratic society. According to the former Chief Justice of the Supreme Court of Canada, public confidence in the justice system is fostered through: (1) protecting the fundamental values that structure the system (such as the presumption of innocence); (2) striving to ensure that the process and outcomes of the system are of high quality; and (3), safeguarding access to the system (McLachlin 2003). While all three elements are important, the third goal of ensuring public access (or access to justice) must be met before securing the other two. Without the opportunity for public participation, there is no public justice system (CBC v. New Brunswick; Trial Lawyers Association of B.C. v. British Columbia).

Access to justice is routinely identified as one of the most important issues facing the Canadian justice system (Canadian Bar Association 2013; Hausegger, Hennigar, and Riddell 2015). Most often, the concerns regarding access to justice focus on the costs of using the system and the access to information about the system. The other equally important element of access to justice concerns the transparency of the system, also known as the principle of open court. Open court relates not only to the dissemination of what courts do through the publication of legal decisions and access to court records, it also means “open doors” and the right of the public and the media to attend court proceedings (McLachlin 2002; Paciocco 2005). Ensuring an open court is of critical importance for two central reasons. First, publicizing court proceedings can help to protect the due process rights of those brought before the court and ensure that the proceedings are conducted with integrity (Paciocco 2005, 387). Second, and more significantly, the principle of open court creates a check on the power of the state and its courts, providing an important form of accountability (CBC v. New Brunswick; Vancouver Sun (Re); Paciocco (2005))

Since most individuals cannot physically attend court, media reporting on court proceedings is essential to maintain an open court (Paciocco 2005; Edmonton Journal v. Alberta). The media acts as a “proxy” or intermediary for the public by attending court and reporting on the process and outcomes, thereby providing individuals with the opportunity to make informed judgements about their legal system (McLachlin 2002; Hall-Coates 2015). The ability of the news media to attend and report on court proceedings has been deemed so vital to the Canadian legal system that it has been protected under Section 2(b)–freedom of the press and other media–found in the Charter of Rights and Freedoms (CBC v. New Brunswick (A.G.)). Media have traditionally engaged in this proxy role through reporting courtroom stories in newspapers and, in some jurisdictions, by broadcasting court proceedings on television. Although common practice in many American courtrooms,² televised broadcasts of trial court proceedings are relatively rare in Canada and photographs are expressly prohibited (McFeat 2010; Scotti 2016). While the Supreme Court of Canada is seen as a pioneer in providing live broadcasting of its proceedings on CPAC and through its website, this openness is not extended to trial courts. Indeed, an important distinction has been made between appeal and trial courts because the latter often involve the participation of witnesses (McFeat 2010). The decision to grant access to television cameras resides with the discretion of the trial judge, as an extension of administrative independence.

The spread of digital technology into the judicial system has created a novel set of challenges for a system that is notoriously slow in adapting to change. ³ Our focus here is the use of Twitter in the courtroom to share trial proceedings in real time. There is a normative debate in the literature about the use of Twitter, and social media more broadly, in the courtroom. While this literature focuses on the potentials and perils of digital technologies in courts by both the media and the public, we focus on the former, given that only the media is permitted to live-tweet in the trials examined in this article. Some suggest that Twitter in court creates new opportunities to enhance the principle of open court. When journalists live-tweet a trial, they can provide the public with a low-barrier means to access the trial proceedings, as the recipient or follower only needs to have access to the Internet.⁴ Viewed in this way, live-tweeting can be understood as a means to open the courtroom doors to create a link between the news-creator and news-reader, producing a democratizing effect and improving accessibility (Chadwick 2013; Packer 2012; Lambert 2011; Hewett 2015; Comrie and Fountaine 2016). Twitter can provide an important element of transparency to journalistic practices by linking the public directly to sources and by providing insight into the reporting process, including the correction of facts and reporters’ editorial practices (Chacon, Giasson, and Brin 2015; Chadwick 2013; Vis 2013; Swasy 2016).

In a similar vein, live-tweeting allows for greater transparency of the actions of trial participants and proceedings, allowing for reaction and debate in real time, which creates an opportunity to increase accountability in the criminal justice process (Goehler, Dias, and Bralow 2010; Winnick 2014; Comrie and Fountaine 2016). This increased access can not only improve public confidence in the system, it can also provide an educative function by communicating how the trial process works. This educative function can be especially valuable for individuals who lack the confidence or the means to physically attend courtroom proceedings (Lambert 2011; McGowan 2013; Synodinou 2012; Packer 2012). Finally, while many of these advantages could be achieved through televising court proceedings, the resources required to live-tweet are comparably quite minimal—simply a reporter with a computer or cellphone—making the media coverage much less intrusive and distracting to the ongoing proceedings, compared with a television camera (Hall-Coates 2015).

The increased access to the courtroom provided by Twitter does carry some disadvantages. Live-tweeting essentially requires that journalists operate without editorial oversight, meaning that reporters must not only produce tweets about the case as it is unfolding, they must also make judgement calls about what is appropriate to share with the public on the spot (Hewett 2015; Hall-Coates 2015). Importantly, unlike a traditional news article, once information is shared through a tweet, it is difficult to issue a meaningful retraction as authority over the information is surrendered (Hall-Coates 2015; Synodinou 2012; Chacon, Giasson, and Brin 2015). The lack of editorial oversight may be a greater issue in criminal trials where there is potential for sharing overly graphic material with the public (English 2012). The loss of control over a tweet can be more problematic in the context of a criminal trial because journalists may share information that is later decided immaterial or prejudicial against the accused, undermining the right to a fair trial (Goehler, Dias, and Bralow 2010; Janoski-Haehlen 2011; McGowan 2013; Schutz and Cannon 2013). The challenges of live-tweeting are heightened when considering that many journalists do not have legal training (Sauvageau, Schneiderman, and Taras 2006).

The nature of Twitter not only gives rise to problems regarding control over information, the micro-blogging format also poses challenges with regard to context. With a limit on the number of characters per tweet, it can be difficult for live-tweeters to explain complex court proceedings, and legal outcomes (Schutz and Cannon 2013; Sossin and Bacal 2013; Winnick 2014). This inability to provide context can essentially reduce live-tweeting by reporters to little more than stenography, which lacks the potential insights that could be provided by an experienced journalist. According to critics, because tweets prevent meaningful context and journalistic analysis, the educative possibilities of live-tweeting are undermined (Winnick 2014; McGowan 2013; Schutz and Cannon 2013). Instead, real-time tweeting provides a form of entertainment and contributes to the sensationalism of a trial—a fact that is most pronounced when the accused is a public figure or celebrity. This focus on sensationalism could serve to undermine the gravity of the trial, including the consequences for those involved, especially the accused.

These questions are not just academic discussions. Rather, the issue of open court and live-tweeting raises important questions for the public and journalists working in the field. For example, in 2012, the Toronto Star publicly considered the paper’s responsibility in covering the Rafferty trial in light of the existing publication bans and graphic nature of the crimes.⁵ The Star’s public editor noted that the trial raised serious legal and ethical questions that required careful editorial review prior to publication and would not be live-tweeted (English 2012).

Despite substantial debate regarding the role of Twitter in court and the impact on access to justice and the principle of open court, there is a considerable gap in the empirical literature that examines the nature and use of live-tweeting of criminal trials or live-tweeting in general. A notable exception can be found in Scott’s content analysis of live tweets from the Oscar Pistorius murder trial in South Africa, which found that reporters relied on traditional journalistic practices when live-tweeting, as most tweets were used to communicate information and to inform the public (Scott 2016). Another exception can be found in Harada and McGuire’s (2016) analysis of the tweeting during the sexual assault and murder trial of Russell Williams. The authors found that detailing the graphic content of the Williams trial through Twitter presented unique challenges that were fundamentally distinct from traditional court reporting, leading the authors to urge the Canadian journalistic community to adopt best practices when live-tweeting trials including guidelines on the appropriate use of Twitter in court.

The Trials

This study is based on a content analysis of two high-profile criminal trials. R v. Ghomeshi received significant attention from the public and the media due to the celebrity status of the accused. As mentioned, social media was especially important in this trial. R v. Saretzky similarly garnered a great deal of attention because of the heinous nature of the crimes. Because of the substantial public interest, both trials were live-tweeted by several professional journalists. We briefly outline each trial to provide the necessary context for the analysis that follows.

R v. Ghomeshi

In late 2014, Jian Ghomeshi, a high-profile and popular CBC personality, was charged with four counts of sexual assault and one count of overcoming resistance by choking contrary to sections 271 and 246(a) of the Criminal Code of Canada. The allegations arose from three different complainants. In January 2015, Ghomeshi was charged with three additional counts of sexual assault. The allegations against Ghomeshi span from 2002 to 2008. Ghomeshi pleaded not guilty to one count of choking and four counts of sexual assault, and the trial began on February 1, 2016, lasting nine days. During the trial, the identity of two of the complainants (referred to as L.R. and S.D.) were protected from identification, while the third, actor Lucy DeCoutere, publicly came forward as one of the complainants. On March 24, 2016, Ghomeshi was found not guilty on all five counts. In reaching his decision, Justice Horkins found that the evidence presented in the case, including complainant testimony, raised reasonable doubt and therefore did not satisfy the burden of proof placed on the Crown in a prosecution (R v. Ghomeshi 2016 ONCJ 155).

R v. Saretzky

On September 14, 2015, Crowsnest Pass Alberta RCMP were called to investigate the sudden death of Terry Blanchette. Through the investigation, the police learned that Blanchette’s daughter, two-year-old Hailey Dunbar-Blanchette was missing and appeared to have been abducted. Blanchette’s death was ruled a homicide, and the following day, the police discovered the remains of Hailey Dunbar-Blanchette (Mertz 2016). On September 16, 2015, Derek Saretzky was charged with two counts of first-degree murder. Saretzky would later confess to killing Blanchette and his daughter, and to killing Hanne Meketech, a neighbour of Saretzky’s grandparents (Mertz 2016).

Saretzky’s triple first-degree murder trial took place in June 2017. A jury found Saretzky guilty of the three killings and of an additional charge of committing an indignity to a dead human body. On August 9, 2017, Justice Tilleman of the Alberta Court of the Queen’s Bench sentenced Saretzky to three consecutive life sentences or seventy-five years of parole ineligibility (R v. Saretzky).⁶ The Saretzky investigation and trial garnered a significant amount of public attention, making national headlines (Grant and Fletcher 2017; Martin 2017). The public became quickly attentive to the case as a result of the public alert issued for the missing toddler, and the public attention was sustained due to the seriousness of the offences and the violent nature in which they were carried out.

R v. Ghomeshi and R v. Saretzky were selected for analysis as both cases are examples of recent, highly salient criminal trials, that received substantial attention from the public. Both cases were trials (rather than appeals) heard in lower courts of first instance, with Ghomeshi taking place in the Ontario Court of Justice and Saretzky in the Alberta Court of the Queen’s Bench. The accused in both cases faced the possibility of incarceration upon a finding of guilt. The Saretzky case was heard by a judge and jury, while Ghomeshi was heard by a provincial court judge. Similar to other studies of live-tweeting,⁷ in selecting Ghomeshi for analysis, we include a high-status or high-profile accused; however, in including Saretzky, we also analyze the trial of a private citizen.

Data Selection and Methodology

In this section, we explain our sources of data and the methodology employed. The unit of analysis is a tweet by an individual journalist. The focus is on the journalist rather than the news agency or organization because it is the journalist who is physically in the courtroom, and journalists who live-tweet do so from personal, rather than organizational, Twitter accounts. While the unit of analysis is a tweet by a journalist, this is not a study of journalists or journalism per se. Our focus is on what live-tweeting looks like and how it operates in criminal trials and its relationship to open court. By necessity, journalists are the object of study because the administrative policies of the courts in Alberta and Ontario restrict the use of digital technology to journalists (Puddister and Small 2019). Since the Twitter data comes from 2016 and 2017, all tweets have a character limit of 140 (rather than the current 280 limit).

The Ghomeshi trial was nine days in length and took place between February 1 and February 11, 2016, with the verdict announced on March 24, 2016. ⁸ The Saretzky trial took place over fifteen days between the dates of June 7 and June 29, 2017, with sentencing held on August 29, 2017. Because we wanted to capture live-tweeting throughout the trial process, we examined the tweets of journalists who live-tweeted at least 50 percent of trial days. This results in the examination of tweets by two journalists in the Ghomeshi trial and five journalists for the Saretzky trial.⁹ The Ghomeshi trial journalists live-tweeted a total of 3,683 times over the entire trial, and in the Saretzky trial, the journalists live-tweeted a total of 5,536 times (see Table I) .

Table I Live-Tweeting Journalists

The analysis that follows is based on a sample of tweets from the trial. ¹⁰ A sample was necessary because the tweets were human coded. Human coding of online texts is time-consuming compared with computer coding. However, we believe that human coders are superior to computer-based coding in this case in their capacity to understand “latent meanings or the subtleties of human language” (Lewis, Zamith, and Hermida 2013, 35). This is crucially important on Twitter, where hashtags, links and emojis might be used to convey important qualifying content, which computer-based coding may fail to capture. We sampled tweets by day, rather than conducting a random sample. This purposive sampling was necessary because analyzing live tweets must be done in chronological order. A single tweet is often part of a larger narrative chain (or thread), and understanding the context of tweets that precede and follow is essential. Put differently, when users rely on a narrative chain to live-tweet, a single tweet cannot be understood in isolation, rendering random sampling impractical.

Our selection of trial days was also intentional because we wanted to ensure that our analysis included the key procedural days in trials. We included the first date of trial to capture opening statements by counsel, the day of the verdict, and sentencing (where applicable). Beyond these key days, we include a random selection of the remaining days of the trial (resulting in half of the total trial dates). For the Ghomeshi trial, the sample includes a total of six days (day 1 (February 1), verdict day (March 24), and February 2, 4, 6, 11). The sample for the Saretzky trial includes a total of nine days (day 1 (June 7), verdict day (June 29), sentencing day (August 9) and June 8, 15, 21). Table I shows the total number and percentage of tweets from the sample by journalists. In both trials, our sample includes more than 60 percent of all tweets, allowing us to be confident that our analysis provides a thorough overview of the live-tweeting during the two trials.

We conduct a content analysis to determine the nature of live-tweeting during the two criminal trials. Content analysis is a method of “measuring or quantifying dimensions of the content of messages,” which is useful for describing and making inferences “about the sources who produced those messages, or [to] draw inferences about the reception of those messages by their audiences” (Benoit 2011, 258–59). It is also a very popular method of analyzing political communication on Twitter (Jungherr 2016). As such, this analysis is an instance of supply research. In the digital politics literature, there are generally two main types of research: supply and demand (Vaccari 2013). Using methods such as content analysis, supply research examines the online content produced by political actors (e.g., political parties, interest groups, the media). By contrast, demand research focuses on the use of political digital content by citizens or others. This analysis is focussed on the supply of political content by the media. If one wanted to understand the demand side or how the public consumes live tweets, different research questions and different methodological techniques (such as surveys) are necessary. This is a fruitful place for future research.

The coding scheme was developed both deductively and inductively. It draws inspiration from Katy Scott’s (2016) content analysis of live tweets from the Pistorius trial, and other studies of live-tweeting (Vis 2013; Lawrence et al. 2014; Harada and McGuire 2016) and Twitter journalism (Lasorsa, Lewis, and Holton 2012; Hewett 2015). We also develop categories that addressed the principles of an open court as it relates to Twitter. The final coding scheme was also refined after two pre-tests.

The variable choice is based on the existing scholarly literature. As in Scott’s work, tweets were first assessed for their primary purpose (Table II). For a more nuanced understanding of live-tweeting, primary purpose was sub-divided into providing information about the scene, quotes and information from the trial. Tweets may also provide information about the trial that was created by another user and retweeted by the tweet author. Retweeting, according to Lasorsa et al. (2012, 26) is “an indication of a journalist’s ‘opening the gates’ to allow others to participate in the news production process.” Responding is another instance where Twitter can be used to create greater accountability and transparency in news-making, through the responding to questions and comments posed by followers about the trial. It is worth pointing out that the analysis of retweeting and responding is often used in digital politics studies to assess the extent to which political actors make use of the interactive capabilities of Twitter (for instance, Parmelee and Bichard 2011; Small 2014). Job talking is a final form of transparency (Lawrence et al. 2014). By tweeting about the process of live-tweeting while doing it, journalists provide a peek behind the curtain (Chadwick 2013). The primary purpose categories are mutually exclusive.¹¹

Table II Primary Purpose Definitions

In addition to primary purpose of tweets, the coding scheme considers several other variables that allow the tweet author to provide greater context: opining, background, editorial decisions and the inclusion of web content (photos, links, and hashtags). These variables allow us to consider the normative debate about Twitter in courts more fully. The pre-tests of the methodology indicated that most of these variables occur in conjunction with one of the primary purposes discussed. The coding scheme first considers opining about the court case. Previous studies have found some evidence of journalists providing their personal opinion in blog posts or in tweets (Singer 2005; Scott 2016). On one hand, this could be considered problematic in that journalists opining appears to deviate from traditional professional norms of objectivity found in Western journalism (Lasorsa, Lewis, and Holton 2012). On the other, Farida Vis (2013) suggests that the opining by journalists on Twitter contributes to increased transparency and audience value. As discussed above, a related concern in the literature is the lack of editorial oversight of journalist tweeting. Whereas a print article about a court case would be subject to review by an editor, live tweets are not, which can be a concern because a retraction of a tweet is nearly impossible (Hall-Coates 2015). That said, flagging corrections and highlighting mistakes are very much a part of live-blogging that might be translated in microblogging (Thurman and Walters 2013). Indeed, Harada and McGuire’s (2016) examination of tweets from the Russell Williams sexual assault and murder trial found examples of journalists making explicit editorial choices. Some journalists indicated that they would not be describing the contents of a video or photograph. Others included the words like “warning” or “graphic content” at the beginning of tweets and/or when resuming live-tweeting of the trial. In the pre-tests, we also found evidence of such editorial decisions, including journalists providing corrections to previous tweets. Finally, we consider whether the user included other web content in tweets. While tweets have a character limit, users can communicate additional information through the inclusion of photos or other multimedia elements. Additionally, links to online web content related to the case may be included in a tweet through the use of a URL shortener. We coded whether a journalist linked to his or her own media organization, another media organization, or another source. Hashtags have become a central component of Twitter. A hashtag is a keyword assigned to a tweet that describes it, providing additional information and/or aides in searching. The inclusion of such items can result in greater context for the tweet (Artwick 2013).

Two trained undergraduate students coded the tweets. Cohen’s kappa was used to estimate intercoder reliability, which measures the extent to which independent coders make the same coding decisions (Freelon 2010). Cohen’s kappa is a conservative measure of intercoder reliability because it does not give credit for chance agreement between the coders (Lombard, Snyder-Duch, and Bracken 2002). Cohen’s kappa ranges from 0 to 1.00, with larger values indicating better reliability; 0.80 represents very high intercoder reliability while 0.60 represents acceptable intercoder reliability (Lombard, Snyder-Duch, and Bracken 2002). We attained a very high level of intercoder reliability between coders (that is, above 0.80).¹² Cohen’s kappa for the entire coding scheme was calculated at 0.95. The variable Primary Purpose had the most disagreement between coders; Cohen’s kappa was still high at 0.84.

Results: Play-by-Play Justice

Before exploring the findings of the content analysis, it is worth considering how much content was produced by the journalists live-tweeting in these two trials (Table I). The two journalists in Ghomeshi live-tweeted 3,683 times over the duration of the trial, an average of 368 tweets per day. Chatelaine’s Sarah Boesveld tweeted more often than Alyshah Hasham at the Toronto Star; Boesveld produced 711 tweets more than Hasham over the same time period. Journalists in Saretzky live-tweeted less than in Ghomeshi. The five journalists tweeted a total of 5,536, an average of 369 tweets per day. Less than a hundred tweets were produced by the journalists each day. Overall, we can see that live-tweeting produces a considerable amount of content. Using a crude approximation that one character equals 0.5 words, then Bill Graveland, who tweeted the least frequently, wrote about 4,375 words per day. At the other end, the Star’s Sarah Boesveld wrote around 15,379 words per day during the Ghomeshi trial. Whereas the public may read a newspaper article summarizing the trial or watch a short clip on television, Twitter allows for extensive and specific details of the court proceedings to be shared.

A sports broadcast analogy is useful in explaining the findings of this article. In the coverage of live sports, there are typically two broadcasters: the play-by-play commentator and the colour commentator. As the name implies, the play-by-play commentator calls the action on the field, move for move, as the play happens.

The colour commentator, on the other hand, adds context to the plays—why a person or team does what they do, the intended effects of a play and other relevant information such as statistics. In the context of a court proceedings, when tweeting is play-by-play, the live-tweeter would largely describe what is going on in the courtroom, whereas a colour commentator would include additional information about court room actors (e.g., emotions, background, dress), criminal law, and the trial process more broadly. As discussed below, the findings of the content analysis (Tables III and IV) indicate that live-tweeting in these two trials is primarily play-by-play in style, confirming the assumption that live-tweeting is essentially a form of courtroom stenography (Jobb 2010). Despite opportunities to provide colour commentary, we find little evidence of this from the two trials. This, we argue, confirms the assumptions made by the critics of Twitter in court, that the medium does not allow for context (Schutz and Cannon 2013; Sossin and Bacal 2013; Winnick 2014). On one hand, live-tweeting does increase the principle of open court by providing extensive detail of a trial not found in pre-Twitter news coverage. On the other hand, live-tweeting contributes little to increased engagement, education and accountability as suggested by some commentators in the literature (Lambert 2011; McGowan 2013; Synodinou 2012; Packer 2012).

Table III Primary Purpose of Live-Tweets

Table IV Contextual Elements of Live-Tweets

Table III provides the results of the content analysis for the variable of primary purpose. Taken together, the reporting function (quotes, scene, and trial) make up about 86 percent of all tweets in both trials. This supports the contention that the live-tweeting journalists were essentially play-by-play commentators. Following the live tweets of a criminal trial would result in extensive knowledge about what was said, what happened, and what it looked like in the courtroom. By a wide margin, however, the most common purpose of live tweets was reporting quotes—that is directly or indirectly describing what court actors said during the trial (66.2 percent overall across both trials). The following series of tweets by CTV’s Kaella Carr is illustrative of reporting quotes. Here, Carr describes the testimony of Terrence Megli (Saretzky’s grandfather and long-time friend of Meketech, one of the victims):

• Megli says Meketech came into about 90,000 dollars. Says she kept it in her trailer because she didn’t trust the banks. #Saretzky

• Megli says Meketech and her ex-husband went through a bad divorce and often got into fights. #Saretzky

• Megli says Meketech had a knee replacement and several other minor procedures. He wasn’t aware if she had any prescription drugs. #Saretzky

• Megli says he never witnessed any fights or arguments between his grandson and Meketech. #Saretzky

Tweets also describe other aspects of the trial, such as visuals and photos being shown in the courtroom. For instance, from Patrick Burles’ (Lethbridge News Now) Twitter feed:

• Photos are now moving into Blanchette’s home, after Cpl. Ridding covered the exterior #Saretzk

• Red stain shown on the deck outside the side door of Blanchette’s home #Saretzky

Given the nature of the Saretzky trial, with an identifiable crime scene, there was significantly more trial description than in Ghomeshi (13.5 percent compared with 2.7 percent). Overall, we find what Hall-Coates (2015, 120) calls “gavel-to-gavel coverage,” where tweets provide a detailed unfolding of the court proceedings. Warnick and Heineman (2012) suggest that Twitter is an inadequate tool in courts due to an inability to provide information or analysis beyond the facts of the case. Our findings concur, but not because a single tweet is too short. Instead, the issue is that a single tweet is often part of a longer narrative chain of tweets and cannot be understood in isolation. This then requires followers to be constantly following the live tweets in the court room in real time. Reading a single tweet here or there would lack context. That said, when it comes to the amount of content produced by those who live-tweeted, Twitter can increase the principle of open court by providing far more detail than traditional news reporting.

Despite the opportunities for interaction between journalists and followers during the trial, there is little take-up. This is clear from the low number of both retweeting and responding (together, 4 percent). While there was some retweeting in Ghomeshi, it was practically non-existent in Saretzky. The vast majority of Ghomeshi retweets came from Alyshah Hasham, who typically retweeted the tweets of other journalists covering the trial, including the other journalists in this study. As such, she engaged with other reporters—not the public. In the case of responding, just under 2 percent, on average, of all tweets from the two trials were of this kind. Similar to Hasham’s retweets of other journalists, closer inspection of tweets coded as responding shows that many of the responses were to other journalists, not followers or the public. Our data is slightly lower than Scott’s analysis of the Oscar Pistorius trial, where 5.4 percent of the sample were responses and 4.7 percent were retweets. Nevertheless, this opportunity to create a more open court by allowing reporters to engage more directly with the public does not appear to be taking place in either criminal trial.

In addition to assessing the primary purpose of each tweet, we also consider other variables that allow live-tweeters to provide greater context to their tweets. Table IV provides further evidence supporting the notion that live-tweeting is analogous to play-by-play in sports. It should be pointed out that these categories are not mutually exclusive. That is, a journalist could provide background information about the trial while providing an opinion. Did the live-tweeter go beyond play-by-play in the live-tweeting of these two trials? Do they provide extra context or “colour”? In short, the answer is no. With the exception of adding hashtags, very little contextual information was added to live tweets in either criminal trial. Just over 71 percent of tweets included a hashtag. That said, our data shows that the hashtags used do not enhance the content of tweets. Perhaps not surprisingly the most common hashtags were #Ghomeshi and #Saretzky. These are useful in placing the tweets within a broader online discussion of the trials on social media and allowing for searching, but do little to describe the tweet or provide additional contextual information. Two interesting hashtags were used by the two journalists in Ghomeshi: #IBelieveLucy¹³ and #WeBelieveSurvivors.

• “Wherever we go, however we dress, no means no and yes means yes!” Another chant. High energy despite cold #WeBelieveSurvivors #Ghomeshi

• “Our bodies! Our lives! We Will not be victimized!” women chant. #WeBelieveSurvivors #Ghomeshi cars honking in support

These hashtags can be considered “hashtag activism,” where social media users raise public awareness of a political issue using some clever or biting keyword on social media. In general, these two hashtags were used to counter the perceived lack of credibility of the complainants by the defence and the judge. However, their use by the journalist, as in the above example, is more descriptive of the situation than engaging in some sort of political activism on the part of the press. Here Sarah Boesveld is reporting the quotes from a rally outside of the courtroom supporting victims of sexual assault, including the complainants in the Ghomeshi trial. Again, hashtags add little additional context to the live tweets.

As discussed, many scholars suggest that Twitter in the courtroom could allow for enhanced public knowledge and education about court functions (Packer 2012; Lambert 2011). There are two ways this could be accomplished in the context of live-tweeting. First, journalists could explain who court actors are, what their roles are, explain court procedures and provide background to the trial or even refer to earlier parts of the trials while tweeting. Alternatively, they could provide links to other online materials that have more detailed information such as news stories or court documents. There is some evidence of this in our sample: 8.9 percent of tweets provided background information within the tweet. The following example from Kevin Martin (Postmedia) during the Saretzky trial demonstrates why adding context can matter.

• “I’m not going to judge you even if (Hailey’s) gone,” McCauley says to #Saretzky.

The addition of the word “Hailey’s” (one of the victims) makes the quote make more sense. Without knowing the details of this trial, it is unclear who McCauley is and what purpose he serves in the trial. This tweet is of a quote mid-testimony by RCMP Staff Sgt. Mike McCauley. Martin tweeted more than 200 times during McCauley’s testimony but his title is mentioned less than five times. This means if the reader is not well ensconced in the trial (e.g., knowing the background of the trial or the relevant actors), or if the reader started following mid-day or mid-trial, they may not know what is going on and what is being communicated by the journalist. There appears to be little consideration by live-tweeters that some followers might only login to Twitter for a short period of time and consume information in short bursts. The lack of basic colour commentary in this context renders tweets difficult to use. Almost nine percent of tweets included links to other digital content. About half of those were links to stories written by the journalist or their news organization about that specific trial. The other half were photos related to the trial, though not taken during the trial.

While other studies have found some evidence of journalists providing their personal opinion in blog posts or in tweets (Lasorsa, Lewis, and Holton 2012; Scott 2016), there is little of it in the live tweets of the Ghomeshi and Saretzky trials. The following is an example from Sarah Boesveld from Ghomeshi, where she provides an opinion on a quote during the testimony of complainant Lucy DeCoutere

• Lucy is staying very calm and bright. It’s incredible… #Ghomeshi

As Table IV shows, adding extra content in the form of an opinion was very uncommon, with less than 2 percent of the sampled live tweets doing so. Even though journalists have the option to opine in the venue, this possibility is not routinely exercised.

Another concern in the normative literature is the lack of editorial control over what is posted by journalists on Twitter. As mentioned, Harada and McGuire (2016, 142) found evidence of journalists making “editorial decisions on the fly,” indicating in their tweets that they would not be describing the contents of a video or photograph or using phrases like “warning” or “graphic content.” There is some evidence of this in the two trials in our study. Editorial decisions were evident in 5 percent of the Ghomeshi trial and just under 10 percent in Saretzky. Here are two different of examples:

• Correction to earlier tweets that say the complainant had an interview with Global -- it was with CBC’s The National.

• “Day 7 of #Saretzky murder trial to start shortly. WARNING: Some tweets will contain graphic details. Mute #Saretzky… https://t.co/zfcwsBdDCk”

The first, from Sarah Boesveld, is a correction to a previous tweet, while the second is from Patrick Burles and is a warning applied to a suite of forthcoming tweets. It is not possible to know what information about the trials the journalists did not share as a result of “on the fly” editorial decisions. Most often, journalists made editorial decisions to provide warnings prior to tweeting graphic content or to correct a spelling error in a previous tweet, which supports Sossin and Bacal’s assumption that professional journalists will actively engage in fact-checking and the filtering of content (Sossin and Bacal 2013). Overall, these data do lend some evidence to the assumptions of Hewett and Hall-Coates regarding editorial oversight, to the extent that a live-tweeting journalist must engage in self-editing and oversight (Hewett 2015; Hall-Coates 2015).

Discussion and Conclusions

In speaking about the challenges raised by the digital communication revolution, Beverly McLachlin (2012), former Chief Justice of the Supreme Court, notes that the judicial system and the legal community will have little choice but to confront the challenges and opportunities presented by the expansion of digital technology into the courtroom. The greater integration of digital technology in the legal system raises many issues, such as the use of electronic records by litigants, the ethics of social media use by judges, and the admissibility of electronic documents (Eltis 2011; Sossin and Bacal 2013; Coughlan and Currie 2013). By focussing on live-tweeting, this paper attempts to understand one aspect of these challenges. While we focus on journalists, much remains unknown in understanding the role of live-tweeting in the courts. Depending on the trial and the court, non-journalists might live-tweet. While Puddister and Small (2019) found that most Canadian courts only allow journalists to live-tweet with judicial permission, some courts also allow the general public to do so. For instance, it has become practice for legal experts, such as law professors, to live-tweet high-profile Supreme Court of Canada proceedings (Devlin and Dodek 2016). Whether the live-tweeting by a law professor is similar in style to that of journalists remains a question for future research. It seems reasonable to think that law professors will focus more on providing colour commentary, given their training. Others, including those who are non-experts or those who have a vested stake in the trial outcome (such as activists), might also live-tweet, which could also have an impact on the principle of open court and may give rise to a different set of concerns regarding live-tweeting not applicable to professional journalists. Indeed, it is unlikely that our findings about opining and linking would hold up for this group of actors. Finally, the other group of participants worthy of study is the general public, those who are following the live-tweets and simply have opinions about a trial. The Ghomeshi trial was notable for the amount of Twitter attention it received. The following are just a few examples of tweets after the verdict:

• Imagine if we reacted with this level of interest, compassion and emotion to sexual assault cases that don’t involve a celebrity. #Ghomeshi

• not guilty by reason of patriarchy. #ghomeshi

• Feels wrong to selfie but my 10yr old daughter & I came to old city hall to stand with the women & all who want change. End #VAW #ghomeshi

Hence, future research should also consider how the general public engages with the court system through Twitter and the effect of live-tweeting on individuals and activists who have a vested interest in the outcome and the process of the trial (i.e., demand research). While the examination of live-tweeting by journalists provides an important starting place for considering the empirical and normative role of social media in the courts, further analysis is needed.

To that end, a central contribution of this article has been to provide a methodological framework for the analysis of live-tweeting of courtroom proceedings. As noted above, the stream-like nature of live-tweeting requires researchers to eschew random sampling of tweets, and instead rely on a purposive sampling. Purposive sampling allows for the analysis of an entire thread of tweets, ensuring that each tweet can be understood and examined in relation to the content that precedes and succeeds it. While such choices preclude random sampling, they ensure that necessary context is not lost, allowing for a full appreciation of the work of the live-tweeter. Moreover, the coding scheme developed could be readily used in the analysis of live-tweeting by journalists in other court contexts, such as high-profile civil or constitutional trials, and could be applied to the other types of live-tweeters discussed above.

Our analysis provides a starting point for future research by putting an empirical face on an overly normative discussion amongst scholars, jurists, and commentators about the ways in which digital technologies enhance or hinder court processes. Overall, we find that the live-tweeting of criminal trials in Canada is akin to play-by-play by sports broadcasters. The live-tweeters in our study spent much of their time tweeting details of the trial as they happened in real time, including verbatim quotes and actions in the courtroom just like a play-by-play broadcaster would do in a baseball game. Little extra content or “colour” was added to these tweets. The data shows that the journalists did not frequently write tweets that provide analysis and opinion of the courtroom proceedings, nor did the live-tweeters routinely interact with their followers by responding. Interestingly, when comparing the live-tweeting of Ghomeshi and Saretzky, the results are notably quite similar. Despite several differences between the two trials, including the seriousness of the crime (and the severity of the potential sentence), status of the accused, level of court and jurisdiction, there were remarkable similarities in how these trials were live-tweeted. Our analysis demonstrates that live-tweets provide a very accurate description of who said what, and what happened during the trials, but little else.

Reflecting on the normative debate regarding the use and appropriateness of Twitter in courtrooms, we find that the advantages and disadvantages are overstated in the literature. Many of the criticisms regarding live-tweeting centered on the lack of editorial oversight and loss of control over a tweet once it is posted. While these concerns hold merit, our data did not provide evidence to support these concerns in the two trials studied. The results provide evidence to suggest that the concerns regarding the inability to provide context and explain the complexities of a criminal trial through live-tweeting are real. We found that due to the stream-like nature of live-tweeting, a single tweet often does not contain the necessary context to be understood in isolation from the tweets that preceded it. As a result, along with the criticism that context is not possible in live-tweeting, the claims that twitter can provide an educative function (a perceived benefit) are also overstated. That being said, the fact that a trial is being live-tweeted may cause the court actors, such as the judge or prosecutor, to monitor their behaviour, knowing any missteps could be instantly shared on social media. This could provide a form of oversight and accountability that cannot be captured by our analysis.

In denying the media’s request to televise the trial of Gerald Stanley in the death of Colten Boushie, Justice Popescul explained that since the media had the ability to “live-tweet and post or engage in other live text-based communication from court,” the principle of open court was maintained (R v. Stanley [2018] S.J. No. 95, para 75). Given our conclusions, live-tweeting should not been seen as an alternative to arguments for live video streaming or television broadcasting of court proceedings. Though live-streaming or televising criminal trial proceedings, where witnesses are involved (as in the trials examined here) certainly presents challenges for courtroom decorum, sensationalism, and the privacy of participants, these concerns are not applicable in the case of appellate courts, which are primarily concerned with issues of law and rarely hear from witnesses (Puddister and Small 2019; McLachlin 2012). Our findings suggest that the use of live-tweeting should not foreclose future debates regarding the merits and challenges that these media present for Canadian courtrooms, both at the trial and appellate level. At best, the live-tweeting of criminal trials only minimally enhances the open court principle. The thoughtful analysis and contextualization that is possible in routine reporting by court journalists cannot be replaced by live-tweeting.