Journal of Mason Graduate Research Volume 5 Number 2 Year 2018 © Author ISSN: 2327-0764 Pages: 108-127 Assault and Accusation Without Agents: Verb Voice in Media Narratives of Campus Sexual Assault RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ George Mason University Previous studies of sexual assault have analyzed the connection between the narrative of events surrounding sexual assault allegations and the agency of the narrative’s characters, especially the assailant and the accuser. To contribute to this growing literature, we conducted a mixed methods word-level content analysis, testing whether the voice of verbs associated with the actions of the assailant and the accuser indicated an absence of agency. This study found that verbs associated with assailants were primarily written in passive voice and verbs associated with accusers were primarily written in active voice in both campus and non-campus sexual assault news reports. Implications for the research and practice of mass media reporting of sexual assault are discussed. Keywords: Agency, sexual assault, grammar, mixed methods INTRODUCTION News reports of sexual assault cases are frequently accused of engaging in victim blaming—assigning more responsibility for the assault to the victim than to the assailant. Victim blaming is visible at the macro level of news stories, such as in the construction of the narrative and the representation of the characters, or agents, in that narrative (Barnett, 2008, 2012; Worthington, 2005; 2008a, 2008b). Perhaps more subtly, victim blaming also appears at the micro level of a news story, in the choice of words and grammar of the sentences comprising the story. A common micro device associated with victim blaming is the use of passive voice when describing the assault. When someone writes that “a victim was attacked,” readers disassociate the assailant with the action and will even hold the subject of the phrase—a victim—responsible RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ 109 for the action—attacked (Bohner, 2001; Henley, Miller & Beazley, 1995). Although victim blame is generally accepted as a common rape myth—a prevailing but false belief associated with rape causes and factors—the appearance of victim blame changes when applied to news stories about campus sexual assaults. Campus sexual assaults—assaults that occurred on a college campus, almost exclusively involving students enrolled at the college—have become a focus of national concern in the United States in recent years. Politicians have formally addressed the campus sexual assault as a problem that requires new policy solutions (The White House, 2014; McCaskill, 2014). Not only have campus sexual assault trials gained national media coverage, but criticism of how news reporters cover these trials has also gained national attention (Beck, 2013; Hargis, 2016; Henneberger, 2013; LaChance, 2016). In a particularly controversial news report of a campus sexual assault, Rolling Stone published the story “A Rape on Campus” which was retracted within months and led to lawsuits against Rolling Stone (Haag, 2017; Sisario, 2017). To better understand the news reporting of campus sexual assault, a 2016 study of 93 campus sexual assault stories in The Washington Post found that most news stories diminished or dismissed the agency of both victims of assault and assailants (Lawrence, Stabile & Fernandez, 2016). That is, when referring to the actions—or even the ability to act—of the primary actors in the story, most media narratives failed to account for the assailant’s agency in the action of the assault or the victim’s agency in reporting the assault or accusing the assailant. For this study, we extend the work of the 2016 study to determine whether the use of verb voice in the news reporting of sexual assault and campus sexual assault ascribes agency to the assailant or the victim when referring to actions associated with either the assault or the accusation. Furthermore, we analyze whether the frequency of active or passive voice use in news reports of sexual assault changed before or after the publication of the Rolling Stone article “A Rape on Campus,” which highlighted and politicized campus sexual assault and the way it is reported. Finally, we offer recommendations for future research in the area, including other avenues of research that might help illuminate how media narratives of sexual assault and campus sexual assault are currently constructed. LITERATURE REVIEW National Response to Campus Sexual Assault ASSAULT AND ACCUSATION WITHOUT AGENTS 110 In the last several years, campus sexual assault has become the focus of increased government and media attention in the United States. With the “Dear Colleague” letter of 2011, the U.S. Department of Education’s Office for Civil Rights (OCR) notified all institutions subject to Title IX, including school districts, colleges, and universities, that OCR would consider the sexual harassment of students, including acts of sexual violence, a form of sex discrimination prohibited by Title IX of the Civil Rights Act of 1964 (US Department of Education). In September 2014, the Obama administration formed the “Task Force to Protect Students from Sexual Assault.” In 2015, the Campus Accountability and Safety Act, championed by Senator McCaskill (D-MO) was passed. Also in 2015, the Association of American Universities (AUU) conducted a survey of 27 campuses, which collected 150,000 responses and concluded that “more than 20 percent of female undergraduates at an array of prominent universities said this year they were victims of sexual assault and misconduct” (Anderson & Svrluga, 2015). These initiatives and activities demonstrate a growing public awareness of the problem of campus sexual assault and a demand for greater accountability on the part of universities in preventing sexual assault crimes. In addition to the increased focus from the highest levels of the federal government, high- profile sexual assault cases at Harvard, Columbia, Princeton, the Naval Academy, and James Madison University brought renewed attention to the institutional responses of universities in cases of sexual assault. Recent high-profile cases, such as Brock Turner’s 2016 conviction and subsequently lenient criminal sentence for the assault he committed at Stanford University (Meixler, 2017), have highlighted not only how universities respond to sexual assault allegations but also how the judicial system can fail to adequately respond to cases of campus sexual assault. While many argue that universities and judicial systems do not do enough to prevent and deter campus sexual assault, others have claimed that universities are ill-equipped to handle sexual assault. In her 2015 article, “The College Rape Overcorrection,” Emily Yoffe argues that the methodologies used to compile commonly used statistics are flawed and require more attention. Yoffe also argues that the foundational studies that are often used to characterize perpetrators are methodologically flawed and give an exaggerated impression of perpetrators. This exaggeration leads universities to overreact, by decreasing evidentiary standards and adjudicating cases in ways that violate the rights of the accused and do not address the needs of RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ 111 the accuser. Similarly, in an editorial in The Washington Post, George Will suggested that the problem was exaggerated (Will, 2014). Media Narratives of Sexual Assault Media attention to high-profile sexual assault cases, particularly with scandals involving universities, such as the 2006 Duke lacrosse case, is not new; reporting of campus sexual assault cases has been previously studied, particularly in communication and feminist studies. Barnett (2012) and Worthington (2005; 2008a, 2008b) found that media narratives of campus sexual assault describe university administrators as covering up incidences of sexual assault to protect the reputation of their campuses. Worthington (2005; 2008b) also shows that the universities’ mishandling of cases and the inadequate punishment of male perpetrators are common themes in media reporting. Research has been conducted on the role that fraternities and athletic teams play in the perpetuation of rape culture (Krakauer, 2015; Sanday, 2007) and on the role that the race of the perpetrator plays in sexual assault narratives (Patton & Snyder-Yuly, 2007). Several studies trace the narratives of sexual assault that are prevalent in media coverage. Barnett (2012) shows that news reports portray women as innocent and victimized or as wanton and deserving, while men are portrayed as sex fiends. In addition, Barnett (2012) finds that news accounts of rape are often selective and superficial, which adds to the confusion surrounding the terminology used to cover rape stories, for example, when reporters use terms for lovemaking to describe assaults (Barnett, 2012, p. 14). Worthington (2005) describes some of the constraints investigative reporters face when reporting on campus sexual assault stories; for example, a college’s litigious reputation might impact how the story is represented (p.10). However, Worthington (2008a, 2008b) states that reporters can meet the requirements for progressive journalistic reporting of gender violence by applying four criteria that emerge from critical feminist literature: a) the story selection reflects the types of crimes that occur, especially an emphasis on acquaintance rape, which is more prevalent than stranger rape; b) the narrative avoids stereotypes that either blame the victim or mitigate suspect responsibility; c) more attention is given to the role of social structures in causing and normalizing gender violence; d) and the perspectives of victims and advocates are more fully incorporated. Building on this literature of media narratives in sexual assault and with the goal of informing campus sexual assault policy, Lawrence et al. (2016) conducted a narrative analysis of ASSAULT AND ACCUSATION WITHOUT AGENTS 112 two years of sexual assault reporting within the Washington Post. Lawrence et al. compiled 458 articles focused on sexual assault and analyzed the rhetorical narratives of 93 campus sexual assault stories. They found that these stories consistently de-emphasized the agency of both the assailant and the accuser in the sexual assault, while emphasizing the agency of the university as bearing both blame and responsibility. That is, the study found that universities were either praised or blamed for how they addressed sexual assault allegations and that there was an expectation for universities to take action. In most news reports, there was no similar expectation of the students—either assailants or victims—to take action. Verb Voice and Agency Whereas the Lawrence et al. (2016) study analyzed how agency in sexual assault stories is represented in narrative, other studies have analyzed how agency is represented at the language level, such as verb voice. Henley et al. (1995) explained how the use of passive voice in descriptions of attacks upon women influences readers’ perceptions of victim responsibility in the attack. In “Syntax, Semantics, and Sexual Violence Agency and the Passive Voice,” Henley et al. described three studies they conducted as a part of an overarching project that examined the prevalence and impact of verb voice in reports of violence against women. The first and second study analyzed the content and semantics of news stories about violence against women, published in two six-month periods in a major U.S. newspaper. These two studies found that descriptions of sexual violence and nonsexual violence committed by male perpetrators were predominantly constructed with passive voice. The third study in Henley et al.’s (1995) research was an experiment that tested the influence of verb voice in mock news stories on readers’ perceptions of violence and its effects. This study found that when news stories about sexual violence predominantly used passive voice, readers were more likely to assume lower levels of harm suffered by the victim and to assign lower levels of responsibility to the perpetrator than to the victim. The findings of these studies suggest that the grammar used in news stories about violence against women can and do perpetuate negative conceptions of victim responsibility in cases of sexual assault. In addition to analyzing the influence of verb voice upon the reader, a study by Bohner (2001) correlated the use of passive voice to describe sexual assault with the writer’s own beliefs about the responsibility of the victim in the assault. Bohner (2001) conducted an experiment and survey RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ 113 that found that writers who primarily use passive voice to describe sexual assault are more likely to believe in rape myths about the responsibility of the victim in the rape. This experiment also found that participants used passive voice when referring to the rape itself significantly more often than they used passive voice to describe other actions by either the perpetrator or the victim immediately before or after the rape (Bohner, 2001). Although the participants in the Bohner (2001) experiment were not professional journalists, it is worth noting that a writer’s bias influences the writer’s grammatical choices in the reporting of sexual assault, which can then contribute to the development of bias among its readers. It is also worth noting that the style guide for American journalists, The Associated Press Stylebook and Briefing on Media Law (2016), provides no guidance for which verb voice to use when reporting on sexual assault, except that leads be written in active voice. Our study, therefore, sheds light on an important topic that has been understudied in the literature. Although previous studies have analyzed the major narrative elements of news coverage of rape and sexual assault, both generally and specifically for campus sexual assault, and while other studies have analyzed linguistic characteristics of sexual assault narratives, no study has yet used the same dataset to analyze both macro and micro features of media narratives of sexual assault. In addition, although there have been linguistic analyses of sexual assault generally, there is no study of verb voice in media narratives specific to campus sexual assault. METHODS We conducted a word-level content analysis of the dataset from study by Lawrence et al. (2016), with a focus on both general and campus-specific sexual assault stories, in order to provide a more nuanced analysis of media narratives of sexual assault. Understanding how media narratives construct agency at the macro and micro level can provide additional insight into how this reporting shapes public perceptions of victims and assailants. Four hypotheses guide our study. The first hypothesis tests whether the results consistent in the literature on verb voice in sexual assault descriptions can be replicated in the dataset of 458 articles on sexual assault from The Washington Post: ASSAULT AND ACCUSATION WITHOUT AGENTS 114 H1: In the news reporting of sexual assault, verbs associated with assault will primarily be written in passive voice. The second hypothesis seeks to confirm whether the results of the macro-level 2016 analysis are visible in a micro-level analysis as well. That is, because verb voice indicates the assignment of agency at a micro level, and because the previous study found that the narrative of campus sexual assault elided the agency of both the assailant and the accuser, this hypothesis tests whether the dataset of The Washington Post articles on campus sexual assault predominantly use passive voice for verbs that describe the actions of the assailant and of the accuser: H2: In the news reporting of campus sexual assault, verbs associated with assault and accusation will primarily be written in passive voice. Furthermore, to analyze a gap in the literature on verb voice in sexual assault, the following hypothesis tests whether the lack of agency ascribed to accusers in the campus sexual assault narrative is particular to campus sexual assault: H3: There will be a difference between news reports on campus sexual assault and news reports on non-campus sexual assault in the use of passive voice for verbs associated with accusation. Finally, to continue a line of inquiry from the 2016 Lawrence et al. study, the following hypothesis tests whether the frequency of passive voice changes around the time of publication of the Rolling Stone article “A Rape on Campus”: H4: When describing actions associated with either the assault or the accusation, news reports published in two periods preceding and two periods following the “A Rape on Campus” article will use passive voice to different extents. RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ 115 Data We use a dataset compiled from articles on sexual assault and campus sexual assault in The Washington Post from January 1, 2013, to January 26, 2016 because the news outlet has national readership and covers both national and international news. We selected the dataset used for this study for two reasons. First, this study sought to confirm findings from a previous study using the same dataset, which found that for the subset of campus sexual assault, blame and responsibility were often assigned to the university, while the assailant and accuser often were not assigned agency. Rather, the assailant and accuser were acted upon, primarily in their role as students (Lawrence et al., 2016). Therefore, this existing dataset provided the opportunity to extend prior research on media narratives of sexual assault. Second, UVA, the basis for the Rolling Stone article, is located about 80 miles from the Washington DC metropolitan area, and researchers expected that the story would receive consistent coverage from The Washington Post. The articles used in the dataset were published a little more than a year before and after Rolling Stone published “A Rape on Campus.” The key terms used in the dataset were rape, sexual violence, sexual assault, and campus sexual assault. Excluded from this dataset were police crime reports, articles that included the phrase sexual assault but did not cover instances of sexual assault, and AP-generated reports. The remaining dataset consisted of 458 articles about sexual assault. The researchers categorized these 458 articles according to the main focus of the story, including campus, UVA/Rolling Stone (UVA/RS), military, and international. To complete our analysis on campus sexual assault articles, we combined two categorized subsets: 93 campus (non-military) and 51 UVA/RS. These combined subsets consist of 144 articles that dealt directly with campus sexual assault. Analytic Procedure To supplement the findings of the previously conducted narrative analysis, we applied quantitative tools and concepts derived from sociolinguistic and communication scholarship. We used an Excel macro to identify words associated with the assailant and the accuser1 and 1 Much of the literature on sexual assault and verb voice uses the term victim to refer to the person who is assaulted and reports the assault. However, for this study, the authors use the term accuser, to clarify the agency enacted by this character in the campus sexual assault narrative. Consistent with the literature, the authors use the term assailant to refer to the person who commits the assault and is accused of the assault. ASSAULT AND ACCUSATION WITHOUT AGENTS 116 searched the dataset of 458 articles on sexual assault for six target verbs. We searched for three target verbs associated with the actions of the assailant: assaulted, raped, and attacked. We also searched for three target verbs associated with the actions of the accuser: accused, reported, and claimed. Drawing on the study by Henley et al. (1995), this study limited target verbs to the past tense form because this form remains the same regardless of verb voice. We conducted word-level content analysis to determine whether the identified verbs appeared in passive voice, active voice, or other (e.g., participles). See Table 1 for examples of identified and coded phrases. We inputted the data into IBM SPSS 24, noting the coded verb voice type for each target verb. We also categorized the target verbs by three additional variables: ● Agent Type: whether the target verb referred to the action of an assailant or accuser. ● Campus Focus: whether or not the target verb originated in a story specific to campus sexual assault. ● Period of Publication: whether the verb originated in an article published in the first or second half of the time preceding or following Nov. 19, 2014, the date of publication of “A Rape on Campus.” We then used IBM SPSS 24 statistics software to analyze the frequency of passive and active voice usage and to determine whether verb voice correlated with agent type, campus or non- campus focus, and period of publication. In addition, we noted the number of target verbs that appeared in each of the 458 articles. RESULTS OF A QUANTITATIVE WORD-LEVEL CONTENT ANALYSIS Of the 458 articles in the dataset, 363 articles contained between one and 21 uses of at least one of the six verbs, with a total of 1,079 words identified. Of these, 597 words were coded as active (N = 218) or passive (N = 379) and as assailant (N = 397) or accuser (N = 200), for a total of four agent and voice combinations: active assailant verbs (N = 115), passive assailant verbs (N = 282), active accuser verbs (N = 103), and passive accuser verbs (N = 97). RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ 117 Table 1 Examples of Coded Sentences Voice Verb Example Active Assaulted “...he overcame her and sexually assaulted her…” Raped “...took her to a remote spot and raped her…” Attacked “He allegedly attacked Outland three months later.” Reported “...this person reported the assault soon after the incident…” Accused “...victims have accused high-profile men…” Claimed “...two women claimed he had sexually assaulted them...” Passive Assaulted “...one in five women will be sexually assaulted...” Raped “...how Jackie was raped…” Attacked “...the victim was attacked after being followed…” Reported “Fifteen assaults were reported…” Accused “...they had been falsely accused.” Claimed N/A The coders identified the remaining 482 words as Other: words that were not used as verbs (including participles, gerunds, and adjectives) or verbs enacted by an agent other than the assailant or accuser (such as police officers, lawyers, university staff, and news sources). Table 2 details the frequency of each of the six coded verbs. Table 2 Frequency of Six Verbs Agent Verb N Percent Assailant Assaulted 159 26.6 Raped 191 32.0 Attacked 44 7.4 Accuser Reported 105 17.6 Accused 86 14.4 Claimed 12 2.0 Total 597 100.0 To test whether verbs associated with assault were primarily written in passive voice (H1), we compiled descriptive statistics on the 397 assailant verbs. Descriptive statistics showed that 71% of assailant verbs were in passive voice, whereas 29% of assailant verbs were in active voice. ASSAULT AND ACCUSATION WITHOUT AGENTS 118 To test (H2), we analyzed descriptive statistics to determine whether assailant and accuser verbs in news reports of campus sexual assault were primarily written in passive voice. Of the 182 verbs in this data set, 29 (15.9%) were active verbs associated with the assailant, 22 (12.1%) were active verbs associated with the accuser, 113 (62.1%) were passive verbs associated with the assailant, and 18 (9.9%) were passive verbs associated with the accuser. Furthermore, a chi-square test was conducted to assess whether the different uses of active and passive and assailant and accuser verbs were significantly different than expected for a random distribution. The result for this test was significant: X2 (1, N = 182) = 18.5, p < .001. To test H3, we conducted a chi-square test to determine whether news reports on sexual assault use passive voice for verbs associated with accusation to a different extent than news reports on campus sexual assault. The result for this test was not significant, demonstrating that there is no significant difference in the use of passive voice for accuser verbs among all stories on sexual assault: X2 (1, N = 200) = .25, p > .05. To test H4, we conducted a chi-square test on all 597 verbs, categorized by their publication within one of four periods of time: • Period 1: January 1, 2013 to December 10, 2013 • Period 2: December 11, 2013 to November 18, 2014 • Period 3: November 19, 2014 to June 23, 2015 • Period 4: June 24, 2015 to January 26, 2016 Combined, these periods mark the total timespan of publication of the news reports in our dataset. We divided this timespan by the date of publication of “A Rape on Campus,” November 19, 2014, and we then divided these two halves of the timespan by half again. The chi-square test was significant. That is, the rate of active and passive voice for both assailant and accuser verbs in these four time periods was significantly different than what one would be expect from a random distribution (X2 (9, 597) = 31.97, p < .001). More importantly, using Bonferroni-adjusted Z tests, we found that passive accuser verbs were less likely to appear in the third period and passive assailant verbs were more likely to appear in the third period than in any other period (See Figure 1). There were no other differences in proportional verb use for RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ 119 any of the active verbs in the four time periods. This indicates that the greatest discrepancy between the actual and expected counts occurred in the third period, when passive assailant verbs were most frequent and passive accuser verbs were least frequent. This result is meaningful because the third period includes all of the relevant dates for the publication of “A Rape on Campus,” including when the investigation of the article began and when Rolling Stone officially retracted the piece. Figure 1 shows the rate of usage of each agent and voice type over the four time periods, highlighting the sudden changes in passive and active voice usage in the third period. Figure 1. Percentage of verbs by agent and voice type over four time periods. THE IMPACT OF VERB VOICE IN MEDIA NARRATIVES OF SEXUAL ASSAULT To extend the research of verb voice and agency in stories of sexual assault, this study examined a larger dataset over a longer period of time than previous studies, the last of which was conducted more than 15 years ago (Bohner, 2001). In testing H1, the results of this study ASSAULT AND ACCUSATION WITHOUT AGENTS 120 aligned with previous studies, finding that most news reports of sexual assault primarily write about the actions of assailants in passive voice. In fact, the difference between the use of passive and active voice for verbs associated with actions of the assailant was greater than 2 to 1. This finding is meaningful in part because the results of our larger dataset align with previous smaller studies, and because the finding demonstrates that the patterns of verb voice usage in writing about sexual assault have not changed in more than 15 years. It is understandable, to a degree, that journalists writing about sexual assault avoid using active voice when they do not know or do not assuredly know (that is, without a legal conviction) who committed a sexual assault. However, there are ways to write about sexual assault in active voice without directly assigning blame to a specific individual (such as, “the assailant attacked”). In addition, as previous literature shows, the preponderance of passive voice usage to refer to the actions of assailants leads readers to assign more blame to victims of assault than to assailants (Henley et al., 1995). The findings of the previous studies and this current study indicate that there is a problematic trend in the grammar of news stories about sexual assault. While it is important that more attention be paid to the grammar used to depict assailants, a micro-level analysis provides further insight into the implications for use of voice for both the assailant and the accuser in campus sexual assault narratives. While the macro-level analysis of campus sexual assault media narratives showed an absence of agency of both assailants and accusers (Lawrence, et al., 2016), in this micro-level analysis, we find that the accuser is assigned agency through use of active voice. This finding (H2) demonstrates that the micro-level analysis of the same dataset differs in a very specific way from the macro-level analysis; writers do assign some level of agency to accusers. The actions of assailants are still overwhelmingly depicted with passive voice, which confirms findings from previous studies (Henley et al., 1995; Lawrence et al., 2016). However, at the word level, the depiction of accusers is more problematic and might indicate a degree of victim blaming that is not captured by the macro-level analysis of the narrative. Furthermore, from a methodological standpoint, this nuanced but important difference between the findings of the narrative analysis and the word-level content analysis supports the claim of Sandberg and Ugelvik (2016) that narrative analysis can benefit from an increased use of sociolinguistic and quantitative methods. Because the findings from testing H2 did not align with the findings from the macro-level RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ 121 analysis (Lawrence et al., 2016), the findings from testing H3 are meaningful because they continue to highlight this discrepancy between what the narratives in sexual assault stories reveal and what the grammar of sexual assault stories imply about the agency of accusers. Ultimately, in an attempt to address a gap in the literature on the agency of accusers, this finding suggests that both campus sexual assault stories and non-campus sexual assault stories assign agency to accusers in much the same way. Further research is needed to determine whether non-campus sexual assault stories elide the agency of the accuser at the narrative level (as campus sexual assault stories do, according to the study by Lawrence et al.) despite assigning agency at the word level. Such research would indicate the extent to which implicit victim blaming happens in the reporting of sexual assault stories. Finally, the findings from testing H4 indicate that a controversial article on sexual assault seems to impact how other stories on sexual assault are written. In their investigation of the reporting failures in “A Rape on Campus,” Coronel et al. (2015) expressed concerns that the issues surrounding the faulty reporting of the story would negatively impact the extent to which readers believe claims of sexual assault: “[T]he magazine’s failure may have spread the idea that many women invent rape allegations” (p. 52). Although this current study did not investigate how readers perceive the responsibility of assailants and accusers, previous studies show that when writers assign less responsibility to assailants and assign more blame to accusers, these biases are reflected in the writers’ uses of verb voice (Bohner, 2001). Therefore, because the findings from testing H4 indicate that a change in grammatical choices of sexual assault reporting was associated with the publication and retraction of “A Rape on Campus,” one can infer that “A Rape on Campus” impacted how writers perceived assailant and accuser responsibility, which ultimately influences how readers perceive assailant and accuser responsibility. LIMITATIONS While this study has shown the importance of conducting micro-level analysis that informs and expands on macro-level analysis, and while it has confirmed that assailants are overwhelmingly depicted in passive voice, the study has at least three limitations. First, the dataset is large but not comprehensive. Further studies that include a wider variety of publications, a different time period, or news reports on different types of crime might shed light ASSAULT AND ACCUSATION WITHOUT AGENTS 122 into variations in the use of passive and active voice in news reporting. Second, although journalistic standards recommend writing in active voice, we have no way of knowing whether passive voice is used in similar ways for reporting of non-sexual assault crimes. Future research is needed to understand and compare verb voice in general reports of crime to reports of sexual assault. Third, the coding scheme used in this analysis must also be carefully analyzed, as the verbs chosen to indicate actions of the accuser—attacked, reported, and claimed—might have generated a higher number of instances of active voice, because the words themselves are more easily constructed when using active voice. This is most true for claimed. For example, “the rape was claimed to have happened” is an awkward phrase but one of the few ways in which the verb claimed could appear in passive voice. Then again, there were so few instances of claimed, relative to the rest of the verbs in the dataset, that the impact of this verb on the analysis is likely minimal. Additionally, although the assailant is always the assailant, the accuser is not always the victim, as in the case of someone who was assaulted and then murdered. Given the nature of passive voice, we could not definitively say who completed the action of accusing in every instance. However, this actually lends more credence to the findings, because it suggests that some of the passive actions of the accuser were not in fact the actions of the victim. Therefore, if those instances were controlled for, active verbs associated with the victim as accuser would constitute a higher percentage. CONCLUSION This study expands on previous studies on narrative analysis of crime by using quantitative methods to uncover nuanced but enduring patterns in the reporting of sexual assault crimes. This study also expands on previous studies of verb voice in sexual assault stories by using the largest dataset for this type of analysis yet. We believe that applying a quantitative method to the study of narratives of sexual assault crimes provides a more complex picture of how journalism reporting practices participate in the construction of problematic narratives of sexual assault crimes. The findings of this study show that the way news reports on sexual assault are written now retains journalistic habits that cross decades. That is, practices for reporting on sexual assault have not changed in the more than 15 years since the research conducted by Henley et al. (1995) and Bohner (2001). Despite calls to reduce victim-blaming in news reporting RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ 123 (Worthington, 2008a, 2008b), the grammar of the sentences that describe sexual assault cases— which has been shown to influence readers’ perceptions of victim and assailant responsibility in the crime (Henley et al., 1995)—also has not changed. In addition, this study shows that the publication of “A Rape on Campus” may have affected how journalists used language (either consciously or subconsciously) to characterize the agency of both the assailant and the accuser in sexual assault news stories. Further research must examine more closely the impact of salacious narratives on subsequent reporting of related stories. When victims of sexual assault are already so frequently discredited, it becomes that much more important that the people communicating these stories do so with care—from the nature of the story itself (for example, journalists must accept the challenge of reporting on complex and nuanced accounts rather than seek narratives with easily identifiable villains), down to the grammar of the sentences. Finally, the impact of social media in representations of crime, and particularly sexual assault, has been studied extensively in recent years (Dodge, 2016; Powell, Overington, & Hamilton, 2017; Salter, 2013). While studies of new media provide important insight into the narrative constructions of victims, perpetrators, and sexual violence, the findings from this study show that we must continue to analyze traditional news coverage to trace how sexual assault narratives construct the victim and the assailant. The year 2017 represented a pivotal moment in the conversation surrounding sexual assault in U.S. media outlets and in media outlets around the world. Notably, the #metoo campaign brought attention to the stories of women from many different backgrounds, and The Silence Breakers were named TIME’s 2017 Person of the Year. Although the trending #metoo hashtag was itself a social media phenomenon, the exigency for the trend was the traditional news coverage conducted by American journalists throughout the year as well as the publication of high-profile survivor stories in The New York Times and other news sources. This movement shows the relevancy of professional journalism for American perceptions of sexual assault and sexual harassment. Given the importance of news coverage for cultural shifts in attitudes toward sexual assault, it is critical that the news coverage of sexual assault avoid victim-blaming patterns themselves, patterns that this and previous studies have shown to be deeply ingrained in the way sexual assault news stories are reported. Furthermore, this relevancy of traditional news ASSAULT AND ACCUSATION WITHOUT AGENTS 124 coverage of sexual assault shows the importance for researchers in criminology, media, and cultural studies to continue to investigate the impact of this medium on the public and on social media movements. REFERENCES Anderson, N. & Svrluga, S. (2015, September 21). What a massive sexual assault survey found at 27 top US universities. The Washington Post. Retrieved from https://www.washingtonpost.com/ Barnett, B. (2008). Framing rape: An examination of public relations strategies in the Duke university lacrosse case. Communication, Culture & Critique, 1, 179–202. doi: 10.1111/j.1753-9137.2008.00018.x Barnett, B. (2012). How newspapers frame rape allegations: The Duke university case. Women and Language, 35(2), 11–33. Beck, L. (2013, March 18). Here’s what CNN should’ve said about the Steubenville rape case. Jezebel. Retrieved from https://jezebel.com/ Bohner, G. (2001). Writing about rape: Use of the passive voice and other distancing text features as an expression of perceived responsibility of the victim. British Journal of Social Psychology, 40, 515–529. doi: 10.1348/014466601164957 Coronel, S., Coll, S., & Kravitz, D. (2015). The rolling stone report. Columbia Journalism Review, 54(2), 48–64. Dodge, A. (2016). Digitizing rape culture: Online sexual violence and the power of the digital photograph. Crime, media, culture, 12, 65–82. Haag, M. (2017, April 11). Rolling Stone settles lawsuit over debunked campus rape article. The New York Times. Retrieved from https://www.nytimes.com/ Hargis. C. (2016, June 6). How sexual assault stories should be covered: A tale of two news reports. Media Matters. Retrieved from RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ 125 https://www.mediamatters.org/blog/2016/06/06/how-sexual-assault-stories-should-be- covered-tale-two-news-reports/210757 Henley, N. M., Miller, M., & Beazley, J. A. (1995). Syntax, semantics, and sexual violence agency and the passive voice. Journal of Language and Social Psychology, 14, 60–84. doi: 10.1177/0261927X95141004 Henneberger, M. (2013, March 18). Steubenville, the media and ‘rape, essentially.’ The Washington Post. Retrieved from https://www.washingtonpost.com/ Krakauer, J. (2015). Missoula: Rape and the justice system in a college town. New York, NY: Anchor Books. LaChance, N. (2016, September 2). Media continues to refer to Brock Turner as a “Stanford swimmer” rather than a rapist. The Intercept. Retrieved from https://theintercept.com/ / Lawrence, H., Stabile, B., and Fernandez, L. (2016). The rhetoric of campus sexual assault: Media constructions of student advocacy and campus citizenship. Poster presentation: Association of Teachers of Technical Writing Conference, Houston, TX. Meixler, E. (2017, December 4). Brock Turner is trying to get his sexual assault convictions overturned. TIME. Retrieved from http://time.com/ Patton, T. O., & Snyder-Yuly, J. (2007). Any four black men will do: Rape, race, and the ultimate scapegoat. Journal of Black Studies, 37, 859–895. Retrieved from http://www.jstor.org/stable/40034959 Powell, A., Overington, C., & Hamilton, G. (2017). Following #JillMeagher: Collective meaning-making in response to crime events via social media. Crime, Media, Culture, Advance online publication. doi: 10.1177/1741659017721276. Salter, M. (2013). Justice and revenge in online counter-publics: Emerging responses to sexual violence in the age of social media. Crime, Media, Culture, 9, 225–242. ASSAULT AND ACCUSATION WITHOUT AGENTS 126 Sanday, P. K. (2007). Fraternity gang rape: Sex, brotherhood, and privilege on campus. New York, US: NYU Press. Sandberg, S., & Ugelvik, T. (2016). The past, present, and future of narrative criminology: A review and an invitation. Crime, Media, Culture, 12, 129–136. Sisario, B. (2017, September 19). Rolling Stone faces revived lawsuit over campus rape article. The New York Times. Retrieved from https://www.nytimes.com/ The Associated Press. (2016). The Associated Press stylebook and briefing on media law. New York, NY: Basic Books. The White House, Office of the Press Secretary. (2014, September 19). Remarks by the president at “It’s On Us” campaign rollout [Press release]. Retrieved from https://obamawhitehouse.archives.gov/the-press-office/2014/09/19/remarks-president-its- us-campaign-rollout/ McCaskill, C. (2014, July 30). Bipartisan bill takes aim at sexual assault on college and university campuses [Press release]. Retrieved from https://www.mccaskill.senate.gov/media-center/news-releases/campus-accountability- and-safety-act U.S. Department of Education (2011). Dear Colleague Letter. Retrieved from http://www2.ed.gov/about/offices/list/ocr/letters/colleague-201104.html. Will, G. (2014, June 6). George Will: Colleges become the victims of progressivism. The Washington Post. Retrieved from https://www.washingtonpost.com/ Worthington, N. (2005). Negotiating news representations of rape: Reporting on a college sexual assault scandal. Media Report to Women, 33(4), 6–14. Worthington, N. (2008a). Encoding and decoding rape news: How progressive reporting inverts textual orientations. Women’s Studies in Communication, 31(3), 344–367. doi: 10.1080/07491409.2008.10162546 RACHAEL GRAHAM LUSSOS & LOURDES FERNANDEZ 127 Worthington, N. (2008b). Progress and persistent problems: Local TV news framing of acquaintance rape on campus. Feminist Media Studies, 8, 1–16. doi: 10.1080/14680770701824878 Yoffe, E. (2014, December 7). The college rape overcorrection. Slate. Retrieved from http://www.slate.com/ Literature Review Data Results of a Quantitative Word-Level Content Analysis The Impact of Verb Voice in Media Narratives of Sexual Assault Conclusion