Abstract
This study compares the news media’s visual framing of the Los Angeles Riots of 1992 and the Ferguson, Missouri, riots of 2014. A visual content analysis of 387 news images published in the New York Times, Los Angeles Times, Chicago Tribune and Atlanta Journal-Constitution was conducted. Results show that newspapers’ visual portrayals of civil unrest have shifted focus from the confrontation between protesters and the police to an emphasis on the depiction of morality, human interest and civil emotions.
Keywords
In Spring 1991, a video showed an African American man, Rodney King, was beaten excessively by police officers following a high-speed police chase in Los Angeles. A series of riots between police and civilians and an increase in racial tensions in the United States followed. Twenty-two years later, Americans experienced a civil-rights déjà vu, as Michael Brown, an unarmed African American and recent high-school graduate, died after being shot six times by a white Ferguson, Missouri, police officer. A theft of cigarillos sparked the initial confrontation. Civil unrest erupted in the town and the nation, as U.S. news media covered the fallout using imagery of protests and violence, as well as tears and vigils. Tensions between police and civilians across different races have since intensified, exemplified more recently by, among other events, the shooting of unarmed and innocent black male caregiver Charles Kinsey while he lay on the ground with his hands in the air.
While the United States is 50 years into what some call a post-civil-rights era, questions remain about the state of civil rights. This study examines the news media’s role in and reaction to this evolution, specifically how visual portrayals of these confrontations differ in two time periods.
Images are chosen as the subjects of this study because they are an important form of communication.1,2 Images are a first stop in eye-scanning for readers,3,4,5 increase readers’ memories of an accompanying text, 6 and create an emotional response to the news.7,8,9,10
This study compares the visual frames in news coverage of the Los Angeles 1992 Riot, the post-civil-rights movement and the Ferguson, Missouri, 2014 Riot. The study specifically addresses image frames, subjects portrayed, emotion, the balance of power, the races of subjects in the riots and the portrayal of weapons.
The passage of time and subsequent societal changes between the Rodney King and Michael Brown events would suggest a change in news media framing of this topic. Despite the contemporary relevance of the topic given tensions between civilians of all races and police officers of all races, and the power of images to form opinions on important societal and political issues, no one has yet studied these changes from a visual standpoint. This study attempts to fill this gap by examining four newspapers that represent the East, West, North and South of the United States.
This article presents a brief history of civil rights and police brutality in America. Media framing and visual framing literature, specifically, are discussed. Following that, a review of scholarly literature focusing on variables measured in this study is undertaken, specifically, literature on the image framing of racism and protests in America, framing of subjects portrayed in riots and image scheme and balance of power.
Literature Review
Civil Rights in America
The adoption of the 13th Amendment on December 18, 1865, ended legal U.S. slavery. This was the first time a constitutional amendment respecting individuals’ freedom was directed at the federal government, the states and, most significantly, to private individuals. 11 The 39th U.S. Congress considered many proposals designed to strike down the Black Codes, 12 laws passed by Southern states that restricted African Americans’ freedom after the Civil War. The civil rights movements of the 1960s, with its acts toward legal recognition of and federal support for African Americans, procured a ban on discrimination based on race, color, religion or sex. While the Civil Rights Act of 1964 and subsequent laws worked toward equality for African Americans, racial discrimination is still a perceived reality in the United States. 13 The police brutality suffered by Rodney King, Michael Brown and other African American men and women has shown that police brutality is a national phenomenon (with not only African Americans as victims) and that the U.S. Department of Justice should take action for these victims. 14
Police Brutality in America
“Police brutality” refers to unwarranted actions of police, including the use of profane and abusive language, threats to use force if not obeyed, prodding with a nightstick or approaching with a pistol and the actual use of physical force or violence itself as well as commands to move on or get home, stopping and questioning people on the street or searching them and their cars. Police brutality has a long history in the United States.
15
Frank Moss, a New York City (NYC) police commissioner, stated in 1903, There has been through the courts and the streets a dreary procession of citizens with broken heads and bruised bodies against few of whom was violence needed to effect an arrest. Many of them had done nothing to deserve an arrest. In a majority of such cases, no complaint was made. If the victim complains, his charge is generally dismissed. The police are practically above the law.
This police brutality largely remained a reality through the civil rights and post-civil-rights eras. In 1968, Reiss 16 stated there was “precedent in American history—but never before has the issue of police brutality assumed the public urgency it has today.”
More recently, examples of the beating of Rodney King, and the deaths of Amadou Diallo in 1999 and Michael Brown in 2014, showed how African Americans have been assaulted by the members of the police system.17,18,19 Even though a recent study indicates that the greater race and gender diversity in the NYC area has made the New York City Police Department (NYPD) behave better, African Americans, among all races, still are more prone to be victims of police brutality. 20
However, the real difference between “proper” and “brutal” is hard to define. For citizens in a democratic society, “police brutality” is the judgment of whether a person has been given full rights and treated with dignity by law enforcement. 21
Framing and Visual Framing of Riots and Police Brutality
Media frames’ power on shaping the social world has long been recognized by media effects researchers. 22 Framing is the process of mass media in selecting, emphasizing and endowing specific values, facts and other considerations to the audience for making related judgments.23,24 Gitlin explains that “media frames are persistent patterns of cognition, interpretation, and presentation, of selection, emphasis and exclusion, by which symbol-handlers routinely organize discourse, whether verbal or visual.” 25 He first noted that, particularly in visual depictions, news outlets are not simply holding up a mirror to reality but are, instead, at least in part composing reality, and the composition was entering into our own deliberations—and more, our understandings of who we are and what we were about. 26 Addressing this issue, visual framing 27 “looks at the selection of what content is included in the photograph, why a photographer chose this content over other content, and what effect the content has on views of the content.” More specifically, visual framing “involves the selective prioritization of some images to promote a specific interpretation of events conducted either consciously or subconsciously.” 28 While photographs are often perceived by audiences as a complete and objective representation of an event, they tell only a portion of the whole story, and are thus capable of obscuring and overwhelming the truth. 29
Studying the visual framing of movements or protests concerning civil rights issues is important because of the communicative and emotional properties of visuals as well as the prominence of visuals in the contemporary news environment. Photographs have a greater capacity to “shock” than written language. 30 Visual images in news media teach the public what is valuable and what is appropriate. 31 Photography can define who is worthy and who is unworthy, societally, and inspire people to make judgments about the characters and personalities of others.32,33 Images can create more impressions for readers than texts in the current “skimmers” culture in which readers spend little time on actual text.34 35,36 Thus, what many readers learn about an issue is based largely on the images they view, as well as the emotional response that images evoke. In a seminal study, Paivio et al. found that photographs are recalled more quickly and for a longer period of time than other forms of communication. 37 Photographic images are powerful tools in raising emotions and concerns among audiences in controversial issues.38,39,40,41
Visual Emotional Portrayals in News about Riots and Police Brutality
The history of the United States is “filled with images of protest and expressions against an established group, beginning with the Revolutionary War and continuing today.” 42 As Iyengar and Simon suggested, framing “should be particularly significant as a determinant choice when the choice problem involves politics.” 43
Previous Visual Studies of Racism and Protests
Visual Studies of Emotions
Emotions are described as “discrete and consistent responses to internal or external events which have a particular significance for the organism.” 44 Emotion has at least three dimensions—evaluation-pleasantness, potency-control and activation-arousal as suggested by early research, 45 while recent emotion studies have focused on the valence-arousal model. 46 The valence-arousal model divides emotions into two dimensions: valence, the negative or positive feeling the experience brings and arousal, the energized or enervated state that the experience brings. Basically, emotions can be categorized into multidimensions of activation/deactivation and pleasant/unpleasant. Four categories of emotions comprise the most basic 12 emotions. For example, the unpleasant activation category includes tense, nervous, stressed and upset; the pleasant activation consists of alert, excited, elated and happy; the unpleasant deactivation contains sad, depressed, bored and fatigued; the pleasant deactivation category comprises contented, serene, relaxed and calm. 47
Image Frames
When news media frame wars, protests and conflicts, photographs “most often offer prompts for prevailing government versions of events and rarely contribute independent, new, or unique visual information.” 48 Many cultural studies-based examinations have been qualitative studies or just essays rather than quantitative examinations, which the present study undertakes. Griffin examined how U.S. photos in news magazines covered the “War on Terrorism” in Afghanistan and the invasion of Iraq, and the results showed that these photographs were primarily to establish “narrative themes within official discourse.” 49 Gitlin examined how news media framed protests during the Vietnam War and found the antiwar messages supported by the protestors were ignored by news media and that protestors were framed as the radical students. 50 Certain studies have undertaken quantification of an element of emotion: tone, using varying definitions of positive, negative and neutral. For example, a study investigated the news media coverage of the anti-Vietnam War protest in two different aspects: amount and tone of news media coverage. 51 It shows that the extent of the news media coverage varied over the war period with almost unnoticeable coverage at the very beginning but never reached a very high level; even when the stories did appear, news media’s focus was on the sensational. 52 The authors also noticed the valence—“the overall impression created by a news story and by its juxtaposition with other news stories” 53 —of news stories on the antiwar protests was usually negative, defined as depictions of protest as a social disruption. 54
More recently, Douai investigated Canadian news media’s visual framing of eruption of violence and questionable police tactics during the G20 Toronto summit in both print and online media outlets. 55 Douai argued the prevalence of “lawlessness and violence” 56 frame in Canadian mainstream news media, and the explanation of this attention, was driven by newsworthiness considerations and the more sensational nature of lawlessness and violence than peaceful demonstrations. The researcher also addressed valence, defining positive images as those that glamorize subjects and negative images as those that portray subjects as violent or threatening.
Balance in Power Schema
A content analysis of how the New York Times’ editorials used reasonableness discovered five image schemas embedded in news paragraphs that can invoke reasonableness: vision, path, foundation, balance and force.57,58 In Hicks’s study of image schemas in the New York Times coverage of the Sean Bell shooting, he examined the image schema of balance. 59 The balance schema “connects the experience of balancing with our understanding of balanced emotions, views, amounts, power and justice.” 60
Previous research on police brutality has focused on (1) sociological or situational, (2) psychological or individual and (3) organizational factors of police brutality.61,62 However, rarely have previous studies examined these frames on news media image coverage of riots associated with police brutality. As the Ferguson, Missouri, riot in 2014 raised public attention on civil rights and police brutality again in the postracial society, a historical comparison of news image framing on two riots that were caused by police brutality is important to understanding the civil rights movement in the United States.
Portrayal Weapons
A study of investigating policing of antiglobalization protest summarized the weapons used by police against the protesters in 2000 and 2001. 63 Tear gas, water cannons and concussion grenades were used by police in the antiglobalization protest in the Czech Republic, and similar weapons were also used in another protest in the Quebec City. 64 However, studies on news photos’ portrayal of weapons used in the police-brutality-provoked riots have been limited. Examining the coverage of weapons for police and protestors in both riots is useful to see whether the situations of civil right issues and police brutality have become more severe and whether conflicts and violence between two subject groups have risen over two time periods.
Summary and Research Questions
This literature review shows the various ways in which newspapers have portrayed violence and rioting in the United States at given points in time by measuring certain variables. What is missing in the literature, however, is a more comprehensive examination of multiple variables, and an examination to newspaper visual portrayals of riots—particularly, the relationship between police and rioters—have changed over time. To fill these gaps, this study poses the following research questions to examine how violence and police brutality in regard to minority groups was framed in newspaper photos of these two riots:
Method
Sample
This visual content analysis focuses on four U.S. newspapers: The New York Times, Los Angeles Times, Chicago Tribune and Atlanta Journal-Constitution. These newspapers were selected based on regional representativeness (representing the East, West, North and South of the United States), accessibility and circulation. The time period sampled is five weeks after the beginning of each incident to obtain an adequate sample of images for analysis. For the 1992 Los Angeles riots, the exact time period is April 29, 1992, to June 6, 1992; for the 2014 Ferguson incident, the exact time period is August 9, 2014, to September 5, 2014. Only the first wave of unrest has been covered; this is because the duration of the two cases was largely different, and it was impossible to determine the real “fading away” moment of the Ferguson unrest.
Every page of the four newspapers in the selected time period was examined, and any image relevant to the two incidents was sampled, resulting in a total of 387—55 from The New York Times (zero Los Angeles, 55 Ferguson), 182 from the Los Angeles Times (175 Los Angeles, seven Ferguson), 105 from the Chicago Tribune (42 Los Angeles, 63 Ferguson) and 45 from the Atlanta Journal-Constitution (14 Los Angeles, 31 Ferguson), 231 from the Los Angeles riots and 156 from the Ferguson riots (images collected from the four U.S. newspapers).
Variables
Emotion of subjects
Emotions are coded based on subjects’ facial expressions. See the codebook appendix for details. Emotions of protesters and police, the subjects in the photos, were categorized into the basic 12 specific emotions (unpleasant activation: tense/nervous/stressed/upset; pleasant activation: alert/excited/elated/happy; unpleasant deactivation: sad/depressed/bored/fatigued; pleasant deactivation: contented/serene/relaxed/calm). As a component of emotion, valence stands for whether the news photo frames the subject positive, neutral or negative, which could include the posture or action of the subjects, or even the interactions between the subjects with the opposite group. A “positive valence” image refers to those that “glamorize” the subjects, either or both the police and protesters. If police were portrayed glamorously while carrying out their duties or being friendly with the protestors, this image would be coded as “positive.” A “positive” image of protestors portrays protesters behaving in an orderly manner or being nice to the police. A “negative valence” image portrays the subjects as violent or threatening in disorderly gatherings. For example, if the protesters are looting, this image would be coded as “negative.” Other images that could not be clearly identified as either positive or negative were coded as “neutral.”
Visual frames
Three major visual frames—(1) Law and Order, (2) Lawlessness and Violence and (3) Civil Liberties—that dominated the coverage of protesters’ encounters with police in the anticorporate globalization movement were used to explore the frames in the visual coverage of the two riots. 65
Framing of subjects
As both incidents and the following riots were associated with minority races and police brutality, examining the framing of subjects is essential to understanding the overall framing of the two riots. Again in Douai’s study of news media visual coverage of the protestors and police encounters in the G20 summit, subjects were divided into several majors groups: (1) protestors, (2) police, (3) world leaders and (4) other, and each subject’s visual tone was examined as well. In this study, the subject group of “world leaders” was replaced with “politicians.” 66
Balance of power schema
Balance in power has been one of the major three balance schemas in the investigation of the coverage of the Sean Bell shooting. 67 In this study, balance of power between police and protestors is a useful factor to see the extent of police brutality and riots. 68
Portrayal of weapons
According to the Review of Department of Justice’s Use of Less-Lethal Weapons, 69 weapons examined in the news photos included nonprojectile weapons and nonincapacitating impact weapons, incapacitating agents, projectile impact weapons (detailed definitions and examples of each type were given below). Other types of weapons and body armor for police group and protestors group were also examined.
Coding Process
Two independent coders engaged in the coding of all the images for (1) Image Valence (Positive/Negative/Neutral), (2) Race (Black/White/Mixed/Other), (3) Subject Portrayal Valence (Positive/Negative/Neutral), (4) Image Frame (Lawlessness/Order/Civil Liberty), (5) Weapon (Impact/Projectile/Incapacitating), (6) Emotion (Relaxed/Fatigued/Eagerness/Anxious/Sadness/Fear/Anger/Painful for protestors; Relaxed/Indifferent/Worried/Anxious/Sadness/Fear/Anger/Friendly for police). Emotion variables were set differently for police and protesters as the two groups are naturally portrayed differently in newspapers pictures and (7) Balance of Power (protestors’ power over police/police’s power over protestors/even power).
Image descriptions were examined as part of the images themselves during the coding process. The construction of weapon categories was informed by the Review of Department of Justice’s Use of Less-Lethal Weapons, and that of emotion categories was informed by an emergent coding process. A detailed instruction was developed for the coding process based on initial analysis of the images. After one training session and several re-calibrations of the coding instructions, two coders were able to reach a good level of an averaged reliability of 0.82 measured by Krippendorf’s Alpha (see Table 1 for individual alphas).
Intercoder Reliability for Used Variables (N = 69)
Results
Image Frames
Most of images from both riots (45 percent) present lawlessness and violence, while 31 percent present law and order and 21 percent present civil liberties. The depiction of civil liberties clearly increased, while the depiction of lawlessness decreased between riots. Specifically, depiction of lawlessness and violence is dominant in the Los Angeles Riot (55 percent, which decreased to 31 percent in the Ferguson Riot), but civil liberties were more frequently present in the Ferguson Riot (42 percent, up from only 7 percent in the Los Angeles Riot). These differences were statistically significant. Table 2 provides the full details of these results.
The Presence of Image Frames in the Pictures
Subject Frames
Images depicting the Ferguson riot tend to depict both more negative and positive, in general, but photos depicting the Los Angeles Riots contained fewer images of protesters. Broadly, images showed more protesters as the subjects, and more of protesters with identifiable tones were identified from the Ferguson riot’s images. On the contrary, the percentage of images depicting police remained broadly unchanged, but the tone of these images was statistically significantly different. Specifically, images took on a less positive and more negative tone from the Los Angeles Riots to the Ferguson Riots. In the Los Angeles Riots images, those depicting police were about 19 percent positive, while only about 10 percent were positive in the Ferguson Riots. Similarly, in the Los Angeles Riots, about 3 percent of images were negative, but almost 8 percent were negative in the Ferguson Riots. Neutral images, too, increased, from about 13 to 18 percent whereas police were associated with more negative attitudes. These results were statistically significant. See Table 3.
The Presence of Subject Tones in the Pictures
Emotions in images were analyzed to answer RQ3. In both protests, the dominant emotions of depicted among protesters in the images were eagerness (8 percent), sadness (7 percent) and anger (7 percent). The dominant emotions depicted among police were seriousness (13 percent), worriedness (7 percent) and friendliness (6 percent).
Among images that depicted protesters, the emotional portrayal differed statistically significantly between the two riots. Protesters more frequently were shown eager (from about 4 to 14 percent), sad (from about 1 percent to about 17 percent) and fearful (from about 1 percent to about 5 percent) in the Ferguson Riot than in the Los Angeles Riot. These changes were statistically significant (see Table 4). There were no significant differences between the two incidents regarding the depiction of police emotions.
The Presence of Subject Emotions in the Pictures
Dominant Races of Subjects
In both protests, most (31 percent) protesters present in the pictures were African American, compared with only 2 percent who were white. Conversely, the largest group of identifiable police (28 percent) were white, and very few (5 percent) were black.
Race becomes much more prominent, in general, from the Los Angeles riot to the Ferguson riot, at least in the depiction of protesters (45 to 75 percent). The frequency of African American protesters depicted rose from 18 percent in the Los Angeles riot to 50 percent in the Ferguson riot. These findings were statistically significant (χ2 = 52.675, df = 5, p < .001). Differences in the depiction of the races of police officers are not significant.
Emotions
With respect to subject emotions, Table 4 shows that of the 387 pictures examined, eagerness, sadness and anger were the frequent emotions protestors had in the pictures while police were described as worried and serious. The results of a chi-square test suggest that the two incidents differed significantly in terms of the presence of protestor emotions (χ2 = 66.21, df = 8, p = .000), but no differences between the two incidents regarding police emotions present in the pictures (χ2 = 8.024, df = 8, p = .431). Across time, news media seemed more likely to sympathize with protestors. Table 3 shows full results.
Valence
The dominant valence in the pictures was negative for both incidents (58 percent for the Los Angeles riot and 49 percent for the Ferguson riot, respectively), but the two riots differed significantly. Fewer negative (49.4 percent) and positive (12.8 percent) images were present in the Ferguson riot than in the Los Angeles riot (58 and 21.2 percent, respectively). Overall, images were more neutral in the most recent riot (35.9 percent) than in the Los Angeles riot (20.3 percent). These results are statistically significant (χ2 = 15.417, df = 3, p < .001).
Balance of Power
The balance of power is in favor of police more frequently than the protesters in Ferguson riot (20 percent) than in the Los Angeles riot (10 percent). Similarly, the balance of power in favor of the protesters shows less frequently in the Ferguson riot (6.4 percent) than in the Los Angeles riot (10 percent). These results are statistically significant (χ2 = 10.464, df = 3, p < .05).
Weapons
The two incidents differed significantly in terms of the presence of weapon holders (χ2 = 14.393, df = 4, p < .01). In the pictures of Ferguson riot, only 3.8 percent of protesters held weapons while 22 percent of police held weapons; however, in the images of the Los Angeles riot, 13.4 percent of protesters were shown with weapons, and 28 percent of police carried weapons. Of 387 pictures, about 91 percent protesters did not hold weapons and 74 percent of police did not carry weapons.
The two incidents differed significantly in terms of type of weapons used by protesters and also by police (see Tables 5 and 6). Both groups held less weapons in the pictures of the Ferguson riot (5 percent for protesters and 26 percent for police) compared with the pictures of the Los Angeles riot (12 and 29 percent, respectively). Especially for projectile impact weapons such as guns, both protesters and police were showed less frequently with this type of weapons in the Ferguson riot (0 percent for protestors and 15 percent for police) than in the Los Angeles riot (2 and 22 percent, respectively).
Who Are the Weapon Holders in the Pictures?
The Types of Weapons in the Pictures
Discussion
This study’s results suggest some plausible implications. First, newspapers’ portrayals of civil unrest shifted their focus from the confrontation between protesters and the police to an emphasis on the depiction of the morality, human interest and civil emotions.
This change of dominant frames portraying the two riots might be due to the different coverage and the nature of the riots in the two different times. The 1992 Los Angeles riot was quickly turned into citywide chaos as property damage and robbery rapidly spread to other locations. For example, Los Angeles’s Koreatown was attacked and nearly destroyed during the incident. However, the 2014 Ferguson event was less violent in many respects. The different nature of the incidents may somehow suggest that news media would frame the events in distinct ways, and this, of course, includes the way they select images and put them in print. More than 20 years passed between the two riots. So perhaps, police training might have improved, to meet the needs of a more diverse society, to produce better-behaved law enforcement over time. (On the contrary, U.S. police during the period increasingly were trained to use even extreme measures to protect themselves, regardless of what happened to others, and to integrate sharply increasing amounts of military equipment and tactics into police work.)
The positive valence that depicted police decreased from one time period to the other while the negative valence has increased. This is surprisingly opposite of the conclusion that the visual frame used in portraying police-brutality-provoked riots has shifted from “confrontations” to “civil liberties.” The reason behind this might be that police brutality to the minority group has been more intolerable in the society over time, which made newspaper editors or journalists focus more on the negative image of the police to express their anger to the police and compassion to the victims.
Beside this, subject emotions are different in the two events as well. The “friendly” facial emotion of the police, often depicted in the photos of the 1992 Los Angeles riots, decreased largely in the case of 2014 Ferguson riots, suggesting newspapers today seem to use fewer images trying to tell the audience how good the relationship was between the police and “good citizens”; also, the “sadness” and “eagerness” of protesters was portrayed much more in the Ferguson riots. In the 1992 Los Angeles riots, 78 percent of images did not show any identifiable emotion of the protesters, as the subjects were absent in the examined pictures, photos were taken from long shots or perhaps the journalists’ background or news editing skills might also play a role in the different visual portrayals of police-brutality-provoked riots.
Moreover, during the Ferguson riots, image valence tended to be more neutral, weapon possession/usage and the power of the protesters appeared to be weaker. The use of weapons by both parties decreased from the 1992 event to the 2014 event; however, police, consistent with the general militarization of U.S. police, used increasingly advanced weapons against a group of protesters who used nonprojectile impact weapons in the 2014 riot. In comparison, during the 1992 riot, the protesters used projectile impact weapons. In short, the images showing weapons held by either party decreased in numbers, however, police weapons became more advanced (or more “civilized” with the increased use of tear gas or other sorts of incapacitating agents) while the protesters were more vulnerable.
That the samples examined in this study were collected in microfilm format might be another limitation for this study, as this format limited coders’ ability to identify every detail of people depicted in the picture from a broad angle or a long shot. However, because coders have endeavored to get images in their best quality and those microfilms with the similar resolution were used for both time periods, the comparison might still be valid for the present study.
Future studies should include more similar incidents such as the 1964 Harlem, 1965 Watts, 1967 Newark, 1980 Miami and 2016 Charlotte ones, from a more qualitative perspective to further enrich the knowledge toward the topic itself. For example, an examination of the 1964 Harlem riots in comparison with these two incidents could be another critical exploration. Explanation of the image content would surely provide more depth and understanding toward the contemporary print news and perhaps lay the basis for a historical study with special concern on the transformation of Americans’ collective memory of African Americans’ pursuit of rights in the last century.
Footnotes
Editors’ Note
This article was accepted for publication under the editorship of Sandra H. Utt and Elinor Kelley Grusin.