Content-Type: text/html

The Symbolic Convergence of Color on "Cops"

by David B. Franz, Master's Candidate
and
William R. Davie, Ph.D.
Department of Communication
University of Louisiana, Lafayette
Lafayette, LA  70504-3650
<a href="/cgi-bin/wa?LOGON=A3%3Dind0309D%26L%3DAEJMC%26E%3D8bit%26P%3D3958233%26B%3D--%26T%3Dtext%252Fplain%3B%2520charset%3Diso-8859-1" >[log in to unmask]</a>
337/482-6140 voice, 337/482-6104 fax

Submitted to the 2003 Student Research Competition
Minorities and Communications Division
Association for Education in Journalism and Mass Communication


Professor Emmanuel Onyedike
Norfolk State University
Dept. of Mass Communications and Journalism
700 Park Ave.
Norfolk, VA 23504


"The Symbolic Convergence of Color on "Cops"


Abstract:

Reality television has brought forth a variety of slices of life, but
perhaps none more powerful to the public perception of crime and justice
than the reality police drama.  This content analysis based on substantial
data sets from "Cops" applies the symbolic convergence theory to
demonstrate how the show establishes patterns of what constitutes a
rhetorical vision based on the behavior of police officers toward minorities.

The Symbolic Convergence of Color on "Cops"

Introduction

In today's world of media monopolies, cheap and accessible video, and
digital technologies, distinctions between actual and mediated realities
are increasingly blurred.  The trend toward "reality" television
programming that began in the late 1980s has seized momentum over the past
few seasons, fueled by the fact that it creates compelling drama for
viewers, and it is relatively easy to produce.  It has become a staple
among both television production companies and networks searching for an
elusive collection of loyal viewers with desirable demographics.
Among the reasons for reality television first becoming a subject of
interest, and then concern among scholars in mass communication is the
genre's claim to present "real life."  On all levels of production, choices
are made with respect to what content is chosen, how it is compressed, and
presented in an entertaining fashion.  The result of these editorial
decisions creates a version of "un-reality" designed to achieve narrative
goals that places emphasis upon certain characters and moral themes. These
strategies however, have social consequences for women and minorities,
which make the program choices worthy of investigation.
The study undertaken here consists of a detailed content analysis of one
reality crime drama on television, Langley Production's "Cops," which has
enjoyed success in its first run of programming on the Twentieth-Century
Fox Network (Fox) affiliates, and is also syndicated on the United
Paramount Network (UPN) affiliates.  In its context, we applied theoretical
elements relevant to reality crime dramas.  First, in order to assess the
overall social message generated by the "reality" program, we used
Bormann's Symbolic Convergence Theory based on its three-tiered
perspective.  It calls first for understanding the shared human
communication in which (1) ideas about the nature of the world (rhetorical
visions) stem from (2) repeated stories (fantasy themes) in which (3)
meaningful symbols or actors interact in expected and understood ways -- in
other words, narrative paradigms.  These narrative paradigms fulfill a
psychological need for group members through key characters and themes
(e.g. Santa Claus's rewarding good girls and boys).  They also create myths
for audiences to make sense of their common experiences (Wyatt,
1993).  Sharing these visions allows group members to communicate
effectively with each other; to base decisions on common criteria; to meet
shared objectives, and to initiate new members (Bormann, 1985:
133-134).  The object here was to apply the concepts of SCT to this
particular genre in order to reveal the rhetorical vision it conveys to its
audiences.





Review of Literature

In many ways, "Cops" is a pioneer and leader in the genre of reality
programming.  Since its premiere on March 11, 1989 on the Fox network,
"Cops" has become the longest running of reality police dramas, and one of
the most popular according to Nielsen ratings (Kooista et al., 1998).  It
has been nominated for four Emmy awards and has won the 1993 American
Television Award for Best Reality Show (Langley, 2001).  It appears
syndicated, or marketed to local stations, in 90 percent of the United
States television markets (Langley, 2001).  "Cops" also has been criticized
for its audience appeal by those who dismiss it as "no-brow vérité," a
class of "blue-collar" programs which expose society's underside in
exchange for ratings (Rowen, 2001; 2).  Yet, it has become a syndication
staple.  "Cops" is used as "filler" material when other prime-time Fox
programs are cancelled leaving a hole in the schedule.  The show's low-cost
and efficient production schedule have made it a favorite among programming
executives.  "Cops" costs on average about $200,000 per episode, or roughly
one-forth of the price of a typical half-hour sitcom (Doyle, 1998).

"Cops" and Reality Crime Television
"Cops" was chosen for this study not only because of its popularity and
industry accolades, but also because of its stark cinema-vérité style. The
"Cops" Web site (www.tvcops.com) boasts that it presents "the unvarnished
truth…with no elaborate editing" (Langley, 2001; 2).  In the context of
reality television, this vividness of video pictures and reality montage
may easily overshadow the audio track (cf. Smith, 1996).  Given the absence
of voice-over narration, this TV style of cinema-vérité poses a special
venue.  Without additional commentary voicing over the visuals, the
producers may effectively masque their message through the sense of
unedited video reality.  Presumably, footage that also appears untouched
with regard to special effects and music more effectively conveys this
verisimilitude of reality, while concealing artifices of the editing
process.  For this study, this video style narrowly focuses our attention
to the sequence and symbolism of the video montage, thus avoiding the
complication of layers of music, sound effects, and narration.

Rhetorical Premise
In the entertaining and easily produced context of reality television, the
"Cops" of law enforcement could be either an heroic portrayal of the life
of police officers or a tool to maintain the power of authority figures
over minorities.  Doyle (1998) maintained that "Cops" was shot from the
police officer's point of view and edited to "fit traditional media
templates" (97) of law and order.  In every episode, the audience is
introduced to an officer they can identify with, while other characters,
including suspects and perpetrators, go unnamed and have their faces
pixilated in order to conceal personal though not their ethic identity.
On the surface, it appears the narrative focuses only on the issue of an
immediate complaint involving an alleged criminal violation (drug
infractions, domestic disputes, traffic calamities, etc.) while refraining
from addressing larger issues of poverty, unemployment, social and cultural
factors, and their potent chemistry as catalysts of crime.
As in the ancient Greek theater and classic narrative, "Cops" presents
heroes whose virtues can be easily understood and lauded by the
audience.  These protagonists are challenged by evil circumstances and
forces of crimes, which must be defeated in order to restore order to civil
society.  Given this exigency, it is not surprising that police officers
are shown exhibiting a variety of aggressive tactics to meet the challenge
of defeating criminal actors.  A 1994 study by Oliver found 51 percent of
the police officers depicted were committing aggressive acts toward
criminal suspects.  In contrast, only 19.4 percent of the depicted suspects
were retaliating with aggressive behaviors toward the police.
The question of ethnicity has not been overlooked.  In addition to the
familiar portrayal of aggressive police officers, Oliver and Armstrong
(1998) discovered what appeared to be a bias against minorities.  Based on
a previous study by Oliver (1994), their conclusion was that "Cops"
over-represented blacks as criminals while whites appeared on the side of
justice.  In that study, Caucasians accounted for 70.7 percent of all
characters but only 54.3 percent of the total number of criminal
suspects.  Kunkel et al. (1996) similarly found in a National Television
Violence Study that 51 percent of suspects in reality television programs
were viewed as Caucasian, while only 17 percent of the aggressors were
black.  In other words, suspects of color were more prominent among
perpetrators of crime than their share of the population would
indicate.  Oliver and Armstrong (1998) concluded that reality crime
programming portrays a bias against minority offenders by over sampling
video clips of minority rather than Caucasian suspects.
Recent FBI data (2001) indicate that Caucasians make up 69.7 percent
(n=6,324,006) of the total number of the reported criminal suspects for the
year, with African Americans second at 27.9 percent (n=2,528,368).  While
the FBI figures seem to support the conclusion of an anti-minority slant in
this genre, there has been some debate.  The FBI data in contrast to Oliver
(1994) counted more Caucasians by including Latinos in that category,
whereas Oliver coded them as a separate minority group.
More startling was the finding by Oliver (1994) with respect to aggressive
behavior.  Fifty-one percent of police officers in reality television
exhibit aggression toward suspects.  A 1991 report (Dobrin et al., 1996)
estimated actual city police officers use bodily force in order to capture
suspects only 27.2 percent of the time, or about half of what reality crime
dramas would lead viewers to believe.  Similarly, Kunkel et al. (1996)
discovered 61percent of 170 violent sequences sampled from police reality
programs present "credible threats" (n=103) against police by criminal
suspects.  Thus, the genre's slant appeared to be directed toward over
sampling scenes of violent interaction between police and suspects.
Underscoring this conclusion was the evidence collected from production
personnel, such as Debra Seagal's (1993) account of the network pressure
applied to create a narrative conflict through violence.  Seagal, a former
"tape logger" for "American Detective," a Cops-style show on ABC, was
advised to create complete, exciting stories, even if certain scenes had to
be "fixed" (52) with stock footage or additional sound.  Seagal, in
documenting her show's thirst for blood, referred to an office bulletin
reminding the production crew to look for "Death, Stab, Shoot,
Strangulation, Club, Suicide" (51).  To create an entertaining program,
exciting footage was culled and over sampled to maintain audience interest,
and it necessitated violent, or at least potentially violent, confrontation.
In order to make general applications of such prior research, this study
drew upon a relevant but overlooked theory.  Bormann's (1985) Symbolic
Convergence Theory (SCT), which is an extension of his earlier work in
fantasy theme analysis, was chosen to show how dramatic elements within a
message elicit a targeted response from the audience.  This style of
"reality" television programming that follows police pursuing criminals
constitutes a form of rhetorical vision yet to be examined in terms of its
ethnic symbols and themes within the SCT paradigm (Bormann, 2000).  Because
SCT has been applied in previous research to various forms of mass media,
it would improve our understanding to apply a visual vocabulary of rhetoric
to police dramas, such as "Cops," or "America's Most Wanted".   Schaefer
(1996) contended there was a special need for further investigation to
create such a vocabulary in video and film study using Bormann's analysis.
Research conducted using Symbolic Convergence Theory is generally
qualitative, often with the aim of substantiating the theory's versatility
rather than generating new ideas (Hsia, 1988) or revealing the nature or
causes of particular phenomena.  In this light, SCT has been applied
broadly to a variety of communication experiences from the messages of
non-profit organizations (cf. Ford, 1989), public relations campaigns
(Duffy, 1997), to broader social movements (Smith and Windes, 1993; Huxman,
1996).  Originally designed for small group communication, SCT migrated to
evaluating various forms of mass media.  The romance novels of Barbara
Cartland (Doyle, 1985), for example, contain symbols of men and women who
interact according to a particular set of themes for that genre.  The
rhetorical vision created in that example suggested aggressive women are
undesirable by men whereas nurturing women are alluring.
        SCT first addresses the players or symbols within an act of
communication and then assesses how these elements interact to form
"fantasy themes," common story elements influencing a group's collective
mythology.  The overall meaning established for an audience through these
shared stories is not only dependent on actors or actions, but moreover in
how they repeatedly interact or converge in the narrative.  Thus, the
larger aim is to go beyond counting the simple elements of the rhetor's
visual message, and explore the narrative's themes and draw meaningful
conclusions based on those interactions.
Accepting Bormann, Knutson, and Muslof's (1997) conclusion that group
consciousness owes its effectiveness to the rhetorical skill found within
the message, the goal here then is to apply SCT measures to the genre under
study.  This analysis will reveal the underlying ethnic message of "Cops"
in order to make a reasonable assessment of its impact upon the collective
social consciousness of the audience.

Documenting Reality
Does the mere act of videotaping crime scenes distort reality with regard
to law enforcement?  Zoglin (1992) cites several reasons to explain why the
presence of television crews prompts police to behave differently.  First,
police officers may want to appear more humane and citizen-friendly, and as
a gesture of public relations that encourages departments to cooperate with
camera crews.  In return, the producers of "Cops" allow police departments
to screen each episode before it airs.  They also are granted access to
footage for prosecuting the videotaped suspects.  Some police departments,
however, (notably those in Chicago and San Diego) (63) discovered that
camera crews were more of an inconvenience, and withdrew access to reality
TV producers.
Kooistra et al. (1998), believe the presentation of police work on
television actually is a public service since it helps to demystify the law
enforcement community. Nonetheless, certain police officers enjoy becoming
"stars," and so "perform" for the cameras.  Seagal observed those officers
who were "flattered by the recognition" of television and would turn to
face the camera while trying to elicit information from criminal
suspects.  Seagal noticed the line between television production and police
work was further blurred in that some police departments allowed camera
crews to carry guns and badges.
On the "Cops" Web site, executive producer John Langley (2001) describes
his show's purpose as one directed at depicting the lifestyle of police
officers, and he declares "every cop in the world is a hero" (2).  Spectral
Ministries' COPSwatch Web site (www.specmind.com), however, expressed a
somewhat different view.  An attorney, Richard Glen Boire (2001) routinely
evaluates the program at that site, and offers his criticism and commentary
on the "unlawful police tactics routinely captured on episodes" of what he
calls "the media parasite" (1). Boire has compared it to Orwell's Big Brother.
In order to shed light on this debate, we investigated the rhetorical view
of reality crime dramas as seen through the eyes of its leading product,
"Cops".  Although the bias of reality crime television has been briefly
summarized (cf. Doyle, 1998; Oliver 1994), this study proposes a clearer
measure for addressing patterns of aggression and ethnicity this narrative
employs by applying SCT.
First, we established a measure based on screen time devoted to particular
characters.  For example, what is the mean length of time devoted to
protagonists, antagonists, and their interactions within the context of the
crime-fighting episode?  Second, we assessed the content of media messages
not only by the symbols they contain but also by the types of interactions
using these symbols.  Thus, this paradigm would be applied to account for
the dramatic complexities within the message of reality crime programming.
The first level of Symbolic Convergence Theory research deals with the
identification of symbols appearing within a message.  For the purposes of
this study, the symbols contained within reality crime television
programming are defined as protagonists and antagonists who appear in this
network syndicated program.
This study focuses on two types of characters, police officers and criminal
suspects.  Each suspect was coded individually by ethnicity and gender in
order to assess the construction of character types within the
narrative.  Depicted police officers were coded in two categories, first
according to the ethnicity and gender of the primary or responding officer,
and second according to the same variables for other officers who appear
within the broadcasts.


Hypotheses

In considering symbolic content, it was hypothesized that following Oliver
(1994): Caucasians would appear more frequently as suspects than
African-Americans, or other minorities, and the relative frequency of
suspects' ethnicity would approximate the national FBI crime
statistics.  Following the concept of "news slant" proposed by Entman
(1989) and Kaniss (1991), a bias within the sample would be measured by the
relative amounts of coverage given to different actors.  The hypothesis
proposed to investigate whether a bias against minority suspects is
apparent through the amount of screen time devoted to non-Caucasian
criminal suspects.  Stated simply, we predicted that the mean length of
time devoted to video sequences depicting criminal suspects of color would
be longer than those depicting Caucasian suspects hence reinforcing the
vision of minorities as criminals.

H1: mcs < mncs (The average amount of screen time for Caucasian suspects
will be less than the average amount of screen time for non-Caucasian
suspects.)

The level of thematic content in SCT research identifies common recurring
types of interactions, or convergence between symbols.  In this study,
variables included both the characters' aggressive role in the interaction,
and whether the action was taken by police officers or criminal
suspects.  We also investigated the physical, verbal or non-verbal, and
psychological nature of coded acts of aggression.
Interaction was further coded with respect to atmosphere, seriousness, and
clarity, in order to establish an overall rating of intensity for the
symbolic interaction (after Mustonen and Pulkkinen, 1997).  For the
purposes of this study, atmosphere was defined by the overall tone of the
scene in which an act of aggression takes place.  The tone of the scene was
operationally defined as quarrelsome, frightening, exciting, unclear or comic.
The seriousness of an act was defined by both the potential amount of harm
an act could inflict on the victim and the actual harm incurred.  For
example, a protagonist shooting at a suspect has the potential of
physically harming him or her, though the actual consequences might range
from the death of the suspect if he or she was struck, to no harm at all,
if, for example, the bullets miss the suspect.  Clarity was defined as the
amount of verbal and/or visual detail, whether graphic or vague, in
depicting an aggressive act.  Following previous studies by Oliver (1994)
and Kunkel et al. (1996), it was hypothesized that a larger number of
actors committing aggressive acts would be members of the law enforcement
community than would be criminal suspects or other characters.

H2: fpa > fsa (The frequency of police aggression will be greater than the
frequency of suspect aggression.)

In order to verify Oliver's (1994) observation of aggressive law
enforcement, we predict that the mean amount of time given to this theme
would be greater than time allotted to other themes in which police
officers are shown to be less aggressive or remiss in apprehending
dangerous subjects:

H3: mpa > msa (The average screen time of police aggression will be longer
than the average screen time of suspect aggression.)

In contrast to Doyle's (1998) and Oliver and Armstrong's (1998) findings on
reality television, we predicted the physical acts of law enforcers'
aggression against minorities would both occur more frequently and be
allotted more on-screen time.  Simply stated, audiences will see more
police aggression against suspects of color than against Caucasian suspects
in both occurrence and average length of time on the television screen.

H4: fpancs > fpacs (The frequency of aggression against non-Caucasian
suspects will be greater than the frequency of aggression against Caucasian
suspects.)

H5: mpancs > mpacs (The average length of screen time of aggression against
non-Caucasian suspects will be greater than the average length of screen
time for aggression against Caucasians.)

SCT research at the rhetorical vision level seeks to interpret an overall
worldview established by the creators of a message based on recurring
themes presented within a narrative pattern.  For this study, the
rhetorical vision of the program was defined by the accepted structure of
interactions between police officers and criminal suspects viewed as
universal in the sampled episodes of "Cops".  When the interactions of
protagonist and antagonist form symbolic roles "converging," as they did
within this context, then the recurring themes reinforce Bormann's (1985)
thesis with respect to establishing a rhetorical vision for society.
The interaction of character roles produces variables that translate into
the audience's perceived glamorization (Mustonen and Pulkkinen, 1997), or
characterization, of the protagonist and antagonist as "heroes," "villains"
or "neutral" characters, based on their portrayal within the
narrative.  The perceived effectiveness and justification of the aggression
establish an overall idea of the "attractiveness" (Mustonen and Pulkkinen,
1997) within the context of the narrative.
At this level, we chose to look beyond quantitative studies by Oliver
(1994) and Kunkel et al. (1996) to qualitative sources, such as Seagal
(1993) and Boire (2001), particularly since they have been largely
overlooked on this issue of reality television and character depiction.
Video editor Seagal (1993) and attorney Boire (2001) claim an abundance of
police aggression in reality crime programming is served to the audience as
acceptable behavior by the authorities.  It was thus hypothesized that acts
of aggression by police officers against suspects would appear to be acts
of reasonable force.

H6: fjust > funjust  (The frequency of acts of justified police aggression
will be greater than the frequency of unjustified acts of police aggression.)

Televised scenes in which officers' aggressive behavior appears for no
apparent reason, or for purely instrumental reasons (i.e. in situations in
which police officers act in an offensive role to actively "fight" crime)
would outnumber the examples of force used that would appear to be
appropriate as an act of self-defense, or to protect citizens from
harm.  Consequently, the average amount of time producers allot to video
sequences of these unjustified acts of force by police would be briefer in
screen time than those appearing necessary within the scope of an officer's
duty. Given the law enforcement community's cooperation with "Cops," (cf.
Seagal, 1996; Zoglin, 1989), it was hypothesized that the program would
depict acts of police aggression independent of the aggressor, and as such
would more frequently present such officers as heroes rather than as
neutral characters, or as villains (e.g. rogue cops).

H7: fph > fpn (The frequency of police officers presented as heroes will be
greater than the frequency of police officer's presented as neutral or
antagonistic characters.)

The above hypotheses would either substantiate or deny previous studies'
findings of reality television's abundance of violent interaction with
ethnic minorities as criminal suspects.  In addition, the data would serve
to determine the importance of the rhetorical vision with regard to the
interactions of authorities with citizens suspected of crimes.

Sampling Methodology

Two data sets were collected.  Seven episodes of "Cops" were broadcast per
week at the time the first sample was drawn.  Fox broadcast two first-run
episodes or episodes repeated from earlier in the 2001 season at 7:00 p.m.
CST on Saturday nights.  Meanwhile, UPN televised older, syndicated
episodes weeknights at 10:30 p.m.  Each episode generally consisted of
three distinctive vignettes of crime and law enforcement.  These vignettes
presented sequences that were edited together to create a coherent story
line concluding with a resolution, often revealed in the form of an arrest
or final commentary by a police officer.
For the purposes of this study, all seven weekly episodes of "Cops"
throughout the month of April 2001 were recorded.  These programs were
videotaped by one of the researchers and students in an introductory course
to electronic media at a state university.  Another series of programs were
recorded by throughout the month of June 2001.  This sample of programs
provided a pool of 56 episodes from which 13 vignettes were selected.  An
additional two constructed weeks of network programming (40 vignettes) were
randomly drawn to reconcile the study samples.  This procedure was
recommended by Riffe et al. (1996).
The videotape format created a time code through the VCR dubbing process to
ensure accurate timing of each shot's length, as well as key features that
would facilitate a detailed analysis such as slow motion and the use of
digital freeze frames.  Schaefer (1997) observed that recordings of an
entire broadcast, rather than just the video segments, have an added
advantage. The original commercial breaks and announcer lead-ins (78)
influence the overall pacing and presentation of the show.
Coders segmented each selected show into three vignettes separated by
commercial breaks.  The vignettes were coded shot by shot.  For the purpose
of this study, a "shot" was defined as one continuous piece of footage
without interruption or edited change in perspective.  Using the individual
shot as the unit of analysis held two advantages: (1.) It distilled the
vignettes into discrete sequences encompassing several actions and
organizing them in the order of presentation.  By breaking each vignette
into its distinct cuts, video sequences containing no interaction between
characters could be culled from ones that do containing interaction.  (2.)
As each shot represents an editing choice, the vignettes represent the sum
total of juxtaposed scenes and compressed moments of action.  It is in this
interaction between actors engaged in conflict that the true nature of the
narrative is revealed.  While this may not affect the frequency count for
either character actions or interactive themes, coding in this manner
accounts for the amount of time devoted by the show to both themes and
characters

  Interpreting SCT as a Coding Scheme
The coding scheme was adapted from Mustonen and Pulkkinen's (1997) analysis
of violence in Finnish television.  Mustonen and Pulkkinen's precise
definition of violent acts in television was judged too narrow for the
purposes of this study, which assumes a broader application of the concept
of "aggression."  Even though many scenes on "Cops" contain violence, there
are those that do not.  It can be said that within most interactive
sequences between any two characters, however, one usually assumes a more
aggressive role.
When a police officer chases a suspect, for example, the officer assumes
the aggressive role; in committing a crime, a suspect would assume the
aggressive role.  Coding at this level focuses on the nature of an act of
aggression, defined as "an interaction between actors in which one takes an
offensive posture in order to achieve a goal."  For the purposes of this
study, the term "aggression" was defined as either physical or
non-physical/verbal, and covered any number of types of interaction based
on a protagonist's intent and the status of the protagonist and antagonist.
The rhetorical vision level of analysis was adapted from Mustonen and
Pulkkinen's (1997) "attractiveness of violence" (p.185-186) since it
focuses on the audience's perceived justification, glamorization and
demonstrated efficacy (i.e. the overall meaning) of an aggressive
act.  Mustonen and Pulkkinen (1997), in characterizing violent acts, do not
differentiate among the actors committing them.  An additional coding
section (after Kooistra et al., 1998) assessed both the ethnicity and
gender of police officers and criminal suspects who act in the "Cops'"
narratives.  This information relates to the symbolic level of
communication and reveals information about the nature of interacting
symbols such as the "white cop" symbolizing law and order, or "two black
males" representing crime.  Schaefer (1997) defines several editing
techniques that serve as an additive to the actual video and audio. Coders
analyzed these elements in terms of the constructed "reality" they give the
piece.

Coding Definitions
The ethnicity of central characters (protagonists) and criminal suspects
(up to four actors) were coded as either "Caucasian," "African-American,"
"Latino" or "Other," according to FBI guidelines.  An additional "Unknown"
category was added for video sequences in which the ethnicity of a
character was not identifiable.  The category of "Mixed" also was provided
for scenes in which actors of more than one ethnic origin were presented.
Each shot was then coded on the basis of whether it contained an aggressive
interaction between police officers and suspects, and if so, whether an
aggressive act was undertaken by law enforcement authorities, suspects, or
by both parties.  These acts of aggression were categorized as verbal,
physical, or psychological based on certain behaviors, borrowing from
Oliver and Armstrong's research (1998).  Researchers also included scenes
where police are chasing another party or there is damage to property,
underscoring the nature of the aggressive activity.  Verbal aggression was
defined as "angry talk, insults, threats and/or humiliation."
Psychological aggression consisted of "non-physical or unspoken aggression
conveyed through tone of voice, pressure, intimidation, irony and
scorn."  Given that Mustonen and Pulkkinen included violations of personal
space (e.g. unwarranted search of property) in this category, we discerned
these acts may be demeaning to citizens, although no spoken or direct
physical contact occurs.  For example, if a police officer decides to look
through a woman's purse without her permission, this act may be void of
direct physical contact -- lacking even in a verbal exchange -- but it can
be intimidating through the intrusion of personal belongings and space.
Thus, assessing implicit aggression required coders to interpret the
narrative beyond the explicit dialogue and actions by also giving attention
to an actor's intonations, movements, and any consequential reactions.
If a video scene contained aggressive interaction between characters, then
coders were asked to rate it according to its seriousness by accounting for
both the act itself and its consequences, including the atmosphere and the
level of detail in this video depiction of the aggression.  Acts of
aggression were defined in five ways: (1.) playful act, (2.) hostile
gesture, (3.) insult, (4.) attempt to hurt, and (5.) hurt.  The depicted or
implied results of an act of aggression were evaluated in terms of the
level of harm inflicted.  These included (1.) no harm, (2.) material harm,
(3.) mild injury, (4.) moderate injury (medical assistance), (5.) severe
injury (hospital/physician care), and (6.) death.
We operationally defined atmosphere by assessing and coding the scene in
one the following categories: (1.) humorous, (2.) quarrelsome, (3.)
exciting, (4.) frightening, or (5.) neutral.  Coders also assigned a
general level of explicitness to each videotaped act, from unclear (0), to
a graphic depiction of violence or aggression (3).
Coders gave an overall intensity rating to each act of aggression.  This
assessment was based on their response to variables measuring the
seriousness of the aggressive act, and the mood or atmosphere surrounding
it, based on the explicitness of graphic detail in footage.  Coders
assigned a summative measure for the perceived intensity of each video
sequence (mild, moderate, and extreme) based upon their evaluation of the
variables in each shot. Mild shots generally did not contain any
interaction between police officers and suspects.  Moderate and extreme
scenes were identified by the level of emotional excitement generated by
the actors involved.
For an analysis of the rhetorical vision of each video clip, coders noted
the justification for and effectiveness of law enforcement acts of
aggression, as well as the characterization of role players within the
narrative structure.  In this analysis, coders assessed whether or not acts
committed by a protagonist should be perceived as seemingly justified, and
if it meant they were necessary to apprehend criminals.  These
considerations logically would influence the audience's perception as to
whether such tactics appeared to be an acceptable means for exacting
aggressive law enforcement.  Characterization (cf.  "glamorization,"
Mustonen and Pulkkinen, 1997) of the characters as either heroes or
villains was determined by how the presentation of the characters
themselves had a bearing on the direction of the program's
message.  Because police actions are designed to curtail criminal acts,
which is central to the theme and narrative structure of this program, only
sequences containing police-suspect interactions were analyzed for this study.
Justification of aggression was defined at three levels (after Mustonen and
Pulkkinen, 1997), and required coding decisions based on the following
criteria: (1.) Intentionality -- whether the aggression was unintended,
intended for external reasons and social norms such as a policeman "doing
his duty," or internally motivated, such as by personal gain or greed; (2.)
Motivation -- whether the aggression was in defense to help others or save
oneself; offensive; both offensive and defensive, or purely reactionary;
and (3.) Planning -- whether or not the aggression appeared to be
premeditated.
The difference between motivation and intention may be small but it is
significant. Intention can be viewed in the context of particular events
and/or circumstances that prompt the interaction between police officers
and the suspect(s).  Motivation, on the other hand, was defined as the
drive behind a character's decision to behave in certain ways as the
narrative unfolded.  For example, the intention of a law enforcement
officer may be viewed as one of reasonable force to apprehend a suspicious
suspect.  However, the officer's motivation to chase, strike or shoot the
suspect may be offensive, defensive or purely reactionary, depending on how
the scene played out on-screen.
Coders also defined the efficacy of aggression on two levels: (1.) degree
to which the consequences of an aggressor's violence were suggested, shown,
or emphasized by the footage, or not depicted at all, and (2.) degree to
which an act of aggression is depicted as successful in achieving its goal
or purpose (e.g. police capture of the criminal).  Characterization is the
audience's (or coder's) perception of the police officers or suspects based
on their portrayal within the narrative by the producers.  That is, whether
or not the aggressor is depicted "in the right" as a hero, for example, or
"in the wrong" as a villain.  It also concerns whether or not the victim is
presented as deserving the use of aggression, or as undeserving.  The
program could make this distinction in several ways; for example, when
producers opt to pixilate (blur) a suspect's face, his or her innocence
seems questionable to the viewer, whereas officers shown congratulating
each other after an arrest would visually underscore their success in
dealing with crime.  In conducting the subsequent analysis of these data, a
chi-square analysis was applied to the variable of occurrence, and
independent t tests of unequal variance, in cases of determining average
shot length, were used to evaluate the principal hypotheses.

Expected Results
In order to validate the research methodology, a pilot study was conducted
using 13 vignettes randomly selected from programs televised in April of
that year. Student coders were given two hours of instruction, including a
sample vignette to view and evaluate for practice. Coders listed and rated
each video sequence within the randomly chosen vignette according to
the  distinct variables previously defined.  The selected episodes were
first given an arbitrary program number, then coders listed the date and
time of the broadcast and the station (UPN or Fox).  Coders noted whether
the program was a first-run, repeat (1-2 years old), or a syndicated
broadcast (2+ year old), in addition to the shooting locales.  Coders
evaluated the audio techniques employed in the recorded episodes.
These techniques followed Schaefer's (1997) definition as edits that
manipulate the "reality" of the piece.  These production techniques
included the use of asynchronous audio for voice-overs, where narration
from an announcer elaborates upon the subject and action of the
footage.  They also included music and natural (ambient) noise, which is
added in the post-production process, but does not appear to be part of the
actual audio track associated with the scene of the crime.  Typically,
audio tracks for "Cops" include narration by a police officer explaining
external shots of buildings, patrol cars or helicopters.  These editing
techniques also include "non-straight cut" transitions, and
"non-continuity" shots.  Non-straight-cut transitions were video effects
such as fades and wipes that tend to draw attention to the edits in scene
shifts.  Effects of this type show that the footage is not seamless and
there are edits between shots.
Non-continuity shots are ones that depart from the main setting or
principal sequence of the action and included montage cuts, (shots edited
together for setting or tone, but not for plot) and stock footage, which
are shots taken from other sources beyond the primary location of the
dramatic action.  Coders recorded these visual elements as well as provided
short descriptions of the vignettes.
The 13 vignettes chosen yielded 150 video shots for this first phase of the
study.  Of these sequences, 100 percent of all primary responding police
officers were coded as white males. Caucasians also made up 61.5 percent of
the primary criminal suspects.  Male suspects accounted for 78.8 percent of
the study sample.
Physical interaction between police officers and suspects occurred in 14.7
percent of the total number of video segments.  Of the entire assortment of
scenes, police officers assumed the role of an aggressor in 26 percent of
the shots.  The entire sample of "Cops "episodes contained 51 cases in
which a primary suspect was interviewed or apprehended by police
investigators, and thus qualified for this analysis.
As expected, Caucasian criminal suspects appeared within the sampled scenes
more often than minority suspects, and accounted for 62.7 percent (n=32) of
the sampled criminal suspects (Table 1). Minority criminal suspects,
however, were on-screen for longer periods of time than white
suspects.  The mean time difference was 44 seconds for minority actors
compared to 38 seconds for Caucasian actors as indicated in Table 1.

Table 1: On-screen time and Ethnicity of Primary Suspects
Frequency               Average Screen-Time
Caucasian       32 (62.7%)                      38.34   (secs)
Minority        19 (37.3%)                      43.53   (secs)
Total                   51 (100%)

Of the 51 analyzed shots, 27 contained an identifiable aggressor, while 24
were found to have no identifiable aggressor.  In only one of these shots
did the criminal suspect take an aggressive posture against the arresting
police officer.  Scenes of aggressive law enforcement averaged 55 seconds
for the 26 scenes, significantly higher than the 15 seconds devoted to the
single aggressive suspect.  The pilot study found a balance of police
aggression against Caucasian and minority suspects.  Yet, the largest share
(77%) of police aggression against Caucasians appeared to be moderate in
nature, while police aggression against minority suspects was just as
likely to be extreme in intensity.

Table 2: Overall Intensity of Police Aggression
Mild            Moderate        Extreme         Total
Against Caucasians      2 (15.4%)       10 (76.9%)      1 (7.7%)        13
Against Minorities      2 (14.3%)       6 (42.9%)       6 (42.9%)       14

First Data Set Analysis

The results of this initial study produced some interesting results,
particularly a 50 percent split in terms of the overall frequency of
aggression committed against Caucasians and minority subjects. In terms of
our initial expectations, we found Caucasians comprised a larger portion of
the sample in "Cops" programs than did subjects of color. We also expected
that the average screen-time devoted to minority suspects would be greater
than it would be for Caucasians. That expectation was supported by our
pilot sample, but not by a significant percentage.  We suspected that a
larger sample size proposed for a general study would enable us to test the
data more precisely. On a symbolic level, however, pilot data suggested
that even though Caucasian suspects appeared more frequently, on average,
"Cops" devoted more time to minority suspects.
The hypotheses that intense and physical aggression will be committed
against minority suspects found mixed support at first. Overall, acts of
aggression were committed equally against Caucasians and
non-Caucasians.  However, findings indicated a tendency toward extreme
aggressive law enforcement especially in the pursuit of ethnic minorities.
In terms of the rhetorical vision, initial results showed a bent toward
violent interactions between police officers and criminal suspects.
The findings of the pilot study inspired the sampling of a larger pool of
data to apply statistical tests to create a data set that could be
generalized to the larger population of programs, such as in Kooistra et
al.'s (1998) study of 43 episodes of "Cops."  Also, coding sheets prepared
by the researchers for the main study were revised in order to refine
discrete variables. Finally, coders were given improved written and oral
instructions to enhance the reliability of the instrument.

Main Data Set
In October 2001, 14 randomly selected episodes of "Cops" from the months of
April and June 2001 were assigned to six students in an electronic media
course. Coders were given a two-hour training session and reliability test
consisting of a randomly selected vignette coded by all the students
individually for correlation and comparison; for the purpose of comparison
the researchers viewed and rated the vignette with the other coders.  An
initial reliability test consisted of a randomly selected vignette coded by
all the students individually correlated for comparison.
The coders were carefully informed of the definitions of the variables and
relationships between them. A reliability test was conducted involving all
students watching the same "Cops'" vignette by coding each of the 11 shots
according to the identified variables. The reliability test results for the
six coders whose data was sampled for the study were correlated to assess
the instrument's consistency. The data were found to be in general
agreement about what was presented in the vignette.
   Coders were in 100 percent agreement with respect to the gender and
ethnicity of the police officers and the gender of the criminal suspect in
the vignettes.  There was minor disagreement as to which shot revealed the
ethnicity of the suspect, however, a significance level of p< .003 was
easily achieved. Coders were also in agreement with regard to the identity
of the aggressor and the nature of the physical, verbal and psychological
aggression depicted.
In addition, a 1:1 agreement was observed for the variables measuring the
efficacy and seriousness of the aggressive acts, and the degree of
glamorization of the object of aggression. While most study variables
showed significant agreement, variables measuring intent and motivation of
aggression produced less consistency, and had to be abandoned. The
intensity and attractiveness summative ratings, considered peripheral to
the study, were also found unreliable by this test.
In order to test a larger portion of the study data, two of the 14 episodes
were coded in their entirety by the researchers. The Coefficient of
Reliability values for the variables in both programs were calculated at
above 80 percent in all but two cases. As in the previous reliability test,
the variables for intent of aggression and motivation of aggression, failed
to produce reliability in one of the episodes.  In accordance, hypotheses
regarding these variables were not statistically tested.
The sample of 42 "Cops" vignettes produced a large variety of complex
police-criminal suspect interactions, spanning a variety of response
situations. Table 3 contains a list of the crimes, emergencies and
complaints confronting police officers in the sampled programs.
The sample consisted of 526 individually coded video sequences running four
hours, 18 minutes and 43 seconds. Of these scenes coded, 164, or 46.9
percent of the sample, contained aggressive interaction between police and
criminal suspects, representing two hours, one minute and six seconds of
network programming.

Table 3. Police Officer Response Situations in Main Sample
Suspicious activity/ traffic stop:               9
Illicit Narcotics:                               7
Theft/ shoplifting:                      5
Public disturbance:                      3
Victim stabbed:                                  2
Public nudity:                           2
Fist fight:                              2
Fatality/ suicide:                       2
Medical assistance:                      2
Drunk driving:                           1
Domestic violence:                       1
Mugging:                                 1
Warrant served:                          1
Hit and run:                             1
Kidnapping:                              1
Ghost sighting:                                  1
Unknown:                                 1

In addition to suspicious activities, drug-related crimes, domestic
disturbances, calls for medical assistance, and other typical "Cops" fare,
the sample programs covered story lines as mundane as a public disturbance
provoked by a couple wanting take-out food at a restaurant where it was not
available (11 April 2001, vignette #2).  At the other end of the spectrum,
the police call could be as bizarre as a response to a report of a ghost
fondling a New Mexico man's daughter on Halloween (30 April 2001, vignette
#3).
"Cops" vignettes also ranged from the frightening to the
humorous.  Officers in the sampled programs respond to two mysterious
deaths. In one vignette (6 June 2001, vignette #3), police in Las Vegas
found the body of an apparent suicide victim in a pick up truck in the
desert. In another instance (14 April 2001, vignette #3), Pierce County
officers investigating a burglary discovered the suspect dead at the scene.
A darkly amusing scene reminiscent of Stanley Kowalski in "A Street Car
Named Desire" exemplified the humorous vignettes.   In this echo of Marlon
Brando, an illegal Mexican immigrant, blinded by crystal methamphetamine,
yells his girlfriend Lorena's name in the courtyard of their apartment
building (13 June 2001, vignette #3).
The single vignette in which the officer's intent for accosting a suspect
was unexplained illustrates Schaefer's (1997) concept of montage: In this
vignette (#1, 13 June 2001), a police officer is shown discussing the
department's search for a murder suspect immediately before he suddenly
takes off to chase a vehicle.  No explanation is given for the subsequent
chase, and so the audience is left with the impression that the officer has
found the murder suspect he had described. The vignette ends with the
pursued suspect crashing his vehicle and his arrest for evading the
police.  The suspect is never identified as the wanted murderer, and
responding officers cite only his failure to pull-over at a traffic stop as
the reason behind his arrest yet the guilty cloud of murderer is never
lifted from his head.
Of 526 video clips, 466 (95.1%) showed the primary police officer to be
Caucasian (Table 4).  Latinos were coded as a distant second at 3.5 percent
(n=17), followed by police of unidentified ethnicity (n=5, 1%), with
African-American officers (n=2, .4%) figuring to be the least likely
protagonists.

Table 4: Ethnicity of Police Officers
Ethnicity of Police Officer     Primary         Additional
Caucasian                       466 (95.1%)     257 (84.5%)
Latino                                  17 (3.5%)           6 (2.0%)
Unknown                             5 (1.0%)        6 (2.0%)
African-American                    2 (.4%)         0
Mixed                              NA             35 (11.5%)
Total                                     490            304

This sample contained 247 video shots (47%) that contained criminal
suspects.  Of these, the majority of primary suspects (56.7%, n=140) were
Caucasian, while 29.1 percent (n=72) were African-American. Among the 76
instances in which the ethnicity of secondary suspects was coded, 69.7 percent



Table 5: Ethnicity of Suspects in the Study Sample
Ethnicity of Suspect    First                   Second          Third           Additional
Caucasian               140 (56.7%)             44 (62.0%)          6(100%)     3(100%)
African-American        72 (29.1%)              15(21.1%)
Latino                          23 (9.3%)               7 (9.9%)
Other                           12 (4.9%)               1 (7.0%)
Total                           247                     67                 6            3

The data show Caucasian criminal suspects did appear within the sampled
video segments more often than suspects of color.  As shown in the table,
primary Caucasian antagonists appear almost twice as often as primary
African-American suspects. In 239 recorded appearances of suspects, 93.4
percent of minority suspects interacted with white police officers. When
accounting for aggressive interaction, a similar 88.1 percent of the
minority suspects interacted with white police officers.

Table 6: Frequency of Interaction of Police Officers and Suspects by
Aggression:
Caucasian suspect       Minority Suspect        Total
Caucasian officer       127 (95.5%)             99 (93.4%)              226 (94.6%)
Minority officer            6 (4.5%)             7 (6.6%)                       13 (5.4%)
Total                           133                     106                     239

Table 6a: Physical aggression only:
Caucasian suspect       Minority Suspect        Total
Caucasian officer       49 (90.7%)              52 (88.1%)              101 (89.4%)
Minority officer          5 (9.3%)              7 (11.9%)               12 (10.6%)
Total                           54                      59                      113

Table 7: Mean On-screen Time and Frequency of Suspect
Suspect                 Frequency       Avg. Screen-Time
Caucasian               56              37.9    (secs)
Minority                59              48.5    (secs)
Total                           115

The second hypothesis predicted that more on-screen time would be devoted
to minority criminal suspects than Caucasian ones. In the 115 cases in
which police officers aggressively accost ethnically identifiable suspect,
minority criminal suspects were not only found to be greater in frequency
than white suspects, but were given longer on screen-time than Caucasian
ones, averaging a ratio of 49 to 38 seconds, respectively (Table 7).

Table 8: Type of Police Aggression
Physical  Shooting              1 (.6%)
Threatening with weapon         10 (6.1%)
Fighting                        7 (5.4%)
Hitting with weapon             1 (.6%)
Chasing                         47 (28.8%)
No physical interaction                 63 (38.7%)
Other                                   34 (20.9%)
Total                                   163

Table 8a: Psychological Aggression   16 (9.8%)
Intimidation                     28 (17.1%)
Violation of rights                1 (.6%)
Irony/ scorn                     12 (7.3%)
Other                              5 (.3%)
No psychological aggression     102 (64.0%)
Total                                   164

Table 8b: Verbal Aggression
Angry talk                      32 (19.5%)
Personal hurt                           16 (9.8%)
Threat/humiliation              15 (9.1%)
No verbal aggression            101 (61.6%)
Total                                   164

The third hypothesis predicted that most observed acts of aggression would
be committed by the law enforcement protagonists, and not criminal
suspects. Police officers were depicted as initiating aggression in 96.3%
(n=158) of the 164 cases where such interaction occurred. In 68.8 percent
(n=362) of the sample, however, no aggression was recorded at all. In six
cases where police were not the sole aggressors, five incidents featured
fighting and arguing that could be attributed to both parties, while only
one case depicted a suspect attempting to physically hurt an officer.  Of
the 92 cases in which physical aggression occurred, 51.1 percent (n=42)
consisted of police officers actively pursuing suspects (Table 8). In
addition to 62 acts of psychological aggression (Table 8a) were coded along
with verbal aggression (Table 8b) that occurred in 63 cases. Verbal
aggression came largely in the form of police officers becoming
argumentative with suspects (50.8%, n=32).  The most common incidents of
psychological aggression (45.2%, n=28) came in the form of police officers
speaking in an intimidating manner with suspects. These actions include a
raised level of voice, shouted commands, and the use of names other than
the suspect's identity in questioning.
The fourth hypothesis predicted that a longer period of screen-time would
be allotted to video sequences featuring police officers as protagonists
actively and aggressively pursuing criminal offenders. Of the 159 shots
that featured an identifiable aggressor, in only one of these scenes was a
criminal suspect found to assume an aggressive role against a police
officer. Such aggressive law enforcement activity averaged 44.4 seconds per
clip, slightly higher than the 30.0 seconds devoted to the single depiction
of a defiant suspect. With such an imbalance between cases of police and
suspect aggression, a test for equality of variance could not be properly
calculated. However, the raw numbers speak for themselves: In more than
four hours of analyzed footage, only about half a minute of footage shows
law enforcement officers challenged in their duty.
A test of equal variance comparing the mean screen time devoted to
aggression against white suspects and ones of color found more time on
average spent on minority suspects: 48.5 seconds compared to 37.9 seconds.
This ratio that weighted against suspects of color was not only maintained
but also increased when calculating scenes containing physical aggression
only.  On the whole, shots that contained any aggression at all were found
to last on average twice as long as those containing no aggression at 44 to
22 seconds respectively.
A test of unequal variance comparing the mean length of video shots
containing aggression compared to those without such actions showed
aggressive acts received twice the on-screen time as scenes containing no
aggression between police officers and suspects. In only nine cases (.05%),
did the police appear to be trying to hurt or actually hurting an
individual. In one of these instances (2 June 2001, vignette #1), five
white male officers spend over two minutes to not only pin, strike and kick
an immobile black male, but also to release a police dog, which continues
to bite him as he is handcuffed. The suspect, who had been under suspicion
for the expired plates on his vehicle (legitimately borrowed from a
friend), attempted to elude the police. The officers assured the audience
that the suspect initiated the chase in order to buy time to dispose of
illegal narcotics, although they offer no evidence of such.  In another
scene (13 June 2001, vignette #3), an off-duty police officer catches a
subject leaving the scene of a burglary with a handful of silverware, and
runs him over with his car.
Regardless of intent, coders cited injury as a consequence of police
aggression in 13 (.08%) cases. Death was discussed as a consequence of
crime twice (.01%). In fact, only one actual suspect fatality occurred
among the sampled vignettes.  In that particular case (6 June 2001,
vignette #3), police reported to a burglary and found the suspect dead at
the scene. In the second case (2 June 2001, vignette #3), no death was
depicted but was discussed by the police at the scene. The responding
officers chase a vehicle that is deemed suspicious due to a cracked
windshield, and recover an illegal handgun. They then discuss how being
"forced' into chasing a suspect at high speeds could result in the death of
either the suspects, bystanders or themselves.
We predicted that police officers would be depicted as heroic protagonists
more often than they were depicted in a negative light. For the purposes of
this study, glamorization was defined as the perceived portrayal of the
aggressor or the victim of his aggression.  Based on the context of the
action the aggressor may appear as a hero in serving the community or as a
villain if he is perceived as inflicting unnecessary punishment. The
antagonist can also be perceived as a hero or villain depending on whether
he is depicted as deserving or undeserving of his treatment. Often both
protagonist and antagonist, however, can appear as simply "neutral"
characters, i.e. those who do not act outside of the normal law enforcement
role in the case of the police or those suspect who are depicted in being a
credible dangerous threat to others.   In no case were the aggressors
(police officers in all but one instance) seen as villains, and consistent
with this finding, the subject of aggression was depicted as a villain or
neutral character in all cases.
Table 10 contains the frequencies of "glamorization" (Mustonen and
Pulkkinen, 1997) for an act of police aggression as compared with the
portrayal of the suspect. Coders found police portrayed in the light of a
hero, 58.3% of the suspects were depicted as a villain. Coders also found
suspects were neutral 66.2% in cases in which the police officer was
portrayed as neutral.

Table 10: Glamorization of the Police Aggressor vs. the Suspect
Villainous suspect      Neutral Suspect                 Total
Neutral officer                 30 (41.7%)      57 (66.2%)              87 (55.1%)
Heroic officer                  42 (58.3%)      29 (33.8%)              71 (44.9%)

In measuring the perceived efficacy of consequences and gratification of
police aggression, coders found their aggressive acts were presented
without consequence in 133 of 164 cases (81%). Consequence was
operationally defined as the degree to which the ultimate effect of an
aggressive act on the target subject is depicted, whether it is actually
shown, or suggested visually or verbally.
Gratification was defined as the degree to which the protagonist's goals
are achieved, by his aggressive tactics. Coders found consequences to
aggressive acts depicted only 10.4 percent (n=17) of the time. Viewers
found aggression was a successful means for achieving goals (29.3%, n=48),
almost as often as it was depicted as failing to do so (29.9%, n=49).
Finally, the attractiveness of each shot, defined as the determination of
whether the aggression depicted was perceived as a useful and justified
means of achieving a goal, or unattractive act, that is unsatisfactory and
with negative consequence. Though this variable did not reach a valid level
of significance in the reliability test, it was also peripheral to the
study hypotheses and was not statistically tested.



Conclusions

In terms of our hypothesies, Caucasians comprised a larger portion of the
study group than minority actors. However, our chi-square analysis testing
the interaction between suspects and police officers by ethnicity failed to
produce a statistically significant result.  Nonetheless, the ratio of
Caucasian police to minority officers was clearly greater than the ratio of
Caucasian to minority suspects. The second hypothesis postulated that the
average screen-time devoted to minority suspects conversely would be
greater than for Caucasians.  This prediction was supported in both the
sample means and the t-test of equal variance.  These data suggest a bias
on the part of the program's producers does exist within the program, and
that it is against minority citizens that does not appear by calculating
simple frequencies of a suspect's race.  In cases, where aggression is
committed against a suspect of an identifiable ethnicity, the frequency of
aggression against suspects of color is larger. The mean length of these
shots is also longer.
This ethnic slant against minority offenders could be said to exist within
the law enforcement community itself, if not the producers of
"Cops".  Episodes of the program represent a series of conscious production
choices, each of which narrow the amount of footage that will be eventually
broadcast from the total amount of potential footage producers have access
to edit. In the field, videographers must decide what officers and suspects
are most interesting to shoot.  In post-production, decisions are made as
to which sequences are to be used, and how long each sequence should run in
the episode.  Actual racial profiling by police officers certainly may
color and even magnify the anti-minority message of "Cops," especially in
that the officers depicted within it approve each episode.  To say that the
program simply mirrors this bias, however, is to say that the producers
have shot footage of all possible police activity in the cities they are
invited into (effectively defeating the program's low-cost advantage), and
then randomly sampling footage to use in the show.
For example, in one vignette (2 June 2001, vignette #3) police officers
stopped a vehicle driven by a Caucasian male with an African-American
passenger because of a cracked rear windshield. The videotaped interaction
between police and suspects ran six minutes and 39 seconds. The
African-American passenger however was on-screen five minutes and 58
seconds of the vignette. The Caucasian suspect was found in possession of
an illegal handgun, but was on-screen only 41 seconds, and was seen in only
three out of ten video shots. Analysis by counting just the frequency of
suspects by ethnicity would have missed such an obvious racial slant.
However, taken shot-by-shot, it becomes clear that this vignette
exaggerated the significance of questioning the African-American passenger,
who though charged with no crime, appears to be more dangerous by the sheer
length of the interrogation.
In a similar instance viewed earlier in the same episode, police stopped a
Caucasian woman in order to serve her a warrant. Although this woman is
found in possession of crack cocaine, she is quickly taken into custody
off-screen. Meanwhile, the audience's attention is refocused on the
African-American passenger in her truck.  Police have him lie in the
street, before beginning an extensive interrogation, while he protests his
innocence.  This particular episode contains a scene of five Caucasian
police officers pinning down one African-American male, who shows no sign
of resisting them. Again, possible police bias in their treatment of the
African-American suspects may be magnified by the producer's decision to
shoot and broadcast footage of the man's interrogation, while dismissing
the story line of the woman, who is arrested for both an outstanding
warrant and the possession of a controlled substance.
These data also show the ethnic slant of the show obviously on the side of
protagonists and against the suspects presented, with the overwhelming
percentage of white male police officers included in the sampled programs.
Clearly, the protagonist roles are predominately filled by white actors,
while the antagonists are more diverse in ethnicity. This leaves the clear
impression that an old axiom from the western cinema has been rewritten:
"Good guys wear white (skin)."
In both data sets, the narrative themes depict an aggressive law
enforcement community, championed and colored almost exclusively by white
male officers. When minority police officers do appear on-screen, which is
rare, they are relegated to secondary roles, standing in the shadow of
Caucasian police leading the investigation. Additional law enforcement
personnel of color, represent 15.5 percent of cases where additional police
officers appear. At the rhetorical vision level of analysis, police
aggression appeared more frequently to be reactionary in nature and as a
necessary means of law enforcement. The greatest share of all aggression
was found to be in the actual chasing down of suspects. Verbal and
psychological aggression, while not as common, also was found within the
study sample. Verbal aggression, usually heard as angry talk, was most
often in evidence as one technique to get suspects to comply and/or confess.
At the implicit psychological level, officers in many cases not only
assumed an air of intimidation but one of scorn as well, when dealing with
their suspects.  In one case, (13 June 2001, vignette #2) an officer
referred to a couple found having sex in a parked car as "chief" and
"doll." The female suspect, while dressing, was thoroughly questioned about
how promiscuous she was and repeatedly reminded that she did not know the
name of her lover.
 From the initial data set, "Cops" appeared to be tolerant of police
aggression given their difficult and dangerous responsibility of capturing
criminals. Since the syndicated program is screened by the police officers
that appear in it before airing, it somehow must be deemed as acceptable to
the law enforcement community. Given the repeated fantasy theme that police
officers aggressively fight crime, these portrayals influenced by the
cinema-vérité style, suggest law enforcement officers in reality represent
a group of altruistic heroes, working not for themselves but for the good
of the community.
Schaefer (1997) suggested that scholarly research into film, video and
broadcast formats, is hampered by the lack of a unified vocabulary. This
study has demonstrated that Bormann's Symbolic Convergence paradigm offers
such a vocabulary. SCT has been used to define the content of a television
program as a series of symbols that have no meaning out of context. The
symbolic content, here a group of law enforcement officers and their
antagonists, interact in a series of repeated narrative plots, called
fantasy themes.
Taken together, these narratives are negligent of larger truths but
repeated over and over, or "chained out" (Bormann, 1985), as a staple of
both Fox and UPN network affiliates, they form a powerful rhetorical
vision, a perceived message of prejudice for the audience. The "fantasy"
that is presented to the "Cops" viewers is that they are under constant
protection by an aggressive, mainly white, male law enforcement community
successfully battling an often dangerous and diverse collection of
losers.  The danger of such a vision is that an ethnically slanted criminal
population and an over-abundance of police aggression against them is taken
to be revered as just. With new forms of low-cost reality programs
appearing regularly, and with "Cops" continuing into its fifteenth season,
the influence of this form of "televised reality" skewing perceptions of
"actual reality" shows no sign of waning, and the damage to citizens of
color no signs of abating.

References

Boire, R. G. (2001) The COPSwatch report. . I. Retrieved February 20, 2001
from World Wide Web : http://www.specmind.com

  Bormann, E. G. (1985). Symbolic convergence theory: A communication
formulation. Journal of Communication, 35, 128-138.

  Bormann, E. G. (2000). Personal communication.

  Bormann, E. G., Cragan, J. F. and Shields, D. C. (1996). An expansion of
the rhetorical vision component of the symbolic convergence theory: The
cold war paradigm case. Communication Monographs, 63, 1-28.

  Bormann, E. G., Knutson, R. L. & Muslof, K. (1997). Why do people share
fantasies?: An empirical investigation of a basic tenet of the symbolic
convergence theory. Communication Studies, 48, 254-276.

Dobrin, A., Wiersema, B., Loftlin, C. and McDowall, D. (1996) The
Statistical Handbook on Violence in America. Phoenix: Oryx Press.

Dominick, J. R., Wurtzel, A., and Lometti, G. (1975). Television journalism
vs. Show business: A content analysis of Eyewitness news. Journalism
Quarterly, 52. 213-218.

Doyle, A. (1998). "Cops": television policing as policing reality. In
Fishman, M. and Cavender, G. (eds.) Entertaining Crime Hawthorne, NY:
Aldine De Gruyter. pp. 95-116.

Doyle, M. V. (1985). The rhetoric of romance: A fantasy theme analysis of
Barbara Cartland novels. Southern Communication Journal, 51, 24-48.

Drew, D. G., and Cadwell, R. (1985). Some effects of video editing on
perceptions of television news. Journalism Quarterly, 62. 828-831, 849.

Duffy, M. (1997). High stakes: A fantasy theme analysis of the selling of
riverboat gambling in Iowa. Southern Communication Journal, 51, 24-48.

Ellis, J. C. (1989). The Documentary Idea. Englewood Cliffs, NJ: Prentice Hall.

Entman, R. M. (1989). Democracy Without Citizens. New York: Oxford
University Press.

Entman, R. M. (1993). Framing: Toward clarification of a fractured
paradigm. Journal of Communication, 43 (4), 51-58.

Federal Bureau of Investigation (2001) Uniform Crime Report: crimes in the
United States, 2001 I. Retrieved October 24, 2001 from the World Wide Web
http://www.fbi.gov/ucr/cius_00/00crime3.pdf.

Ford, L. A. (1989). Fetching good out of evil in AA: A Bormannean fantasy
theme analysis of the "Big Book" of Alcoholics Anonymous. Communication
Quarterly, 37, 1-15.

Henderson, J. L. "Ancient myths and modern man." In Jung, C., ed. Man and
his Symbols. New York: Penguin Books, pp. 104-157.

Hsia, H. J. (1988). Mass Communications Research Methods: A Step by Step
Approach. Hillsdale, NJ: Erlbaum.

Huxman, S. S. (1996). Mary Wollstonecraft, Margaret Fuller and Angelina
Grimke: Symbolic convergence and a nascent rhetorical vision. Communication
Quarterly. 40, 16-28.

Kaniss, P. (1991). Making Local News. Chicago: University of Chicago Press.

Kooistra, P. G., Mahoney, J. S. and Westervelt, S. D. (1998). The world of
crime according to "Cops." . In Fishman, M. and Cavender, G. (eds.)
Entertaining Crime Hawthorne, NY: Aldine De Gruyter. pp. 141-158.

Kunkel, D., Wilson, B. J., Linz, D., Potter, J., Donnerstien, E., Smith, S.
L., Blumenthal, E., and Gray, T. (1996). Violence in television programming
overall: University of California, Santa Barbara Study. In National
Television Violence Study: Scientific Papers. pp. 1-172.

-Langley, J. (2001) Cops Fact-sheet. I. Retrieved October 2, 2001 from the
World Wide Web http://www.wvah.com/programs/cops/.

Langley, J. (2001) Cops. I. Retrieved October 2, 2001 from the World Wide
Web http://www.tvcops.com (5).

Levin, G. R. (1971). Documentary Explorations. Garden City, NY: Doubleday.

Lichter, L. S. and Lichter, S. R. (1983) Prime Time Crime. Washington, DC:
The Media Institute.

Lippmann, W. (1922) Public Opinion. New York: Simon and Schuster.

Mustonen, A. and Pulkkinen, L. (1997). Television violence: A development
of a coding scheme. Journal of Broadcasting and Electronic Media, 41. pp.
168-189.

Oliver, M. B. (1994). Portrayals of crime, race, and aggression in
"reality-based" police shows: A content analysis. Journal of Broadcasting
and Electronic Media, 38 . pp. 179-192.

Oliver, M. B. and Armstrong, G. B. (1998). The color of crime: Perceptions
of Caucasians' and African-Americans' involvement in crime. In Fishman, M.
and Cavender, G. (eds.) Entertaining Crime Hawthorne, NY: Aldine De
Gruyter. pp. 19-35.

Riffe, D. Lacy, S., Nagovan, J. and L. Burkum. (1996). The effectiveness of
simple random and stratified random sampling in broadcast news content
analysis. Journalism and Mass Communication Quarterly, 73. Pp.159-168.

Rowen, B. (2001). History of reality TV. I. Retrieved February 20, 2001
from World Wide Web : http://www.infoplease.com

Schaefer, R. J. (1997). Editing strategies in television news
documentaries. Journal of Communication, 47 (4), 69-88.

Seagal, D. (1993) "Tales from the cutting-room floor," Harper's Magazine.
Nov. pp.50-57.

Smith, H. (1996). The Power Game: How Washington Works. New York: Random House.

Smith, R. R. & Windes, R. R. (1993). Symbolic convergence and abolitionism:
A terministic reinterpretation. Southern Communication Journal. 59, 45-59.

Swallow, N. (1966) Factual Televsion. London: Focal Press.

Zoglin, R. (1989). "The cops and the cameras," Time Apr 6. pp 62-63.