This paper was presented at the Association for Education in Journalism and
Mass Communication in San Antonio, Texas August 2005.
If you have questions about this paper, please contact the author
directly. If you have questions about the archives, email
rakyat [ at ] eparker.org. For an explanation of the subject line,
send email to
[log in to unmask] with just the four words, "get help info aejmc," in the
body (drop the "").
(Feb 2006)
Thank you.
Elliott Parker
====================================================================
Limiting the Warrior Woman on Prime-Time
Paper submitted to the Association for Education in Journalism & Mass
Communication: Commission on the Status of Women, San Antonio Conference, 2005
Limiting the Warrior Woman on Prime-Time: Using Content Analysis to
Examine the Ambiguous Messages of Empowerment and Containment
Jennifer M. Fogel
Syracuse University
S. I. Newhouse School of Public Communications
28770 Ramblewood Drive
Farmington Hills, MI 48334
(248) 851-2377
[log in to unmask]
Abstract
Within the past decade, television has produced an abundance of
action heroines that exhibit a duality of gender traits, confusing
gender stereotypes. Moreover, their power often comes at the price of
their femininity or is only considered a pretender to the culturally
established male authority. A content analysis of televised female
crime fighters examined the relationship between the origination
(whether human or metahuman) of the character and variables such as
personal problems, masquerade, relationships.
Limiting the Warrior Woman on Prime-Time: Using Content Analysis to
Examine the Ambiguous Messages of Empowerment and Containment
The representation of women in the media has been a concern in
cultural studies for quite some time. Due to the interaction between
society and culture regarding the so-called "feminine ideal," media
representations of female characters have sparked considerable
scrutiny, particularly after the Women's Movement in the 1970s. Over
the past 34 years, audiences have viewed both progressive and
regressive images of feminism on television. Moreover, within the
past decade television has produced an abundance of female action
heroes that serve as icons of female empowerment.
Currently, generations of young women search for female role models
on shows like Xena: Warrior Princess, Buffy the Vampire Slayer, and
Le Femme Nikita, all of which have spawned numerous academic
criticisms. Since 1994, there have been 21 female-led crime dramas on
prime-time (excluding those in syndication). These shows revolve
around female leads that exhibit a duality of gender traits,
confusing typical notions of gender roles. This apparent confusion
has sparked contention among feminist critics concerning the
transgressive nature of these warrior women, as their power often
comes at the price of their femininity or is only considered a
pretender to the culturally established male authority (Early &
Kennedy, 2003; Inness, 1999, 2004).
The intent of this study was to further understand conflicting images
of female crime fighters as warrior women on television. A content
analysis of televised crime dramas featuring women in lead or major
supporting roles premiering in the last decade, examined the
relationship between the origination (whether human or metahuman) of
the character and variables such as success rate, masquerade, and
personal life problems – including familial relationships,
friendships with both men and women, and romantic relationships.
Most of the feminist academic study regarding warrior women has opted
against quantifying data, instead choosing to explore ambiguities
through textual analysis of scenes and situations. However, I
believed a content analysis was warranted to measure the manifest
content of the crime drama. It also helps to quantify the occurrence
of variables from themes that critics using textual analysis have
cited as problematic, such as issues regarding job performance,
masquerade and the negotiation of professional and private lives.
Therefore, I created and tested hypotheses that relate to more
general issues regarding warrior women and ambiguous messages of
empowerment. With the average viewing diet on the rise to over seven
hours a day and children continuing to learn social cues from the
media, it is extremely important to ensure that the messages the
media produce positively reinforce gender equality.
Literature Review
While changing societal values during the 1970s saw an expansion of
visible roles for women on television, it also activated an immediate
concern over gender stereotyping. Early feminist television criticism
focused on quantitative data, exploring the issue of gender
stereotypes through content analysis and surveys. Since the 1970s,
content analyses of programming have shown that women are
under-represented in television programming and sex role images have
remained quite stable and conventional (Signorielli, 1989).
Signorielli and Bacue (1999) conducted a content analysis of
prime-time network dramas across three decades. Realizing that
television has become a "socializing agent" in our society, there is
an "underlying assumption that television's images about gender roles
have kept pace with the changes in the roles, status, and aspirations
of women that have taken place since the sixties" (p. 528). In
comparing programs of the 1990s to those of the 1970s and 1980s, they
found that males outnumbered females in network prime-time
programming in each decade (p. 532). However, they did not find
evidence to support that women were necessarily relegated to
domesticity, but rather both men and women were cast in gender-typed
occupations.
Elasmar, Hasegawa and Brain (1999) performed another content
analysis of portrayals of women on prime-time network programs during
the 1992-1993 season. The analysis looked at the general trends of
portrayals of women and how these portrayals relate back to society's
acceptance of women outside the home. The results showed more women
working on television compared to previous decades and 44% of the
female characters measured were employed outside the home. However,
women as a whole were still under-represented in prime-time
programming, and they still held lower status positions than men (p.
22). The study concluded that society's greater acceptance of women
led to the creation of more female roles on television, particularly
single women with professional occupations (p. 28).
Furthermore, Glascock (2001) conducted a content analysis of
prime-time programming during the 1996-1997 television season hoping
to gauge whether gender roles reflected more progressive social
mores. Glascock not only compared demographic variables between the
genders, but he also analyzed behaviors for both sexes in terms of
verbal/physical aggression, altruism and affection. He found that
characters portrayed behaviors typically associated with male and
female stereotypes (e.g. aggression and empathy) (p. 667). However,
he also found women to be more verbally aggressive, especially in comedies.
Within the last decade, female characters in prime-time have not
only assumed traditionally masculine occupations like cops and
lawyers, but they also are portrayed as more powerful and assertive.
Mirroring a slight evolution of gender roles in today's society,
these characters transgress patriarchal control by possessing
narrative agency. Oppenheimer, Adams-Price, Goodman, Codling and
Coker (2003) recently conducted a survey to examine audience
reactions to these new strong female characters, specifically gauging
how the audience viewed the femininity or masculinity of the
character. The study acknowledges that female characters were
"trapped in an either/or situation: maintain femininity by staying at
home…or become a masculinized woman in the working world" (p. 164).
In the end, changing social ideology about women's roles in society
was seen as a contributing factor to the acceptance of these characters.
In the 1970s, television began to offered strong female role models
in varying forms of crime fighters, yet these heroines provided a
limited view of gender disruptions; still stressing the importance of
keeping in line with cultural expectations of femininity if one
wanted to transgress the boundaries of patriarchy-defined behavior
(Helford, 2000). The Bionic Woman (1976-1978) and Wonder Woman
(1976-1979) may have overpowered the male dominated generic
conventions of the action drama, but they still conformed to typical
notions of feminine beauty ideals and submitted to male bosses and boyfriends.
It was not until the 1990s that action heroines regained their
presence in prime-time, this time in the form of warrior women. This
led to a greater number of female-led science fiction and fantasy
programming, along with more females taking up male dominated roles
in the justice system (p. 7). However, feminists criticize that while
these warrior women with their "girl power" rhetoric and
gender-bending behavior offer images of female empowerment, they
still play into patriarchal containment.
Yvonne Tasker (1993) was one of the first to use the term "warrior
women" to define the proliferation of strong female characters in
action hero roles in the cinema during the late 1980s and early
1990s. These women were no longer the cartoonish camp of the
televisual action heroine who needed to compensate for her
assertiveness by emphasizing her sexuality and traditional notions of
femininity. Instead, these women possessed not only narrative agency
but their muscularity and independence allowed them to break free of
male control, often placed in situations in which all the surrounding
males were either incompetent or irrelevant to the narrative action.
This challenge to the typically "masculine" genre has seen warrior
women change viewers' expectations of the hero and allowed for an
accommodation of female agency (Early & Kennedy, 2003, p. 3).
As previously indicated, current renditions of the warrior woman have
sparked debate over the transgressive nature of the action heroine.
Issues of sexuality (e.g. dress and exploitative femininity),
relationships (e.g. lack of relationships with the opposite sex, as
well as other women) and masquerade (e.g. the need to hide abilities
or engage in undercover work) are all themes feminist scholars have
found problematic. Yet, even while rupturing the conventions of the
action genre by controlling the narrative, warrior women often long
to be normal like other girls (Owen, 1999). Warrior women find their
position and obligation as protector/warrior overwhelms the ability
to maintain a "personal" life. Many warrior women are single, failing
to sustain a workable relationship with men due to work obligations
or hidden identities. Additionally, the guilt that warrior women
often feel because of their heroic abilities and their desire to be
like "normal" girls alienates them from the patriarchal world in
which they live (Crosby, 2004). Thus, warrior women as female crime
fighters often experience more personal problems than expressed by
their male counterparts, because they must negotiate their public
hero life with their private female desires. However, even though
male heroes are often portrayed as loners in order to assert their
individualistic tendencies, warrior women find strength in their
emotional connections to others, especially other women. The
interdependency of warrior women on female (and male) sidekicks
inverts the notion that the hero must work alone and also portrays a
form of feminist community bonding (Ross, 2004). Although isolated by
duty, warrior women find relationships integral to resistance of
patriarchal control and an emotional outlet, again redefining generic
conventions.
Hypotheses and conceptual definitions
Drawing upon past research, this study explores how warrior women
perform within a typically male environment—the crime drama—and
whether their empowerment is limited by conventional patriarchal
gender structures. First, it is important to define "crime fighter"
because it is the role that situates the warrior woman in the place
formerly occupied by the male hero. The crime fighter is also the
main source of the point of view, narrative agency and audience
identification (Heinecken, 2003). To simplify the definition, I will
use Scharrer's (2001) definition of "good guy" ("those who work to
uphold law, order, and justice often by enforcing the law and
capturing criminals") to refer to crime fighters (p. 618). Therefore,
crime fighters can refer to police officers, private investigators,
those who work in conjunction with the police (i.e., forensics
investigators), government agents, and characters that work against
those typically deemed as "evil." In addition, recent female-driven
crime dramas have focused on warrior women that possess certain
superpowers or supernatural gifts that increase their ability to
function as a crime fighter. For the study, crime fighters will also
be categorized as either superheroines or everyday crime fighters.
As discussed earlier, a warrior woman's physical prowess and
strength is often deemed a traditionally masculine trait. The
sexuality and femininity of the warrior woman are overemphasized in
order to compensate for these masculine qualities. According to
Inness (1999), "[The warrior woman's] tougher and more masculine
image suggests that a greater variety of gender roles are open to
women; at the same time, however, her toughness is often mitigated by
her femininity, which American culture commonly associates with
weakness" (p. 5). In other words, unlike the action hero, the warrior
woman will experience more personal problems in the course of her
work. Personal problems can greatly affect the job performance of the
crime fighter, and thus the warrior woman's success rate. Since the
warrior woman is still inscribed by patriarchal dichotomous gender
definitions, she is often read as unable to be both male and female
and this confusion causes failures in other portions of her life.
Personal problems include: negative effects on the well-being of
self, friends and family such as death, injury, loss of significant
other, capture of self or others, and arguments with family, friends,
and co-workers. Additionally, because superheroines often face
personal problems due to their need to hide their talents,
superheroines are more likely to experience more personal problems
than everyday crime fighters. However, their supernatural abilities
and resignation to their duty afford them a greater success rate than
everyday crime fighters:
H1: Superheroines will experience more personal problems than
everyday crime fighters.
H2: The success rate for superheroines will be larger than the rate
for everyday crime fighters.
Like earlier incarnations of the action heroine, contemporary warrior
women still perform forms of masquerade while carrying out their
duties. Masquerade can occur as either hiding a secret identity in
the form of a public identity, or actually going undercover in a new
form of costume to infiltrate certain situations. Warrior women often
use masquerade to conform to sociological gender expectations, thus
allowing them to use patriarchal gender conventions to their benefit
as they play down their more "masculine" type physical abilities.
Then, the element of masquerade could be seen as a way for warrior
women to overpower their enemies when they are inadvertently
underestimated. Moreover, this cross-gender behavior also shows an
interdependence between gender identities. Presently, superheroine
characters are more apt to undergo elements of masquerade because
they must keep their abilities hidden, while everyday crime fighters
might find that they can perform their duties without having to go
undercover:
H3: Superheroines will undergo a form of masquerade in order to
perform their duties more often than everyday crime fighters.
Whereas traditional action hero conventions call for the hero to be
separated from and elevated above the rest of the population to
encapsulate his rugged individualism, warrior women are increasingly
interdependent on others. Although for some warrior women it becomes
imperative to construct themselves as unlike other "girls," they also
work better when they are part of the community in which they
protect. Unlike past incarnations of action heroines, warrior women
desire companionship in the form of friendships as well as romantic
relationships. In general terms, men are typically more oriented
towards traits of individualism, assertiveness, and self-confidence
while women are more inclined to expressiveness and concern for
others (Bem, 1974). Moreover, Gardner and Gabriel (2004) concluded
that women are more interdependent and find social bonds more
important than men. However, Ross (2004) codes female friendships as
a form of feminist bonding to usurp patriarchal authority. Thus,
warrior women's interdependence on other women becomes beneficial to
their heroism. The same can be said for familial relationships, as
many warrior women find motivation and guidance from parents,
siblings and children. On the other hand, romantic relationships
often hinder the progress of warrior women and they are forced to
choose between love and duty. In some cases, these men harass and
disdain the position of warrior woman as an independent woman. Even
more, romantic relationships can become unhealthy or an emotional
weakness for warrior women. Superheroines are more likely to
establish relationships with others because they rely on close
friends and family to maintain an emotional outlet and share the
burden of a secret identity. On the other hand, everyday crime
fighters are more likely to experience unsuccessful relationships
because they are dedicated to their work, often finding that outside
attachments inhibit their ability to perform their duties:
H4: (a) Superheroines are more likely to have familial relationships
than everyday crime fighters. (b) Superheroines will have more
familial relationships than everyday crime fighters.
H5: (a) Superheroines are more likely to form friendships than
everyday crime fighters. (b) Superheroines will have more friendships
than everyday crime fighters. (c) Superheroines are more likely to
befriend women than men more so than everyday crime fighters.
H6: (a) Superheroines are more likely to have romantic relationships
than everyday crime fighters. (b) Superheroines will have more
romantic relationships than everyday crime fighters
H7: Superheroines will have more successful relationships overall
than everyday crime fighters
Some critics of warrior women have called "girl power" a new form of
containment for female empowerment (Fudge, 1999). Disguised as towers
of strength, these women exude extreme sex appeal making them more
economically viable for the media as objects of entertainment. Women
can adopt the characteristics of the action hero, but she must also
remain a female. By adhering to dichotomous gender restrictions,
these critics fail to see gender as socially constructed categories
that have become fluid. Even though the traits listed as typically
feminine or masculine as defined by the Bem Sex Role Inventory (Bem,
1974) have stayed constant (Auster & Ohm, 2000), Auster and Ohm
(2000) found that contemporary men and women desire to have a mix of
feminine and masculine traits for themselves. Therefore, warrior
women as "eye candy" and a construct of girl power, affirms strength
as a "masculine" quality yet "reveals the falsity of this assumption"
(Heinecken, 2003, p. 29). Through the negotiation of conventional
gender traits, current renditions of warrior women have created a
more realistic depiction of changing gender roles within contemporary
social contexts.
Method
Typically, previous feminist criticism has relied on textual
analyses and audience reception studies in order to allow for a
deeper investigation of the interplay between media images and
cultural contexts. While content analyses were able to quantify the
representation of women on television, they could not further
interpret audience reactions or the meanings that could be drawn from
particular texts. However, this study uses a content analysis to
measure identifiable variables concerning aspects of power issues and
gender relationships that feminists have explored in their textual
analyses of warrior women texts. Therefore, the content analysis aims
to show that these themes are clearly present and repetitive in the
text. This closer reading allows me to further examine the
ambiguities that have arisen in the feminist debate surrounding the
warrior woman as transgressive of patriarchal control.
Sampling
A content analysis allows me to investigate the "manifest"
(observed) content and connect it to the broader context of the
phenomenon of warrior women. The sampling frame consists of principal
female crime fighter characters appearing in weekly prime-time dramas
offered on the six major television networks within the last decade
(i.e. since September 1994). Principal characters are defined as
those who appear in the main title credits and have the largest
percentage of screen time. The unit of analysis is character and from
a sampling frame of 110 female crime fighters, a convenience sample
of 61 characters was chosen based on the availability of broadcasts
during the time the study was conducted. Within the last decade,
network crime dramas have depicted only 15 female crime fighter
characters that can be classified as superheroines, and 11 of them
are represented in the sample. Superheroines account for 18% of the
total sample and include all the superheroine characters presently on
network television. In order to qualify as a crime drama, the program
must include principal, recurring characters in pursuit of criminal
suspects or villains and aired more than five episodes before
cancellation. These characters can be classified as police officers,
private investigators, those who work in conjunction with the police
(i.e., forensics investigators and medical examiners), government
agents, and characters that work against supernatural "evil." Neither
Law & Order nor any other program that featured lawyers as the
principal characters were included in the original sampling frame
because their focus is on upholding the law and not the active
pursuit of criminals. Characters chosen for the sampling frame come
from shows that were listed on Fall television schedules since 1994
(www.inthe90s.com).
Three episodes per character from each character's respective show
were used for the content analysis. Since the crime dramas used to
obtain the sampling frame had to have aired at least five episodes
before cancellation, three episodes can potentially represent a large
percentage of the show's run. The episodes included were chosen at
random based upon the selection of episodes available for analysis by
October 2004. These episodes were either videotaped from their
original broadcast or syndication, or available through a DVD
collection. Collectively, the sample of characters represents 32
crime dramas (all one hour in length), and encompasses 119 hours of
programming.
Coding/Operationalization
Each character was placed into the nominal categories of an everyday
crime fighter (1) or superheroine (2). The superheroine variable was
operationalized as any character that displays abilities beyond
normal human capability. Therefore, abnormal characteristics of
physical, mental and emotional prowess were coded for each character.
Examples included abnormal strength, increased perception of the five
senses, magical powers and genetic abnormalities.
Another variable coded was job performance. This was measured by a
character's success rate. In order to investigate whether there is a
difference between everyday crime fighters and superheroines, the
success rate of each character was found by dividing the total number
of successes each character had in each episode by the total number
of problems. Success was defined as the number of problems solved in
each of the three episodes analyzed. Success could be a result of a
villain's incarceration, arrest, confession, injury or death, finding
the cause of death, and finding solutions to any problems that may
come up within the course of the investigation. Problems could be
finding evidence, looking for witnesses, missing criminals, and
unanswered questions. Next, the number of personal problems for each
character was coded for all three episodes. Personal problems were
defined as all issues of a personal nature that could affect job
performance and could include: injury, misunderstandings with
co-workers or friends, break-ups, and death.
Additionally, each character was coded for relationships. Subordinate
characters that were present in the episodes were placed into the
categories of familial, friendship, or romantic if the relationship
had significance to the character (e.g. parent, child, coworker,
etc.). Each of these relationships was further broken down into the
type of relationship based on the three general categories. Familial
relationships were labeled parent (1), sibling (2), child (3), or
other (e.g. nephew) (4). Friendships were labeled by gender (male =
1; female = 2) and whether or not the character was a coworker (Yes =
1; No = 2) of the crime fighter. Romantic relationships were labeled
as husband (1), steady boyfriend/girlfriend (2), or date (3).
In addition to the type of relationship, the overall success of the
relationship was important to determine. One measurement of a
successful relationship was whether or not the subordinate character
is aware (aware = 1; not aware = 2) of the crime fighter's job.
Another measurement coded the frequency of work-related and
non-work-related problems that affect the relationship between the
subordinate character and the crime fighter.
As shown from the previous literature review, the issue of masquerade
warrants study. Therefore, each character was also coded for the
presence of masquerade in which the character has to sustain a hidden
identity or use undercover work in order to perform their duties.
Masquerade included: lying to friends and family, dressing in
disguise, working under an assumed name, or leading a double life.
Each hypothesis was tested using either Pearson's Chi-square test of
proportions or an independent t-test, which tested the nominal
independent variable of character type (everyday or superheroine)
against all dependent variables coded earlier. The dependent
variables of personal problems, success rate, masquerade, number of
familial, friendships, and romantic relationships, work-related and
non-work-related problems were all ratio variables taken by
frequency. All other dependent variables were nominal and were
appropriate for proportional analysis. From the data, I then
determined whether there was a statistically significant difference
between everyday crime fighters and superheroines for each variable coded.
Reliability
A two-step process was used to establish intercoder reliability for
each variable. First, the author and a trained second coder who was
female (a graduate media studies student) coded 10% of the sample (6
characters) as a pre-test. The episodes used for this pre-test were
not included in the study. The pre-test found that the operational
definitions of many of the variables were not specific enough to
adequately and reliably test. After a thorough discussion between the
author and the second coder, new operational definitions were put
into place in the coding index, particularly more detailed
measurements within the relationship categories. With the new coding
index, 15% of the sample (9 characters) was then coded, this time
with episodes featured in the study. Intercoder reliability and
agreement was calculated using Scott's pi and Pearson's r. For the
variables type, presence of family, friends and romance, and
awareness Scott's pi = 1.0. Pearson's r was found for personal
problems (r = .99), success rate (r = .93), masquerade (r = 1.0),
number of familial relationships (r = 1.0), number of friendships (r
= .97), number of romantic relationships (r = 1.0), work-related
problems (r = .98) and non-work-related problems (r = .92).
Validity
As a convenience sample, the inclusion of certain characters was
based on their availability and therefore the study only contained
55% of the total 110 characters depicted on television within the
last decade. There could be an issue of generalizability to the
entire population of crime dramas. However, I believe that the
specific phenomenon under examination (warrior women) is well
represented by the sample, even though there is greater emphasis on
the most current renditions of crime fighters. Moreover, the sample
only includes 11 superheroines while the other 50 are everyday crime
fighters. There may be problems in the analysis with the low number
of superheroines when compared to everyday crime fighters, although
this is a reflection of the small amount of superheroines (15)
present on network television in the last decade. Additionally, there
could be problems with structural validity. Although each
operationalization appears on its face to accurately represent the
respective variables, there could be some definitional problems that
may affect the results.
Results
A total of 61 characters were coded in terms of their type of
character, personal problems, success rate, masquerade, and presence,
frequency and type of relationships. Of the 61 characters in the
sample, 50 (82%) were everyday crime fighters and the remaining 11
characters were categorized as superheroines. Superheroines accounted
for 18% of the total sample and included all of the superheroine
characters presently on broadcast television. Table 1 shows the
average number of personal problems, the mean success rate and the
mean frequency of masquerade. As expected, the average success rate
(82.61%) is relatively high. Crime fighters generally solve all of
their problems within the span of an episode. Table 1 also shows the
mean number of personal problems per character was 7.51, the mean
number of work-related problems was 2.01 and the mean number of
non-work-related problems was 1.31.
Since the number of the three types of relationships (familial,
friendship and romantic) is important in examining the hypotheses, it
was first important to identify whether the three types of
relationships were present overall for each character. Therefore, a
second level of analysis was performed to further investigate the
frequencies between crime fighters and the subordinate characters
coded in the three episodes (see Table 2). From the total sample of
61 characters, only 28 (46%) had some form of familial relationship
present within the three episodes coded. Within the sample, 44 (72%)
of the characters maintained a friendship and 36 (59%) displayed some
type of romantic relationship. Additionally, the gender of each
subordinate character (N = 193) was coded and 68% (132) were male
while 32% (61) were female (see Table 3). Table 3 also shows 68% of
the subordinate characters had a relationship with an everyday crime
fighter, while 32% had a relationship with a superheroine. Of the 193
subordinate characters coded, 102 maintained a friendship with a
crime fighter and the breakdown of the gender of these friendships
was 68 male and 34 female. In terms of whether or not the subordinate
character was aware of the crime fighter's job, 180 of the 193
subordinate characters were aware.
Hypothesis Tests
The first three hypotheses examined the relationship between the
type of crime fighter and issues related to performance. Hypothesis
1, predicting that superheroines would experience more personal
problems than everyday crime fighters, was supported by an
independent t-test. Table 4 shows that superheroines experienced
significantly more personal problems than everyday crime fighters.
The second hypothesis predicted that superheroines would have a
larger proportional success rate than everyday crime fighters. An
independent t-test did not find a significant difference between the
success rates of the superheroine in comparison to the everyday crime
fighter, thus Hypothesis 2 was not supported. Therefore, I conclude
that superpowers do not lead to a difference in job performance.
Hypothesis 3 predicted that superheroines would undergo a form of
masquerade while performing their duties more often than everyday
crime fighters. As Table 4 shows, this hypothesis was supported by an
independent t-test.
The next series of hypotheses dealt specifically with types of
relationships each crime fighter had with subordinate characters
present in any of the episodes coded in the study. The first type of
relationship examined was familial relationships. The total number of
familial relationships present in the episodes were counted and coded
by type of familial relationship (parent = 1, sibling = 2, child = 3,
other = 4). A category for multiple presences was added to the type
of familial relationship (multiple = 5) when a crime fighter had more
than one type present in the episodes coded (i.e. a character had
both a sibling and a child). Hypothesis 4a predicted that
superheroines were more likely to have familial relationships than
everyday crime fighters. Chi-square analysis revealed no significant
differences, thus Hypothesis 4a was not supported (see Table 5).
However, Hypothesis 4b (superheroines will have more familial
relationships than everyday crime fighters) was supported by an
independent t-test (see Table 9).
Additionally, Table 6 examines the type of familial relationship
each character had and the differences between superheroines and
everyday crime fighters were significant. Everyday crime fighters had
more parent only or child only relationships while superheroines had
more siblings only or multiple family members. Yet, because of the
low cell frequency due to the sample size, this test is used more for
descriptive purposes than statistical inference.
The second type of relationship examined was friendships. A
chi-square analysis found that Hypothesis 5a (superheroines are more
likely to form friendships than everyday crime fighters) was not
supported (see Table 7). Hypothesis 5b predicted that superheroines
have more friendships than everyday crime fighters. An independent
t-test showed that even though neither crime fighter is more likely
to have friends, of those that do, superheroines have significantly
more friendships than everyday crime fighters (see Table 9).
Therefore, Hypothesis 5b was supported. Chi-square analysis was also
used to test Hypothesis 5c, which predicted superheroines are more
likely to form friendship with women than men and more so than
everyday crime fighters. Table 10 shows that superheroines are more
likely to have female friendships than everyday crime fighters, but
they also have a significantly higher percentage of friends from both
genders. Thus, Hypothesis 5c was also supported.
The third type of relationship examined was romantic relationships.
Romantic relationships were coded by type (husband = 1, steady
boyfriend/girlfriend = 2, date = 3). Romantic relationships were also
coded by frequency of presence in the episodes used for the study. If
a subordinate character happened to be a steady boyfriend in one
episode and a husband in another, the character was coded as a
husband since a husband would have a more significant relationship to
a crime fighter. Hypothesis 6a predicted that superheroines are more
likely to have romantic relationships than everyday crime fighters.
Table 8 shows that a chi-square analysis found that a higher
percentage of superheroines had romantic relationships than everyday
crime fighters, thus Hypothesis 6a was supported. In fact, all of the
superheroines in the sample had some form of romantic relationship in
the episodes coded, while only half of the everyday crime fighters
displayed a romantic interest. However, even though all of the
superheroines had a romantic relationship present, there was no
significant difference between superheroines and everyday crime
fighters in terms of the number of romantic relationships present for
each character (see Table 9). Therefore, Hypothesis 6b was not supported.
Hypothesis 7 examined the overall success of the relationships crime
fighters have with subordinate characters. The success of the
relationship was based on the number of work- and non-work-related
problems inherent in the relationship with the subordinate character,
and whether or not the subordinate character was aware of the crime
fighter's job. Therefore, a successful relationship with a
subordinate character would have the fewest number of problems and
the character would be aware of the crime fighter's job. Hypothesis 7
predicted that superheroines would have more successful relationships
overall than everyday crime fighters. However, two independent
t-tests and a chi-square analysis found that everyday crime fighters
have more successful relationships. Table 11 shows that superheroines
have significantly more work-related problems than everyday crime
fighters. Another independent t-test found no significant differences
between everyday crime fighters and superheroines in regards to
non-work-related problems. However, a chi-square analysis shows that
everyday crime fighters had more relationships with people who were
aware of their job (96.9%) than superheroines (85.5%) (see Table 12).
Therefore, the data do not support Hypothesis 7 but do show that
everyday crime fighters have more successful relationships overall
than superheroines.
Discussion
The intent of this study was to further understand conflicting
images of female crime fighters as warrior women on television. The
study explored issues regarding the complexities inherent in the
transformation of women into positions of power and authority
historically reserved for men. In order to successfully navigate this
transition, warrior women are often portrayed with both masculine and
feminine traits. However, this mixture of traits tends to become
extremely problematic for women because they are still expected to
adhere to patriarchal conventions. Therefore, even though these
warrior women are given more agency within the narrative, their
empowerment is often mitigated by ambivalent and conflicting
messages. These strong and assertive women must also project the
"ideal" female body and constantly weigh their professional duties
against personal life desires. Thus, I argue that warrior women are
more likely to experience problems because they are forced to
negotiate between patriarchal conventions and feminist aspirations.
By analyzing significant differences among female crime fighters,
particularly those who are portrayed as having added superhuman
abilities, I can more fully understand the limitations and
empowerment of warrior women on television. Therefore, I examined the
relationship between the origination (whether human or metahuman) of
the character and variables such as success rate, masquerade, and
personal life problems—including familial relationships, friendships
with both men and women, and romantic relationships. The analysis
found that there were significant differences between everyday crime
fighters and their superheroine counterparts. As expected,
superheroines had more personal problems than everyday crime
fighters. Since warrior women are inscribed by patriarchal gender
definitions, they are often forced to exhibit either female or
"figuratively" male traits, which in turn cause failures in having
successful personal lives. Additionally, superheroines often face
personal problems due to their need to hide their talents, and thus
are more likely to have more personal problems than everyday crime
fighters. However, the analysis found no significant differences
between the proportional success rate of everyday crime fighters and
superheroines. Even though personal problems can greatly affect job
performance, neither type of crime fighter displayed evidence of
lowered success rates. Even supernatural abilities did not
significantly affect the success rate of superheroines. Overall, in
measuring job performance I found that the average success rate was
high (82.6%), most likely a benefit of the generic conventions of the
televised crime drama wherein problems are often solved within the
span of the program. Moreover, the analysis showed that superheroines
did undergo a form of masquerade while performing their duties more
often than everyday crime fighters. This is mostly due to the fact
that superheroines have dual identities in order to keep their
supernatural abilities hidden from the world.
Additionally, in terms of relationships, superheroines had a
significantly higher presence of familial characters and friends than
everyday crime fighters. Superheroines also had more siblings-only or
multiple family members present while everyday crime fighters had
more parent-only or child-only relationships. This is important
because many of the familial relationships that everyday crime
fighters had were relegated to the background of the storyline unlike
family members who had a relationship with a superheroine.
Furthermore, having multiple familial relationships could lead to
more problems overall for the crime fighter, particularly if they are
central to the actions of the warrior woman. Moreover, friendships
were more likely to be with both men and women, although
superheroines did have a higher percentage of female friends than
their counterparts. However, superheroines were no more likely to
have familial and friend relationships than everyday crime fighters.
This is once again due to generic conventions that use these types of
relationships to fully flesh out warrior women characters and allow
the audience to see them from different perspectives.
Another equally used plot convention is the addition of a romantic
interest. It was not surprising to find that all of the superheroine
characters sampled had a romantic relationship of some type. However,
the analysis found no significant difference between superheroines
and everyday crime fighters in terms of the number of romantic
relationships present for each character. Again, because the
portrayal of a romantic interest for a female character is heavily
relied upon in this genre, it is not surprising that there would be
no differences found.
Even though secondary characters are less likely to be aware of a
superheroine's job and abilities, I believed that the overall success
of the warrior woman's relationship with other characters would be
higher than her everyday counterpart. Moreover, many relationships
evolve to the point where superheroines feel comfortable in confiding
in these secondary characters. However, the analysis found that
subordinate characters had more work-related problems with
superheroines and were less aware of their job. Therefore, my
prediction that everyday crime fighters are more likely to experience
unsuccessful relationships because they are dedicated to their work
and find outside attachments a nuisance in their ability to perform
their duties was incorrect.
Unlike previous content analyses that have concentrated on the
frequency of warrior women on television and their respective job
duties (Elasmar, Hasegawa, & Brain, 1999; Signorielli & Bacue, 1999),
my analysis looked more at the frequency of problems and relationship
issues that are inherent in the portrayal of the warrior woman.
Moreover, these issues concentrate more on character portrayals than
audience reactions to strong female characters. Thus, the content
analysis, using variables drawn from issues explored in previous
textual analyses, could both support previous literature, as well as
inform the specific situations that are used to demonstrate themes of
agency and gender conflation in later textual analyses. However,
findings should be considered with some caution due to sample size,
particularly with the low number of superheroines compared to
everyday crime fighters. Additionally, the sample was a convenience
sample drawn from the characters that were available at the time the
study was conducted, which may also have an effect on the results.
References
Auster, C. J., & Ohm, S. C. (2000). Masculinity and femininity in
contemporary American
society: A reevaluation using the Bem Sex-Role Inventory. Sex Roles,
43(7/8), 499-528.
Bem, S. L. (1974). The measurement of psychological androgyny.
Journal of consulting
and clinical psychology, 42, 155-162.
Crosby, S. (2004). The cruelest season: Female heroes snapped into
sacrificial heroines.
In S. A. Inness (Ed.). Action chicks: New images of tough women in
popular culture (pp. 153-180). NY: Palgrave Macmillian.
Early, F., & Kennedy, K. (2003). Introduction: Athena's daughters. In
F. Early, & K.
Kennedy (Eds.). Athena's daughters: Television's new women warriors
(pp. 1-10). Syracuse, NY: Syracuse University Press.
Elasmar, M., Hasegawa, K., & Brain, M. (1999). The portrayal of women in U.S.
primetime television. Journal of Broadcasting & Electronic Media, 43(1), 20-34.
Fudge, R. (1999, Summer). The Buffy effect or, a tale of cleavage and
marketing. Bitch
Magazine, 10.
Gardner. W. L., & Gabriel, S. (2004). Gender differences in
relational and collective
interdependence. In A. H. Eagly, A. E. Beall, & R. J. Sternberg
(Eds.). The Psychology of Gender (pp. 169-191). NY: The Guilford Press.
Glascock, J. (2001). Gender roles on prime-time network television:
Demographics and
behaviors. Journal of Broadcasting & Electronic Media, 45(4), 656-669.
Heinecken, D. (2003). The warrior women of television: A feminist
cultural analysis of
the new female body in popular media. NY: Peter Lang Publishing, Inc.
Helford, E. R. (2000). Introduction. In E. R. Helford (Ed.). Fantasy
girls: Gender in the
new universe of science fiction and fantasy television (p. 1-9).
Lanham, Maryland: Rowman & Littlefield Publishers, Inc.
Inness, S. A. (1999). Tough girls. Philadelphia, PA: University of
Pennsylvania Press.
Inness, S. A. (2004). Introduction: "Boxing gloves and bustiers:" New
images of tough
women. In S. A. Inness (Ed.). Action chicks: New images of tough
women in popular culture (pp.1-17). NY: Palgrave Macmillian.
Oppenheimer, B., Adams-Price, C., Goodman, M., Codling, J., & Coker,
J. D. (2003).
Audience perceptions of strong female characters on television.
Communication Research Reports, 20(2), 161-172.
Owen, A. S. (1999). Buffy the Vampire Slayer: Vampires, postmodernity, and
postfeminism. Journal of Popular Film & Television, 27(2), 24-31.
Ross, S. (2004). "Tough enough:" Female friendship and heroism in
Xena and Buffy. In
S. A. Inness (Ed.). Action chicks: New images of tough women in
popular culture (pp. 231-256). NY: Palgrave Macmillian.
Scharrer, E. (2001). Tough guys: The portrayal of hypermasculinity
and aggression in
televised police dramas. Journal of Broadcasting & Electronic Media,
45(4), 615-634.
Signorielli, N. (1989). Television and conceptions about sex roles:
Maintaining
conventionality and the status quo. Sex Roles, 21(5/6), 341-360.
Signorielli, N., & Bacue, A. (1999). Recognition and respect: A
content analysis of
prime-time characters across three decades. Sex Roles, 40(7/8), 527-544.
Tasker, Y. (1993). Spectacular bodies: Gender, genre and the action
cinema. London:
Routledge.
Table 1
Means and Standard Deviation for Crime Fighter Personal Problems,
Success Rate, Masquerade, Number of Familial, Friendship and Romantic
Relationships, and Work-Related and Non-Work-Related Problems Variables
Variables
Mean
Std. Deviation
N
Frequency of personal problems
7.51
5.39
61
The success rate
82.61%
12.78
61
Frequency of masquerade
1.00
.50
61
Number of familial relationships
.87
1.23
28
Number of friendships
1.74
1.48
44
Number of romantic relationships
.77
.76
36
Frequency of work-related problems
2.01
1.98
61
Frequency of non-work-related problems
1.31
1.80
61
Table 2
Percentages for Crime Fighter Character Relationship Variables
Variables
Percent
Type of crime fighter
Everyday
82.00
Superheroine
18.00
100.00%
(N = 61)
Presence of familial relationships
Yes
46.00
No
54.00
100.00%
(N = 61)
Presence of friendships
Yes
72.00
No
28.00
100.00%
(N = 61)
Presence of romantic relationships
Yes
59.00
No
41.00
100.00%
(N = 61)
Table 3
Percentages for Subordinate Character Relationship Variables
Variables
Percent
Gender of subordinate character
Male
68.00
Female
32.00
100.00%
(N = 193)
Relationship to type of character
Everyday
68.00
Superheroine
32.00
100.00%
(N = 193)
Gender of the friend to crime fighter
Male
67.00
Female
33.00
100.00%
(N = 102)
Awareness of crime fighter's job
Yes
93.00
No
7.00
100.00%
(N = 193)
Table 4
Independent t-tests for Personal Problems, Success Rate and
Masquerade (Everyday or Superheroine)
Variables
Everyday
mean (& SD)
(N=50)
Superheroine mean (& SD)
(N=11)
t value
df
Significance
Frequency of personal problems
6.30
(5.13)
13.00
(2.24)
-6.76
59
p < .05
The success rate
82.80%
(13.41)
81.73%
(9.92)
.25
59
NS
Frequency of masquerade
.72
(2.19)
2.27
(2.15)
-2.13
59
p < .05
Table 5
Cross-tabulation of Presence of Familial Relationship and Type of Character
Type of Character
Presence of family
Everyday
Superheroine
Yes
42.0%
63.6%
No
58.0
36.4
100.0%
(N = 50)
100.0%
(N = 11)
X2 = 1.70, df = 1, NS
Table 6
Cross-tabulation of Type of Family Member and Type of Character
Type of Character
Type of family member
Everyday
Superheroine
Parent
33.3%
14.3%
Sibling
9.5
28.6
Child
38.2
0.0
Other
9.5
0.0
Multiple
9.5
57.1
100.0%
(N = 21)
100.0%
(N = 7)
X2 = 10.89, df = 4, p < .05
Table 7
Cross-tabulation of Presence of Friendship and Type of Character
Type of Character
Presence of friendships
Everyday
Superheroine
Yes
68.0%
90.9%
No
32.0
9.1
100.0%
(N = 50)
100.0%
(N = 11)
X2 = 2.35, df = 1, NS
Table 8
Cross-tabulation of Presence of Romantic Relationship and Type of Character
Type of Character
Presence of romantic relationships
Everyday
Superheroine
Yes
50.0%
100.0%
No
50.0
.0
100.0%
(N = 50)
100.0%
(N = 11)
X2 = 9.32, df = 1, p < .05
Table 9
Independent t-tests for Number of Familial, Friendship and Romantic
Relationships (Everyday or Superheroine)
Variables
Everyday
mean (& SD)
Superheroine mean (& SD)
t value
df
Significance
Number of familial relationships
1.48
(.60)
N = 21
3.14
(1.57)
N = 7
-2.74
6.59
p < .05
Number of friendships
2.12
(1.04)
N = 34
3.40
(1.17)
N = 10
-3.34
42
p < .05
Number of romantic relationships
1.24
(.44)
N = 25
1.45
(.69)
N = 11
-.95
13.67
NS
Table 10
Cross-tabulation of Gender of Friend and Type of Character
Type of Character
Gender of friend
Everyday
Superheroine
Male
44.1%
.0%
Female
8.8
10.0
Both Sexesa
47.1
90.0
100.0%
(N = 34)
100.0%
(N = 10)
a Since the gender of the friend was only noted once, a character
could have only male, only female or both types of friends.
X2 = 6.93, df = 2, p < .05
Table 11
Independent t-tests for Work-Related and Non-Work-Related Problems
(Everyday or Superheroine)
Variables
Everyday
mean (& SD)
(N = 131)
Superheroine mean (& SD)
(N = 62)
t value
df
Significance
Number of work-related problems
1.74
(1.82)
2.58
(2.19)
-2.80
191
p < .05
Number of non-work-related problems
1.17
(1.73)
1.60
(1.92)
-1.55
191
NS
Table 12
Cross-tabulation of Awareness and Type of Character
Type of Character
Awareness of job
Everyday
Superheroine
Yes
96.9%
85.5%
No
3.1
14.5
100.0%
(N = 193)
100.0%
(N = 193)
X2 = 8.80, df = 1, p < .05
|
