College Students’ Multiple
Stereotypes of Lesbians:
A Cognitive Perspective
Wendy Geiger, PhD
Central Missouri State University
Jake Harwood, PhD
University of Arizona
Mary Lee Hummert, PhD
University of Kansas
ABSTRACT. This paper examines stereotypes of lesbians held by col-
lege students. Multiple stereotypes are elicited from a free response trait
listing task, followed by a sorting task. The results of the sorting task are
submitted to cluster analysis and multidimensional scaling to reveal the
complexity of cognitive representations of this group. Eight types are
described, reflecting underlying distinctions between positive percep-
tions (e.g., lipstick lesbian, career-oriented feminist) and negative per-
ceptions (e.g., sexually deviant, angry butch) and also between relative
strength and weakness. The research is discussed in terms of cognitive
Wendy Geiger (PhD, University of Kansas) is Assistant Professor of Communica
tion at Central Missouri State University. Jake Harwood (PhD, University of Califor
nia, Santa Barbara) is Associate Professor of Communication at the University of
Kansas. Mary Lee Hummert is Professor of Communication at the University of
The authors express gratitude to Greg Shepherd, Howard Sypher, Virgil Adams,
and Haley Nevins for their assistance during the project.
Correspondence may be addressed: Wendy Geiger, Department of Communica
tion, 127 C Martin Hall, Warrensburg, MO 64093 (E-mail: firstname.lastname@example.org).
Journal of Homosexuality, Vol. 51(3) 2006
Available online at http://jh.haworthpress.com
© 2006 by The Haworth Press, Inc. All rights reserved.
perspectives on stereotyping and gender inversion theory. Suggestions
for future research are provided.
doi:10.1300/J082v51n03_08 [Article cop
ies available for a fee from The Haworth Document Delivery Service: 1-800-
HAWORTH. E-mail address: <email@example.com> Website:
<http://www.HaworthPress.com> © 2006 by The Haworth Press, Inc. All rights
KEYWORDS. Homophobia, lesbians, stereotypes, cognition, traits, at
titudes, intergroup, gay, cluster analysis
MULTIPLE STEREOTYPES OF LESBIANS
In recent years, our culture has begun to acknowledge gay men and
lesbians, and their functioning as a group within society. Silence and de-
nial has been replaced by conversation. The confirmation that gay men
and lesbians are a part of society is reflected in political, media, and
religious organizations. Legislation regarding gays in the military, same-
sex marriage (and proposed constitutional amendments denying such
rights), and federal protection against discrimination and hate crimes
has marked a significant increase in political talk. Movies are increas-
ingly exploring gay and lesbian themes and including more homosexual
characters. Prime-time television programs such as Will and Grace,
Spin City, ER, and Ellen include either leading or supporting gay and
lesbian characters, and many of those shows have explored themes spe-
cifically associated with being gay (e.g., “coming out”). In addition, re-
ligious organizations have faced controversy concerning how to deal
with men and women in same-sex relationships (e.g., United Methodist
News Service, 1999). This discussion of gay issues was absent from the
mainstream only 10 or 15 years ago, and illustrates the changing place
of lesbians and gay men in today’s society.
In the context of this increasing public discourse, it is important to
understand how individuals view gay men and women, but the research
in this area is somewhat limited. What does exist suggests that attitudes
are overwhelmingly negative (Herek, 1988, 1991). In addition, much of
the existing research has grouped gay men and lesbians (Kite & Whit
ley, 1996). Given that stereotypes of men and women are very different
(Eckes, 1996), we believe that stereotypes of gay men and women are
worthy of separate examination. To illustrate the value of this approach,
the study reported here examines the complexity in cognitive represen
tations of lesbians.
166 JOURNAL OF HOMOSEXUALITY
The most prevalent theory driving research on gay stereotypes is gen
der inversion theory (Kite & Deaux, 1987; Taylor, 1983). This theory
hypothesizes that gay men will be perceived as being similar to hetero
sexual women, and lesbians will be viewed as more like heterosexual
men. Both Taylor and Kite and Deaux have applied this perspective in
their research on stereotypes of homosexuals. For example, Taylor’s re
spondents rated gay men as more feminine than lesbians, and lesbians
as more masculine than gay men. Kite and Deaux (1987) also found
evidence supporting gender inversion theory. Their participants used
feminine traits (e.g., high voice, sensitive) to describe gay men and mas
culine traits (e.g., short hair, tough, like sports) to describe lesbians.
Such research has been useful, as have more simplistic “positive versus
negative” studies of attitudes towards lesbians and gay men ( Schwan
berg, 1996 ). However, past research in general has failed to examine
the complexity and variability in stereotypes of lesbians or gay men.
The cognitive perspective on stereotyping provides a framework that
helps to address this issue (Ashmore & Del Boca, 1981; Hamilton &
Trolier, 1986). The cognitive perspective asserts that stereotypes are
multiple and individuals may hold several stereotypes regarding any
given group (Brewer, Dull, & Lui, 1981; Hummert, 1990). These are ar-
ranged hierarchically, with superordinate categories having several
subcategories or substereotypes at lower levels. For example, stereo-
types of college students may include student athlete, sorority/fraternity
member, overly studious, party-loving student, and the like (Ashmore,
Del Boca, & Beebe, 2002). These subcategories exist within a broader
cognitive representation of the group as a whole, providing the perceiver
with a variety of information to apply in any given situation. The majority
of research on such hierarchical representations suggests that particular
stereotypes will be activated depending upon situational cues and the
characteristics of specific out-group targets (Hummert et al., 2004).
Research on age (Brewer et al., 1981; Hummert et al., 1994) and gen
der (Eckes, 1994; Six & Eckes, 1991) stereotypes supports the cognitive
perspective on stereotypes. For instance, Hummert et al. (1994) found
positive and negative stereotypes of older adults (e.g., the “perfect
grandparent” and the “golden ager” versus the “shrew/curmudgeon” and
“severely impaired”). Similarly, Eckes’ (1994) study of gender stereo
types described multiple stereotypes of women. He suggested that ste
reotypes of women are split along a traditional (e.g., the “Chick” and
“Housewife”) versus unconventional (e.g., “Career Woman” and the
“Women’s Libber”) divide. This research has helped us understand
cognitive representations of groups, the links between such representa
Geiger, Harwood, and Hummert 167
tions and intergroup communication, and the ways in which stereotypes
can be changed, or resist change, as a result of intergroup contact (Hew
stone & Brown, 1986; Hummert et al., 2004).
Our study explores stereotypes of lesbians from the cognitive per
spective, building upon Kite and Deaux’s (1987) research. Lesbians
have rarely been the sole focus of empirical research, and have often
been grouped with gay men for research purposes. We argue that these
groups are distinct and worthy of separate consideration. Given work
from the cognitive perspective, we aimed to uncover the content and
structure of stereotypes of lesbians. Based on previous stereotype re
search, as well as gender inversion theory, we expected to uncover both
positive and negative stereotypes of lesbians, and we expected to un-
cover stereotypes that varied in terms of their masculinity, with at least
some subtypes that reflected considerable masculine elements.
The study occurred in two stages. In the first stage, participants gener-
ated traits that they associated with lesbians as a group (trait generation
task). In the second stage, a separate group of participants sorted these
traits into piles intended to represent types of lesbians (trait-sorting task).
This procedure is commonly employed within the literature identifying
substereotypes of various groups (Hummert et al., 1994; Eckes, 1994).
Hierarchical cluster analysis and multidimensional scaling were then
used to identify the structure of cognitive representations of lesbians.
All respondents were enrolled in an introductory speech communica
tion course at a large mid-western university. They received course credit
for participation. The course fulfills a college requirement, and hence at
tracts a wide variety of majors. In the trait generation task, 61 people
participated (45 females, 16 males, age M = 20.34, SD = 3.15, range =
18-34). The trait-sorting task included 63 participants (39 females, 24
males, age M = 20.34, SD = 2.54, range 18-25). The majority of partici
pants in both studies were white (92%).
Trait Generation Task Materials and Procedures
Participants in the trait generation task were instructed to “Think about
what you know about lesbians, what you think about lesbians, what you
168 JOURNAL OF HOMOSEXUALITY
have heard about lesbians and what you have read about lesbians.” Par
ticipants were encouraged to write down as many traits as they could
think of, both positive and negative. They were instructed that the traits
did not have to reflect their own opinions, but simply associations with
lesbians with which they were familiar. The reason for including this
caveat was to get the most complete and broad description of lesbians
and to encourage participants to include traits (either positive or nega
tive) that they may not have felt comfortable including otherwise. The
average number of traits generated was 7.75 (range = 0-27). The re
spondents listed 355 characteristics and traits in total. To obtain a man
ageable number of traits for a sorting task, the researchers established
criteria for inclusion of traits in the final list.
First, synonyms of, and derogatory terms for, being a lesbian were
eliminated from the list (e.g., “gay,” “whorebag”). Second, very similar
trait descriptors were grouped (e.g., “manly,” “male-like”), and the re-
searchers jointly decided on one trait that was the best exemplar. Fi-
nally, responses that were not traits or were obscure were deleted from
the list (e.g., “lesbians should not be allowed to adopt”). Our litmus test
here was that each term on the final list should make sense in the sen-
tence “Lesbians are ______.” Two authors performed these procedures
independently, and then met to discuss and reach consensus on differ-
ences of opinion. The process resulted in a list of 94 traits for the sorting
task in the current study (see Figures 1 and 2 for the full list). Narrowing
down the trait lists was essential in order to provide a manageable set of
traits for the sorting task. The presence of obscure terms or synonyms
would have negative repercussions both methodologically and statisti-
cally (Everitt, 1993; Aldenderfer & Blashfield, 1985).
Trait-Sorting Task Materials and Procedures
Procedures for the sorting task were based on methods used by
Hummert et al. (1994) and Schmidt and Boland (1986) in studying ste
reotypes of older adults. Participants were given a stack of 94 randomly
ordered index cards, each displaying one trait. The participants were
asked to sort the cards into groups that represented types of lesbians of
which they were aware, either from personal contact, the media, friends,
family, or others. They were instructed that they did not need to believe
that these characterizations were true or accurate. The participants were
instructed to create 3-25 groups. Once the groups were formed, partici
pants were asked to provide a name or label for the type of lesbian that
each group represented. No time limit was imposed on the task and par
Geiger, Harwood, and Hummert 169
ticipants could change their groupings while they worked. The task took
most respondents 20-40 minutes. Groups that were labeled as “miscel
laneous” or unlabelled were not included in the analysis. Hence, some
of the participants had only 2 groups recorded, even though the instruc
tions called for a minimum of 3 groups. The number of groups gener
ated ranged from 2 to 15 (M = 5.87, SD = 2.71; 60.3% of the participants
generated 5 or more groups, 39.7% generated 4 or fewer groups).
The results of the sorting task were transformed into a 94 × 94 similar
ity matrix, with entries in the matrix reflecting the frequency with which
each pair of traits was placed into the same group. The data were analyzed
using hierarchical cluster analysis (HCA: Everitt, 1993) and multidimen
sional scaling (MDS: Kruskal & Wish, 1978). HCA has been used exten-
sively to identify substereotypes from a variety of groups, including age
(Hummert et al., 1994; Schmidt & Boland, 1986), gender (Eckes, 1994;
Edwards, 1992; Six & Eckes, 1991), and the mentally ill homeless (Mow-
bray, Bybee, & Cohen, 1993). The combined use of HCA and MDS pro-
vides statistical triangulation in terms of interpreting the structure of the
lesbian stereotypes, as well as providing a richer picture of their cognitive
organization (Kruskal & Wish, 1978).
Results of HCA using the average-linkage between-group method
produced two high-level clusters, one of positive traits and one negative
of negative traits. Determining that there were two high-level clusters
was based on the large percentage change (62%) between clusters indi
cated by the agglomeration coefficients (Hair et al., 1992). Each of the
high–level clusters had equal numbers of traits (N = 47 traits for the pos
itive and negative clusters). Examination of the dendrogram and the
agglomeration schedule also indicated a distinct set of clusters at a
lower level-eight clusters. Using Schmidt and Boland’s (1986) and
Hummert’s (1990) guidelines, only mid-level categories with more than
two traits were considered as representing subcategories. These subcat
egories or stereotypes are described below. They were labeled by the re
searchers who consulted with the respondents’ trait-sorting labels, and
who also used general cultural knowledge of the lesbian community.
170 JOURNAL OF HOMOSEXUALITY
The dendrogram for the positive subtypes is shown in Figure 1. Four
positive subcategories emerged: Lipstick lesbian, career-oriented femi
nist, soft-butch, and free-spirit. Fourteen traits were associated with the
lipstick-lesbian category. This is a stereotype associated with very femi
nine qualities such as beauty, sensitivity, and maternal instincts. The ca
reer-oriented feminist lesbian incorporated 24 traits, and was the broad
est of all the subtypes. This was a woman who is a proud, successful
professional with a strong sense of self. She is also open-minded and
creative. Three traits were associated with the soft-butch lesbian. This
subtype was of an athletic, powerful feminist, but not one who is explic
itly masculine (cf. the angry butch stereotype, below). Finally, the free-
spirit (six traits), was characterized as an eccentric and mysterious per
son who is a nonconformist.
The dendrogram for the negative subtypes is shown in Figure 2. As
with the positive clusters, four mid-level clusters emerged here: Hyper-
sexual, sexually confused, sexually deviant,andangry butch.Thehyper-
sexual lesbian was defined by three traits and characterized as a woman
who enjoys having sex with a variety of people, including both men and
women. The sexually confused lesbian had a base of six traits, and sug-
gested a woman who is “closeted,” confused, and uncomfortable re-
garding her sexual orientation. The sexually deviant lesbian, with a base
of 13 traits, characterized an immoral or mentally ill woman who chooses
to have the “wrong” type of sex. This appeared to be the most negative
subtype. Finally, 25 traits were associated with the angry butch lesbian.
These traits characterized an angry, dominating, defensive, and humor
less person, who was also seen as being masculine and unattractive in
appearance (i.e., stocky, muscular, and/or overweight).
In addition to the hierarchical cluster analysis, a multidimensional
scaling (MDS) procedure was performed on the 94 × 94 matrix. A
two-dimensional solution provided the best fit to the data (Stress = .10,
= .97). The MDS solution of the trait-sorting matrix is illustrated in
Figure 3, overlaid by the cluster analysis solution. As can be seen, the
subtypes of lesbians that were identified in the cluster analysis were ap
parent in the MDS analysis. The groups on the left side of the model are
all negative stereotypes, while the groups on the right are positive clus
ters. Hence, we interpret the horizontal dimension of the model as re
Geiger, Harwood, and Hummert 171
172 JOURNAL OF HOMOSEXUALITY
FIGURE 1. Dendrogram of Traits from Sorting Task: Positive Subtypes
Geiger, Harwood, and Hummert 173
FIGURE 2. Dendrogram of Traits from Sorting Task: Negative Subtypes
FIGURE 3. Multidimensional Scaling Solution of Traits Overlaid by Cluster Analysis Solution
flecting a negative-positive dimension. The vertical dimension in
Figure 3 appears to reflect a strong-weak dimension. Strong personality
characteristics predominate at the top of the model (e.g., dominant,
strong-willed, outspoken, powerful), with weak characteristics at the
bottom of the model (e.g., shy, submissive, insecure). As can be seen
from Figure 3, the three clusters with implications for specific sexual
behaviors are fairly closely entwined, as might be expected (center and
left at bottom of the model). The other negative stereotype (angry
butch) seems somewhat distinct, and to be associated with the more
“powerful” pole. On the positive side, the career-oriented feminist ste
reotype appears to be central, with the other positive stereotypes exist
ing as satellites. The career-oriented feminist type is the broadest
stereotype that we uncovered, incorporating elements of the other three.
Traits like “vegetarian” or “natural” might typically be associated with
the free-spirit type, traits like “assertive” and “outspoken” are perhaps
also characteristic of the soft-butch, and a trait like “creative” might be
seen as typical of the lipstick lesbian. Put differently, the MDS solution
suggests that negative stereotypes of lesbians are somewhat more di-
verse and distinct in their content (and hence perhaps more complex)
than positive types.
Finally, it is interesting to note the areas in which particular stereo-
types are most closely related in the multidimensional space. The “free-
spirit” stereotype is spatially located not far from two of the sexually re-
lated negative stereotypes (sexually confused and hypersexual). Like-
wise, the soft-butch stereotype is relatively close to (although apparently
distinct from) the angry butch. In both cases we would suggest that
these are possible routes for recategorization from positive to negative
stereotypes or vice versa. For instance, an individual who is initially
classified positively into the free-spirit stereotype might easily be
recategorized into a closely related negative stereotype (e.g., hyper
sexual). Recategorization from, for instance, the lipstick-lesbian cate
gory to the angry butch category is less likely according to Figure 3.
Our analysis reveals support for complex and diverse cognitive rep
resentations of lesbians. The cluster analysis revealed a high-level dis
tinction between positive and negative stereotypes, and lower-level
stereotypes of each high-level category. The dendrogram indicated
clear distinctions between these clusters, and the solution was stable
Geiger, Harwood, and Hummert 175
when tested with different clustering algorithms. The distinctions be
tween clusters were replicated in the multidimensional scaling analysis
where meaningful underlying dimensions were also found. Most im
portant, the emerging clusters have face validity. They reflect culturally
prevalent representations of lesbians within and outside the lesbian
community (e.g., the angry butch clearly has cultural currency as evi
denced by “Dykes on Bikes” and representations of such in the straight
community: Newton, 1984; Pisankaneva, 2002).
The number of trait groupings generated in the sorting task provides
evidence supporting the existence of substereotypes of lesbians. Many
of the participants distinguished many more types than the minimum
(three) that the sorting task required. Participants were aware they
would need to provide labels for the groups they created, hence they
were motivated to only create groups that were meaningful to them. In
this context, it is interesting to note that the number of groups generated
varied greatly within our sample (from 2 to 15). These results suggest
that the complexity in cognitive representations of lesbians varies
greatly, although they appear to be centered on some shared basic repre-
sentations. As a final note here, we would reiterate the evidence from
the MDS solution suggesting that negative stereotypes exhibit more
differentiation than positive stereotypes.
The current research adds to the existing support for considering ste-
reotypes as hierarchical and multiple in nature (Brewer et al., 1981;
Eckes, 1996; Hummert et al., 1994). Examinations of numerous groups
have now suggested this, and the knowledge has provided useful twists
on theory in this area. The extent to which social psychologists can con
tribute to changing prejudicial attitudes depends upon our ability to un
derstand the ways in which those attitudes are represented. The clear
evidence, that stereotypes are hierarchically arranged cognitive struc
tures, has already influenced theorizing about attitude change (e.g., Brewer
& Miller, 1988; Hewstone, 1996; Rothbart & John, 1985). The current
paper demonstrates that such issues are relevant to stereotypes of lesbi
ans, and hence we suggest that interventions to address homophobia
will need to incorporate knowledge about the multiple stereotypes of this
group. These findings also suggest that challenging homophobia will
need to occur in gender-specific ways. There is no reason to suggest that
cognitive representations of gay men will resemble those from this
176 JOURNAL OF HOMOSEXUALITY
study in any way, and hence attempts to adjust such cognitive represen
tations would need to proceed somewhat independently.
The results of the MDS analysis enhance our comprehension of such
issues. Understanding dimensions underlying cognitive representations
of groups is crucial in understanding perceptions of those groups. Judg
ments of positivity and strength appear fundamental in guiding catego
rization of lesbians. Hence, attempts (either individual or societal) to
change perceptions should be directed towards those dimensions. Ex
aminations of media portrayals of lesbians might productively focus on
these dimensions in evaluating portrayals, and future work on interper
sonal perception of lesbians might also capitalize on these as evaluative
parameters. In addition, those wishing to measure attitudes in a more tra
ditional fashion (e.g., semantic differential scales) might pay attention
to these as particularly important underlying dimensions (and also see
Osgood, Suci, & Tannenbaum, 1957).
The presence of positive subtypes is undoubtedly a double-edged sword
for those attempting to address issues of prejudice. At one level, these
subtypes reflect liberated and generous evaluations of out-groups. How-
ever, positive subtypes have often been implicated in cognitive strate-
gies to resist attitude change. An individual with prejudicial attitudes
towards lesbians may use a positive subtype to “fence off” exceptions
to their generally negative feelings about the group (Allport, 1954;
Rothbart & John, 1985). Considerable previous work has examined ho-
mophobia, and it is clear from recent hate crimes that prejudicial atti-
tudes towards homosexuals have serious consequences (Herek, 2000).
The understanding of cognitive representations of lesbians outlined in
this paper will further our understanding of stereotype development and
change in this context. More generally, we hope that this work and other
work in this vein contributes to an understanding of the relationship be
tween substereotypes and overall attitudes towards groups. We need to
develop clearer articulations of the ways in which representations of
“subgroups” are tied to representations of groups as a whole (Maurer,
Park, & Rothbart, 1995). Positive subtypes that are tied to an overall
group representation are of obvious value in reducing prejudice. Those
that are detached from the overall representation of the group are proba
bly harmful to intergroup cooperation.
This study has implications for gender inversion theory (Kite &
Deaux, 1987). Gender inversion theory has suggested that stereotypes
Geiger, Harwood, and Hummert 177
of lesbians include traits that are typically associated with males (and
vice versa for gay men). While there is certainly evidence in this study
that traits typically associated with males were included in several of the
subtypes of lesbians, our subjects also generated many typically femi
nine traits (e.g., caring, maternal), and a number of the substereotypes
were not at all masculine in nature (e.g., lipstick lesbian). Therefore, we
suggest that gender inversion theory addresses only one (primarily neg
ative) facet of lesbian stereotypes. As a result, invoking gender inver
sion theory may well restrict the study of homosexual stereotypes,
limiting the ability of researchers to uncover other representations. That
said, it is interesting that the strong-weak dimension is clearly important
in our study, and it would be possible to interpret this dimension as at
least somewhat gendered.
We were surprised by the sophistication of representations uncov
ered in the current study. Both the trait generation and the sorting task
revealed considerable breadth and depth in people’s representations.
This is particularly surprising given the taboo nature of homosexuality
until relatively recently. Part of the explanation may be the college con-
text of the research–homosexuals have traditionally been more “out”
and more accepted on college campuses than in many other locations.
However, the increasing discussion of gay issues in the media is an ad-
ditional factor. We would advocate investigation of cognitive represen-
tations of these groups outside of college campuses, and investigations
of the diversity of such representations in the media. Such examinations
would yield additional information about the sophistication of represen-
tations of this group in the broader culture.
A final implication of this research is in terms of issues of gender ste
reotyping. Although no direct comparisons can be made between the
two studies, it is interesting that the groupings of lesbians in the current
paper did not correspond to the groupings of women found in Eckes
(1996) study. This may suggest that although lesbians are women, they
are seen as a unique group with unique qualities and traits. More precise
delineations of the links between multiple stereotypes of gay and straight
members of both sexes would be interesting for future research, includ
ing perhaps offering a more sophisticated development of gender inver
First, future work should examine situational features that might trig
ger the various subtypes uncovered in this paper. For instance, a number
178 JOURNAL OF HOMOSEXUALITY
of the subtypes in our study included physical descriptors. To what ex
tent are these subtypes activated by physical cues? Useful research
could be done examining responses to photographic, trait-based, or
communication-based stimuli, in order to understand their relative
strength in predicting stereotype activation. Theoretical perspectives
examining such activation cues have been extremely productive in other
areas (e.g., the activation of various age stereotypes: Hummert, Gar
stka, & Shaner, 1997). Also, we would advocate research examining the
relative accessibility of positive and negative subtypes of lesbians among
different groups–particularly individuals high in homophobia, and gay
people themselves (Hajek & Giles, 2002; Herek, 1991; Kite & Whitely,
1996). Homophobic individuals might have less complex representa
tions of lesbians, or they might demonstrate quite sophisticated and var
ied negative stereotypes (Linville, 1982). Lesbians fitting into the
positive subcategories might be “subtyped” into a group of “excep-
tions,” rather than being treated as a meaningful subgroup of lesbians
(Richards & Hewstone, 2001).
Second, future research might include quantitative evaluations of
these subtypes. It would be useful to examine evaluative ratings of each
subtype to confirm our conceptualization of them as negative-positive
and strong-weak, as well as differences in such evaluations (e.g., be-
tween homosexual and heterosexual respondents, males and females,
etc.). In addition, we might examine perceptions of the origins of these
subtypes. Some subtypes may be grounded in particular exemplars
(e.g., Ellen DeGeneres, a friend, a family member), while others are
grounded in more abstract prototypes (Smith & Zarate, 1990).
Ultimately, the development of this research should address contact
and communication between lesbians and heterosexuals (Herek & Capi
tanio, 1996). Social identity (SIT) theory has interesting insights to offer
on such processes (Tajfel & Turner, 1986; Hajek, Abrams, & Murachver,
2005). For example, SIT suggests that group members seek to establish
positive distinctiveness from out-groups. From this perspective, positive
lesbian subtypes (e.g., lipstick lesbians) might be particularly threatening
to a heterosexual woman, since they threaten the distinctiveness of her
group membership. On the other hand, a lipstick lesbian might be per
ceived positively as gender-confirming if the categorization is based on
sex rather than sexuality (Crisp & Hewstone, 2000; Gaertner et al.,
2000). Hence, the extent to which being a lesbian is disclosed becomes
central, and the issue of “coming out” as a fundamentally important
communication issue is raised. Lesbians who “fit” into certain of the
categories described in this paper may find “coming out” to be a rela
Geiger, Harwood, and Hummert 179
tively unproblematic event–their friends and colleagues may have al
ready categorized them. However, for individuals who present a less
good fit, “coming out” may be considerably more thorny. As a result,
the contact between homosexuals and heterosexuals may be more or
less intergroup in nature, with consequences for attitude changes that are
becoming better documented in the literature (e.g., Richards & Hew
The current study has demonstrated that stereotypes of lesbians are
well developed in both breadth and depth. These results support the so
cial psychological position that stereotypes are hierarchical and multi
ple in nature. In addition, this paper challenges and expands previous
research grounded in a homophobia perspective. We suggest that ho
mophobia towards lesbians may indeed be widespread, however it must
be understood in the context of complex cognitive representations that
include positive elements. Such a recognition will provide more sophis-
ticated answers to questions concerning how to address homophobia.
Aldenderfer, M. S., & Blashfield, R. K. (1985). Cluster analysis. Beverly Hills, CA:
Allport, G. W. (1954). The nature of prejudice. Reading, MA: Addison-Wesley.
Ashmore, R. D., & Del Boca, F. K. (1981). Conceptual approaches to stereotypes and
stereotyping. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and in-
tergroup behavior (pp. 1-35). Hillsdale, NJ: Erlbaum.
Ashmore, R. D., Del Boca, F. K., & Beebe, M. (2002). “Alkie,’ ”Frat Brother,’ and
“Jock’: Perceived types of college students and stereotypes about drinking. Journal
of Applied Social Psychology, 32, 885-907.
Brewer, M. B., Dull, V., & Lui, L. (1981). Perceptions of the elderly: Stereotypes as
prototypes. Journal of Personality and Social Psychology, 41, 656-670.
Brewer, M. B., & Miller, N. (1988). Contact and cooperation: When do they work?
In P. Katz & D. Taylor (Eds.), Eliminating racism: Means and controversies
(pp. 315-326). New York: Plenum Press.
Crisp, R. J., & Hewstone, M. (2000). Multiple categorization and social identity. In
D. Capozza & R. Brown (Eds.), Social identity processes: Trends in theory and re
search (pp. 149-166). London: Sage Publications.
Eckes, T. (1994). Explorations in gender cognition: Content and structure of female
and male subtypes. Social Cognition, 12, 37-60.
Eckes, T. (1996). Linking female and male subtypes to situations: A range-of-situa
tion-fit effect. Sex Roles, 35, 401-426.
Edwards, G. (1992). The structure and content of the male gender role stereotype: An
exploration of subtypes. Sex Roles, 27, 533-551.
Everitt, B. S. (1993). Cluster analysis (3rd ed.). New York: Edward Arnold.
180 JOURNAL OF HOMOSEXUALITY
Gaertner, S. L., Dovidio, J. F., Nier, J. A., Banker, B. S., Ward, C. M., Houlette, M., &
Loux, S. (2000). The common ingroup identity model for reducing intergroup bias:
Progress and challenges. In D. Capozza & R. Brown (Eds.), Social identity pro
cesses: Trends in theory and research (pp. 133-148). London: Sage Publications.
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1992). Multivariate data
analysis. New York: Macmillan.
Hajek, C., Abrams, J., & Murachver, T. (2005). An intergroup analysis of sexual and
gender identity. In J. Harwood & H. Giles (Eds.), Intergroup communication: Mul
tiple perspectives (pp. 43-64). New York: Peter Lang.
Hajek, C., & Giles, H. (2002). The old man out: An intergroup analysis of inter
generational communication among gay men. Journal of Communication, 52(4),
Hamilton, D. L., & Trolier, T. K. (1986). Stereotypes and stereotyping: An overview of
the cognitive approach. In J. G. Dovidio and S. L. Gaertner (Eds.), Prejudice, dis
crimination, and racism (pp. 127-163). Orlando, FL: Academic Press.
Herek, G. M. (1988). Heterosexuals’ attitudes toward lesbians and gay men: Correlates
and gender differences. The Journal of Sex Research, 25, 451-477.
Herek, G. M. (1991). Stigma, prejudice, and violence against lesbians and gay men. In
J. Gonsiorek & J. Weinrich (Eds.), Homosexuality: Research implications for pub-
lic policy (pp. 60-80). Newbury Park, CA: Sage.
Herek, G. M. (2000). The psychology of sexual prejudice. Current Directions in Psy-
chological Science, 9, 19-22.
Herek, G. M., & Capitanio, J. P. (1996). “Some of my best friends”: Intergroup contact,
concealable stigma, and heterosexuals’ attitudes toward gay men and lesbians. Per-
sonality and Social Psychology Bulletin, 22, 412-424.
Hewstone, M. (1996). Contact and categorization: Social psychological interventions
to change intergroup relations. In C. N. Macrae, C. Stangor, & M. Hewstone (Eds.),
Stereotypes and stereotyping (pp. 323-368). New York: The Guilford Press.
Hewstone, M., & Brown, R (1986). Contact is not enough: An intergroup perspective
on the ‘contact hypothesis.’ In M. Hewstone & R. Brown (Eds.) Contact and con-
flict in intergroup encounters (pp. 1-44). Cambridge, MA: Basil Blackwell.
Hummert, M. L. (1990). Multiple stereotypes of elderly and young adults: A compari
son of structure and evaluation. Psychology and Aging, 5, 182-193.
Hummert, M. L., Garstka, T. A., Ryan, E. B., & Bonnesen, J. L. (2004). The role of age
stereotypes in interpersonal communication. In J. F. Nussbaum and J. Coupland
(Eds.), Handbook of communication and aging research (pp. 91-121). Mahwah,
NJ: Lawrence Erlbaum.
Hummert, M., Garstka, T. A., & Shaner, J. L. (1997). Stereotyping of older adults: The
role of target facial cues and perceiver characteristics. Psychology & Aging, 12,
Hummert, M. L., Garstka, T. A., Shaner, J. L., & Strahm, S. (1994). Stereotypes of the
elderly held by the young, middle aged, and elderly adults. Journal of Gerontology:
Psychological Sciences, 49, 240-249.
Kite, M. E., & Deaux, K. (1987). Gender belief systems: Homosexuality and the im
plicit inversion theory. Psychology of Women Quarterly, 11, 83-96.
Geiger, Harwood, and Hummert 181
Kite, M. E., & Johnson, B. T. (1988). Attitudes toward older and younger adults: A
meta analysis. Psychology and Aging, 3, 233-244.
Kite, M. E., & Whitley, B. E. (1996). Sex differences in attitudes toward homosexual
persons, behaviors, and civil rights: A meta-analysis. Personality and Social Psy
chology Bulletin, 22, 336-353.
Kruskal, J. B., & Wish, M. (1978). Multidimensional scaling. Newbury Park, CA:
Linville, P. W. (1982). The complexity-extremity effect and age-based stereotyping.
Journal of Personality and Social Psychology, 42, 193-211.
Maurer, K. L., Park, B., & Rothbart, M. (1995). Subtyping versus subgrouping pro
cesses in stereotype representation. Journal of Personality and Social Psychology,
Mowbray, C. T., Bybee, D., & Cohen, E. (1993). Describing the homeless mentally ill:
Cluster analysis results. American Journal of Community Psychology, 21, 67-93.
Newton, E. (1984). The mythic mannish lesbian: Radclyffe Hall and the new woman.
Signs, 9, 557-575.
Osgood, C. E., Suci, G., & Tannenbaum, P. (1957). The measurement of meaning. Chi-
cago: University of Illinois Press.
Pisankaneva, M. (2002). Reflections on butch-femme and the emerging lesbian com-
munity in Bulgaria. Journal of Lesbian Studies, 6(2), 135-144.
Richards, Z., & Hewstone, M. (2001). Subtyping and subgrouping: Processes for the
prevention and promotion of stereotype change. Personality and Social Psychology
Review, 5, 52-73.
Rothbart, M., & John, O. P. (1985). Social categorization and behavioral episodes: A
cognitive analysis of the effects of intergroup contact. Journal of Social Issues, 41,
Schmidt, D. F., & Boland, S. M. (1986). The structure of impressions of older adults:
Evidence for multiple stereotypes. Psychology and Aging, 1, 255-260.
Schwanberg, S. L. (1996). Health care professionals’ attitudes toward lesbian women
and gay men.; Journal of Homosexuality, 31(3), 71-83.
Six, B., & Eckes, T. (1991). A closer look at the complex structure of gender stereo-
types. Sex Roles, 24, 57-71.
Smith, E. R., & Zarate, M. A. (1990). Exemplar and prototype use in social categoriza
tion. Social Cognition, 8, 243-262.
Taylor, A. (1983). Conception of masculinity and femininity as a basis for stereotypes
of male and female homosexuals. Journal of Homosexuality, 9, 37-53.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In
S. Worschel & W. G. Austin (Eds.), The social psychology of intergroup relations
(2nd ed.) (pp. 7-24). Chicago: Nelson-Hall.
United Methodist News Service (1999, August 10). Committee hear Dells appeal; de
cision expected in 60 days. 1-4.
182 JOURNAL OF HOMOSEXUALITY