From Appealing to Appalling: Disenchantment with a Romantic Partner

Abstract

This study examines the tendency for thereto be similarities between the qualities that initially attract individuals to romantic partners and those they later dislike, that is, "fatal attractions." Approximately 44 percent of the individuals in this sample of 125 dating persons experience fatal attractions. Individual cases illustrate opposing themes, such as "nice to passive" and "strong to stubborn." One-third (33.7%) of the respondents themselves identify similarities between attracting and disliked partner characteristics. This disenchantment occurs in ongoing, as well as previous, relationships, suggesting that it is not simply sour grapes but is associated with the dissipation of infatuation. Dissimilar or extreme qualities in a partner are significantly more likely to become disliked.

Full text

Sage Publications, Inc.
is collaborating with JSTOR to digitize, preserve and extend access to
Sociological Perspectives.
http://www.jstor.org
Sage Publications, Inc.
FROM APPEALING TO APPALLING: DISENCHANTMENT WITH A ROMANTIC PARTNER
Author(s): Diane H. Felmlee
Source:
Sociological Perspectives,
Vol. 44, No. 3 (Fall 2001), pp. 263-280
Published by: Sage Publications, Inc.
Stable URL: http://www.jstor.org/stable/10.1525/sop.2001.44.3.263
Accessed: 14-10-2015 18:07 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

Sociological Perspectives, Volume 44, Number 3, pages 263–280.
Copyright © 2001 by Pacic Sociological Association. All rights reserved.
Send requests for permission to reprint to: Rights and Permissions, University of California Press,
Journals Division, 2000 Center St., Ste. 303, Berkeley, CA 94704-1223.
ISSN: 0731-1214; online ISSN: 1533-8673
FROM APPEALING TO APPALLING:
DISENCHANTMENT WITH A
ROMANTIC PARTNER
DIANE H. FELMLEE*
University of California, Davis
ABSTRACT : This study examines the tendency for there to be similarities
between the qualities that initially attract indiv iduals to romantic partners
and those they later dislike, that is, “fatal attractions.” Approximately 44
percent of the individuals in this sample of 125 dating persons experienc e
fatal attractions. Individual cases illustrate opposing themes, such as “nice
to passive ” and “strong to stubborn.” One-third (33.7% ) of the respondents
themselves identify similarities between attracting and disliked partner
characteristics. This disenchantment occurs in ongoing, as well as previ-
ous, relationships, suggesting that it is not simply sour grapes but is asso-
ciated with the dissipation of infatuation. Dissimilar or extreme qualities
in a partner are signicantly more likely to become disliked.
Two people are drawn to each other romantically and become involved in a rela-
tionship. Over time they may discover annoying or disturbing aspects of the
other ’s behavior or personality. The main thesis of this research is that there is a
link between these seemingly disparate processes of romantic attraction and dis-
enchantment. Like a moth to a ame, people can be drawn to the very aspects of
another person that they eventually nd troublesome. “Fatal attraction” is one
term for this type of disenchantment, where “fatal” is dened as “prophetic” or
“foretelling a sequence” rather than deadly; this sequence begins with attraction
to a partner quality and ends in disillusionment with that quality.
Fatal attractions occur in both terminated dating relationships (Felmlee 1995)
and marriages (Pines 1997). Examples include a woman who was attracted to a
“relaxed” man whom she eventually found to be “constantly late” and a man
who found a woman’s “shyness and timidity” initially appealing but later
thought she was too “insecure” (Felmlee 1998a).
Previous research on fatal attractions has several limitations that the current
study is designed to address. To date, fatal attractions have been examined among
* Direct all correspondence to: Diane H. Felmlee, Department of Sociology, University of California, Davis, Davis, CA
95616; e -mail: dhfelmlee@ucdavis.edu.
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

264 SO CIOLOG ICAL PERSPECTIVES Volume 44, Numbe r 3, 2001
only two samples of individuals, and the rst purpose of this research is to assess
the incidence of fatal attractions in a new data set and to provide illustrations of
common themes in such attractions with these new data. In addition, earlier
research on this topic was limited to terminated relationships (Felmlee 1995).
Fatal attractions also may occur in ongoing relationships, however, and determin-
ing whether this is the case could help in examining the plausibility of various
explanations for this phenomenon. Thus the second purpose of this research is to
examine whether partner disillusionment appears in current relationships or
whether it occurs only in retrospective accounts of terminated liaisons. Further-
more, in previous work (Felmlee 1995) research coders, rather than respondents,
inferred the presence of a fatal attraction (i.e., similarity between attracting and
disliked traits in a partner). It is not clear whether respondents themselves also
see such similarities or whether fatal attractions are only in the eye of the researcher.
The third purpose of the present study, therefore, is to examine respondents’ own
opinions as to whether there are similarities between the qualities that once
attracted them to their partners and those they later disliked (i.e., a fatal attrac-
tion). Finally, relatively little research to date has attempted to identify what types
of attractions are likely to be “fatal.” The fourth purpose, then, is to analyze such
tendencies in a multivariate, logistic regression analysis.
FATAL ATTRACTION
In previous empirical work on fatal attraction, individuals recalled their most
recent romantic relationships that had ended and described qualities that initially
attracted them to their former partners and the characteristics they later disliked
(Felmlee 1995). Three of the most common types of attractors were physical (e.g.,
attractive, eyes), fun (e.g., good sense of humor, fun), and caring (e.g., thoughtful,
nice). Qualities they did not like, on the other hand, were most frequent in the cat-
egories Selsh (e.g., insensitive), Insecure (e.g., possessive, insecure), and Unde-
pendable (e.g., dishonest, irresponsible). Fatal attractions occurred in approxi-
mately 29.2 percent of the 301 terminated romantic relationships in the study.
Fatal attractions were found to be signicantly more likely when individuals were
attracted to characteristics in a partner that were extreme in nature, in the cate-
gory Different (e.g., “differences”), or in the category Unique (e.g., “one-of-a-kind”),
but less likely when attracting characteristics were in the category Similar (e.g.,
“common values”) (Felmlee 1998a). The likelihood of fatal attractions also was
found to be greater when the individuals themselves (rather than their partners)
initiated a breakup. Finally, in previous research (Felmlee 1998b) individuals were
signicantly less likely to experience disenchantment with the physical, as opposed
to the personality, characteristics of a partner.
Pines (1997) identied fatal attractions among both members of married cou-
ples in clinical samples from the United States and Israel, demonstrating that this
phenomenon is not limited to United States dating relationships. According to
Pines, there were connections between the problem that brought couples in her
sample to therapy and what attracted them to each other when they rst met. For
example, one husband in Pines’s sample was distressed with his wife’s “unfair
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

Disenchantment with a Romantic Partner 265
criticism” and threatening “outbursts.” He said that what most attracted him to
her when they rst met, however, were her “powerful personality,” “directness,”
“cynicism,” and “sharp intelligence, “ qualities that are closely related to those he
later disliked. The wife also appeared to be disturbed by aspects of her husband’s
personality that initially appealed to her (i.e., his easy going nature).
Potential Explanations
There are several possible explanations for fatal attractions. One concerns the
notion that people’s virtues and vices are one and the same, as suggested by clini-
cal psychologists (Goldberg 1993; Jung 1973) and the popular literature (e.g., Purdum
1996). In her discussion of the dark side of love, for instance, Goldberg (1993:8)
argues, “Protectiveness can easily turn into possessiveness; concern into control;
interest into obsession.” If people are romantically drawn to the strengths of
another and strengths have corresponding weaknesses, then it is not surprising
that individuals come to dislike aspects of the qualities that they initially found
appealing; that is, they dislike the vices associated with their loved one’s virtues.
A second explanation for this process of partner disenchantment pertains to
opposing relational forces. Sim mel (1955) argues that small groups encounter
strains between the forces of anomie and solidarity. People desire the solidarity
that results from group membership and at the same time want to maintain their
autonomy. Similarly, dialectical theorists (e.g., Altman, Vinsel, and Brown 1981;
Baxter and Montgomery 1996) maintain that couples face pairs of contradictory
forces, such as autonomy and connection, openness and closedness, and novelty
and predictability. Fatal attractions may occur because individuals are drawn to
characteristics in another that exemplify one dimension of these opposing forces
(e.g., novelty), but then they nd their relationship lacking in the corresponding
dimension (e.g., predictability). Examples of opposing forces that appear in fatal
attractions are fun versus seriousness, connection versus autonomy, and strength
versus vulnerability.
Finally, the social psychological literature on motivated cognition in romantic
relationships (e.g., Miller 1997; Murray, Holmes, and Grifn 1996) is useful for
understanding fatal attractions. It suggests that individuals need to sustain
condence in the belief that they are in the right relationship with the right person
and that they will use various cognitive tactics to maintain satisfaction and com-
mitment. Processes of motivated construal or relationship enhancement are
engaged that help individuals to dispel doubt and sustain condence in their
romantic partners. For example, individuals tend to believe that their relation-
ships are superior to those of other people (Van Lange and Rusbult 1995) and
satised lovers are relatively inattentive to desirable alternatives to their relation-
ship (Miller 1997). Individuals also tend to idealize their partners and see virtues
in their partners’ faults (Murray and Holmes 1993), and this idealization is associ-
ated with greater relationship satisfaction (Murray, Holmes, and Grifn 1996).
However, presumably once a relationship has ended, the motivation for the mask-
ing of weaknesses in a romantic partner is removed. The faults of a partner that
were once transformed into virtues are now seen as vices.
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

266 SO CIOLOG ICAL PERSPECTIVES Volume 44, Number 3, 2001
Predictors of Fatal Attractions: Hypotheses
Several factors are expected to be related to fatal attractions. The rst is “differ-
ences.” Differences in a romantic companion may be attractive initially, either
because involvement with a dissimilar other makes a person feel special (Snyder
and Fromkin 1980) or because it increases the potential for self-expansion (Aron
and Aron 1986). Nevertheless, several arguments suggest that differences between
members of couples can be problematic. Most research nds that similarity, not
dissimilarity, is a powerful basis of attraction (e.g., Byrne 1971; Smith, Byrne, and
Fielding 1995; Surra 1991). In addition, according to balance theory (Newcomb
1961), differences between individuals are stressful cognitively. Dissimilarity also
is likely to be a negative reinforcer, because similarity acts as a positive reinforcer
by validating one’s own perspectives (Byrne and Clore 1970). Finally, differences
are often cited as a reason for the demise of a relationship (Cleek and Pearson
1985; Hill, Rubin, and Peplau 1976).
Because of the problems differences pose for couples, romantic attractions to
differences in a partner are expected to be particularly susceptible to fatal attrac-
tions. An attraction to another can be “different” in two ways: (1) different from
self (i.e., dissimilarity) or (2) different from average (i.e., unusual or extreme).
Both are hypothesized to be prone to disillusionment.
Previous research nds some support for these hypotheses but only in a rudi-
mentary manner. For example, individuals who stated that the quality that
attracted them to their partners was “dissimilarity” (or a related trait such as
“different” or “unique”) were found to be more likely than those with other types of
attractions to later dislike that particular attracting characteristic (Felmlee 1998a).
This type of attractor was mentioned only occasionally (in 5.9% of the cases),
however. In the data collected here, respondents are asked to assess similarities
and dissimilarities between themselves and their partners for all the characteristics
that initially attracted them . A variable measuring dissimilarity is constructed
from this information, and it is used to conduct a more precise examination of
the hypothesis that dissimilar attractors are overrepresented in fatal attractions.
Earlier research nds support, too, for the hypothesis that attractions to
extreme characteristics of another (i.e., qualities that are different from average)
are signicantly more likely than attractions to more moderate characteristics to
be “fatal” (Felmlee 1988a). Extreme characteristics are those that are described by
the respondent with intense adjectives such as “very” or “extremely.” This
hypothesis also will be investigated here with the new data.
Whether a relationship is intact or terminated is hypothesized to relate to fatal
attractions, too. Relationships that have ended are those that may have been espe-
cially problematic and thus most liable to result in disenchantment. Individuals
also may be more likely to negatively recast a former partner’s positive qualities
than those of a current partner. Therefore, fatal attractions are expected to be more
prevalent in previous, as compared to current, relationships.
Finally, whether or not an attraction is physical is expected to be relevant to the
disenchantment process. Most physical characteristics are less likely th an person-
ality characteristics to have an obvious downside, and many are relatively
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

Disenchantment with a Romantic Partner 267
difcult to reinterpret over time. What is the vice associated with the virtue of
possessing “beautiful eyes” or a “nice smile,” for instance? There are exceptions
to this rule, of course. For example, physical attractiveness has negative as well as
positive connotations (e.g., snobbish, vain) (Freeman 1985), and sexiness may
have its corresponding downside, especially for female partners, according to
some respondents (i.e., “slut”). Nevertheless, unlike personality characteristics,
many of the physical attracting qualities mentioned by respondents in this sample
lack a clear dark side (e.g., eyes, smile, hair). Thus a nal hypothesis is that fatal
attractions are less likely when an individual is drawn to the physical, as opposed
to the personality, qualities of another.
METHOD
Sample
To examine instances of fatal attractions, I collected data from 125 undergradu-
ate students enrolled in a large introductory course that fullled a General Educa-
tion requirement for students across the university. The class was relatively
diverse ethnically, with approximately 44.7 percent whites, 33.6 percent Asian
Americans, 11.2 percent Hispanics/Latinos, 4.6 African Americans/blacks, and
5.9 percent other ethnicities. Two individuals in the sample reported on a homo-
sexual relationship. The average relationship length was 1.5 years, and the average
age of participants in the class was 19.4 years. An advantage of this nonmarital
sample is that it is easier to focus on the social psychological process of partner
disenchantment than it would be in a sample in which the contingencies of mar-
riage, parenthood, and economics play a major role.
Procedure
Participants were given a questionnaire that began: “Think of your current
romantic relationship. If you are not currently in a relationship, then think of the
most recent, serious romantic relationship you have had.” Slightly more than half
the respondents (56.7%) reported on a current relationship, and the rest (43.3%) dis-
cussed a previous one. Participants were asked to respond to a series of open-
ended questions, two of which were used to determine whether a fatal attraction
had occurred. The rst was: “Describe the specic qualities that rst attracted you to
that individual.” The second question, which was later in the questionnaire, read:
“What are (were) the qualities about that individual that you nd least attractive?”
To identify fatal attractions, two independent coders were provided with a
denition and several illustrations of fatal attraction. A fatal attraction was
dened as occurring “when a quality listed as least attractive was similar to (e.g.,
a synonym), or a negative interpretation of, a quality reported as initially attract-
ing.” Sample cases of fatal attractions, taken from previous research (Felmlee
1995), were discussed in detail with the coders. The sample cases included the fol-
lowing pairs of attracting and disliked qualities: shy/too shy, funny/constant silli-
ness, strange/too different, nurturing/smothering, older/age, condent/arrogant.
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

268 SO CIOLOG ICAL PERSPECTIVES Volume 44, Number 3, 2001
Each of the 498 attracting qualities and the corresponding 439 disliked qualities
were read by the coders to determine the occurrence of a fatal attraction. An inter-
coder reliability of .77 (kappa value) was obtained for the total sample.
To investigate respondents’ own assessments of fatal attractions, they were
asked the following open-ended question: “Are the qualities that you nd least
attractive in your partner very different from, or very similar to, those that rst
attracted you to that person?” The word “similar” was placed after “different” to
provide a conservative test of the hypothesis. The intercoder reliability for catego-
rization of the responses to this question (i.e., similar/different) was .93. The nal
question for those in an ongoing relationship was: “How long do you think your
current relationship will last?”
Measures
Fatal Attracting Characteristic
The dependent variable in the regressions is a binary variable. If an attracting
quality is “fatal” (i.e., subsequently disliked), it is coded 1. If it is “not fatal” (i.e.,
not disliked), it is coded as 0.
The independent variables are as follows.
Dissimilar
This variable was derived from answers to two open-ended questions: “In what
ways are the qualities that rst attracted you to your partner similar to your own
qualities?” And, “In what ways are the qualities that rst attracted you to your
partner different from your own qualities?” Answers to these questions were used
to create three categories for this variable: (1) an attracting partner quality was
similar to that of the respondent; (2) it was similar in some respects and different
in others; (3) it w as different. The intercoder reliability was .86.
Extrem e
An attracting quality is categorized 1 (8.4% of the cases) if the quality is intense
or exaggerated (e.g., beautiful, wild) or if one of the following adjectives is used
as a modier: very, extremely, totally, or great. Otherwise it is 0. The intercoder
reliability was .85 (kappa value).
Physical
This variable is 1 if an attracting quality is a physical or a sexual characteristic;
otherwise it is 0. The intercoder reliability was .92 (kappa value).
Past Relationship
This variable is coded 1 for relationships that have already ended; ongoing rela-
tionships are coded 0.
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

Disenchantment with a Romantic Partner 269
Female
This is coded 1 if the respondent is female and 0 if the respondent is male.
Months Together
This is the total number of months that the relationship lasted.
Percent Fatal Qualities
This variable measures the percentage of “fatal” attracting qualities listed by
each respondent. It is the number of fatal traits for each respondent, divided by the
total number of attracting traits (excluding the current case). This variable is
included in the model to control for the tendency of individuals to have multiple
fatal attracting qualities, which could lead to correlated errors.
RESULTS
Occurrence of Fatal Attractions
Fatal attractions occur in a substantial proportion (44%) of the sample. This
means that at least one of the qualities listed as “least attractive” is directly
related to one or more of those reported as initially attracting for well over one-
third of the participants. In addition, fatal attractions are not limited to individ-
uals describing terminated relationships. Approximately 39.1 percent of the
participants in ongoing, current relationships report at least one fatal attracting
quality, although the fatal attraction rate is about one and a half tim es higher
for those reporting on a re lationship that has broken up (58.8%) (x2 5 4.6 , p 5
.03). The ve most common general themes in these fatal attractions and the
percentage of attracting qualities in each are discussed below. Table 1 presen ts
ten such themes.
Nice to Passive (17.6%)
One of the most common fatal attractions occurs when a person mentions the
nice qualities of a partner as appealing but then dislikes aspects of that niceness.
One woman, for example, reports that she was attracted to her former boyfriend
because he was “nice” and “considerate.” In describing what she found least
attractive about him, however, she says: “[I] couldn’t tell what he was feeling. He
didn’t say because he didn’t want to upset me.” It appears that perhaps he was
too nice and considerate.
Strong to Stubborn (17.6%)
Participants become disenchanted with the initially attractive qualities of a
partner that are indicative of a strong nature. For example, one man was attracted
to a woman whom he describes as “headstrong and independent.” Nevertheless,
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

270 SO CIOLOG ICAL PERSPECTIVES Volume 44, Number 3, 2001
the fact that she “never tells [him] what she is upset about,” is what he nds least
attractive about her.
Funny to Flaky (13.5%)
This theme involves cases in which sense of humor, or ability to have fun, is an
attracting quality in a partner. In such cases the partner’s weakness is that he is
“aky,” according to one woman, or that she is “immature,” in the words of a
man. In another case, a woman says that sense of humor attracted her to her boy-
friend but then reports that he “doesn’t always take other people’s feelings seri-
ously (jokes around too much).”
TABLE 1
Illustrations of Fatal Attractions by General Type (Percentage of Fatal Attractions):
Attracting Partner’s Quality—Disliked Partner’s Quality.
Nice to Passive (17.6%)
“accommodating”—“changes opinion to t environment”
“sensitive”—“insecure”
Strong to Stubborn (17.6%)
“aggressor”—“stubborn, strong-willed”
“independent”—“independence”
Funny to Flaky (13.5%)
“doing fun things”—“ake”
“sense of humor”—“jokes too much”
Outgoing to Over the Top (10.8%)
“outgoing”—“so active”
“friendly”—“too friendly”
Caring to Clinging (9.4%)
“caring”—“jealous; insecure”
“attentiveness”—“controlling”
Quiet to Closed (9.4%)
“quiet”—“would get quiet and shut me out”
“shyness”—“too shy”
Exciting to Scary (8.1%)
“exciting”—“short temper; unfaithful”
“adventurous”—“lack of responsibility”
Physically Attractive to High Maintenance (5.4%)
“looks”—“high maintenance”
“sensuous and amorous”—“unfaithful”
Laid-back to Lazy (4.1%)
“easy-going”—“forgetfulness”
“easygoing”—“cares too little”
Successful to Workaholic (4.1%)
“motivated”—“too busy”
“hard-working”—“too busy”
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

Disenchantment with a Romantic Partner 271
Outgoing to Over the Top (10.8%)
This is an instance in which an individual becomes disillusioned with his or her
partner’s outgoing nature. In one case a man was interested in a woman because
of her “friendliness” but now complains that she “talks too much (too friendly).”
Caring to Clinging (9.5%)
Here one woman was attracted to a man who was “very attentive” and persis-
tent but disliked that he “tries to be controlling.” Another woman described a
former partner as “caring,” “sensitive,” and someone who listened to her. Yet she
did not like that he also got jealous very easily and “he hated it when [she]
wanted to spend time with other friends.”
Individuals’ Assessments of Disenchantment
We still do not know individuals’ own opinions as to whether the qualities they
dislike in a companion are similar to those they found initially appealing. To
investigate this issue, respondents were asked whether the qualities that they
least liked about their partners were “very different from or very similar to” those
that initially attracted them.
A sizable minority of those answering the question report that there are similar-
ities between disliked and liked partner characteristics (approximately one-third,
or 33.7%). Respondents and external coders agree signicantly in this assessment
of similarity/dissimilarity; cases in which respondents mention similarity in
response to this question are signicantly more likely to be coded independently
as a fatal attraction (i.e., similarity) than those where they report differences (x2 5
8.47, p 5 .004).1 Furthermore, the content of these responses is indicative of fatal
attractions. In one instance, for example, a woman was attracted to a persisten t
and attentive man who was also “too controlling.” When asked to compare these
sides of his personality, she replied: “They were very similar. When I rst met
him, I was younger and attered by his atten tion, so at the time I didn’t realize
that he was trying to control me.” In another instance, a man says that the pros
and cons of his girlfriend were “very similar.” He was drawn to his girlfriend’s
openness but did not like that “[s]he acted that way with everyone.”
In some instances (10.6% of the respondents), an attraction does not appear to be
“fatal” to outside coders, but the participant’s response to the comparison question
suggests that it may have been. For example, a man reports that he was drawn to a
woman because of her personality, looks, and independence. He dislikes, however,
that “[s]he is shy and not able to express her ideas,” qualities that do not seem to be
the negative side of the attractors. In comparing attracting qualities with least liked
ones, however, the man says: “They are sort of similar because she is a quiet, nice
girl.” Apparently, the aspects of her personality that attracted him were that she
was quiet and nice. This example illustrates one of the main reasons for disagree-
ments between coders and respondents: respondents use a vague term, such as
“personality,” to describe speci c traits that they liked or did not like in a partner.
The majority of participants who answer this question, however, state that the
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

272 SO CIOLOG ICAL PERSPECTIVES Volume 44, Number 3, 2001
qualities they dislike in a partner are very different, or somewhat different, from
those they liked initially (66.3%). For example, one participant replied: “Different,
that’s why those qualities were less attractive.” “Yes, opposite,” says another.
In some cases (24.5%), respondents report that the attracting and disliked
aspects of their partner are different, but the external coders disagree with the
respondents’ assessment and instead see some similarities. In one such instance, a
man was drawn to a woman because “[s]he enjoyed talking to her plant and her
pet lizard” and she was “mentally disturbed.” What he least liked about her, how-
ever, was that she was “depressed.” He said that her negative qualities were dif-
ferent from those that rst attracted him. Yet to an outside observer it seems plau-
sible that being depressed is a symptom of mental disturbance, a quality he
claimed to nd initially appealing.2
Finally, these ndings are probably a conservative measure of the degree to
which respondents see similarities between attractive and disliked partner quali-
ties. The notion that there could be similarities between these qualities (i.e., fatal
attractions) is counterintuitive. A number of participants point out that common
sense says that appealing and disliked traits of a partner are “obviously” or “of
course” very different. Others appear to view the question of whether there are
differences or similarities between attracting and disliked partner qualities as
strange (22.7% of all responses). Illustrations of similarities between these quali-
ties may be necessary for many individuals to even entertain the possibility in
their own situations.
The Disenchantment Process
In their open-ended responses to the question asking them to compare liked
and disliked partner qualities, several respondents described how it was that they
came to dislike certain aspects of their partners. These responses also speak to the
process involved in becoming disillusioned with a romantic partner, whether that
disillusionment is with characteristics that are different from or similar to those
that were originally deemed compelling. Four possible scenarios were reported,
and I label these: (1) Time will tell, (2) Rose-colored glasses, (3) People change,
and (4) Sour grapes. Note, however, that only a minority of respondents offered
this additional, unasked for information (28.8%), and therefore the ndings are
only suggestive. Furthermore, there are not enough of such responses (4%) from
those who saw similarities between attracting and disliked partner qualities (i.e.,
fatal attractions) to enable distinctions to be made between the disenchantment
process for “fatal” and “nonfatal” attractions.
The “Time will tell” scenario implies that the downsides of a partner’s virtues
are not apparent from the beginning of a relationship. Vices take time to appear,
either because romantic partners hide their weaknesses at the initiation of a rela-
tionship or because situations do not occur immediately in which weaknesses
become evident. For instance, one man says: “I did not really know that these
unattractive qualities existed until we were well into our relationship.” Individu-
als may acknowledge the limitations of an intimate partner only when a state of
infatuation has waned, as suggested by the second scenario, “Rose-colored
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

Disenchantment with a Romantic Partner 273
glasses.” People are aware of the limitations of a loved one from the beginning,
according to research on infatuation (McClanahan et al. 1990; Tennov 1979), but
interpret those weaknesses in a more positive light while infatuated. As one
woman said when comparing her companion’s virtues and vices: “They are very
different. I think it’s because when you meet someone you are so overwhelmed
by their good qualities that you overlook the bad ones, but eventually time will
make those bad qualities surface and you won’t be as likely to oversee them.” In
the third scenario, “People change,” a partner develops a new liability over the
course of a relationship. One woman complains about her former boyfriend, for
example: “When I rst met him he was tone and buff, now he’s gaine[ed] weight
and is somewhat abby.” The nal scenario, “Sour grapes,” suggests that fatal
attractions occur with time because one is more prone to nd fault with one’s
romantic companion’s qualities once a relationship has ended, whether or not
these “faults” reect an actual vice. Disparaging a former partner’s strengths may
help to reduce the cognitive dissonance (Festinger 1957) that is generated during
a breakup. For example, one participant whose girlfriend would not make a com-
mitment to him comments: “I’m so mad at her that little things bother me now.”3
Predictors of Fatal Attractions
We saw earlier that there is evidence that fatal attractions can occur, both on the
basis of the researcher’s investigation of the data and from the observations of the
participants themselves. The hypotheses outlined earlier suggest that not all qual-
ities are equally vulnerable to such attractions, however. In the following analysis,
I examine these hypotheses in a multivariate analysis of the likelihood of a fatal
attraction, where the occurrence of fatal attraction is identied by coders. Because
the dependent variable in such an analysis is a binary variable (a fatal attraction;
no fatal attraction), I use logistic regression analysis. The unit of analysis is the
attracting quality or qualities mentioned by the participants for the total sample.
Bivariate correlations and descriptive statistics for the variables in the analysis are
shown in Table 2.
The ndings in Table 3 provide a good deal of support for the fatal attraction
hypotheses. For example, the perception that a partner’s quality is dissimilar, rather
than somewhat dissimilar or similar to the individual’s ow n qualities, signicantly
increases the probability of a fatal attraction. The effect is substantial; an attracting
characteristic that the respondent views as dissimilar from his or her own is more
than 1.7 times (exp(B) 5 1.77)4 as likely to be later disliked as a quality seen as
somewhat dissimilar and more than 3.38 times as likely as one seen as similar.
Attracting qualities that are described in an extreme manner also are particularly
prevalent in fatal attractions, as hypothesized. Once again the effect is large.
Extreme attractors are more than three times as likely to be fatal as their more mod-
erate counterparts (exp(B) 5 3.09). In addition, physical attracting characteristics
are less likely than personality characteristics to be later disliked. The chances of a
fatal attraction are reduced by 96 percent (exp(B) 5 .04) if an attracting quality is a
physical, rather than a personality, characteristic. Moreover, fatal attractions are
signicantly more likely in past, as opposed to current, relationships, as predicted.
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

274 SO CIOLOG ICAL PERSPECTIVES Volume 44, Number 3, 2001
Such attractions are more than two and a half times as prevalent in past relation-
ships (exp(B) 5 2.69). Finally, the effects of the control variables Months Together,
Female, and Percent Fatal Qualities do not reach statistical signicance.
In an analysis not shown here, I use a xed effect model (Kessler and Greenberg
1981) to control for the possibility of correlated errors within individuals, instead
of the control variable Percent Fatal Qualities. I include 124 dummy variables in
the model, one for each individual (minus one). The ndings are almost identical
to those in the model presented in Table 2, and the conclusions regarding the sub-
stantive variables remain the same.
DISCUSSION
Fatal attractions occur in 44 percent of the relationships examined here, a nding
that provides evidence that this is a common phenomenon in intimate relation-
TABLE 3
Coefcient and Antilog Coefcient (Odds Ratio) in a Logistic Regression of the Likelihood
of a Fatal Attraction, N 5 498 Attracting Qualities (standard errors in parentheses)
Independent Variable B Exp (B)
Dissimilar quality .57*** (.17) 1.77
Extreme quality 1.13** (.39) 3.09
Physical quality 23.17** (1.02) .04
Past relationship .99*** (.30) 2.69
Months together 2.02 (.00) .98
Female 2.24 (.29) .79
Percent fatal qualities 21.32 (.81) .27
Intercept 22.15 (.59)
x261.63***; df 5 7
*p # .05; ** p # .01; * **p # .001; two-tailed test.
TABLE 2
Bivariate Correlations among Variables and Descriptive Statistics (N 5 498)
Variables 1 2 3 4 5 6 7 8
(1) Dissimilar quality 1.00 .05 2.04 2.08 .04 2.04 2.03 .16**
(2) Extreme quality 1.00 2.04 .05 .00 2.02 .06 .18**
(3) Physical quality 1.00 .08 2.07 2.09* .02 2.18**
(4) Past relationship 1.00 2.23** .06 .25** .12**
(5) Months together 1.00 .14** 2.13** 2.09
(6) Female 1.00 2.04 2.02
(7) Percent fatal qualities 1.00 2.03
(8) Fatal quality 1.00
Means 1.31 .08 .19 .46 16.18 .66 .14 .15
SD .69 .28 .39 .50 17.54 .48 .19 .35
*p # .05; ** p # .01.
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

Disenchantment with a Romantic Partner 275
ships. The proportion of fatal attractions in this sample is higher than that
reported in a previous sample, however, where it was about 30 percent (Felmlee
1995). In the earlier study, more than twice as many participants reported attract-
ing qualities that were very vague, such as “personality,” and all of these were
coded as nonfatal. In this study, unlike the previous one, participants were asked
to be very specic in their responses, and as a result there is a higher proportion of
fatal attractions.
There are a variety of themes that arise in the fatal attractions in this sample,
indicating that many different partner characteristics are susceptible to disillu-
sionment. These themes also demonstrate that fatal attractions arise for opposite
personality characteristics. Individuals who are liked for their strength of charac-
ter, for example, often are viewed as stubborn. Those who are attractive because
of their gentleness of character, on the other hand, tend to be faulted for their pas-
siveness. In another example, individuals whose drive and motivation make them
appealing are later seen as too driven, but those who are liked for their easygoing
tendencies are viewed as lazy.
These patterns in fatal attractions also provide further evidence of opposing
social forces in dyads. Contradictions between solidarity and anomie (Simmel
1955), for instance, are apparent in the fatal attractions labeled “Caring to Cling-
ing” (9.4% of the cases). Here respondents demonstrate a need for solidarity by
being drawn to the caring elements of a partner’s personality. Their need for inde-
pendence, on the other hand is reected in their rejection of that same person’s
apparent jealousy and possessiveness, and this desire for independence could
lead to anomie (i.e., a lack of solidarity).
Contradictions between other social forces identied by dialectical theorists
(e.g., Baxter and Montgomery 1997) are also evident. In fatal attractions that
t the “Exciting to Scary” theme, for example, individuals appear to be enticed
by the novel, exciting aspects of another but nd the unpredictability disturb-
ing, thus demonstrating tensions between novelty and predictability. The “Out-
going to Over the Top” theme, on the other hand, illustrates aspects of the
openness/closedness contra diction. Individuals are drawn to the open, out-
going nature of another and yet are repulsed by that person’s tendency to be too
open and outgoing (i.e., lacking in closedness) . Additional con tradictory
themes that are not discussed directly in dialectical work are evident, too, in
these data. Examples of such themes are: “Laid-Back to Lazy,” “Successful to
Workaholic,” “Nice to Passive,” and “Physically Attractive to High Mainte-
nance.” These themes do not t clearly into the three main dialectical contradic-
tions described earlier, and they suggest possible routes of expansion to the per-
spective, such as the inclusion of an opposing theme based on motivation
(motivated vs. laid-back).
One other important nding uncovered here is that a signicant proportion of
respondents (33.7%) identify similarities between the qualities that attracted them
to their partner and those they dislike about that person. In other words, the
notion of fatal attraction is not only observed by external coders; individuals
themselves recognize this pattern of disenchantment in their relationships. That
there is a signicant level of agreement between coders and respondents concern-
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

276 SO CIOLOG ICAL PERSPECTIVES Volume 44, Number 3, 2001
ing similarity between attracting and disliked qualities of a partner provides fur-
ther evidence of the validity of this phenomenon.
The Disenchantment Process
An important avenue for future research is to examine the process of disen-
chantment in more detail. Certain responses obtained here suggest several pos-
sible disenchantment scenarios: (1) Time will tell (a partner ’s liabilities surface
with time); (2) Rose-colored glasses (infatuation produces an inated assessment
of a partner); (3) People change (people can develop weaknesses); and (4) Sour
grapes (a partner is negatively evaluated after a breakup).
Various ndings reported herein speak to these scenarios as possible explana-
tions of fatal attractions. For example, we see here that although fatal attractions
are more prevalent in past relationships, they occur in a substantial proportion of
those that are current. “Sour grapes” is not as good an interpretation of partner
disenchantment in ongoing dyads, because dissonance should be relatively low
when compared to that in dyads that have broken up. Of course, current relation-
ships with fatal attractions could be destined to end quickly, and additional infor-
mation, in which respondents are asked to predict the length of their ongoing
relationships, is used to examine this possibility. These predictions are not corre-
lated signicantly with the likelihood of a fatal attraction, however, according to
analyses not shown here. Thus current relationships with fatal attractions may not
be particularly likely to end abruptly.
Another nding that concerns theoretical explanations for fatal attractions is that
relationship length is not related signicantly to the likelihood of such an attraction.
Presumably individuals in lengthy relationships have had more opportunities to
uncover the vices associated with their loved ones’ virtues. Because the occurrence
of fatal attractions does not increase with time, however, the “Time will tell” expla-
nation of partner disenchantment is challenged. On the other hand, this explanation
cannot be ruled out entirely because relationships progress at varying speeds and
certain weaknesses in an intimate companion may be apparent more quickly than
others. Furthermore, it is difcult to draw inferences, one way or the other, regarding
the “People change” scenario. It is possible that partners alter, or are perceived of as
altering, their personality or behavior in such a way that it becomes bothersome.
Findings from this and other studies, nevertheless, do provide some support
for the “Rose-colored glasses” interpretation of fatal attractions that evolves out of
research on infatuation (e.g., Tennov 1979). For instance, we nd here that indi-
viduals in a current relationship often experience partner disenchantment, which
could occur precisely when infatuation fades and rose-colored glasses are
removed. Presumably infatuation also dissipates at differing rates for various
individuals, and thus it is not surprising that relationship duration is unrelated to
the likelihood of fatal attraction. Finally, initiators of breakups, as opposed to the
recipients, are overrepresented in fatal attractions (Felmlee 1998b), and this pro-
vides additional evidence for the rose-colored glasses thesis. Breakup initiators
are the ones most likely to possess relatively low levels of dissonance and to have
experienced an erosion of infatuation.
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

Disenchantment with a Romantic Partner 277
Predictors of Fatal Attractions
Additional ndings of interest from the multivariate analyses are that individuals
tend to dislike the qualities that initially attracted them to a partner when those
qualities are dissimilar from their own qualities, are extreme in nature, or are per-
sonality, rather than physical, characteristics. Dissimilar attractors may be partic-
ularly susceptible to disenchantment because intracouple differences, although
intriguing, often lead to disagreements. Extreme attractors, on the other hand, are
more likely than their more moderate counterparts to have a negative side. A per-
son who is extremely nice, for example, may not stand up for himself or herself, and
one who is extremely condent may be egotistical. Someone who possesses these
same qualities, but more moderately, may not be as prone to the associated vices.
Moreover, physical attractors are underrepresented in fatal attractions, proba-
bly because, unlike personality characteristics, they often lack a clear downside.
This is not to say that such cases are completely unproblematic. In one case of a
physical attraction, for example, a woman was drawn to a man because of his
“physical features—tall, good-looking, athletic” and because her “friends and
other girls thought he was attractive.” Yet she ended the relationship, because “he
cheated” on her. She goes on to say: “Then I realized all the things that attracted
me to him at rst are what caused us to break up.” In another instance, a man who
was attracted to an extremely pretty woman complains that she “doesn’t like to
read,” and another man disliked that his “physically attractive” partner was
“aky, forgetful, selsh” and “too sexual in the beginning.” In other words, part-
ners initially liked primarily for their physical appearance are later found to be
lacking in salient personality characteristics. Attractions based primarily on a per-
son’s beauty thus may be disenchanting in a broader sense.
The results from the multivariate logistic regression analyses, more generally,
have implications for existing theories of interpersonal attraction. The nding that
dissimilar attracting qualities are particularly likely to lead to disillusionment, for
instance, provides further evidence that dissimilarity, although in some cases
enticing, can be a source of difcu lty in a relationship. Opposites do attract on
occasion, as some have argued (Winch 1955), but eventually they can repel. The
results of the logistic regression analysis also speak to theories regarding the vari-
ous stages through which relationships progress (e.g., Lewis 1973; Reiss 1980).
Here we observe that what brings a couple together at the initiation of a relation-
ship can play a role in what leads to conict or even in what causes a breakup.
According to the perspective presented here, the seeds of relationship discord can
be sown in a variety of ways from the very beginning stage of a romantic pairing.
Fatal Attraction in Ongoing Dyads
The nding that fatal attractions occur in continuing relationships and that they
are not necessarily fatal to the dyad itself also has a number of interesting implica-
tions. It suggests that a person can become disenchanted with the characteristics
of a partner to whom they were initially attracted and yet not become disen-
chanted with the relationship as a whole. One man, when writing about the qual-
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

278 SO CIOLOG ICAL PERSPECTIVES Volume 44, Number 3, 2001
ities that attracted him to his wife of six and a half years, emphasizes her strong
personality: “She says what she believes no matter what.” Nevertheless, the fact
that “she can be extremely stubborn and pig-headed” is what he nds least attrac-
tive about her (i.e., “Strong to Stubborn” type of fatal attraction). In his answer to
the question regarding how long he thinks his relationship with this woman will
last, he answers: “Forever.” This case demonstrates, therefore, that it is possible to
nd the dark side of a loved one’s merits troublesome and yet remain completely
committed to that person.
How is it that this man can dislike aspects of the qualities that once attracted
him to his wife and yet remain satised with the relationship? Some answers to
this question may be found in other parts of his questionnaire. This individual
recognizes that there are connections between the positive and negative qualities
of his mate. When asked whether the characteristics that he nds least attractive
about her are very different from or very similar to those that rst attracted him,
he replies: “They are related. Her stubborn nature comes from her strong sense of
self-worth and her personal beliefs.” Being aware that his wife’s liabilities stem
from her strengths may help him to accept his wife as she is. This respondent also
acknowledges his own similar strengths and weaknesses. Perhaps having short-
comings that are similar to those of his partner also makes it easier to appreciate
his spouse. Thus being aware of the light and dark sides of one’s partner’s charac-
ter, as well as one’s own, may assist in coping with disenchantment in a continu-
ing relationship.
There is a subtle message in fatal attractions, perhaps: in romance, people
“want it all.” They desire a partner who is strong yet gentle, nice but aggressive,
and hardworking yet easygoing. It seems unlikely that the human psyche is exi-
ble enough to meet such contrasting expectations. The occurrence of disenchant-
ment in committed relationships, on the other hand, demonstrates that certain
individuals are aware that they do not have everything in a partner and at the
same time remain invested in their relationship.
This research has some limitations. The sample consists largely of highly edu-
cated, unmarried, although ethnically diverse, young people, and additional
research is needed based on more representative, married, and older samples.
College relationships are apt to have relatively high rates of turnover (the average
relationship duration for this sample is 1.5 years), at least when compared with a
married sample, and disenchantment may be more common in such pairings than
in those with longer commitments. Another possible problem is that fatal attrac-
tions are inferred from open-ended responses, whereas the use of established
scales to identify such attractions could improve standardization. In addition,
information is obtained from only one member of a couple, and in-depth inter-
view data from both members would be valuable in gaining insight into the in-
ternal dynamics of this phenomenon.
These caveats notw ithstanding, this stud y demonstrates that disenchant-
ment with a partner ’s virtues is relatively common in t his sample and that it is
often recognized by individuals themselves. Moreover, disenchantment of this
type does not automatically result in a breakup. The paradox inherent in the
concept of fatal attraction is that the choice one makes in a partner will have a
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

Disenchantment with a Romantic Partner 279
shadow side and that one’s partner ’s choice of mate (i.e., oneself) is apt to have
a downside as well. Perhaps an awareness of this tendency aids in accepting
both a partner ’s and one’s own shortcomings, thus allowing a relationship to
ourish.
Acknowledgments: I would like to thank Scott Gartner, Susan Sprecher, Larry
Cohen, and the anonymous reviewers for their helpful comments on an earlier
draft of this article. Susan Miller, Anna Muraco, Sue Schroeder, and Heather
Kohler provided invaluable help with data and table preparation and coding.
NOTES
1. Both coders and respondents report similarities between attracting and disliked traits
for 23.4% of the respondents. Both report differences in 41.5% of the cases. Coders saw
similarities and respondents saw differences in 24.5% of the cases, whereas respondents
saw similarities and coders saw differences in 10.6% of the cases.
2. Other examples in which coders, but not respondents, thought there were similarities
between attracting and disliked traits included the following pairs of liked/disliked
partner qualities: “a quiet person”/“he would get quiet and shut me out,” “assertive”/
“whines and complains,” “her looks”/“feminine upkeep (facials, sun therapy),” and
“outspoken”/“certain things he says are uncalled for.”
3. Of those respondents who gave a description of their disenchantment experience, the
distribution by type of scenario was as follows: Time will tell (63.9%), Rose-colored
glasses (16.6%), People change (13.9%), and Sour grapes (5.6%). It is difcult to place too
much weight on these statistics, however, given that only a minority of respondents
gave such responses and that participants were not asked directly whether any of these
scenarios occurred.
4. If an independent variable increases by one unit, then the estimated odds of a fatal
attraction is multiplied by the odds ratio, i.e., the antilog coefcient (exp(B)).
REFERENCES
Altman, Irwin, Anne Vinsel, and Barbara B. Brown. 1981. “Dialectic Conceptions in Social
Psychology: An Application to Social Penetration and Privacy Regulation.” Pp. 107–
60 in Advances in Experimental Social Psychology, vol. 14, edited by L. Berkowitz. New
York: Academic Press.
Aron, Arthur and Elaine Aron. 1986. Love and the Expansion of Self: Understanding Attraction
and Satisfaction. New York: Hemisphere Publishing.
Baxter, Leslie A. and Barbara M. Montgomery. 1996. Relating: Dialogues and Dialectics. New
York: Guilford Press.
Byrne, Donn. 1971. The Attraction Paradigm. New York: Academic Press.
Byrne, Donn, and Gerald L. Clore. 1970. “A Reinforcement Model of Evaluative Processes.”
Personality: An International Journal 1:103–28.
Cleek, Margaret Guminski and T. Allan Pearson. 1985. “Perceived Causes of Divorce: An
Analysis of Interrelationships.” Journal of Marriage and the Family 47:179–83.
Felmlee, Diane H. 1995. “Fatal Attractions: Affection and Disaffection in Intimate Relation-
ships.” Journal of Social and Personal Relationships 12:295–311.
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions

280 SO CIOLOG ICAL PERSPECTIVES Volume 44, Number 3, 2001
———.1998a. “Be Careful What You Wish for . . . : A Quantitative and Qualitative Investi-
gation of Fatal Attractions.” Personal Relationships 5:235–53.
——— .1998b. “Fatal Attraction.” Pp. 3–31 in The Dark Side of Relationships, edited by B. H.
Spitzberg and W. R. Cupach. Hillsdale, NJ: Lawrence Erlbaum.
Festinger, Leon. 1957. A Theory of Cognitive Dissonance. New York: Harper and Row.
Freeman, Harvey R. 1985. “Somatic Attractiveness: As in Other Things, Moderation Is
Best.” Psychology of Wom en Quarterly 9:311–22.
Goldberg, Jane G. 1993. The Dark Side of Love. New York: G . P. Putnam’s Sons.
Hill, Christopher T., Zick Rubin, and Letitia A. Peplau. 1976. “Breakups before Marriage:
The End of 103 Affairs.” Journal of Social Issues 32:147–68.
Jung, Carl G. 1973. Memories, Dreams, Reections. Edited by A. Jaffe and translated by
R. Winston and C. Winston. New York: Pantheon Books.
Kessler, Ronald C. and David F. Greenberg. 1981. Linear Panel Analysis: Models of Quantita-
tive Change. New York: Academic Press.
Lewis, Robert A. 1973. “A Longitudinal Test of a Developmental Framework for Premarital
Dyadic Formation.” Journal of Marriage and the Family 35:16–25.
McClanahan, Kimberly K., Joel A. Gold, Ellen Lenney, Richard M. Ryckman, and Gordon
E. Kulberg. 1990. “Infatuation and Attraction to a Dissimilar Other: Is Love Blind?”
Journal of Social Psychology 130:433–45.
Miller, Rowland S. 1997. “Inattentive and Contented: Relationship Commitment and
Attention to Alternatives.” Journal of Personality and Social Psychology 73:758–66.
Murray, Sandra L. and John G. Holmes. 1993. “Seeing Virtues in Faults: Negativity and the
Transformation of Interpersonal Narratives in Close Relationships.” Journal of Person-
ality and Social Psychology 65:707–22.
Murray, Sandra L., John G. Holmes, and Dale W. Grifn. 1996. “The Benets of Positive
Illusions: Idealization and the Construction of Satisfaction in Close Relationships.”
Journal of Personality and Social Psychology 70:79–98.
Newcomb, Theodore Mead. 1961. The Acquaintance Process. New York: Holt, Rinehart and
Winston.
Pines, Ayala M. 1997. “Fatal Attractions or Wise Unconscious Choices: The Relationship
between Causes for Entering and Breaking Intimate Relationships.” Personal Rela-
tionship Issues 4:1–6.
Purdum, Todd S. 1996. “Facets of Clinton.” New York Times Magazine, May 19, 36–41, 62.
Reiss, Ira L. 1980. Family Systems in America, 3rd ed. New York: Holt, Rinehart and
Winston.
Simmel, Georg. 1955. Conict and the Web of Group-Afliation. Glencoe, IL: Free Press.
Smith, Eleanor R., Donn Byrne, and Paul J. Fielding. 1995. “Interpersonal Attraction as a
Function of Extreme Gender Role Adherence.” Personal Relationships 2:161–72.
Snyder, C. R. and Howard L. Fromkin. 1980. Uniqueness: The Human Pursuit of Difference.
New York: Plenum Press.
Surra, Catherine A. 1991. “Mate Selection and Premarital Relationships.” Pp. 54–75 in
Contemporary Families, edited by A. Booth. Minneapolis, MN: National Council on
Family Relations.
Tennov, Dorothy. 1979. Love and Limerence: The Experience of Being in Love. New York: Stein
and Day.
Van Lange, Paul A. M. and Caryl E. Rusbult. 1995. “My Relationship Is Better than—and
not as Bad as—Yours: The Perception of Superiority in Close Relationships.” Person-
ality and Social Psychology Bulletin 21:32–44.
Winch, Robert F. 1955. “The Theory of Complementary Needs in Mate Selection: A Test of
One Kind of Complementariness.” American Sociological Review 20:552–55.
This content downloaded from 146.186.111.127 on Wed, 14 Oct 2015 18:07:17 UTC
All use subject to JSTOR Terms and Conditions