P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ

Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

Archives of Sexual Behavior, Vol. 31, No. 3, June 2002, pp. 275–287 (°
C 2002)

Sexual Motivation and the Duration of Partnership

Dietrich Klusmann, Dr. phil.1

Received October 4, 1999; revision received August 8, 2001; accepted September 8, 2001

The variation of sexual motivation with duration of partnership is analyzed in data from a survey
of German students. The sample of 1865 includes only students aged 19–32 who reported to be
heterosexual and to live in a steady partnership. Main results are (1) sexual activity and sexual
satisfaction decline in women and men as the duration of partnership increases; (2) sexual desire
only declines in women; and (3) desire for tenderness declines in men and rises in women. Because
these results are based on cross-sectional data, a longitudinal explanation is precarious. Individual
differences in mating strategy associated with the probability of having a partnership of shorter or
longer duration at the time of the survey may account for some part of the findings. This possibility
set aside, post hoc explanations for the results as reflecting a modal time course of partnership are
evaluated with regard to habituation, routine, gender role prescriptions, and polarization of roles. In
addition, an explanation from evolutionary psychology is offered, entailing the following ideas: the
psychological mechanisms of attachment in an adult pair bond have evolved from the parent–child
bond. Due to this nonsexual origin, a stable pair-bond does not require high levels of sexual desire,
after an initial phase of infatuation has passed. Nevertheless, male sexual desire should stay at a high
level because it was selected for in evolutionary history as a precaution against the risk of sperm
competition. The course of female sexual desire is assumed to reflect an adaptive function: to boost
attachment in order to establish the bond.

KEY WORDS: sexual desire; partnership; gender differences; evolutionary psychology.

INTRODUCTION Every continued relationship between two persons

leaves a history of emotional and behavioral changes, in-
This paper is part of a larger survey study aimed cluding the history of sexual desire. Nevertheless, the in-
at the description of current sexual behavior in German terest in researching such changes seems to be small. Even
students (Schmidt, 2000). The tabulation of variables re- in large studies with detailed breakdowns of many vari-
flecting sexual motivation by the duration of partnership ables, the duration of partnership was seldom included,
revealed an unexpected association: with increasing du- or if so, analyzed thoroughly as an independent variable.
ration of partnership, a marked decline in sexual motiva- In a recent, large English survey (Johnson, Wadsworth,
tion occurred in women but not in men. The aim of this Wellings, & Field, 1994), which contains a highly differ-
paper is to present this result, analyze it as far as possible entiated body of tables, duration of partnership was in-
with the data at hand, and discuss possible explanations. In vestigated only in one analysis. It revealed a clear trend
order to do so, three sources of knowledge will be consid- of reducing frequency of sexual intercourse with increas-
ered: (1) results from other surveys; (2) social psychology; ing duration of partnership, which was much stronger than
and (3) theories about the evolutionary origins of gender the respective trend for age. There were no differences be-
differences in sexual motivation. tween men and women, but this has to be expected from
logical considerations alone. If a sexual act requires the
1 Klinik
cooperation of a woman and a man in order to be per-
und Poliklinik für Innere Medizin, Abteilung für Medizi-
nische Psychologie, Universitätsklinikum Hamburg-Eppendorf, Mar-
formed, its frequency must be the same for both sexes in
tinistr. 52, 20246 Hamburg, Germany; e-mail: klusmann@uke. the population. Thus, actual sexual behavior cannot differ
uni-hamburg.de. much between women and men as long as the sample is

275
0004-0002/02/0600-0275/0 °
C 2002 Plenum Publishing Corporation
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

276 Klusmann

representative enough for the population of eligible part- occurred; gender differences in the ipsative measure were
ners. Sexual behavior often requires a compromise be- not reported. Women who rated their sexual desire as be-
tween the different proclivities of women and men ing lower than their partners gave lower ratings on sexual
(Symons, 1979), but sexual desire and other emotions, satisfaction and on satisfaction with the relationship com-
which are antecedent or contingent to sexual behavior are pared with women who rated their sexual desire as being
not restricted in this way. Thus, gender differences are higher than their partners. In men, no such trend existed.
expected mainly to appear in the frequency of emotions, Gender differences in motivation are also investi-
not behaviors. In the survey by Johnson et al. (1994), the gated in human mating psychology from the perspective of
relation between duration of partnership and emotions is evolutionary theory. Although this theoretical framework
mentioned only with an attitudinal item: more than two does not deal directly with the question at hand, some
thirds of the sample agreed with the assertion that sexual of its ideas might be useful to understand the findings to
satisfaction generally will increase with the duration of a be presented. Therefore, a short account of evolutionary
relationship. psychology is in place. Those of our ancestors who were
The American survey by Laumann, Gagnon, Michael, equipped with mental features that gave them a reproduc-
and Michaels (1994) differentiated four groups: (1) short- tive advantage were able to spread these features in the
term partnership of less than 1 month; (2) long-term part- gene pool. The evolution of the capacity for emotions and
nership without cohabitation; (3) long-term partnership motives enabling the individual to act effectively in the
with cohabitation; and (4) marriage. Physical and emo- social world, especially in matters pertaining directly to
tional satisfaction rose from about 10–16% in the first reproduction, was also driven by this logic. This includes
group (short-term) to 40–50% in the last group (marriage), the bond between parent and child, the bond between cou-
the other categories ranged in-between. A lack of sex- ples, discrimination in mate choice, and the situational
ual interest was reported by 32% of women, but only by fine-tuning of sexual motivation (Baker & Bellis, 1995;
15% of men (age range, 18–29 years). However, the rela- Buss, 1999; Geary, 1998; Mealey, 2000). In the discus-
tion of these gender differences to duration of partnership sion part of this paper, it will be argued that gender dif-
was not analyzed. ferences in sexual desire as dependent on the duration of
In an investigation of couples, who described them- partnership may have evolutionary roots as well.
selves as happily married (Neiswender Reedy, Birren, & The typical or modal course of changes over time in
Schaie, 1981), emotional security was generally rated to a partnership should be ideally investigated with a lon-
be the most important quality of their relationship and gitudinal design where subjects are observed repeatedly
this emphasis strongly increased with age, especially in as their partnership proceeds. However, it seems that this
women. In both genders, the importance attributed to sex- has not yet been done for the domain of sexual behav-
ual intimacy stayed at a constant level until middle age ior and motivation. Therefore, an approximation with a
(45 years) and then decreased. Communication was given second-best design may be interesting enough as a substi-
most emphasis by the younger couples (around 28 years). tute. This second-best design is the cross-sectional study
A representative survey from Finland (Haavio- with information on the partnership at one single point in
Mannila & Kontula, 1997) revealed gender differences time. The natural variation in the length of time elapsed
in the relation between age and general sexual satisfac- since the beginning of the partnership will be used to
tion. In women the general experience of intercourse as construct a pattern of the modal course. However, some
“very pleasant” was highest at around the age of 30 and precautions must be taken. Only when short-term part-
then dropped steadily until the sixth decade. In men, this nerships have the same characteristics as long-term part-
measure stayed fairly constant throughout the age range. nerships in their early stages, will the cross-sectional study
No measure of duration of partnership was reported. provide the same results as a longitudinal study. As
Davies, Katz, and Jackson (1999) measured the dis- students differ in their willingness and their ability to form
crepancy in sexual desire between the two partners of a enduring relationships, different personality types are ex-
couple. An absolute measure was computed as the dif- pected to characterize the short-term group compared with
ference of self-reported levels of sexual desire of both the long-term group. Thus, any results should be evaluated
partners. In addition, an ipsative measure was obtained with the possibility of self-selection in mind.
by asking the respondents directly if they and their part-
ners had the same levels of sexual desire. Even when de- METHOD
sire levels differed in absolute terms, most respondents
(75%) rated no difference to their partners in the ipsative The survey from which data will be reported is the
measure. With the absolute measure, no gender differences third in a series of surveys about the sexual behavior
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

Sexual Motivation and the Duration of Partnership 277

Table I. Description of the Sample

Men (N = 898) Women (N = 967)

Median Mean SD Median Mean SD p

Age (years, range: 19–32) 26 25.7 3.0 24 24.8 3.3 .001a

Duration of partnership (months) 29 37.9 33.7 33 40.4 33.0 .030a
Lifetime number of sexual partners 3 5.9 7.5 3 5.5 6.3 .499a
Age at first coitus (years)b 18.9 18.2 .001c

a Mann-Whitney U test.
b Median from survival table.
c Wilcox (Gehan) statistic for survival tables.

of German students (Schmidt, Klusmann, Dekker, & had passionate sex within the last week” and “Sex with
Matthiesen, 1998). In contrast to its predecessors from my partner could not be better” (Table II). There was no
1966 to 1981, it includes an elaborated section about part- significant difference between men and women, neither
nership. In the summer of 1996, the questionnaire was sent in the total proportion of yes-responses nor in how these
to 7,500 students from 15 German universities, selected responses decreased with duration of partnership.
for size and location in order to obtain a representative
sample. The response rate of 41% was slightly higher in
women than in men. The sample studied in this paper Desire for Sex
includes only students aged 19–32 who reported to be
heterosexual and were currently committed to a steady Desire for sex was assessed in several ways. The first
partner (71.4% of women and 63.7% of men). In women, measurement was included in a series of statements about
the median duration of commitment was 33 months, in the partnership, introduced by the question: Which of the
men it was 29 months (Table I). Female students were following behaviors and feelings are more characteristic
younger than male students, and they started having sex- of you and which are more characteristic of your partner?
ual intercourse earlier. The median lifetime number of sex- The statement pertaining to sexual desire was “wants to
ual partners was 3 in both genders, 57.4% of the women have sex often.” Answer categories were (a) “more char-
and 58.5% of the men lived in separate flats, 9.5% of the acteristic of me”; (b) “more characteristic of my partner”;
women and 9.4% of the men were married, 7.2% of the (c) “characteristic of both of us”; (d) “characteristic of
couples had children. neither of us.”
With this kind of question, the respondent’s desire
for sex was assessed relative to the perceived desire of the
RESULTS partner. It did not reflect sexual desire in absolute terms. If
the respondent “wanted to have sex often” with the same
Frequency of Coitus strength as his or her partner, this may have occurred on
a higher or a lower level. Nevertheless, an absolute mea-
The median of the reported frequency of coitus in sure of desire could not have been very much at odds with
the last 4 weeks for the total group was 7. It was higher the employed relative measure since it had to produce the
in men (8) than in women (6), but there was no difference observed intrapair differences within partnerships. More-
in the mean. In both genders, the frequency of coitus de- over, a common standard of comparison for the strength
clined as the duration of partnership increased (Table II), of desire is hard to establish.
dropping from a median of 10 in the first year to 5 in the Nearly half of the respondents characterized them-
fifth year. selves as different from their partners, and the other
half used the “both of us” or “neither of us” category
(Table III). If differences were acknowledged, these were
Satisfaction With Sex Life highly gender-specific, with men predominantly attribut-
ing “wants to have sex often” to themselves and women
With a longer duration of partnership, fewer respon- attributing it to their male partners. With a longer duration
dents described themselves as “very satisfied with sex in of partnership, gender-specific characterizations increased
the partnership” (Table II). This drop in sexual satisfaction and use of the “both of us” category decreased. If the sta-
was also revealed in two additional statements: “We have tistical progression with time could be read as reflecting
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

278 Klusmann

Table II. Changes in Sexual Experience and Behavior With Increasing Duration of Partnership (Yes-Responses in % and Odds Ratios From
Logistic Regressiona )

Men Women
<1 year 1–3 years >3 years Exp(B)b p <1 years 1–3 years >3 years Exp(B) p

Experience
Coitus more than seven times in last 4 weeks 67 49 38 0.63 ∗∗∗ 67 43 34 0.61 ∗∗∗

(median)
Very satisfied with sex life 41 27 23 0.72 ∗∗∗ 46 27 27 0.72 ∗∗∗

Passionate sex in last 4 weeks 91 89 84 0.87 ns 93 89 81 0.70 ∗∗

Sex with partner could not be better 48 35 28 0.74 ∗ 54 35 32 0.72 ∗

Desire
Wants to have sex oftenc 76 74 76 1.02 ns 65 44 26 0.53 ∗∗∗

Just wants to be tenderc 70 55 48 0.65 ∗∗∗ 89 89 93 1.21 ns

Sex in partnership not often enough 27 51 51 1.50 ∗∗∗ 24 36 38 1.27 ∗∗

Lack of sexual interest experienced often 1 3 1 0.91 ns 9 14 17 1.45 ∗∗∗

Extrapair sex
Most recent coitus not with committed partner 1 2 2 1.10 ns 2 2 3 1.04 ns
Currently ongoing sexual affair outside 7 5 5 0.87 ns 6 6 7 1.20 ns
partnership
Extrapair sex at least once while in current 12 23 27 1.36 ∗∗∗ 14 20 30 1.55 ∗∗∗

partnership
Wants to have sex outside partnershipc 22 30 33 1.20 ∗∗ 15 19 20 1.14 ns
Masturbation
Masturbation at least one time in the last 82 84 87 1.11 ns 61 63 55 0.82 ∗∗

4 weeks
Masturbation more than four times in the last 44 50 45 0.97 ns 17 16 14 0.89 ns
4 weeks

a For use in logistic regression (Hosmer & Lemeshow, 1989), duration of partnership has been transformed by taking the square root. This reduces the
effect of differences at the high end of the range and normalizes the distribution. Additionally the variable has been z-transformed in order to facilitate
the interpretation of the odds ratios.
b Exp(B) is the odds ratio from a logistic regression with duration of partnership (transformed as described above) as a predictor of the respective

dichotomous row variable (target variable). It reflects the ratio of change in the odds for obtaining a yes-response when duration of partnership
(transformed) changes 1 unit of standard deviation. If the odds ratio is <1, yes-responses decrease with duration of partnership, If it is >1, they
increase. No covariate is introduced into the equation. When age is used as a covariate, results generally stay the same with few exceptions.
c This attribute is counted as a yes-response, if marked by the respondent as characteristic of self either in contrast, or in accordance with the partner.
∗ p = .05. ∗∗ p = .01. ∗∗∗ p = .001.

the modal course of a partnership, this would be a part- Is this drop mediated by age? Age correlates with
nership that starts with a high level of sexual desire in the duration of partnership (r = .32), but it has only a
both partners. Then, after about 1 year, sexual desire in very small effect in a logistic regression to predict sex-
the female partner would drop steadily, whereas it would ual desire. The course of sexual desire with duration of
stay at the same level in the male partner (Fig. 1). partnership is not modified when age is held constant

Table III. The Statement “Wants to have sex often” as Characterizing Me (Respondent), Partner, Both, or Neither: Changes
With Duration of Partnership

Men Women
Duration of
partnership (months) Me Partner Both Neither Total Me Partner Both Neither Total

1–6 16.2 6.9 58.8 18.5 100.0 6.1 16.5 64.3 13.0 100.0
7–18 27.9 8.7 52.5 10.9 100.0 10.8 28.1 43.8 17.3 100.0
19–36 37.9 11.7 34.0 16.5 100.0 12.5 33.0 39.3 15.2 100.0
37–72 47.8 10.8 25.7 15.7 100.0 7.6 47.8 23.0 21.6 100.0
73–212 53.8 5.1 23.1 17.9 100.0 4.7 53.0 22.1 20.1 100.0
Total 37.5 10.0 37.6 15.6 100.0 8.7 37.5 35.8 18.0 100.0
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

Sexual Motivation and the Duration of Partnership 279

Table IV. Sexual Motivationa Predicted by Duration of Partnership

and Other Variables: Examination of Possible Confounding Effects in
Logistic Regression Modelsb

Men Women
β pc β pc

Model 1
Duration of partnership .023 ns −.628d ∗∗∗

Model 2
Duration of partnership .047 ns −.626 ∗∗∗

Age −.120 ns .006 ns

Model 3
Duration of partnership −.006 ns −.629 ∗∗∗

Living together (yese ) −.075 ns .004 ns

Model 4
Fig. 1. “Wants to have sex often.” Percentage of yes-responses by ∗∗∗
Duration of partnership .033 ns −.675
duration of partnership. A yes-response is counted, if the respondent
Imbalance of commitment f .355 ∗∗∗ .397 ∗∗∗
marked this statement as characteristic of himself or herself, either in
Model 5
contrast, or in accordance with the partner (see Table III). Estimation ∗∗∗
Duration of partnership .048 ns −.578
by logistic regression.
Life time number of sexual partners .168 ∗ .239 ∗∗

Model 6
Duration of partnership .061 ns −.613 ∗∗∗
(Table IV, Model 2). Thus, the effect of duration cannot be Age −.128 ns .072 ns
attributed to age differences. Neither can it be attributed Living together (yese ) −.048 ns .081 ns
to the differences between living together in a flat or not Imbalance of commitment f .385 ∗∗∗ .451 ∗∗∗

Life time number of sexual partners .112 ns .189 ∗

(Table IV, Model 3), being married or not, and having
children or not. Age at first coitus −.029 ns −.120 ns
Church attendance (yesg ) −.256 ∗ .038 ns
Political orientation left to right .037 ns .017 ns
Approval for feminist views .004 ns .062 ns
Desire for Tenderness
a Target variable: “Wants to have sex often.” A yes-response is counted,
if the respondent marked this item as characteristic of self, either in
The statement “just wants to be tender” had been contrast, or in accordance with the partner.
presented the same way as the statement “wants to have sex b All continuous variables are z-transformed. For simplicity of presenta-
often.” With this item, gender differences were reversed: tion the constant of the logistic equation is omitted.
c Wald-statistic.
the desire “just to be tender” rose in women and sloped
d Confidence interval for β = −.628: (a = .05, −.773, −.483).
down in men (Fig. 2). Both partners tended to concur in e Deviation coding: (+1) couple lives together; (−1) couple lives in
their desire for tenderness when the partnership had been separate flats.
newly formed; at later stages women and men increasingly f Positive values indicate: own commitment exceeds commitment of
differed. As with the desire for sex, this trend was not partner.
g Deviation coding church attendance: (+1) on a regular basis or some-
confounded with age, cohabitational status, marital status,
or the presence of children. times; (−1) never or rarely.
∗ p = .05. ∗∗ p = .01. ∗∗∗ p = .001.

Lack of Sexual Desire

Coitus Not Often Enough
Lack of sexual desire was experienced by 57.5% of
the women “occasionally” and by 13.3% “often.” With The respondents evaluated the frequency of sex in
men, the respective percentages were 23.9 and 1.0. In their partnerships by choosing from the category labels
women, there was a marked tendency for lack of sexual “too often,” “in just the right frequency,” or “not often
desire to increase with duration of partnership, but not in enough.” At the beginning stage (less than 1 year), the “not
men (Table II). In contrast to the results obtained so far, often enough” category was chosen by women and men
this relationship was modified by age. It was absent in in the same proportion (Table II). With a longer duration,
women around age 22 and most pronounced in women the percentage rose in both genders, but more pronounced
around age 28. in men than in women.
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

280 Klusmann

the men said they did not have sex just for the sake of the
partner. Stated conversely, more men than women report-
edly consented to having sex in spite of an initial lack of
desire. Similar results were obtained when the hypothet-
ical situation was reversed (respondent likes to have sex
and partner does not). Thus, men seemed to be more ea-
ger to accord to the sexual wishes of their female partners
than vice-versa, whether it is to have sex even when they
are somehow not in the mood or to refrain from having
sex when the female partner does not want it. This gender
difference remained constant, regardless of how long the
couple had been together.

Fig. 2. “Just wants to be tender.” Percentage of yes-responses by dura- Sexual Experience at the Most Recent Occasion
tion of partnership. A yes-response is counted, if the respondent marked
this statement as characteristic of himself or herself, either in contrast,
or in accordance with the partner. Estimation by logistic regression. Thus far, it has been shown that sexual satisfaction
diminished with a longer duration of partnership and that
sexual desire diminished too, specifically in women. Does
Sex Outside of the Partnerships
this mean that the most recent sexual experience was
less rewarding in long partnerships compared with short
The occurrence of sexual relations outside of the
partnerships?
steady partnerships could be assessed from a section of
In 90.5% of the respondents, the most recent time
the questionnaire where the last time a respondent had sex
they had sex involved coitus. Nearly half of the events in-
with a sexual partner was described (Table III). In 1.8%
cluded oral–genital stimulation. This kind of stimulation
of the men and 3.1% of the women (difference not signifi-
(active and passive) was reported in men but not in women
cant), the sexual partner at this occasion was not the steady
with decreasing proportions in longer partnerships. Anal
partner. A currently ongoing sexual affair was acknowl-
intercourse was performed seldom and did not vary with
edged by 6.4% of the women and 5.5% of the men. These
duration of partnership. For most respondents, the ini-
two measures of extrapair sexual activity were not related
tiative to have sex had been equal on both sides. When
to the duration of partnership. The stated desire to have
one side was emphasized, both genders reported slightly
a sexual affair significantly increased in men, whereas in
more initiative for men, and this increased with duration
women there was only a nonsignificant tendency.
of partnership.
The emotional experience of the sexual act was gen-
Masturbation erally reported in very positive terms: “I was sexually satis-
fied,” “It was a passionate experience,” and “I was happy”
The level of masturbatory behavior in men was not applied to more than four out of five respondents. Orgasm
influenced by the duration of their partnerships (Table II). was experienced by 92.1% of the men, rising with dura-
In women, a slight decrease could be observed. Both re- tion, and by 65.8% of the women, not rising. For most
sults still prevail after age is controlled. respondents both partners received the same amount of
sexual pleasure. In those who made a distinction, women
and men alike, attributed more sexual pleasure to the male
Dissent About Having Sex
partner. The duration of partnership did not affect this
difference.
There are situations when one partner (in this case
the respondent) does not like to have sex, but the other
(his or her partner) does. This kind of dissent reportedly Further Exploration of the Results
had happened at least once to 82% of the women and to
59% of the men during their recent partnership. Of course, To summarize the preceding results (a) in longer part-
the chance for dissent to occur grows with time spent in nerships, compared with shorter ones, sexual activity and
the partnership; however, it grew faster in women than in sexual satisfaction declined in women and men; (b) sexual
men. After being asked what they usually did after such desire only declined in women; and (c) desire for tender-
a situation occurred, 68.3% of the women and 54.4% of ness declined in men and increased in women. The first
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

Sexual Motivation and the Duration of Partnership 281

finding is not much of a surprise, but the second and third ship because sexual desire is just another aspect of feeling
findings are intriguing. The decline of sexual desire in drawn to someone. If part two prevails (my partner with-
women was strongly revealed in the assessment of sex- draws from me), there might be a negative relationship
ual desire relative to the partner and in an avowed lack of because an uncommitted and perhaps unresponsive part-
sexual desire. It was reflected more weakly in other vari- ner would not engage sexual feelings in the long run. In
ables, but not in all. For instance, the description of the the first position, sexual desire would have the function
most recent sexual act did not vary much with duration of of bridging a gap. In the second, it would be dampened
partnership, but the few existing differences point toward because the emotional exchange is felt to be skewed. The
less female motivation and pleasure in longer partnerships, correlations of “imbalance of commitment” with sexual
compared with shorter ones. desire, as measured in Fig. 1, were r = .15 ( p < .01) in
Therefore, the results can be taken as convergent find- women and r = .06 (ns) in men. Thus the first meaning
ings, deserving explanation—at least in the context of this seems to be operative to a larger extent than the second.
study. Since these findings, especially (b) were not ex- Can the variation of “imbalance of commitment” ex-
pected at the time the survey was planned, no explanatory plain at least some part of the negative relation between
hypothesis had been set up in advance. Nevertheless, the sexual desire and duration of partnership that has been
extensive data of the survey allow a post hoc exploration found in women? In order to do so, it should be negatively
of other variables that may contribute to an explanation of correlated with duration of partnership. Actually the cor-
the results. In the next sections, three steps will be taken relation is positive, but close to 0, in women (r = .08, p <
(1) a provisional investigation of the hypothesis: declining .05) and absent in men (r = −.06, ns). Thus, a lower com-
sexual desire in women reflects a receding commitment in mitment of the male partner cannot provide an explanation
the male partner; (2) a search for confounding variables; for the declining sexual desire in women. To the contrary,
and (3) an investigation of the possibility of self-selection. it would explain an increase in sexual desire, because it
is positively related to sexual desire and at the same time
slightly rises with duration of partnership.
Commitment to the Partner and Sexual Desire

Partners in a couple often differ in their commitment. Confounders and Modifiers

One of them may exceed the other in (1) wishing the part-
nership to endure, (2) wanting to possess the partner for If the relationship between sexual motivation (target)
herself or for himself, (3) wanting to spend much time with and duration of partnership (predictor) could be explained
the partner, (4) wanting to talk to the partner, (5) being jeal- at least partly by a confounding variable, the β coefficient
ous, and (6) being in love. Stated negatively, (7) fear of for duration of partnership should shrink significantly after
intimacy may be elicited more easily in one of the part- the confounder had been introduced into the logistic re-
ners, and (8) he or she may tend to withdraw emotionally. gression equation (for details of the technique, see Hosmer
These eight descriptions were used to build a scale. Each & Lemeshow, 1989). In Table IV, potential confounders
time a statement was marked as describing the respondent are examined (age, living together or separate, imbalance
in contrast to his or her partner, a scale point was given for of commitment, life time number of sexual partners, age
the commitment of the respondent (the negative items re- at first coitus, church attendance, political orientation, ap-
versed). Perceived commitment of the partner was scaled proval for feminist views). The first model contains no
equivalently. The measure to be used was the computed confounder at all. Duration of partnership predicts a con-
difference of both scales. It reflects the relative excess or stant level of sexual motivation in men (β = .023) and
deficit of ego’s commitment after alter’s perceived com- a dropping level in women (β = −.628). The interesting
mitment has been subtracted. A positive value means my coefficient is the β coefficient for women. If one of the
desire to be close to my partner is greater than his desire to candidate variables would reduce the magnitude of this
be close to me. This variable, termed “imbalance of com- coefficient beyond the upper limits of its confidence in-
mitment,” had a normal distribution with a larger mean terval (−.773, −.483), then a confounding effect could be
for women than for men. Positive values carry a mixed assumed. In fact none of the candidate variables showed
message: (1) I feel drawn to my partner; (2) my partner such an effect. Even when all potential confounders were
withdraws from me. included simultaneously (Table IV, Model 6), the regres-
How would “imbalance of commitment” relate to sion coefficient for duration of partnership was not signif-
sexual desire? If part one of its meaning prevails (I feel icantly diminished (from −.628 to −.613). If the variable
drawn to my partner), there might be a positive relation- “imbalance of commitment” was used alone (Model 4) the
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

282 Klusmann

β coefficient even increased from .628 to .675. This in- or less close to one of the extremes. Students who pur-
crease was not significant, but it showed, that “imbalance sue a short-term strategy by definition will enter the study
of commitment” instead of “explaining” the relationship with a short duration of partnership. Students who pursue
between sexual motivation and duration tended to obscure a long-term strategy typically, but not always, will enter
it, acting as a suppressor variable. the study with a long duration of partnership. Therefore,
For brevity’s sake, Table IV does not include the in- sexual strategy and duration of partnership at the time
teraction terms of other predictors with duration of part- of the study are related. If sexual strategy as a sociosex-
nership. When these terms were included, no confounding ual orientation were also related to sexual desire, it could
effect appeared either. The gender-specificity of the rela- play the role of a confounding variable: the relationship
tion between duration and sexual desire could also have between duration of partnership and sexual desire would
been represented by the interaction duration × gender in a reflect differences in sexual strategy associated with both
single logistic regression equation. Instead, two separate variables.
equations for each gender have been chosen, because this How is it to be known which strategy a student pur-
presentation is closer to the raw data and easier to under- sues? The short-time strategy might be revealed by a large
stand. Analyzing the interaction instead of the separate number of sexual partners and perhaps by a general pro-
main effects for men and women basically gave the same clivity to sexual activity, indicated by an early age of first
results. sexual intercourse. If the relationship between duration of
In order to test how specific these results were for partnership and sexual desire would diminish after con-
the chosen target variable (“wants to have sex often”) trolling for these indicators, a case could be made for self-
as an expression of sexual motivation, a second measure selection according to sexual strategy.
of sexual motivation, the variable “lack of sexual desire” The lifetime number of sexual partners encountered
(Table II had been examined as a target variable with the by the respondent correlated with the desire to have sex (as
set of predictors used in Table IV). The results were (in- measured in Fig. 1) by r = .19 ( p < .01) in women and
versely) in accordance with those obtained with “wants to by r = .07 (ns) in men. The correlation with duration of
have sex often.” Using “just wants to be tender” (Table II) partnership equaled r = −.25 ( p < .001) in both genders.
as a target variable yielded roughly a mirror image of the Did the statistical control for the lifetime number of sex-
results obtained for “wants to have sex often” whereby the ual partners alter the relationship between sexual desire
gender effect was exchanged. and duration of partnership? The answer is no (Table IV,
Model 5). A similar result was obtained with age at first
sexual intercourse. These results indicate that large effects
Self-Selection? due to self-selection may not be present, but they are still
not fully appropriate to evaluate this matter.
This is a cross-sectional study. An unknown propor-
tion of the short-term partnerships would not evolve into
DISCUSSION
long-term partnerships and therefore cannot be considered
early stages of a longer course. These “true” short-time
The discussion will focus on the following questions:
partnerships may differ from the forerunners of longer
(1) Why do sexual activity and satisfaction decline with
courses in important respects and thus may undermine the
time in a partnership? (2) Why does sexual desire decline
assumption that a modal time course would show up in
in women but not in men?
the cross-sectional data in the same way as it would in a
longitudinal study. But even when a longitudinal interpre-
tation will always be precarious, at least some probes can Habituation
be made in order to assess the importance of factors that
might be associated with self-selection. Habituation is the process by which a reaction is
To clarify the idea, let us presume a simple model weakened after repeated or prolonged exposure to the
with two types of sexual strategy: (a) many short-term stimulus. In laboratory experiments on humans, arousal re-
partnerships with different mates; (b) one long term- sponses, elicited by sexually stimulating pictures or
partnership with one mate. These strategies can be thought movies, habituate after repeated exposure and gain vigour
to correspond to different sexual temperaments, as in the with new stimuli. This effect is more pronounced in men
distinction of an unrestricted versus restricted sociosex- than in women (Koukounas & Over, 1993; Laan &
ual orientation (Gangestad & Simpson, 1990; Simpson & Everaerd, 1995; O’Donohue & Plaud, 1991). Can habitua-
Gangestad, 1991). Every student’s strategy comes more tion be an explanation for declining sexual desire? It seems
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

Sexual Motivation and the Duration of Partnership 283

to works well in short-term situations encompassing min- intimacy to women. Social construction theory (Hacking,
utes, hours or days, but does it also work in months or 1999) postulates a causal effect of social categories. More
years? Maybe the reaction to a pattern of similarity builds specifically, role models act as moulds for the socializa-
up very slowly in a couple’s sex life; but then our findings tion of gender-specific ways to think to feel and to have
should be reversed: sexual desire should diminish more in motives, including sexual desire. Any behavior or way of
men than in women. Even if our results could be consid- feeling will be reinforced if it is in accordance with the
ered an instance of habituation, the physiological mech- role, and it will be ignored, discouraged, or punished if
anism of habituation would not explain much because it not. This basic idea has been supplemented and refined by
can only elucidate how activation (if sexual desire can the concepts of sexual scenarios and scripts, which more
be called so) diminishes, but not why. The why-question intricately relate the societal level to the level of individual
pertains to possible adaptive functions of habituation in psychology (DeLamater, 1987; Simon & Gagnon, 1987).
sexual responses, an issue that will be discussed later But there is one puzzling complication: the depen-
in the section on evolutionary psychology. dence of gender differences on the duration of partnership.
Gender stereotypes may well rule general propensities of
women and men, but why should gender differences in
Routine sexual desire be virtually absent in the first months of a
partnership and later intensify so markedly? Does the fine-
The decline of sexual activity and satisfaction is not tuning of gender roles include an expectation of female
surprising since most human activities are subject to sur- sexual desire to be high when the couple is just getting
feit when pursued too often and for too long. Long-term together, but not in later stages? This seems to be improb-
partnerships cannot escape some measure of routine, and able in a modern society, and even in a traditional society,
this would “explain” a decline of sexual activity and satis- it would be difficult to conceive. Students in the 1990s are
faction. But is routine really an explanation or just another far removed from submitting to traditional role models,
way to put the question? There are many activities in life at least in their avowed attitudes. Women no longer de-
performed over and over again without ceasing to give valuate themselves when they show sexual desires. To the
pleasure. Why should routine have an effect in this case? contrary, sexual desire is seen as a sign of vigor and per-
And why should it affect women more than men? sonal assertion. Students are generally liberal in their atti-
tudes. But there is still some variation that can be exploited
in order to assess at least superficially whether there is
Receding Commitment a general effect of normative orientations. The following
three variables reflect to some extent how much traditional
The measure “imbalance of commitment” reflects versus nontraditional gender-roles are embraced: political
how partners differ in striving to be close to the other. orientation (left to right), affinity to feminist positions, and
With this measure, male commitment relative to the part- church attendance. None of these variables was related to
ner seems to recede with duration of partnership, whereas sexual desire or desire for tenderness, and consequently
female commitment seems to grow. But this could not none could explain the relation between these variables
explain receding female sexual desire. A lack of commit- and duration of partnership (Table IV, Model 6).
ment in the male partner does not seem to suppress sexual
desire in a woman as long as she herself feels committed
to him, or maybe as long as she sticks to her choice for this Polarization of Roles
partner. Perhaps desire for sex can be conceived as part of
a larger move to get close to the partner and to overcome The discrepancy between men and women in their
his resistance. desire for sex and for tenderness may have resulted from
a polarization of roles whereby one partner assumed the
demanding role and the other assumed the restrictive role.
Gender Roles Role differentiation along this line easily evolves when-
ever people repeatedly interact and mutual expectations
Can the increasing gender-specific differentiation of are shaped and reinforced by past experience. It can be ob-
sexual desire and desire for tenderness with duration of served in many contexts, for example, in work groups, and
partnership be put down to gender role prescriptions? it does not need to be gender-specific. Desires for sex and
There surely exists a general gender stereotype, attribut- tenderness may have been shaped by positive feedback
ing desire for sex to men and desire for tenderness and within the couple, reinforcing small initial differences,
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

284 Klusmann

which eventually led to the sedimentation of complemen- and, according to Zeifman and Hazan (1997), the nonsex-
tary roles. Since role polarization is a general process in- ual component of the bonding process, which most re-
dependent of content it can only explain how roles are dif- sembles the parent–child bond, will take over. Declining
ferentiated, but not why they do so along specific lines. To interest in sexual activity would be an expected conse-
the extent in which such a process took place, the explana- quence of this later stage. As in the parent–child bond, the
tory task is reduced to the question of why small seminal desire for physical proximity would not be at its highest
differences are related to gender. These differences could when the bond is secure, but when it is in the making or
be considered as differences in gender-specific scripts ac- when it is threatened. Thus, this theory predicts committed
quired through a multitude of societal influences, suggest- long-term partnership with a relatively low level of sexual
ing to men that they should be eager for sex and to women desire, but no differences between women and men in this
that they should have a desire to obtain romance and assur- respect. Such differences must be accounted for by other
ance in a partnership with this cleft deepening during the mechanisms.
course of sustained interaction in a partnership. But why
should this polarization process be so overwhelming that
The Coolidge Effect
it be maintained against all the equalizing forces that are
present in student culture? And even more, as Buss (1995)
In many mammals, a selective elicitation of sexual
stated, “culture,” “learning,” and “socialization” represent
urge can be observed, especially in species where males se-
human phenomenon that “explain” only to a certain extent
cure large harems (Symons, 1979, pp. 208–213). In these
and then require explanations themselves (Buss, 1995).
species male sexual activity flags with repeated exposure
Therefore, it may be fruitful to step back from the
to the same female, but can be revived many times by
level of proximal causes considered so far and look at the
the introduction of a new female. This phenomenon is
results from a greater distance, as a reflection of gender-
termed the “Coolidge Effect.” In female animals, no equiv-
specific evolutionary adaptations, that may also be embod-
alent effect has been reported. Although the Coolidge
ied in cultural scripts. In the following section, some ideas
Effect cannot be investigated experimentally in humans,
from evolutionary psychology will be evaluated in order to
some evidence seems to support it: The habituation of
propose an evolutionary account for the decline of female
sexual responses in human males when visual stimuli
desire with duration of partnership and the concomitant
are often repeated, its invigoration with new stimuli, and
constancy of male desire.
anecdotal self-reports. For the task of explaining our re-
sults, the Coolidge Effect makes exactly the wrong pre-
diction because sexual desire is expected to flag in the
Pair-Bond
male instead, as has been found, in the female. Perhaps
the Coolidge Effect works only short-term and mostly
According to Zeifman and Hazan (1997), the adult
reflects a male motivation for sexual variety. It must be
pair-bond and the bond between parent and child has
either absent in a long-term context, or trumped by other
much in common: (a) a similar process of formation: close
mechanisms.
physical proximity, kindness, understanding; (b) similar
reactions to separation and loss; (c) similar physiolog-
ical processes of regulation. These similarities may be Indiscriminating Drive
due to a cooptation of the mental system, regulating the
bond between parent and child by the later evolving pair- Men are generally less discriminating than women
bond system. As the time period of infant dependence was in the choice of sexual partners (Oliver & Hyde, 1993;
growing in our ancestors, a stable bond between parents Symons & Ellis, 1989). More than women, they seem to
probably has furthered reproductive success and thus has experience sexual desire independent of context. There-
been favored by natural selection. The already existing at- fore, it might be argued that men are less sensitive than
tachment mechanism for the parent–child bond could have women to any influence that may impede sexual desire
been the template that was modified into a psychological in longer partnerships. But such an interpretation would
mechanism for the purpose of keeping mates together, be opposed by the Coolidge Effect (should it exists in
a mechanism to ensure the emotional locking of mate humans), since this effect most probably is an adaptation
choice and stable cooperation at least for a sufficient time against indiscriminating sexual activity. It is an example of
span. The first phase of pair-bonding, called infatuation by how fine-tuned the innate programming for sexual arousal
Zeifman and Hazan, is dominated by elation and the de- can be. If it is still true that men choose sexual partners
sire to be close. In later stages, rash infatuation diminishes with less discrimination than women, this only attests to
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

Sexual Motivation and the Duration of Partnership 285

the different contingencies women and men have experi- Safeguarding for sperm competition is solely a male
enced in the evolutionary past after they happened to make concern—a female can always be sure of her paternity,
the wrong choice, but not necessarily to differences in a regardless of whether the father of her offspring is her
general “drive level.” constant partner or not. Sexual infidelity can be a good
At this point, we have a proposition to explain why strategy from a female’s point of view because it would
sexual desire should subside with time in both sexes (the broaden the genetic variety of her offspring and perhaps
nonsexual nature of the bond), but this proposition does improve genetic quality (Gangestad & Simpson, 2000).
not fit the prevailing male sexual desire. Could there be a Helping instead her partner in his precautions against ri-
specific evolutionary advantage for male desire to stay at val sperm would give no immediate advantage to a female.
a high level? But still, securing male cooperation and resources must be
vital whenever female reproductive success depends on
male support. Thus, signals of sexual sincerity and co-
Sperm Competition operation with her male partner’s effort to insure his pa-
ternity should be an adaptive feature of female reproduc-
Sperm competition would explain such an advan- tive strategy in relatively monandrous species (Birkhead,
tage. It can occur whenever different ejaculates meet in the 2000). Perhaps a low level of sexual desire communicated
female reproductive tract. When two males copulate with by the female may count as such a signal. But the oppo-
a female in succession, their chances of fathering an off- site may be more straightforward: If it was important for
spring depend on the interaction between order of copula- a ancestral human female to keep her male partner, why
tion, the interval between two copulations, and the timing should she have been reluctant to have sex anytime he
of the inseminations relative to when the female ovulates wanted to? An evolved motivation for differential sexual
(Birkhead, 2000). In many animals, the frequency of cop- restraint would only make sense if there were advantages
ulation is closely associated with the male’s capacity for to withholding or disadvantages to engaging in sexual ac-
mate guarding. If a male cannot prevent other males from tivity (Thiessen, 1994). Advantages to withholding may be
coming close to his female partner, he can still increase understood in the context of female mating strategy (Buss,
the probability of fathering her offspring by inseminating 1999), but they seem more plausible in the courtship phase
her frequently enough to establish a barrier against rival than in a long-term partnership. A tangible disadvantage
sperm. Thus, copulation as a regular routine can act as an would be the risk of a sexually transmitted disease. Be-
equivalent for mate guarding behavior. The widespread cause these diseases often lead to infertility, a trade-off
occurrence of sperm competition in the animal kingdom between costs and benefits of copulation must be consid-
and its many behavioral repercussions has been recog- ered as a selective force working in evolutionary history
nized by biological research only recently and only against (Baker & Bellis, 1995). When there was not too much at
much resistance (Birkhead, 2000). Then, in a surge of stake, perhaps a moderate level of sexual desire has been
enthusiasm and sensationalism, some exaggerated claims selected for as a compromise.
have been made about the mechanisms of sperm compe-
tition in humans (Baker & Bellis, 1995), which were not
supported by later research (Birkhead, Moore, & Bedford, Sex as Instrumental for the Promotion of Attachment
1997; Moore, Martin, & Birkhead, 1999). In humans, in-
dicators like testes size, sperm number, and sperm length Sperm competition fits in quite logically as a hy-
point toward a relatively modest level of sperm competi- pothesis to explain a constant level of male sexual desire
tion, compared with other primates, and this indicates that (Fig. 1), but it does not account for the course of female
our female ancestors must have been relatively monan- sexual desire. Most readers probably perceive the curve
drous (Birkhead, 2000, pp. 81–83; Gomendio, Harcourt, & for females in Fig. 1 as depicting a drop and try to figure
Roldan, 1998; Smith, 1984), but only relatively. The men- out why the numbers go down. But it could also be in-
tioned reviews indicate some measure of human sperm teresting to reverse the question: Why is female sexual
competition and, most important for the question at hand, desire so high in short (mostly beginning) partnerships?
an evolved disposition in the human male to cope with this Why doesn’t it keep to the same level all over the time
threat to paternity. The results of the present study, which scale? There might be a factor exerting strong influence
show male sexual desire to be high, regardless of the du- in the formative phase of a partnership and subsiding in
ration of partnership, might be explained by an emotional later stages.
preparedness to keep a constant level of insemination in Symons (1979, pp. 253–286) has assembled a mul-
the female partner. titude of findings suggesting the conclusion that women
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

286 Klusmann

more than men use sex instrumentally for the achievement Birkhead, T., Moore, H. D. M., & Bedford, J. M. (1997). Sex, science and
of other goals. Although women generally do not seem to sensationalism. [Review of the book Human Sperm Competition by
R. Robin Baker & Mark A. Bellis. Chapman & Hall, 1995]. Trends
experience less pleasure than men in sexual intercourse, in Ecology and Evolution, 12, 121–122.
in most cultures men, not women, are expected to offer Buss, D. M. (1995). Evolutionary psychology: A new paradigm for psy-
gifts in exchange and to invest materially in courtship. In chological science. Psychological Inquiry, 6, 1–49.
Buss, D. M. (1999). Evolutionary psychology: The new science of mind.
women more than in men, sexual motivation seems to be Boston: Allyn and Bacon.
related to the desire to form an intimate relationship and Davies, S., Katz, J., & Jackson, J. L. (1999). Sexual desire discrepancies:
to secure support. This instrumental nexus presumably Effects on sexual and relationship satisfaction in heterosexual dating
couples. Archives of Sexual Behavior, 28, 553–567.
builds upon a gender-specific psychological preparedness DeLamater, J. (1987). Theories of human sexuality: A sociological ap-
for having specific emotions in specific situations, which proach. In J. H. Geer & W. T. O’Donohue (Eds.), Theories of human
works spontaneously without the intervention of cognitive sexuality (pp. 237–256). New York: Plenum.
Gangestad, S. W., & Simpson, J. A. (1990). Toward an evolutionary
deliberation. In this view, female sexual motivation is seen history of female sociosexual variation. Journal of Personality, 58,
as an adaptation to the way male motivation is designed. 69–96.
It helps to generate heat to establish the newly emerging Gangestad, S. W., & Simpson, J. A. (2000). The evolution of human
mating: Trade offs and strategic pluralism. Behavioral and Brain
pair-bond, and it recedes when it is no longer needed at Sciences, 23, 573–644.
such a high level. This does not imply that there is a point in Geary, D. C. (1998). Male, female: The evolution of sex differences.
asking which gender is adapting more to the other, because Washington, DC: American Psychological Association.
Gomendio, M., Harcourt, A. H., & Roldan, E. R. S. (1998). Sperm com-
the motivational systems of males and females coevolved petition in mammals. In T. R. Birkhead & A. P. Møller (Eds.),
from conflicting demands (Buss, 1999). Sperm competition and sexual selection (pp. 667–756). San Diego:
The reported results would make sense with these Academic Press.
Haavio-Mannila, E., & Kontula, O. (1997). Correlates of increased sex-
views from evolutionary psychology. Nevertheless, at the ual satisfaction. Archives of Sexual Behaviour, 26, 399–414.
end of this discussion the question of how to account for Hacking, I. (1999). The social construction of what? Cambridge, MA:
the gender-specific relation between sexual desire and du- Harvard University Press.
Hosmer, D. W., & Lemeshow, S. (1989). Applied logistic regression.
ration of partnership seems still open to me. If the evolu- New York: Wiley.
tionary legacy of sperm competition were an explanation, Johnson, A. M., Wadsworth, J., Wellings, K., & Field, J. (1994). Sexual
this would be hard to prove. But cultural accounts have attitudes and lifestyles. London: Blackwell.
Koukounas, F., & Over, R. (1993). Habituation and dishabituation of
their difficulties as well. The causal nexus between mech- male sexual arousal. Behavior Research and Therapy, 31, 575–
anisms of gender role socialisation and the way sexual 585.
desire was related to the duration of partnership in this Laan, E., & Everaerd, W. (1995). Habituation of female sexual
arousal to slides and film. Archives of Sexual Behavior, 24, 517–
study may be conceived more convincingly, but at this 541.
point it looks tenuous. The gender-specific effect might Laumann, E. O., Gagnon, J. H., Michael, R. T., & Michaels, S. (1994).
have been the result of the unknown relation between sex- The social organization of sexuality. Sexual practices in the United
States. Chicago: University of Chicago Press.
ual strategy or temperament and the probability of having Mealey, L. (2000). Sex differences: Developmental and evolutionary
entered the study with a specific duration of partnership. strategies. San Diego: Academic Press.
Still, the effect seems to be strong enough to warrant fur- Moore, H. D. M., Martin, M., & Birkhead, T. R. (1999). No evidence for
killer sperm or other selective interactions between human sper-
ther investigation in a longitudinal study. matozoa in ejaculates of different males in vitro. Proceedings of
the Royal Society of London. Series B: Biological Sciences, 266,
2343–2350.
Neiswender Reedy, M., Birren, J. E., & Schaie, K. H. (1981). Age and
ACKNOWLEDGMENTS sex differences in satisfying love relationships across the adult life
span. Human Development, 24, 52–66.
O’Donohue, W. T., & Plaud, J. J. (1991). The long-term habituation of
This survey was conducted by Gunter Schmidt, sexual arousal in the human male. Journal of Behavior Therapy and
Dietrich Klusmann, Arne Dekker, Silja Matthiesen, and Experimental Psychiatry, 22, 87–96.
Oliver, M. B., & Hyde, J. S. (1993). Gender differences in sexuality: A
Siegrid Schäfer. I am thankful to my colleagues for their meta-analysis. Psychological Bulletin, 114, 29–51.
support in the preparation of this paper. Schmidt, G. (Ed.). (2000). Kinder der sexuellen Revolution.
Kontinuität und Wandel studentischer Sexualität 1966–1996.
Gießen: Psychosozial-Verlag.
Schmidt, G., Klusmann, D., Dekker, A., & Matthiesen, S. (1998).
REFERENCES Changes in student’s sexual behaviour: 1966–1981–1996. A first
report on a longitudinal study in West Germany. Scandinavian
Journal of Sexology, 1, 157–173.
Baker, R. R., & Bellis, M. A. (1995). Human sperm competition. Copu- Simon, W., & Gagnon, J. H. (1987). Theories of human sexuality:
lation, masturbation and infidelity. London: Chapman and Hall. A sexual scripts approach. In J. H. Geer & W. T. O’Donohue
Birkhead, T. (2000). Promiscuity: An evolutionary history of sperm com- (Eds.), Theories of human sexuality (pp. 363–383). New York:
petition and sexual conflict. London: Faber and Faber. Plenum.
P1: HAA-GLS/gkj-gxn/HAA/GJF P2: gvt/GKZ
Archives of Sexual Behavior PP457-371755 April 10, 2002 12:37 Style file version July 26, 1999

Sexual Motivation and the Duration of Partnership 287

Simpson, J. A., & Gangestad, S. W. (1991). Individual differences Symons, D., & Ellis, D. (1989). Human male–female differences in
in sociosexuality: Evidence for convergent and discriminant sexual desire. In A. E. Rasa, C. Vogel, & E. Voland (Eds.), The
validity. Journal of Personality and Social Psychology, 60, 31– sociobiology of sexual and reproductive strategies (pp. 131–147).
51. London: Chapman and Hall.
Smith, R. L. (1984). Human sperm competition. In R. L. Smith (Ed.), Thiessen, D. (1994). Environmental tracking by females. Human Nature,
Sperm competition and the evolution of animal mating systems 5, 167–202.
(pp. 602–659). San Diego: Academic Press. Zeifman, D., & Hazan, C. (1997). The bond in pair bonds. In J. A.
Symons, D. (1979). The evolution of human sexuality. New York: Oxford Simpson & D. T. Kenrick (Eds.), Evolutionary social psychology
University Press. (pp. 237–264). Mahwah, NJ: Erlbaum.