The ebb and flow of marital lust: a relational approach.

In general, sexual motivation ensures continuation of the human species. However, on a day-to-day basis, sexual motivation (lust or sexual desire) plays an equally important role as one stimulus for social behavior toward desirable others. Literature on the motivational aspect of sexuality uses three terms somewhat interchangeably--lust, passion, and desire--and preliminary findings suggest these terms overlap to a significant extent in both heterosexual dating and same-sex committed relationships (Hunt et al., 2003; Gargle et al., 2004). In addition, there is empirical evidence that lay people perceive these terms to be semantically similar (Shaver, Schwartz, Kerson, & O'Connor, 1987). For the purposes of consistency with previous research, and due to lack of research findings that would suggest otherwise, we will continue this convention. Thus, our assumption is that lust/sexual desire is the experience of attraction and interest in a desirable partner as a sexual being (Verhulst & Heiman, 1979). The importance of sexual desire in relationships perhaps is best illustrated by the frequency with which people turn to professional help for low sexual desire or discrepancies between partners in level of sexual desire. Approximately 15% of men and 30% of women report having low sexual desire (Laumann, Paik, & Rosen, 1999). In this study, we attempted to shed light on the processes related to lust between married partners.

A Relational Approach to Lust

The central thesis of this research is that lust is best understood in the context of close romantic relationships. Fisher (1998) has suggested three primary emotion-motivation systems in humans that promote social behavior. She proposed that humans (and other mammals) have evolved these basic systems to direct mating, reproduction, and parenting. These systems are the sex drive, attraction, and attachment. The sex drive evolved to ensure general motivation to reproduce. However, the sex drive exists regardless of the presence of a desirable other. On the other hand, the attraction system (e.g., romantic/passionate love, lust) evolved to motivate individuals to seek desirable others with whom mating could be successful. Finally, the attachment system (e.g., feelings of closeness, positive/negative affect toward the partner) evolved to ensure sufficient stability in relationships so that offspring could be successfully nurtured. These latter two systems are only operative when desirable partners are present. Thus, our conceptual approach is a relational one. A relational approach to lust assumes that partners are in a process of trying to work out how to fulfill their respective drives. We believe these more proximal processes, namely the daily affect concerning people's spouses and daily perceived relationship states, may be a fruitful avenue to advance the understanding of lust in marriage. The study of proximal processes may identify within-day couple patterns that could be connected to longer-term outcomes such as sexual/relational functioning or marital stability. Consequently, in this study we address the interplay of daily relationship-related affect and perceived relationship states (e.g., the attachment system) with married people's feelings of lust (e.g., the attraction system) toward their spouses.

Background

Most research on sexual desire has been clinical in nature and focused on understanding hypoactive sexual desire disorders. Only recently has research examined lust in the context of romantic relationships. Unfortunately, the state of research on lust in romantic relationships is limited (Regan & Berscheid, 1999). Much of the previous work uses biological or evolutionary models (see Baldwin & Baldwin, 1997), which ignore socially-based explanations. Research on the ebb and flow of lust/sexual desire in marriage or other long-term relationships is virtually nonexistent. Our intention in this study was to examine lust in marriage from a dally, more detailed, relational perspective.

Our first purpose was descriptive. The initial step in understanding a phenomenon is to describe it. Much research on sexual desire assumes that it is somewhat stable over time. This view ignores the fact that sexual desire can vary from day to day. Consequently, our first purpose was to examine if lust fluctuates from day to day and whether there are identifiable patterns that represent such variation in lust over time.

The second purpose of this study addressed the intrapersonal aspects of lust. First, we examined the extent to which individuals' dally relationship-related affect (positive and negative emotions toward the partner) was linked with their own daily lust levels. For example, we were interested in whether lust was higher or lower on days when people had positive or negative affect toward their spouses. To our knowledge, there are no within-person studies that have examined the association of people's daily affect and relationship states with their own dally lust levels. On the other hand, at the group (between-persons) level, there is evidence of a link between positive affect and lust. For example, research has shown that romantic love, a predominately positive emotion, is related to higher sexual excitement/desire (Regan, 2000). Consequently, our expectation was that on days when people felt relatively positive about their spouses, their lust would be higher.

The association between negative feelings about the partner and lust is less understood. Schacter (1964) theorized that the valence of affect (positive or negative) is determined through a labeling process that follows general physiological arousal to either positive or negative events. In other words, people can be physiologically aroused by a negative stimulus, but they will label the arousal as positive if they are in the context of an attractive other. Some experimental studies (Allen, Kenrick, Linder, & McCall, 1989) have supported this theoretical perspective, showing that general physiological arousal, whether negative or positive, can lead to increased attraction to someone with whom they expect to interact. More relevant to the present study, negative affect has also been shown to decrease sexual desire when the source of arousal is sexual in nature (Bozman & Beck, 1991). Thus, we tentatively speculated that on days when people experience relatively more negative feelings toward their spouses, their lust would be lower.

Next, we examined whether two important basic relationship states--daily feelings of closeness and equality of power between spouses--would moderate the intrapersonal link between affect and lust. Considerable research has established that most behavior in relationships can be characterized along two basic dimensions, closeness/affiliation and power (Wiggins, Trapnell, & Phillips, 1988). However, we know of no research that examined how these relationship states might moderate daily processes in relationships. We expected that perceptions of daily closeness and equality of power would enhance the link between positive affect and lust. We were less certain whether perceptions of daily closeness and equality of power would moderate the intrapersonal link between negative affect and lust.

The third purpose of this study was to explore how one spouse's daily affect concerning the partner was associated with the level of lust experienced by the other spouse. In other words, we were also interested in interpersonal links. A sizable body of scientific evidence shows that human beings can, and will, "catch" the emotional states (e.g., emotional contagion) of others with whom they interact (Bookwala & Schulz, 1996; Thompson & Bolger, 1999). People in close relationships are especially susceptible to emotional contagion because the partners are invested in each other's physical and emotional welfare and are closely attuned to each other's moods (Hatfield, Cacioppo, & Rapson, 1992). This leads to the expectation that one spouse's affect is likely transmitted to the partner. Thus, consistent with our previous reasoning, we expected that positive affect on the part of one spouse would be positively related to the lust of the other spouse, while one's negative affect would be inversely related to lust in the other spouse.

The fourth purpose of this study was to examine the interpersonal link between spouses' levels of lust and whether this interpersonal link would be enhanced or diminished by daily positive and negative affect toward their partners, perceived closeness to their spouses, and the perceived equality of power between spouses. Consistent with an emotional contagion perspective as discussed earlier, we believed that spouses' lust would be positively correlated and that daily positive affect, closeness, and equality of power would strengthen the link, while daily negative affect toward their partners would diminish the link.

In summary, the purposes of this study were to (a) describe and identify common patterns of daily lust over time; (b) examine how daily positive and negative affect toward the spouse were connected with daily lust within individuals and whether this link was moderated by closeness to the spouse and equality of power; (c) examine how one spouse's own daily positive and negative affect toward the partner were connected to the lust of the other spouse and whether this link was moderated by closeness to the spouse and equality of power between spouses; and (d) examine the association between spouses' daily lust levels and whether this link was moderated by closeness to the spouse, equality of power, and positive and negative affect toward the partner.

METHOD

Sample and Procedures

The sample consisted of 48 spouses (24 married couples) who were in a wait-list control group communication skills program. The ethnicity of participants was primarily Caucasian, with only two Latino couples. One spouse in two different couples did not complete the daily diary questionnaires. Consequently, these two couples were not included in the statistical analyses. The participants were generally young (80% under age 50). The majority of participants had been married for 10 years or less (90% less than 20 years). The average number of children was 1.2 (range = 0-4). The participants were generally well-educated (9% with high school degree, 53% with some college or a baccalaureate degree, and 38% with some graduate school or an advanced degree).

Married couples were recruited into an 8-week marital enhancement program through announcements in newspapers and flyers posted in high-traffic areas of a Southwestern U.S. city and major state university. Couples (n = 369) responded to these announcements, but only 154 couples appeared at the initial informational meeting. After we screened couples with significant relationship problems and unwillingness to participate in the research, 68 couples remained. Twenty-four of these couples were placed on a waiting list, and these 24 couples were the sample for this study. These participants were asked to complete, independently from their spouses, a daily questionnaire pertaining to the previous 24 hours. The questionnaires were completed for 56 consecutive days.

Daily Measures

Ten items tapped spouses' reports of their daily levels of lust, positive affect (contentment, satisfaction, affection, happiness) and negative affect (anger, anxiety, irritation, neglect, sadness) toward their partners over the preceding 24 hours. Participants were given this list of emotions and relational states and prompted with the following: "Today while I was with my spouse, I felt ..." (see above list). Participants responded on a 5-point response scale, from 1 (not at all) to 5 (very strongly). The single-item measure of lust has demonstrated discriminant validity in a similar study (Hunt et al., 2003). In that study, lust was more highly positively related to relationship satisfaction for men than for women, supporting previous research showing the higher salience of sexual dimensions in influencing the evaluation of relationships for men versus women (Baldwin & Baldwin, 1997). The items tapping positive and negative affect were summed to create composite scales for both positive (alpha = .97) and negative (alpha = .90) affect. Two additional items measured feelings of closeness and power relative to the spouse (e.g., equality of power) during the preceding 24 hours. These two variables were assessed using stick figure representations. For the measure of closeness, five vertical pairs of stick figures representing the two partners were shown. The first pair of stick figures was the furthest apart and represented the least amount of closeness (scored as 1). The spaces between the remaining four sets of figures were systematically varied to be closer to each other, with the final set being almost totally overlapping and represented the most closeness (scored as 5). This closeness measure exhibited convergent validity in a previous study of married/committed couples (Ridley & Feldman, 2003). The measure was significantly correlated with relationship satisfaction in that study.

The measure of equality of power was also measured with stick figure representations. For the measure of equality of power, five vertical pairs of stick figures representing the two partners were shown. The individual figures in each pair were varied in height so that one figure was taller than the other, representing a difference in power between partners. Each successive pair showed an increased discrepancy in height between the two figures. The first pair showed the largest discrepancy in height (scored as 1) and represented the least equality of power between partners. The final set showed two figures of equal height and represented equality of power between partners (scored as 5). This equality of power measure exhibited criterion-related validity in a previous study of married/committed couples (Ridley & Feldman, 2003). The measure was significantly correlated (r = .31) with relationship satisfaction in that study, such that the more equality of power that was perceived, the higher the marital satisfaction.

Although power has typically been measured as a stable characteristic of relationships, recent advances in statistical techniques (specifically, the development of software that enables the analysis of multi-level data) have allowed researchers to explore the extent to which reports of power vary from day to day in committed relationships. It is interesting to note that for the husbands in our sample, 46% of the variation in reports of equality of power was due to differences between persons and 54% was due to within-person variation. Therefore, husbands varied slightly more from themselves on their reports of equality of power during this study than they did from each other. This finding suggests that although there are individuals who report being higher or lower overall in equality of power from others in the sample, over half of the variation in equality of power was due to situational factors that influence reports of equality of power on a daily basis. Similarly, 43% of the variation in wives' reports of equality of power was due to differences between persons, and 57% was due to within-person variation.

Pattern Properties

In order to examine patterns of lust over the 56 days of this study, we identified two main dimensions on which individuals differed from each other in our sample. A visual inspection by the research team of participants' ratings of lust toward their partners over the 56 days resulted in the creation of five variables which tapped mean level of lust, as well as the extent to which individuals varied from day to day on lust toward their partners. The first variable tapped whether participants were generally high, moderate, or low in lust over this period. This variable was calculated by determining the average daily lust level over the 56 days for each participant in the study. The other four variables were created to reflect the percentage of days on which change occurred and the magnitude of the change from one day to the next for each participant. The four scores were calculated as follows: the percentage of the 56 days when (a) the lust score changed by one scale point from one day to the next; (b) the lust score changed by two scale points from one day to the next; (c) the lust score changed by three scale points from one day to the next; and (d) the lust score changed by four scale points from one day to the next.

Data Analysis

For the purpose of identifying common patterns of change in lust over the 56 days, we performed a hierarchical centroid cluster analysis (SAS, 1990). The five variables outlined above were used to create a small number of clusters. Each cluster represented participants' lust patterns that were similar to each other in average daily level of lust and the degree of fluctuation in lust from day to day. In hierarchical centroid cluster analysis, an agglomerative procedure begins with each observation (pattern in this case) being a cluster. Next, clusters are formed based on squared Euclidian distance measures between group centroids. This yields various groupings of observations/patterns that become clusters, meaning that the patterns contained in the cluster share similar characteristics (e.g., average level of lust and degree of fluctuation in lust). The number of clusters retained for this study was selected using the criteria of parsimony and interpretability.

For the remaining purposes of the study, hierarchical linear modeling analyses (HLM; Bryk & Raudenbusch, 1992) were performed to examine within-person correlations over time in these same variables. "This within [person] analysis is powerful because it controls for numerous extraneous, often immeasurable sources of variance linked to the person, from response tendencies to temperaments" (Larson & Almeida, 1999, p. 9). Within-person analyses (i.e., HLM) examined the association of daily affect, equality of power between spouses, and closeness with daily lust. Cross-spouse analyses (e.g., interpersonal) examined associations between the husbands' and wives' variables. Additional HLM analyses explored how perceived closeness and equality of relational power on a given day might moderate the association of daily positive and negative affect with their own and their spouses' daily lust.

HLM is a statistical method that allows simultaneous estimation of both (a) a separate within-person model of regression slopes and intercepts for each respondent and (b) a between-person model in which the within-person slopes and intercepts are treated as dependent variables regressed on person-level predictor variables. It is important to point out that this estimation procedure takes into consideration the amount of data available from each person, so that missing data on some occasions are taken into account by giving more weight to persons with complete data than those with some missing data (Dempster, Laird, & Rubin, 1977). Based on this feature of the analysis method, data analysis will work with respondent records even if they only complete a portion of the 56 diary days. Missing days in the middle of the series, such as when a respondent completed questionnaires on Days 1-42 and Days 47-56 but missed Days 43-46, can be handled in the same way. Thus, instead of deleting all of the respondent's data due to a missed questionnaire, this approach has the advantage of using all available data from a given respondent. For illustrative purposes, the simple form of an HLM analysis can be conceived of as two separate models: one a within-person model (Level 1) and the other a between-person model (Level 2). To examine the relationship of a person's lust and his or her own negativity, we fit a within-person model that assessed the daily covariation of levels of lust and negativity. This model can be expressed as:

Level 1: [LUST.sub.ti] = [[beta].sub.0i] + [[beta].sub.1i] NEGATIVE [AFFECT.sub.ti] + [e.sub.ti] (1)

where [LUST.sub.ti] is the average level of lust of on [Day.sub.t] of Personi, [[beta].sub.0i] is the intercept indicating [Person.sub.i's] level of lust when all predictors are 0, NEGATIVE [AFFECT.sub.ti] indicates the level of negative affect experienced on [Day.sub.t] by Personi, [[beta].sub.1i] is the slope indicating the lust reactivity of [Person.sub.i] to their own negativity, and [e.sub.ti] is the random component or error on [Day.sub.t] associated with [Person.sub.i]. In order to estimate average effects for the entire sample, the intercepts and slopes of the Level 1 within-person model become the outcomes for the Level 2 between-person equations, as follows:

Level 2: [[beta].sub.0i] = [v.sub.00] + [u.sub.0i] (2)

[[beta].sub.1i] = [v.sub.10] + [u.sub.1i] (3)

Equation 2 shows that Person i's average lust score across the diary days when all Level 1 predictors are zero ([[beta].sub.0i]) is a function of the intercept for the entire sample--the grand mean of the sample--and a random component or error ([u.sub.0i]). Likewise, the equation shows that Person i's reactivity slopes ([[beta].sub.1i]) is a function of the average reactivity of the entire sample (grand mean of reactivity) and a random component or error ([u.sub.1i]).

These models were expanded to include within-person moderators of the lust-negative affect association. At level one the model is expressed as:

Level 1: [LUST.sub.ti]] = [[beta].sub.0i] + [[beta].sub.1i] NEGATIVE [AFFECT.sub.ti] + [[beta].sub.2i] [Moderator.sub.ti] [[beta].sub.3i] NEGATIVE [AFFECT.sub.ti] X Moderator [e.sub.ti] (4)

Level 2: [[beta].sub.0i] = [v.sub.00] + [u.sub.0i] (5)

[[beta].sub.1i] = [v.sub.10] + [u.sub.1i] (6)

[[beta].sub.2i] = [v.sub.20] + [u.sub.2i] (7)

[[beta].sub.3i] = [v.sub.30] + [u.sub.3i] (8)

The moderator coefficient is [[beta].sub.3i] in equation 8.

RESULTS

Cluster Analysis

A cluster analysis using the mean level of lust and the four measures of day-to-day fluctuation over 56 days revealed four interpretable patterns of change in marital lust. The results demonstrated support for our contention that the pattern of lust in married individuals varies widely. Prototypical examples of these four patterns are shown in Figures 1-4. Twelve participants (7 men and 5 women) were identified as having stably low levels of lust. Individuals in this pattern reported little daily variation in lust, generally reporting that they did not feel any lust toward their spouses. In fact, those in this group reported feeling no lust toward their spouses on approximately 90% of the 56 days. Thus, the average level of lust for individuals in this pattern was very low (M = 1.12). Table 1 presents demographic information on each group. However, the small number of participants in each group did not allow for statistical comparison.

[FIGURES 1-4 OMITTED]

Six participants (two men and four women) in the second pattern were identified as having slightly fluctuating, low levels of lust. For the majority of the days, this group reported either no lust (48% of days) or slight lust (41% of days) toward their spouses. Therefore, most fluctuations in daily lust levels were one-unit changes. The average level of lust for individuals in this pattern was also relatively low (M = 1.65).

Fifteen participants (9 men and 6 women) were identified as having moderately fluctuating levels of lust. Individuals with this pattern of lust exhibited more daily variation than the previous two patterns, with almost two thirds of the daily ratings ranging from feelings of slight lust (32% of days) to moderate lust (32% of days) toward their spouses. Individuals in this pattern had fewer consecutive days where they felt the same level of lust and were more likely to have two-unit changes in lust from one day to the next than the previous two patterns. The average level of lust for this group was higher than those in the three other patterns (M = 2.56).

Twelve participants (5 men and 7 women) were identified as having highly fluctuating levels of lust. Although this group spent their highest percentage of time feeling no lust toward their spouses (40% of days), they were best characterized as having frequent fluctuations between feeling no lust to feeling strong or very strong lust toward their spouses. This group was more likely to have three and four units of change in lust from day to day. These frequent, relatively large fluctuations explain why the average level of lust was slightly lower for this pattern (M = 2.15) than for those in the moderately-fluctuating lust pattern.

Intrapersonal Analyses

The next analyses explored (a) the extent to which people's daily affect toward their spouses was connected to their own daily lust and (b) whether daily perceptions of relationship states moderated the link between people's daily affect and their own lust. These analyses used 2,097 days of daily data out of a possible 2,688 total days. Consequently, the average participant completed the daily questionnaires on 78% of the 56 days, which is an acceptable compliance rate. Refer to the methods section for an explanation of how these analyses deal with missing data.

Intrapersonal affect-lust link. On days when individuals had relatively higher positive affect toward their spouses, they were also likely to experience higher lust (see Table 2). Conversely, on days when individuals had relatively higher negative affect toward their spouses, they were also likely to experience lower lust.

Moderation of intrapersonal affect-lust link. Some perceived relational states moderated the link between people's affect toward their spouses and their own levels of lust (see Table 2). We explored moderation effects by graphing equations for those high and low (one standard deviation above and below the mean) on the moderating variables. The link between individuals' positive affect toward their spouses and their own lust was moderated by their own closeness to their spouses. The interaction plots showed that on days when people reported relatively higher closeness to their spouses, the link between their positive affect toward their partners and their own lust was higher. In addition, husbands' perceived closeness moderated the intrapersonal association of wives' positive affect with their own lust. The interaction plots showed that the association of wives' own positive affect with their own lust was higher among husbands who were high versus low on perceived closeness. Wives' perceived closeness did not moderate the intrapersonal positive affect-lust link for husbands.

As above, the link between individuals' negative affect toward the spouses and their own lust was moderated by closeness to their spouses. However, the interaction plots revealed unexpected patterns. On days when husbands and wives perceived relatively higher closeness to their spouses, the negative associations between their own negative affect toward their partners and their own lust was strengthened, not lessened as expected. It is important to note that this association was only marginally significant for wives. The interaction plots also showed that the negative associations of wives' own negative affect with their own lust was strengthened, not lessened, on days when their husbands perceived high closeness. However, wives' perceived closeness did not moderate the intrapersonal negative affect-lust link for husbands.

Perceptions of equality of power between spouses moderated the intrapersonal association of positive and negative affect with lust in several instances. The interaction plots showed that on days when either husbands or wives perceived relatively higher equality of power between spouses, the association of wives' positive affect toward their partners and their own lust was strengthened. However, for husbands, the intrapersonal positive affect-lust association was not moderated by either their own or their wives' perceptions of equality of power.

The interaction plots showed that on days when either husbands or wives perceived relatively higher equality of power, the intrapersonal link between wives' negative affect toward their husbands and their own lust was higher. Similarly, the intrapersonal link between husbands' negative affect toward their spouses and their own lust was higher on days when their wives perceived relatively higher equality of power. However, the intrapersonal link between husbands' negative affect toward their spouses and their own lust was not moderated by their own perceptions of equality of power.

Interpersonal Analyses

The next analyses explored (a) the extent to which spouses' daily affect toward their partners and daily perceptions of relationship states were connected to their partners' daily lust; (b) whether daily perceptions of relationship states moderated the tie between people's daily affect and their partners' lust; (c) the extent to which spouses' daily lust levels were connected, and (d) the extent to which daily perceptions of relationship states and affect toward the partner moderated the interpersonal link between partners' levels of lust.

Interpersonal affect-lust link. Partners' positive and negative affect was associated with the other spouses' lust (see Table 3). On days when partners had relatively higher positive affect toward their spouses, their spouses were also likely to experience higher lust. In addition, on days when partners had relatively higher negative affect toward their spouses, their spouses experienced lower lust.

Moderation of the interpersonal affect-lust link. In only two instances were these interpersonal affect-lust associations moderated by perceived relational states (see Table 3). The interaction plots showed that on days when wives felt relatively higher closeness to their husbands, the association of husbands' positive affect and wives' lust was strengthened. Similarly, on days when wives felt relatively higher equality of power between spouses, the association of husbands' positive affect and wives' lust was higher. Relationship states did not moderate the interpersonal link of husbands' negative affect with their wives' lust, nor did they moderate the interpersonal link between wives' positive and negative affect and their husbands' lust.

Interpersonal lust-lust link. The findings showed a significant positive association between husbands' and wives' lust (see Table 4). On days when one spouse had relatively higher lust, the other spouse was also likely to have higher lust.

Moderation of the interpersonal lust-lust link. One set of these analyses used husbands' lust (predictor) to predict wives' lust (outcome) and the other used wives' lust (predictor) to predict husbands' lust (outcome). Regardless of which partner's lust was used as the predictor, the other partner's feelings of closeness strengthened the association between partners' lust. The interaction plots showed that when wives or husbands reported relatively higher positive affect toward their partners, the link between partners' lust was strengthened. When husbands and wives' lust were the predictors, their own positive affect did not moderate the lust-lust association.

The findings concerning negative affect were not as predicted. Contrary to our expectations, the interaction plots showed that wives' negative affect moderated the link between husbands' (predictor) and wives' lust, reversing the direction of the original relationship. Specifically, on days when wives reported relatively low negative affect, the positive relationship between husbands' lust and wives' lust was weakened to a non-significant relationship. However, on days when wives reported relatively higher levels of negative affect toward their husbands, the association between husbands' lust and wives' lust became negative, while the association between partners' lust was slightly positive when wives reported relatively low negative affect.

Similar moderation by negative affect occurred when wives' lust was the predictor of husbands' lust. The interaction plots showed that on days when husbands reported relatively low negative affect towards their wives, there was not a significant relationship between wives' and husbands' lust; but on days when husbands reported relatively higher negative affect, the association between wives' lust and husbands' lust was reversed to a negative association. Again, it is notable that partners' own negative affect did not moderate the link between their own lust and the lust of the spouses.

Perceived relationship states (closeness and equality of power) also moderated the interpersonal link between spouses' lust. In the link between partners' lust, whether husbands or wives were the predictors of the other partners' lust, the spouse's closeness strengthened the lust-lust link. In addition, wives' equality of power strengthened the link between husbands' and wives' lust. Again, the findings showed that when spouses' lust is the predictor, their own closeness does not moderate the lust-lust link. In addition, the interaction plots showed that on days when wives reported higher equality of power, the interpersonal association of husbands' lust with wives' lust was significantly higher. Husbands' perceptions of equal power did not moderate the interpersonal link between their own lust and wives' lust. Finally, neither wives' nor husbands' reports of equality of power were significant moderators of the link between wives' and husbands' lust.

DISCUSSION

In this study, we examined variations in lust over time and explored interconnections between married individuals' daily affect toward their partners, daily experiences of relationship states, and daily lust. The following discussion focuses on the interpretation and implications of our findings and suggestions for further research.

Patterns of Variation

We found considerable variation in patterns of lust over time in married couples. The small number of participants in each pattern type prohibited formal comparative statistical analyses; however, inspection of the pattern properties suggested substantial similarity on both demographic and social-psychological properties. An exception was that participants in the stable low-level group were married somewhat longer than those in the other patterns. This would be expected, as sexual desire decreases as length of marriage increases. No previous research exists to assist in interpreting these patterns. We speculate that married couples negotiate diverse patterns of sexual interaction that support satisfactory affective and relational states, even when lust is relatively low and stable. Future work using larger samples will allow identification of factors that contribute to variation in lust over time in married couples.

Intrapersonal Processes

As predicted, we found a link between partners' own daily affect toward partners and their own lust. We could not identify a mechanism that accounted for this link. It is plausible that affect toward a partner primes people to think negatively or positively about sex with the partner, thus leading to higher or lower levels of lust. Second, relational affect and sexual factors may be connected via neurohormonal processes. The relational and sexual aspects of relationships may be partially governed by the same neurohormone, oxytocin (Diamond, 2003). Future research should explore such physiological mechanisms.

The intrapersonal link between affect and lust was also moderated by relationship states. The intrapersonal positive affect-lust link between spouses was higher on days when they felt relatively closer to each other. From a cognitive perspective, an atmosphere of closeness may result in heightened accessibility of memory for previous positive events, thus increasing the positive affect-lust link (Ellis & Ashbrook, 1989). On the other hand, we did not expect to find that relational states generally strengthened the intrapersonal link of negative affect and lust. This counterintuitive finding may be due to the difference between global and specific evaluations in relationships (Neff & Karney, 2002). At the daily level, it may be harder for satisfied spouses to justify partners' negative behaviors because there are a limited number of potential explanations for those behaviors on a particular day. On the other hand, when married people are asked to provide global, summative evaluations of their partners, they have numerous experiences with their partners that allow them to rationalize their partners' behavior. It may be fruitful in future research to explore directly how sexual processes differ at the daily versus global levels.

Interpersonal Processes

We expected an interpersonal association between affect and lust, and our expectations were fully supported. The mechanisms by which one spouse's affect may influence the other spouse's lust are largely unexplored. We propose two possible mechanisms. First, the interpersonal link of affect and lust may be due to emotional contagion (Hatfield et al., 1992), as we discussed earlier. Second, people's expectancies about their partners' behaviors may account for the interpersonal link of affect and lust. However, there is little work that examines expectancies at the daily level of analysis in relationship research. For example, if a person experiences positive affect toward the spouse and has positive expectations of the spouse, he or she may communicate positivity to the spouse, thus eliciting a positive reaction (e.g., increased lust).

As expected, we found a positive interpersonal link between couples' individual lust levels. Again, this interpersonal link might be due to emotional contagion or expectancies about the partner. However, the reversal of the valence of the lust-lust link (both husband-wife and wife-husband link) by high negative affect of the other partner is potentially more important. This reversal may be due to contrast effects (Geers & Lassiter, 2005). Contrast effects occur when people are aware that a stimulus is inconsistent with some affective expectation, thus producing a negative reaction. For example, when husbands express lust toward their wives who are experiencing high negative affect toward them, the wives react negatively, thus reducing their lust.

Gender Differences

We did not predict gender differences in the processes examined in this study. The relational states of both husbands and wives moderated the intrapersonal associations of positive and negative affect with lust for women to a greater extent than men. This gender difference may be explained by past research showing that women are better at reading emotional expressions of others (Haviland & Malatesta, 1981) and are more aware of their own emotions than men (Hall, 1984). Thus, wives may be more sensitized than husbands to their partners' relationship states, thus moderating the intrapersonal link of relational affect and lust. Future studies should directly address gender differences in these sexual processes.

Limitations

This study has several strengths, particularly the use of both marital partners and the use of a daily diary approach to examine daily processes. However, as with any study, this one has limitations. First, the generalizability of the findings of this study is limited. The couples in this study were self-selected, had higher-than-average educational attainment, and were predominately Caucasian. Second, the manner in which these data were collected did not allow for causal analysis. In order to examine issues of causality between constructs, it would be necessary to obtain diary data at different points during the day, thus establishing the temporal ordering of relationship events and reports. Third, in this study, as with many self-report studies, common method variance may have been a problem.

The Emerging Story

This study took the initial step in developing a story about sexuality in relationships. A story of this type should include a description of the ebb and flow of lust over time. Since relationships are dynamic and emotional, the ebb and flow of lust would be superimposed onto the ebb and flow of positive and negative emotions occurring daily in the relationship. The person listening to the story would then want to know if the way marital partners treat each other makes a difference, and the meaning of this treatment for sexual and emotional expression.

Based on the results of this study, the story is beginning to evolve. First, there are probably several stories, not just one, to be told. It is striking how different couples can be in experiencing lust in the relationship, from a pattern of no lust to large fluctuations from day to day. Each of the lust patterns identified might have unique stories about the role of emotions in the way couples negotiate sexuality in the relationship. Concerns for power and closeness will likely be implicated in these negotiations, because these are indications of how partners treat each other. Unfortunately, the results presented here leave the story only partially told. For example, a story needs to describe how each lust pattern links to feelings, closeness, and power. We could not complete this part of the story, but our results are suggestive of some aspects of a more complete story. We now know that emotions, both positive and negative, are important in understanding how lust is experienced inside the person and in the relationship. We now know that positive emotions enhance lust and negative emotions dampen it. The evolving story suggests that when partners treat each other well (i.e., are close and equal in power), the link between emotions and lust becomes stronger. The story with negative emotions and lust is probably more complex than the story of positive emotions and lust. A listener to this story would like to know more about the complexities of how emotions (especially negative emotions), closeness, power, and sexual behavior are played out within each sexual lust pattern; so would we.

Manuscript accepted October 13, 2005

REFERENCES

Allen, J. B., Kenrick, D. T., Linder, D. E., & McCall, M. A. (1989). Arousal and attraction: A response-facilitation alternative to misattribution and negative-reinforcement models. Journal of Personality & Social Psychology, 57, 261-270.

Baldwin, J. D., & Baldwin, J. I. (1997). Gender differences in sexual interest. Archives of Sexual Behavior, 26, 181-210.

Bookwala, J., & Schulz, R. (1996). Spousal similarity in subjective well-being: The Cardiovascular Health Study. Psychology and Aging, 11, 582-590.

Bozman, A. W., & Beck, J. G. (1991). Covariation of sexual desire and sexual arousal: The effects of anger and anxiety. Archives of Sexual Behavior, 20, 47-60.

Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models: Applications and data analysis methods. Thousand Oaks, CA: Sage Publications.

Dempster, A., Laird, N., & Rubin, D. (1977). Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, Series B, 39, 1-8.

Diamond, L. M. (2003). What does sexual orientation orient? A bio-behavioral model distinguishing romantic love and sexual desire. Psychological Review, 110, 173-192.

Ellis, H. C. & Ashbrook, P. W. (1989). The "state" of mood and memory research: A selective review. Journal of Social Behavior and Personality, 10, 32-38.

Fischer, H. E. (1998). Lust, attraction, and attachment in mammalian reproduction. Human Nature, 9(1), 23-52.

Gargle, P. B., Herrera, D. E., Hunt, J. E., Ogolsky, B., Collins, D., Ridley, C. A., et al. (2004, November). Lust, passion, and sexual desire: Are these terms synonymous in lesbian intimate relationships? Symposium presented at the National Council on Family Relations, Orlando, FL.

Geers, A. L., & Lassiter, G. D. (2005). Affective assimilation and contrast: Effects of expectations and prior stimulus exposure. Basic & Applied Social Psychology, 27, 143-154.

Hall, J. (1984). Non-verbal sex differences: Communication accuracy and expressive style. Baltimore, MD: Johns Hopkins University Press.

Hatfield, E., Cacioppo, J. T., & Rapson, R. L. (1992). Primitive emotional contagion. In M. Clark (Ed.), Emotion and social behavior (pp. 151-177). Thousand Oaks, CA: Sage Publications.

Haviland, J., & Malatesta, C. (1981). The development of sex differences in non-verbal signals: Fallacies, facts, and fantasy. In C. Mayo & N. Henley (Eds.), Gender and non-verbal behavior (pp. 183-208). New York: Springer-Verlag.

Hunt, J. E., Collins, D. M., Reesing, A. L., Ridley, C. A., Cate, R. M., Banks, S., et al. (2003). The semantics of attraction: The need for semantic clarity in the study of intimate relationships. Symposium paper for the National Council on Family Relations, Vancouver, B.C., Canada.

Larson, R. W., & Almeida, D. M. (1999). Emotional transmission in the daily lives of families: A new paradigm for studying family process. Journal of Marriage & Family, 61, 5-20.

Laumann. E. O., Paik, A., & Rosen, R. C. (1999). Sexual dysfunction in the United States: Prevalence and predictors. Journal of the American Medical Association, 281, 537-544.

Neff, L. A., & Karney, B. R. (2002). Judgements of a relationship partner: Specific accuracy but global enhancement. Journal of Personality, 70, 1,079-1,112.

Regan, P. C. (2000). The role of sexual desire and sexual activity in dating relationships. Social Behavior & Personality, 28, 51-59.

Regan, P., & Berscheid. E. (1999). Lust. Thousand Oaks, CA: Sage.

Ridley, C. A., & Feldman, C. M. (2003). Female domestic violence toward male partners: Exploring conflict responses and outcomes. Journal of Family Violence, 18, 157-171.

SAS. (1990). SAS/STAT User's Guide (Version 6; Vol. 1, 4th ed.) [Computer software manual.] Cary, NC: Author.

Schachter, S. (1964). The interaction of cognitive and physiological determinants of emotional state. In L. Berkowitz (Ed.), Advances in Experimental Social Psychology, pp. 49-79. New York: Academic Press.

Shaver, P., Schwartz, J., Kirson, D., & O'Connor, C. (1987). Emotion knowledge: Further exploration of a prototype approach. Journal of Personality & Social Psychology, 52, 1,061-1,086.

Thompson, A., & Bolger, N. (1999). Emotional transmission in couples under stress. Journal of Marriage & the Family, 61, 38-18.

Verhulst, J., & Heiman, J. R. (1979). An interactional approach to sexual dysfunctions. American Journal of Family Therapy, 7(4), 19-36.

Wiggins, J. S., Trapnell, P., & Phillips, N. (1988). Psychometric and geometric characteristics of the revised interpersonal scales (IAS-R). Multivariate Behavioral Research, 23, 517-530.

Carl A. Ridley, Rodney M. Cate, Dawn M. Collins, Amy L. Reesing, Ana A. Lucero, Michael S. Gilson, and David M. Almeida

University of Arizona

Address correspondence to Carl A. Ridley, University of Arizona, Division of Family Studies and Human Development, 1110 E. South Campus Drive, Tucson, AZ, 85721; e-mail: [e-mail].

Table 1. Means and Standard Deviations of Four Marital Lust Patterns

                                 Slightly
                     Stable     Fluctuating   Moderately      Highly
Variables           Low-Level    Low-Level    Fluctuating   Fluctuating

N (men, women)         12           6            15            12
                       (7, 5)      (2, 4)        (9, 6)        (5, 7)
Lust                    1.12        1.65          2.56          2.15
                        (.12)       (.17)         (.36)         (.25)
Age (in years)         40.91       34.40         41.07         38.17
                      (14.48)      (6.58)       (10.82)        (8.15)
Marital Length
  (in years)           11.75        4.67          4.92          5.25
                      (14.83)      (6.42)        (5.08)        (6.19)
# of Children
  Living at Home        1.27        2.00           .87          1.25
                       (1.19)      (1.23)        (1.19)        (1.49)
Positive Affect        12.02       10.63         12.96         12.01
                       (2.61)      (1.93)        (1.95)        (2.01)
Negative Affect         5.47        6.09          5.96          6.70
                       (1.24)      (1.09)        (1.52)        (1.39)
Closeness               3.02        2.63          3.21          2.85
                       (1.13)       (.86)         (.78)         (.51)
Equality of             4.58        4.50          4.67          4.50
  Power                 (.67)       (.55)         (.62)         (.67)

Table 2. HLM Regression Coefficients For Own Affect
Predicting Own Lust (Intrapersonal) and
Moderators

Association                   Moderators

Positive                          Own             Spouse
Affect [right arrow] Lust      Closeness         Closeness
b                                  b                 b

H [right arrow] H
.50 ****                       .08 ***            .05

W [right arrow] W
.52 ****                       .19 ****           .10 ***

Association                   Moderators

Negative                          Own             Spouse
Affect [right arrow] Lust      Closeness         Closeness
b                                  b                 b

H [right arrow] H
-.34 ****                   -.07 *               -.05

W [right arrow] W
-.22 ****                   -.06 ([dagger])      -.12 ****

Association                   Moderators

                                  Own             Spouse
Positive                       Equality          Equality
Affect [right arrow] Lust      of Power          of Power
b                                  b                 b

H [right arrow] H
.50 ****                       .03            .04

W [right arrow] W
.52 ****                       .12 ****       .05 ([dagger])

Association                   Moderators

                                  Own             Spouse
Negative                       Equality          Equality
Affect [right arrow] Lust      of Power          of Power
b                                  b                 b

H [right arrow] H
-.34 ****                      -.03                .04

W [right arrow] W
-.22 ****                      -.10 ****          -.07 *

Note. H = husbands; W = wives.

([dagger] p < .10

* p < .05

** p < .01

*** p < .005

**** p < .001

Table 3. HLM Regression Coefficients for Own Positive
and Negative Affect Predicting Spouses' Lust
(Interpersonal) and Moderators

Association                          Moderators

Positive                          Own             Spouse
Affect [right arrow] Lust      Closeness         Closeness
b                                  b                 b

H [right arrow] W
.25 ****                          .05             .09 ***

W [right arrow] H
.25 ****                          .03             .01

Association                          Moderators

Negative                          Own             Spouse
Affect [right arrow] Lust      Closeness         Closeness
b                                  b                 b

H [right arrow] W
-.11 ***                         -.03              -.05

W [right arrow] H
-.15 ***                         -.02              -.01

Association                          Moderators

                                  Own             Spouse
Positive                       Equality          Equality
Affect [right arrow] Lust      of Power          of Power
b                                  b                 b

H [right arrow] W
.25 ****                          .02             .08 **

W [right arrow] H
.25 ****                          .01             .01

Association                          Moderators

                                  Own             Spouse
Negative                       Equality          Equality
Affect [right arrow] Lust      of Power          of Power
b                                  b                 b

H [right arrow] W
-.11 ***                         -.01               .02

W [right arrow] H
-.15 ***                         -.02              -.02

Note. H = husbands; W = wives.

([dagger]) p < .10

* p < .05

** p < .01

*** p < .005

**** p < .001

Table 4. HLM Regression Coefficients for Association of
Own Lust With Other Spouses' Lust
(Interpersonal) and Moderators

Association                          Moderators

                                  Own             Spouse
                               Positive          Positive
Lust [right arrow] Lust         Affect            Affect
b                                  b                 b

H [right arrow] W
.29 ****                          .03            .12 ****

W [right arrow] H
.29 ****                          .02            .11 ****

Association                          Moderators

                                  Own             Spouse
Lust [right arrow] Lust       Closeness         Closeness
b                                  b                 b

H [right arrow] W
-.29 ****                        -.04             -.10 **

W [right arrow] H
-.29 ****                        -.01             -.08 **

Association                          Moderators

                                  Own             Spouse
                               Negative          Negative
Lust [right arrow] Lust         Affect            Affect
b                                  b                 b

H [right arrow] W
.29 ****                          .03             .09 ***

W [right arrow] H
.29 ****                          .02             .09 ***

Association                          Moderators

                                  Own             Spouse
                               Equality          Equality
Lust [right arrow] Lust        of Power          of Power
b                                  b                 b

H [right arrow] W
-.29 ****                        -.01               .10 *

W [right arrow] H
-.29 ****                        -.05              -.04

Note. H = husbands; W = wives.

([dagger]) p < .10

* p < .05

** p < .01

*** p < .005

**** p < .001