Abstract
This analysis of previously collected data examined four fitness-relevant issues for their possible role in marital conflict. These were sex, finances, division of labor, and raising children, selected in light of their pertinence to sex differences in reproductive strategies. Over 2,000 couples in five diverse cultures were studied. Marital conflict was assessed by the Problems with Partner scale, which was previously shown to demonstrate measurement invariance across cultures and genders. All four issues were significantly related to perceived marital problems in almost all cases. Thus, conflict tended to arise around issues relevant to reproductive strategies. A few cultural idiosyncrasies emerged and are discussed. In all cultures, wives reported more problems than husbands. Another important issue was kindness. The results suggest that a key factor in marital success or failure may be kindness necessary to sustain this prolonged and intimate relationship of cooperation for raising one's offspring.
Introduction
Evolutionary psychologists tend to agree that marriage enhances reproductive success by recruiting the father to aid in parental care and family protection (Chapais, 2010). Around the world, most adults seek to marry and most of them do marry. This suggests that marriage, or at least biparentalism, constitutes an evolutionarily stable strategy that generally maximizes fitness (for a discussion of marriage from an evolutionary standpoint, see Weisfeld and Weisfeld, 2014).
Studying marital conflict is important because conflict may lead to dissatisfaction and instability in this potentially long-term, beneficial relationship. Even if the marriage remains intact but is filled with conflict that detracts from nurturance of the children, the couple's reproductive fitness may suffer. Although the institution of marriage may provide some feeling of security so that married couples may experience less conflict relative to dating couples (Buss, 1989a) and cohabiting couples (Kenney and McClanahan, 2006), all married couples probably experience some conflict.
Sources of conflict may share some commonalities across cultural groups. Buss's (1989b) cross-cultural research on attributes sought in a mate revealed a universal desire in both sexes for kindness, dependability, and understanding, so a dearth of these attributes might often lead to marital conflict. Sources of conflict may also vary across cultures. Buss et al. (1990) found that culture accounted for 37% of the variance (the highest amount) in the emphasis placed on premarital chastity in mate preferences, indicating potential cultural differences in the role of this factor in marital conflict.
The vast majority of research on marital conflict has concerned various conflict styles and their consequences. Some of this research is cross-cultural (e.g., Cheng, 2010; Feldman, 2010; Fincham, 1999), with different conflict styles being associated with differences in marital satisfaction (Greeff, DeBruyne, and Abraham, 2000), marital duration (Carstensen, 1995), future conflict frequency (Sullivan, 2010), relations with children (Shek, 1998), and collectivism vs. individualism (Holt, 2005). In this paper we will not delve further into this vast literature, but will report on a study involving some possible sources of conflict as perceived by husbands and wives from five highly diverse cultures. In doing so, we aim to shed light on possible evolutionary and cultural aspects of marital conflict.
Given the vast array of disparate findings in the literature on marital conflict, we look to evolutionary theory for some functional explanations that will provide order to those findings. Darwin's (1871) sexual selection theory—and its elaboration, parental investment theory (Trivers, 1972)—may be particularly relevant here. Evolutionary psychologists point out that conflict between husbands and wives may be expected given their contrasting evolved reproductive strategies (Buss, 2003), with males generally exhibiting greater eagerness for sex, and females being choosier.
Thus, conflicts of interest occur over initiation of sexual activity, with the female often rejecting male overtures. Males may be more eager for sex with their partner in order to reduce the risk of her impregnation by a rival, or as a side effect of a low threshold for sex in order to maximize fertilization of additional females. Sexual coercion constitutes another, possibly related source of conflict, usually with the male coercing the female (Miller, 2000; Struckman-Johnson, Struckman-Johnson, and Anderson, 2003). Conflict has also been observed to occur over suspected or acknowledged infidelity, and over mate guarding, or jealousy, as in previous work using the current data set (Dillon et al., 2014). In one cross-cultural survey, marital stability varied greatly across cultures that differed in amount of male parental investment and in degree of male-male competition, suggesting that conflict frequently occurs over the allocation of reproductive effort (Quinlan, 2007). Interestingly, among the Tsimane more conflict occurred over the husband's infidelity and his diversion of resources from his marriage than over the wife's infidelity (Stieglitz, Gurven, Kaplan, and Winking, 2012).
Another factor affecting male-female conflict is the sex ratio. Where the operational sex ratio is low and marriageable males are scarce, as in the contemporary African American community, a low marriage rate, male promiscuity, and low paternal care are favored (Anderson, 1999; Wilson, 1987). Conflict may occur over any of these issues. But even with perfect, lifelong monogamy (no extra-pair sex) and a 1:1 operational sex ratio, each parent benefits from shifting care to the other (Kokko and Jennions, 2008).
Conflict may arise over the number and spacing of children in some cultures (Mason and Taj, 1987) but not in others (Mason and Smith, 2000). When spouses do disagree on family planning, husbands usually desire more children than wives, consistent with women providing more childcare. This sex difference can also be explained by noting that if a mateship dissolves and the woman proceeds to reproduce with other men, her first mate gains no reproductive success from these latter births. So, a man is more interested in a woman's short-term reproduction with him than she is.
Given these centrifugal forces and the importance of an intact marriage (or, more precisely, biparentalism) for children's development in our slowly developing species (Popenoe, 1993), centripetal forces seem to bind spouses together. These binding forces are likely to intensify once reproduction has begun (Hrdy, 2011). These proximate forces include (1) amorousness to promote biparentalism and its fitness value (Jankowiak and Fischer, 1992); (2) sexual receptivity through the menstrual cycle contributing to mate guarding and paternity confidence and, hence, more paternal care (Godkin, 2011); and (3) parent-child attachment promoting parental care and keeping parents together (Bowlby, 1969/1982). Spousal bonding may also be strengthened by (4) the moderate degree of homogamy that tends to characterize mating in humans and many other species (Russell and Wells, 1991), with homogamy perhaps increasing kin altruism between spouses and children by raising consanguinity and reducing conflict (Dunbar and Machin, 2014; Rushton and Bons, 2005); (5) having multiple and healthy children, thereby preferentially sustaining fertile unions (e.g., Goode, 1993); (6) the economic efficiencies of division of labor by gender and economies of scale (Friedl, 1975; Kokko and Jennions, 2008); (7) practice effects enhancing the efficiency of cooperation; and (8) the overrating of spouse's attractiveness and (9) humor (Gottman, 1994; Schaap, 1982). The last two factors have been found to correlate with marital satisfaction cross-culturally (Silveri and Weisfeld, 2008; Weisfeld et al., 2011).
In addition, (10) hormonal factors seem to cement spousal and parental bonds (Insel and Young, 2001). The surge in oxytocin at orgasm tends to enhance bonding (Carmichael, et al., 1987). At marriage and at fatherhood, men's testosterone falls and prolactin rises, apparently enhancing uxoriousness and nurturance (Booth and Dabbs, 1993; Mazur and Michalek, 1998; Storey, Walsh, Quinton, and Wynne-Edwards, 2000).
Another factor is (11) cultural and social pressure to make and maintain a marital commitment and resolve differences. In the U.S., relations between the wife's and the husband's parents seem especially crucial (Bryant, Conger, and Meehan, 2001). Most societies have marriage ceremonies and legal or cultural barriers to divorce (Rosenblatt and Unangst, 1979).
In short, a panoply of proximate factors consolidate marriages and enhance the benefits of spousal support for children. But a consolidated marriage may detract from the fitness of either parent if, for example, better reproductive opportunities are to be found outside of the current marriage.
Despite these marital bonding factors, even newlyweds have been found to experience some conflict. In a U.S. study, wives most often objected to sexual aggression by the husband, and husbands to sexual refusal (Buss, 1989a); furthermore, some spouses were viewed as being inconsiderate or abusive. In later years, disagreements may involve children or finances. For example, marital conflict regarding ideal family size was found to be common in Tanzania (Borgerhoff Mulder, 2009). Betzig (1989) analyzed cross-cultural data on 186 preliterate societies and found that the most common reasons ex-spouses gave for their divorce were sexual difficulties, female infidelity, male economic shortcomings, and male abuse. Other cross-cultural and historical studies have identified women's economic independence as a common factor in divorce (Goode, 1993; Trent and South, 1989). In another cross-cultural analysis, marital instability was associated with economic independence of spouses from each other (Friedl, 1975). U.S. research on marital conflict has focused heavily on the issue of division of labor, but this might be less relevant in more traditional cultures with pronounced division of labor.
Based on this previous research and on evolutionary principles, we hypothesized that important sources of marital conflict would include the following: sex, resources, division of labor, and children. We also included kindness, as Buss (1989b) found that both sexes tended to seek kindness in a mate across cultures. Our reasoning was that lack of kindness would presumably reduce cooperation and raise stress levels for the spouse. In contrast, we reasoned (and confirmed by preliminary analysis) that variables such as religiosity and who makes up first after a quarrel would not predict conflict in the relationship, and we did not include those variables in this study.
We studied five diverse cultures in order to test for potentially universal sources of conflict but also for cultural variability. What is true in one contemporary culture is not necessarily true elsewhere, let alone everywhere (Henrich, Heine, and Norenzayan, 2010). We used data that had been gathered previously to address a variety of questions post hoc. Thus, we did not devise a purpose-built questionnaire to test our hypotheses specifically.
Materials and Methods
Participants
We recruited over 2,600 married couples from the United States, Britain, China, Russia, and Turkey (see Table 1). All couples were recruited from predominantly urban areas in each of their respective cultures (mainly Detroit, London, Beijing, Moscow, and Ankara), and a representative age range of couples was sought. As a method of convenience sampling, this strategy was practical to the extent that urban areas encompass a dense proportion of married individuals living within each of our five selected cultures. However, our samples were nonrandom, and therefore do not represent the entirety of marriage within each culture, or the entirety of subcultures within each culture.
Sample demographics for each culture
The U.S., Turkish, and Russian couples were recruited using modified snowball sampling (Bailey, 1987). The Turkish couples were recruited by psychology students with the assistance of local officials called muhtars. The British couples were recruited through three techniques: placing an advertisement in a women's magazine, hiring a market research company, and convenience sampling by college students; the three methods yielded similar results. The Chinese couples were recruited by giving their children a questionnaire at school to deliver home to their parents. Despite the fact that all the Chinese couples had children, the mean number of children was similar to the figures for the other cultures (see Table 1).
Materials
All participants completed the Marriage and Relationships Questionnaire developed by Russell and Wells (1986). Having completed the questionnaire privately, each spouse placed the completed questionnaire in an envelope and sealed it, and the envelopes of husband and wife were then placed in a larger envelope for collection. The original version of the Marriage and Relationships Questionnaire (MARQ) comprises 235 multiple choice and true/false items, and was developed using British couples in the late 1980s (see Russell and Wells, 1993). The entire questionnaire was translated and back-translated for use in each country, under the supervision of the co-author responsible for data-gathering in that country. Parallel versions of the MARQ exist for its administration to both husbands and wives.
The questionnaire was designed to give a comprehensive picture of the respondents' feelings both about themselves and about their relationship with their spouse. The questionnaire was fashioned with cross-cultural use in mind. The measure is decidedly subjective in its attempt to capture husbands' and wives' perceptions of themselves and each other. However, the MARQ is objective in the sense that spouses may be formally compared to one another, and to other couples, on numerous aspects of marital satisfaction. The validity of the scale is supported by the high level of agreement between spouses on most items. Although it is less well known, the MARQ is similar to many other measures of marital satisfaction to the extent that it has been used to study relationships between spousal perceptions and relationship satisfaction (e.g., Lucas et al., 2004). However, the MARQ is distinguished by the extent to which it defines marital satisfaction multidimensionally. Over a dozen scales assessing various aspects of marital relationships are formally and informally derived from the MARQ.
The MARQ was advantageous in the present study because of this highly multidimensional operationalization, and also because three scales within the MARQ have demonstrated strong measurement invariance for husbands and wives across most of the countries included in our analysis: the Love, Partnership, and Problems with Partner scales (Lucas, et al., 2008). Most other marital satisfaction scales have not been tested for measurement invariance across genders and cultures. For the present study, we used the seven-item “Problems with Partner Scale” as the outcome variable, our measure of dissatisfaction with the spouse and perception of problems in the relationship. The scale consists of the following items:
Does your spouse embarrass you in public?
Do you take your spouse for granted?
Does your spouse get on your nerves?
Does your spouse have irritating habits?
How often do you have a serious row?
Do you feel impatient with your spouse?
Is your spouse really nasty to you?
All items were answered using a five-point Likert scale (higher scores indicating perceiving more problems). Cronbach's alpha values are reported in Table 1 for husbands and wives in each of our five cultures; those values range from .57 to .83 across all groups. Thus, this scale is a fairly robust measure for assessing marital dissatisfaction, in contrast to something like considering divorce, which is not a realistic option in some of our cultures. Further details on the methodology appear in Lucas et al. (2008).
Sources of conflict were assessed by these items: “Do you find sexual fulfillment in your marriage?”; “Is money a problem in your marriage?”; “Do you agree on who does what in your marriage?”; “Do you agree about how the children should be brought up?”; and “Is your spouse kind to you?”
To analyze the data, the following predictor variables were entered into a linear regression: money problems, sex problems, division of labor problems, disagreements about raising children, and spousal lack of kindness.
Results
All predictor variables were entered simultaneously in a linear regression analysis and, as shown in Table 2, the model was significant for all five cultures for both husbands and wives (R2 ranged from .22 to .51). For wives and for husbands, the model significantly predicted increased scores on the Problems with Partner scale by examining financial problems, division of labor, kindness, sexual fulfillment, and child-rearing agreement. Higher scores on predictor variables indicate the item is more of a problem in the relationship. Almost all of the relationships were statistically significant.
Summary of regression analyses for five cultures, predicting Problems with Partner scores for husbands and wives
Note. * p < .05;
p < .01
With regard to wives, all predictors were significant for British and Chinese wives. For American wives, only the child-rearing item was not a significant predictor of their score on the Problems with Partner scale. For Turkish wives, only the division of labor item was not a significant predictor. For Russian wives, only the financial problems item was not a significant predictor (refer to Table 2 for regression weights). Note that the sample sizes shown in Table 2 reflect the numbers of wives and husbands remaining after eliminating subjects with missing responses.
For husbands, the model significantly predicted increased scores on the Problems with Partner scale by examining financial problems, division of labor, kindness, sexual fulfillment, and child-rearing agreement. All predictors were significant for British, Chinese, and Turkish husbands. For American husbands, the sexual fulfillment and the child-rearing predictors were not significant predictors of their score on the Problems with Partner scale. For Russian husbands, the financial problems item was not a significant predictor.
An examination of the partial correlations revealed that “Is your spouse kind to you?” was the most consistent predictor and accounted for the most unique variance in the model across all five cultures for both sexes. This was true in 7 out of 10 cases, and in the remaining cases this predictor was the second highest predictor in the model. For American wives, lower sexual fulfillment was the predictor that accounted for the most variance in the model. With respect to Turkish husbands and wives, perceiving greater financial problems captured the most variance in the model.
All items in the model were examined for mean differences between husbands and wives (see Table 3). In all cultures, the mean score for wives on the Problems with Partner scale was higher, indicating wives perceiving more problems with the marriage than husbands perceived. However, this difference was only significant in the American, British, and Turkish samples. Regarding the item “Do you agree on who does what?” wives reported more disagreement in all five cultures, with British and Russian couples showing a significant sex difference. American and British wives were significantly more likely to report their spouses were kind to them than were the husbands. Chinese and Turkish wives were significantly more likely to say they were less sexually fulfilled in their marriage than were the husbands. American and British wives were significantly more likely to report not agreeing with their husband on how the children should be brought up.
Means and standard deviations for sex differences for all variables
Note. * p < .05;
p < .01
Discussion
Our main findings are that variables related to resources, children, sex, and labor are significant predictors of possible conflict between husbands and wives in these diverse, modern cultures, with some exceptions. These findings confirm our expectation that universal, fitness-salient factors would emerge as common sources of marital conflict, and that these conflicts tend to arise because of basic incompatibilities in reproductive goals between the sexes. Additional sex differences, some of which reflect greater female choosiness, are reported in another paper using the same data set (Weisfeld et al., 2011).
In contrast, various factors have not been found to correlate with marital problems across cultures, including self-fulfillment (Xie and De Frain, 1999) and the arrival of children. Some studies of effects of the arrival of children in Western societies have indicated that marital satisfaction declines (reviewed by Wendorf, Lucas, Imamoğlu, Weisfeld, and Weisfeld, 2011), but the declines are small and are not observed in more collectivist cultures (Dillon and Beechler, 2010). In fact, marital stability rises with number of children in many cross-cultural studies, as expected in evolutionary terms (Goode, 1993).
Another general finding concerns the role of kindness. The importance of kindness is consistent with Gottman's (1994) conclusion from U.S. longitudinal research that insufficient positive communications compared with negative ones predict divorce, and with studies indicating that aggressive, contentious marital communication is associated with less satisfaction. This was true regardless of the degree of self-disclosure exhibited by the couple. Conflict itself can have positive results, depending on how it is waged (Gottman and Krokoff, 1989). Under this interpretation, it is not the amount or style of communication that is crucial; rather, it is a lack of kindness in all interactions, including arguments.
Spousal kindness, which requires moderating one's own selfish needs, may be at least as important as the other variables, and a crucial facilitator of the long-term partnership required by human parental investment (Buss, 1989b). In the U.S., many popular books advise married people on how to secure more benefits from their spouses, but it may be better for people to be more cooperative and less individualistic. Parents, because of homogamy and their shared reproductive interests, are in essence quasi-kin who stand to benefit greatly from cooperation.
The general patterns of results are of great interest since they indicate possible universality, in that threats to reproductive fitness can be seen in problems with sexual access, resources, labor, and children. Conflict is likely to arise in these domains, which are not always easily controlled by the individual. Husbands and wives, however, can act kindly toward each other, thereby mitigating the degree of conflict. These five factors together accounted for a substantial amount of variance in problems with the partner (see Table 2), further indicating their importance.
We found a number of cultural idiosyncrasies. However, explanations for cultural particularities are necessarily tentative and complex, and warrant more extensive treatment than space allows here. The important point is to recognize that, in addition to universals, cultural differences exist that we need to understand in functional and historical terms.
Some of these cultural peculiarities follow. Perceiving greater financial problems accounted for the most variance in the model for the Turkish couples. In a country such as Turkey with high economic mobility and rapid socioeconomic change, couples may be more likely to argue over finances. Russian husbands and wives, in contrast, did not identify financial issues as a significant source of problem with the partner. The Russian author of the study (Marina Butovskaya) suggests that because salary payments are often delayed by the employer, spouses tend to be kind and cooperative toward each other. Couples may tacitly agree not to fight over something beyond their control. Similarly, neither U.S. wives nor husbands identified disagreements over the children as a source of problems with the partner. American couples may know better than to argue about children in an environment that makes family life difficult because rarely is there an extended family nearby to offer assistance.
Couples may, then, internalize cultural realities as they create norms for themselves wherever they are. In this regard, our finding that for Turkish wives division of labor did not appear to be a major source of conflict may be explained with reference to the presence of strong norms favoring gender role specialization. Those forms of family functioning are generally so well internalized that women may be constrained not to disrupt them. “Even in the case of professional women the strains inherent in combining work and domestic responsibilities are seldom allowed to reflect on men who continue to be sheltered from new role demands” (Kandiyoti, 1982, p. 117).
As for sex differences, the fact that wives everywhere reported more conflict is consistent with the female choice principle, suggesting that female choice continues to operate during marriage as well as before it. Female choice may also explain Gottman's (1994) conclusion that it is usually the wife who raises an issue in marital conflict discussions in U.S. marriages, and that the wife exercises considerable influence in U.S. and Dutch marriages (reviewed by Gottman, 1994). In our cross-cultural research, wives were more likely to regret having married their spouse and to have considered divorce than husbands (Weisfeld et al., 2011); this has also been reported in Japan and the U.S. (Akiyama, Antonucci, and Campbell, 1987). Wives' parents were more likely than husbands' to have disapproved of the couple getting married. Husbands were more likely to report love at first sight of the spouse. As the pursuing gender, men may try to be especially kind, as reported by U.S. and U.K. wives in the present study.
One might object that women tend to complain more about things in general, not just about their marriages. However, we found that men complained more about their spouse being sexually unresponsive than did women across cultures (Weisfeld et al., 2011). Thus, men may have some specific complaints, but women may complain more than men about their overall relationship.
Other interpretations of wives' reporting more conflict are plausible. Much literature on gender suggests that women report more problems because they encounter more problems. Historically, being dependent on men for financial security has forced women to accommodate their husband's wishes. Although there seems to be a trend from traditional toward more egalitarian marriages in different parts of the world, still even in dual-career marriages, “despite an essentially supportive orientation, both partners seem to endorse some level of male dominance” and “there is a consistent pattern of inequality in both the domestic and occupational spheres” (Steil, 1983, p. 53).
Of course, cultural factors modify the degree of female choice. When paternal contributions are great, as in patrilineal societies, it is the husband who usually initiates divorce (van den Berghe, 1979). In matrilineal societies such as the Hopi, in which paternal care is comparatively low, the wife typically initiates divorce (Simmons, 1942).
Our findings are limited by the fact that our samples are non-random and drawn largely from urban centers in these various countries. These inconsistencies limit the justification for claims of cultural differences. On the other hand, if a result occurs in a great diversity of samples, this diversity would seem to strengthen the case for the universality of the findings. Also, we did not devise a purpose-built questionnaire to measure the predictor variables by means of scales—although a case can be made for using single items to measure a variable (Wanous, Reichers, and Hury, 1997).
Of course, much more is known about marital interactions and marital satisfaction than has been touched on here, particularly as revealed by Gottman's (1994) longitudinal research on the course of marital dissolution and the typical roles played by husband and wife. Nevertheless, the issues over which conflict arises seem to warrant more research and further evolutionary and sociocultural interpretation.