Поиск по сайту

Пишите нам: info@ethology.ru

Follow etholog on Twitter

Система Orphus

Кросс-культурный метод
Старые форумы
Не в тему

список статей

Culture, ecology, and beliefs about gender in son preference caste groups

Ramaswami Mahalingam

1. Introduction

1.1. Culture, ecology, and evoked culture

1.2. Study descriptions

2. Study 1: son preference and land allocation

2.1. Participants

2.2. Results

2.3. Discussion

3. Study 2: son preference: land versus cash allocation

3.1. Participants

3.2. Results

3.3. Discussion

4. Study 3: culture-of-honor task

4.1. Participants

4.2. Results

4.2.1. Personal insult

4.2.2. Family insult

4.2.3. Caste insult

4.2.4. Village insult

4.2.5. CHS

4.3. Discussion

5. General discussion

6. Conclusions


Appendix A. Culture-of-honor vignettes



1. Introduction

Several cultures have practiced female infanticide to get rid of unwanted girls (Hrdy, 1999, Rousselle, 2001). Recent demographic data indicate that at least such a systemic willful killing of female infants is a thing of the past in many developed countries (Hrdy, 1999). However, extreme gender discrimination persists in countries such as India, China, and Korea. Although female infanticide is now a crime in India, census data of the last two decades show a persisting male bias in sex ratios in several parts of the country (Muthulakshmi, 1997, Natarajan, 1997).

Why does female infanticide persist in India even today? Several explanations have been proposed. Some researchers have argued that structural gender disparities and abject poverty are the causes of female neglect and infanticide. Hence, it was thought that women's empowerment would improve the status of women and subsequently result in better treatment of daughters. Das Gupta and Visaria (1996), however, found that in Punjab, more educated women invested more in their sons than in their daughters. Cultural and ethnographic reports suggest that the role of cultural practices, such as the high costs of dowry (payment to the bridegroom's family) and the prospect of old age care and protection by sons, are the main reasons for the prevalence of female neglect (Miller, 1981). Due to a patrilocal arranged marriage system, dowry for daughters and protection by sons in old age are generally the case for all caste groups; however, because documented evidence suggests that the practice of systemic female infanticide is caste specific (Natarajan, 1997), further research is necessary to determine differences among caste groups.

As noted by Miller (1981), “One fact clearly emerges from the welter of reports and secondary studies from British India: that is, female infanticide was not practiced in every region of India, and not everyone in those areas was involved” (p. 53). Only in certain geographic locations and among certain caste groups that favor male children do people engage in various forms of female neglect, including female feticide and female infanticide. For instance, female neglect in Tamilnadu is reported to be prevalent among Thevars, a warrior caste (Krishnaswamy, 1988, Muthulakshmi, 1997), and among Gounders, a landowning caste (Venkatachalam & Srinivasan, 1993). Sex ratios for children aged 0–6 years in villages predominantly inhabited by these caste groups range from 625 to 750 girls for every 1000 boys (Harriss-White, 2001).

Thevars, called the “warrior race” by the British (Thurston, 1909), are historically known for their valor and chivalry, and for their cavalier attitude and personal honor. Thevars served in the armies of various Tamil kings and in the British army in World Wars I and II. Even today, disproportionately more men from Usilampatti (a town with a high concentration of Thevars) serve in the Indian army than from any other district in Tamilnadu (Harriss-White, 2001). Gounders are a landowning caste. Gounder villages also have one of the most biased sex ratios in the country (625 girls for every 1000 boys aged 0–6 years; 1990 census). Government primary health centers within villages primarily inhabited by Thevars and Gounders have had to create a new category for the cause of death called “social causes” while recording the death of an infant who died from nonmedical reasons (Natarajan, 1997). Goldstein, 1971, Goldstein, 1976 has found that the limited availability of precious land resources and the practice of male inheritance also contributed to differential investment in sons and daughters (also see Levine, 1987, Levine & Silk, 1997). Not surprisingly, a disproportionately higher number of female infants in these villages die of “social causes” (11–27 times higher than male infants; Natarajan, 1997).

Behavioral ecologists point out that ecological factors influence parental investments. According to behavioral ecologists (e.g., Charnov, 1982, Fisher, 1958, Trivers & Willard, 1973), parents will do their best to allocate resources and expend effort for the reproductively most profitable sex. Thus, in polygynous societies, when resources and goods differentially affect men's success in marrying and raising children, inheritance is strongly patrilineal or male biased. Cross-culturally, among traditional societies, women's ability to control resources, inherit property, and hold political office indeed have ecological correlates (Divale & Harris, 1976, Low, 1990a, Low, 1990b, Low, 2000, Whyte, 1978, Whyte, 1979). When land is inherited along patrilineal lines, women control fewer resources and, as a result, have lower status (Low, 1990a).

Ecological constraints affect parental behavior and the differential treatment of sons and daughters. Cross-culturally, female infants and fetuses are at risk only under certain particular and predictable cultural–ecological conditions such as plow agriculture, herding, patrilineal inheritance, dowry, and low resource contribution by women (e.g., see Bugos & McCarthy, 1984, Dickemann, 1979, Dickemann, 1981, Hill & Low, 1991, Hrdy, 1999, Hughes, 1988, Low, 1993, Low, 2000, Torres & Forrest, 1988, Whyte, 1978, Whyte, 1979). These socioecological conditions shape son-biased investment decisions. Several historical findings are also consistent with these claims (e.g., Boswell, 1990, Clarke & Low, 1992, Fuchs, 1989, Ransel, 1988, Sherwood, 1988).

A behavioral ecological perspective sensitive to cultural practices, termed “cultural ecological perspective,” would help us to understand why Thevars and Gounders practice female infanticide. The cultural ecological histories of these caste groups provide some insights about their extreme son bias. Gounders prefer not to have many daughters because daughters take land resource in the form of dowry (paid by the bride's family as land or the land is sold to pay for elaborate wedding expenses; for ethnographic details, see Venkatachalam and Srinivasan, 1993) from the natal family. Since family lineages are patrilineal, having sons maximizes the inclusive fitness of the patrilineal side of the family. A daughter, who ultimately will move to her husband's family, is viewed as a “guest” of the house (Kakar, 1978).

Warrior groups in several cultures have a documented history of female infanticide (Divale & Harris, 1976, Weisfeld, 1993). They prefer to have more sons because they will stand up for the family and kin. Daughters are thus expensive for cultural and material reasons. In addition to requiring expensive marriage costs, daughters are viewed as repositories of family honor. Cross-cultural findings indicate that there is an asymmetry in expectations about male and female chaste behaviors (Broude, 1975, Broude, 1980, Frayser, 1985). Chastity expectations are particularly high among culture-of-honor groups. In intergroup conflict situations, the triumphant warring group can overpower women in the losing group and keep them captive. Such social anxieties are a dominant theme in Thevar cultural narratives (e.g., ballads and folk songs). They valorize women who preserve family and group honor by committing suicide to avoid capture by the warring group. The paucity of women also contributes to male–male competition, ensuring that only the most successful warriors can find a mate (Weisfeld, 1993).

In summary, Thevars and Gounders each have reasons predisposing them to son favoritism, but the reasons relate to their different cultural and ecological backgrounds. Both resource type and cultural attitudes correlate with treatment of women. For Thevars, caste and personal honor are important factors that often result in treating women as objects whose honor needs to be protected. In a cross-culturally typical pattern (Low, 1990b), the landowning Gounders try to avoid dividing land; they, therefore, prefer not to have many daughters, which would lead to the land resources leaving the family in the form of dowry. In both caste groups, however, having a daughter is thought to be better than having no child at all. Consequently, firstborn daughters are rarely killed (Muthulakshmi, 1997, Venkatachalam & Srinivasan, 1993).

1.1. Culture, ecology, and evoked culture

So far, very few studies have examined cultural psychological antecedents to female infanticide. Most hypotheses about female neglect and infanticide are relatively “global” and do not predict differences in infanticide rates with varying cultural conditions (Hrdy, 1999, Keller et al., 2001). Furthermore, hypotheses tend to consider only one set of factors (e.g., cultural but not ecological, or ecological but not cultural); these hypotheses are not true alternatives (see Low, 2000). More than one set of factors could be operating. Until we understand the complex causes and consequences of female infanticide, we cannot successfully discourage its systemic practice.

The phenomenon of evoked culture described by Tooby and Cosmides (1992) is valuable in identifying caste-specific psychological correlates of female neglect. Evoked culture refers to the production of a psychological phenomenon that is shared by only some members of a culture as a result of the triggering of psychological mechanisms by environmental factors of differing degrees within a culture (also see Buss & Kenrick, 1998, Kenrick & Luce, 2000).

The need to transfer ancestral property to sons, in the case of Gounders, and a culture of honor, in the case of Thevars, perhaps act as triggers of an evoked culture of female neglect and infanticide (Cohen & Nisbett, 1994a, 1994b). However, the nature of evoked culture would be different for Thevars and Gounders. Since landowning groups favor sons, Gounders will allocate more resources to their sons than to their daughters and will invest preferentially in those sons who produce more grandsons (Jeffery & Jeffery, 1997). Quite specific predictions can follow from these expectations: for example, even in families with more than one son, the one who produces more grandsons is likely to get a proportionately larger share of the ancestral property.

Although typically unrecognized, scattered evidence suggests a strong relationship between chastity expectations and the prevalence of “cultures of honor,” in which men are quick to take offense and women are objectified (Cohen et al., 1998, Nisbett & Cohen, 1996). Nisbett (1993) has suggested that in pastoral “culture-of-honor” societies, women are likely to be seen as defensible property rather than resource producers. He identifies three major psychological factors that characterize honor cultures: (a) low tolerance of insults; (b) high value on honor and shame; and (c) salience of chastity. Men are expected to cultivate a habit of responding to insults aggressively. Lindisfarne (1998) also provides several cross-cultural examples of honor cultures, where boys are socialized to be “real men” and women are socialized to value chastity.

In a culture of honor, only aggressive men, who are proactive in asserting command over other men and resources, are likely to succeed (see also Chagnon, 1988). Evidence from folklore of warrior castes in India reflects a strong culture of honor in these caste groups (e.g., Miller, 1981). Treating women as objects whose honor must be protected also resonates strongly among Thevars. In their folklore, Thevar women preserve their honor by immolating themselves after a defeat in war (Ramamoorthy, 2002).

In summary, an integrated cultural ecological perspective identifies the following caste-specific psychological antecedents to female infanticide and female neglect. If Gounders' son preference is rooted in land resource concerns, then Gounders would favor sons in resource allocation tasks. If Thevars' son preference is embedded in their honor concerns, they should be higher on measures of culture of honor than Gounders.

1.2. Study descriptions

To test these hypotheses, four field studies were conducted in villages with an extremely male-biased sex ratio that are predominantly inhabited by Thevars and Gounders from the South Indian state Tamilnadu from the districts of Madurai and Salem, respectively. Brahmins, a priestly caste from Madurai district, were included in the study as a comparison group. They are a dominant caste group reported to not have a history of female infanticide. Any study involving social distance measure (Bogardus, 1933) should reveal caste differences among Thevars, Gounders, and Brahmins. I predict that Thevars will have the strongest preference for within-group marriage or marrying up, and will strongly oppose marrying into a lower caste. All studies were conducted by trained research assistants who were native speakers of Tamil but not from the region. Instruments were translated from English to Tamil and backtranslated by two bilingual speakers. The field studies were conducted over a period of several months.

2. Study 1: son preference and land allocation

This study examined whether son preference influences land resource allocation. According to the cultural ecological perspective, son preference leads to differential allocation of land resources to sons. One of the main reasons for son preference among Gounders is to have a son who can take care of the land. In this study, participants were given a moral dilemma where the aging patriarch has to make a decision to distribute his lands among his daughters and sons.

I predict that in allocating land resources, there will be a stronger preference among Gounders to allocate resources to sons than among both Thevars and Brahmins. Since Brahmins are not a landowning caste, they are more likely to distribute their resources equally among sons and daughters.

2.1. Participants

Participants were recruited from a community sample with the help of a former nongovernmental organization that was not influential but familiar to these areas (N=1195; 598 men). There were 391 Brahmins (age: mean=36.5 years, S.D.=14.96), 405 Gounders (age: mean=34 years, S.D.=13.6), and 399 Thevars (age: mean=34.8 years, S.D.=15.75). They were told a story of an aging patriarch who had to make a decision about distributing his agricultural land among his sons and daughters. The participants were given the following three options: (a) equal distribution among sons and daughters; (b) more or all lands to sons; or (c) more or all lands to daughters, and were asked to choose one of these options.

2.2. Results

The frequency of the response by caste and gender are summarized in Table 1. Multinominal logistic regression was performed since the outcome variable, which is allocation of land, was polytomous, consisting of three categories (Table 2 here). The two predictors tested in the model were caste and the gender of the respondent. I expected that since Gounders and Thevars were the extreme son preference caste groups, in comparison with Brahmins, they would allocate more or all lands to sons rather than equally. I also predicted that in comparison with Brahmins, they were likely to allocate land to sons rather than to daughters. Brahmins were the reference category in my SPSS analysis; hence, the results are presented as comparisons between Brahmins and the two other caste groups. Odds ratios (ORs) and the respective confidence intervals (CIs) for each of the predictors are reported in Table 1. I observed that the odds of Brahmins allocating all or more lands to sons versus allocating equally was lower than the odds of Thevars (OR=0.25; CI=0.18–0.34) and Gounders (OR=0.29; CI=0.252–2.38) allocating all or more lands to sons versus allocating equally. Gender differences were seen. The odds of females allocating all or more lands to sons versus allocating equally were greater than the odds of males allocating more lands to sons versus allocating equally (OR=1.611; CI=1.26–2.06). No other differences were statistically significant.

Table 1.

Summary of main findings

Frequency of land distribution choices to offspring, by participants' caste and gender for Study 1
Brahmins Gounders Thevars
Men Women Men Women Men Women
Divide land equally 159 143 112 92 106 79
All or more lands to sons 33 50 87 105 89 115
All or more lands to daughters 3 2 4 5 5 6

The means for each value are shown on a scale from 1 to 5, with higher scores indicating higher tolerance of infidelity.

Table 2.

Multinomial logistic regression coefficients for land allocation task for Study 1 (N=1195)

Contrast β S.E. Wald OR
Equal vs. sons Thevar (1) vs. Brahmin (3) −1.40 0.16 76.00 0.25
Gounder (2) vs. Brahmin (3) −1.24 0.16 60.20 0.29
Males (−1) vs. Females (1) 0.47 0.12 14.59 1.61
Daughters vs. sons Thevar (1) vs. Brahmin (3) −0.12 0.55 0.05 0.89
Gounder (2) vs. Brahmin (3) −0.26 0.57 0.20 0.77
Males (−1) vs. Females (1) 0.18 0.41 0.19 1.20


2.3. Discussion

The Brahmin response pattern was in the expected direction. They were more gender egalitarian than Thevars and Gounders. Surprisingly, Thevars responded like Gounders. Contrary to predictions, most Thevars and Gounders chose to distribute land equally among sons and daughters. However, females allocated more lands to sons than to daughters. Perhaps the women were under pressure to raise their sons to be successful in order to maximize their chances of finding a partner. In addition, because of patrilocal arrangement, they themselves might not have received any land from their parents. Hence, they may be more inclined to believe that sons should get all or more lands.

3. Study 2: son preference: land versus cash allocation

To verify whether the presence of equal distribution as an option might have influenced the response pattern, Study 2 presented a forced-choice moral dilemma that tested son bias in the allocation of land resources. This study presented a scenario in which a patriarch has to decide how to allocate resources between two sons: one of whom has three daughters and one of whom has three sons. The patriarch has assets in land resources and in cash that are of equal value. Participants were asked to decide whether the cash should be given to the son who has all daughters and the land should be given to the son who has all sons, or vice versa. If equal distribution is a sole criterion for allocating resources, there should not be much difference between the choices. This dilemma will test whether there is a bias to keep family lands in the possession of male descendents. This dilemma indirectly tests the central role of land in the desire to have more sons.

I predicted that a significantly higher proportion of Gounders would allocate land resource to sons with male offspring and the cash to the son with daughters, whereas the responses of Brahmins would be equally distributed. I also predicted that the proportion of Thevars' responses favoring giving land to the son with male offspring would be between that of Gounders' and Brahmins'.

3.1. Participants

Participants from the three caste groups were recruited in the same areas (N=647; 56% men). There were 312 Brahmins (age: mean=34 years, S.D.=14), 124 Gounders (age: mean=29.6 years, S.D.=11.9), and 210 Thevars (age: mean=31.2 years, S.D.=14.6). Participants were presented the dilemma and asked to choose one of the two responses as if they were the patriarch.

3.2. Results

The results of Study 2 are summarized by gender and caste of the participants in Table 3. The responses were analyzed using a logistic regression procedure. In the first step, I contrast coded the three caste groups. Specifically, I tested two contrasts: (a) Is there a difference between Thevars and Gounders in the preference for the son with sons to inherit the land? (b) Is there a difference between Brahmins and the other two caste groups? I also tested if the gender of the research participant made a significant difference in the preference for the inheritance of land by sons. Table 4 summarizes the β values, significance levels, ORs, and CIs of the two caste contrasts and the single gender contrast.

Table 3.

Frequency of land distribution choices to offspring based on the gender of offspring's children, by participants' caste and gender for Study 2

Brahmins Gounders Thevars
Men Women Men Women Men Women
Land to son with sons 118 112 69 46 106 88
Land to son with daughters 48 34 8 1 15 2
Table 4.

Logistic regression coefficients for study 2 (N=647)

Contrast β S.E. Wald OR
Contrast 1: Thevars (−1) vs. Gounders (1) −0.039 0.193 0.041 0.96
Contrast 2: Brahmins (2) vs. Others (−1) 0.51 0.084 36.466 1.66
Gender (Females=1 and Males=−1) −0.30⁎⁎ 0.114 6.888 1.74




There were no significant differences between Thevars and Gounders. The odds of Thevars deciding to allocate land to the son with sons was only marginally lower than that of Gounders (OR=0.96; CI=0.66–1.40). The odds of the extreme son preference caste group (Thevars and Gounders) allocating land resources to the sons with sons was greater than the odds of Brahmins allocating land resources to the sons with sons. (OR=1.66; CI=1.41–1.95). In addition, gender differences were observed, as the odds of males allocating land to the sons with sons was lower than the odds of females allocating land to the son with sons (OR=0.74; CI=0.59–0.93).

3.3. Discussion

Participants from all three caste groups prefer to allocate land to the son with more male offspring. Although Brahmins are not a landowning group, perhaps the practice of dowry and patrilocal transfer of wealth among Brahmins may have influenced their decision making. However, there was a stronger son bias among Gounders and Thevars than among Brahmins in allocating land resources to the son who has all-male offspring. These findings suggest that, among Gounders and Thevars, having one or more sons improves one's chances of inheriting ancestral property. A son bias permeates across generations as “a grandson bias,” for a son who produces more sons is in an advantageous position in getting the ancestral property relative to sons who had fewer or no sons.

Consistent with the findings in Study 1, more women preferred to allocate land to the son with all-male offspring than to the son with more female offspring. The gender differences in beliefs about resource allocation reveal the complex predicaments of women in these communities. On one hand, these women are expected to be responsible for maintaining cultural continuity, such as the maintenance of the patrilocal transfer of resources. At the same time, such practices also perpetuate the differential treatment of daughters contributing to male-biased sex ratios. The only way women have access to resources in these communities is by ensuring the patrilocal transfer of family resources. In addition, in patrilineal societies, women gain fitness through grandsons. Hence, preserving land resources for the male lineage may be more critical for women than for men.

4. Study 3: culture-of-honor task

Studies 1 and 2 focused on examining the link between land resources and son bias. In Study 3, I developed a set of culture-of-honor tasks to identify caste-specific underlying psychological correlates between honor culture and female neglect. For Thevars, the warrior caste, honor culture was expected to be stronger than among Gounders and Brahmins. To test this hypothesis, I developed a culture-of-honor task involving four vignettes describing various (personal, family, village, and caste) insults. In all these vignettes, the protagonist responds angrily to an insult (see Appendix A for a description of vignettes). This study examines attitudes toward insults, not about socialization of aggressive behavior as in the case of war situations (Cohen & Nisbett, 1994a, Cohen & Nisbett, 1994b, Ember & Ember, 1994) The respondents were asked to rate the appropriateness of the protagonist's reaction on a 5-point scale (1=strongly agree; 5=strongly disagree). The average scores across the four vignettes were calculated as the culture-of-honor score (CHS; see Table 5 for a summary of means for each caste group). The same participants, in a separate session, were presented with a situation where the protagonist (in one condition a male and in another condition female) responds violently upon finding one's spouse in bed with another person. Participants were randomly assigned to one of these conditions. They were asked to judge the appropriateness of the protagonist's violent reaction (1=strongly agree; 5=strongly disagree).

Table 5.

Descriptive statistics for culture-of-honor dilemmas, by participants' caste and gender for Study 3

Caste Personal honor Family honor Caste honor Village honor CHS
Men Women Men Women Men Women Men Women Men Women
Brahmins 2.16 (1.41) 2.18 (1.35) 2.89 (1.57) 3.00 (1.51) 3.20 (1.59) 3.58 (1.39) 3.45 (1.44) 3.34 (1.37) 2.94 (0.99) 3.03 (0.93)
Gounders 1.89 (1.54) 1.66 (1.31) 3.31 (1.86) 2.89 (1.83) 2.82 (1.82) 4.05 (1.52) 3.30 (1.63) 3.31 (1.31) 2.83 (1.03) 2.98 (0.88)
Thevars 2.38 (1.64) 1.95 (1.51) 3.57 (1.66) 3.38 (1.75) 3.74 (1.59) 4.07 (1.50) 3.67 (1.46) 3.41 (1.49) 3.34 (1.07) 3.20 (1.05)

Lower scores indicate endorsement of a violent response to an insult (1=strongly agree; 5=strongly disagree).

I predicted that Thevars would be more likely to endorse an angry response of the protagonist than would either Gounders or Brahmins. CHS would predict the attitude of spousal response for all caste groups. The relationship between CHS and attitude toward spousal infidelity would be strongest for Thevars. I also predicted that in all groups, female infidelity would be judged more negatively than would male infidelity.

4.1. Participants

Brahmin, Thevar, and Gounder participants were recruited from the same villages (N=1195; 598 men). There were 391 Brahmins (age: mean=36 years, S.D.=15), 405 Gounders (age: mean=34 years, S.D.=13.6), and 399 Thevars (age: mean=34.6 years, S.D.=15.7). They were presented the four insult vignettes. The ratings of the respondents were recorded. Then in a separate session after 2 weeks, the respondents were randomly presented the infidelity story of a male or female spouse.

4.2. Results

A 3 (Caste: Brahmins, Gounders, and Thevars)×2 (Gender of the Participant: male, female) MANOVA of the insult scores for all four vignettes found a main effect for caste [F(2,1174)=14.94, p<.001]. There was no main effect for gender. The two-way interaction between gender and caste was significant [F(2,1174)=9.30, p<.10]. In the following sections, each vignette was analyzed separately.

4.2.1. Personal insult

A 3 (Caste: Brahmins, Gounders, and Thevars)×2 (Gender of the Participant: male, female) analysis of variance (ANOVA) of personal insult scores found main effects for caste [F(2,1192)=9.54, p<.001] and gender [F(1,1192)=6.51, p<.05] (Table 4). No other interactions were found to be significant. Bonferroni post hoc analysis of mean differences between caste groups found that Gounders more strongly endorsed violent response to personal insult than Brahmins (p<.0001) and Thevars (p<.001). Gounder and Thevar women endorsed using violence in response to a personal insult more strongly than Gounder and Thevar men, respectively. However, independent t test found that the gender difference was significant only for Thevars [t(398)=2.72, p<.008].

4.2.2. Family insult

A 3 (Caste: Brahmins, Gounders, and Thevars)×2 (Gender of the Participant: male, female) ANOVA of family insult scores found a main effect only for caste [F(2,1190)=10.05, p<.001]. No other interactions were found to be significant. Bonferroni post hoc analysis found that Brahmins and Gounders endorsed violent reaction to family insult more strongly than Thevars (Thevar vs. Brahmins, p<.001; Thevar vs. Gounders, p<.007).

4.2.3. Caste insult

A 3 (Caste: Brahmins, Gounders, and Thevars)×2 (Gender of the Participant: male, female)×2 (Story Type: cheating husband, cheating wife) ANOVA of caste insult scores found a main effect only for caste [F(2,1191)=12.82, p<.001] and gender [F(1,1191)=50.44, p<.001]. The two-way interaction between caste and gender was significant [F(1,1191)=10.54, p<.001]. Men endorsed a violent response to caste insult more than women in all caste groups, and the gender difference was significant for all caste groups [Brahmins, t(391)=−2.49, p<.05; Gounders, t(403)=−7.40, p<.001; Thevars, t(397)=−2.13, p<.05].

4.2.4. Village insult

A 3 (Caste: Brahmins, Gounders, and Thevars)×2 (Gender of the Participant: male, female) ANOVA of village insult scores found no significant difference between caste groups or between men and women.

4.2.5. CHS

A 3 (Caste: Brahmins, Gounders, and Thevars)×2 (Gender of the Participant: male, female)×2 (Story Type: cheating husband, cheating wife) ANOVA of judgment about cheating scores found a main effect for caste [F(2,1179)=5.19, p<.007] and story type [F(1,1179)=44.11, p<.001]. The interactions between caste and gender [F(2,1179)=21.87, p<.001], between caste and story type [F(1,1179)=4.79, p<.05], and between gender and story type [F(2,1179)=5.92, p<.004] were significant. The three-way interaction between caste, and between gender and story type was also significant [F(2,1179)=5.20, p<.007]. In general, members of all caste groups judged female infidelity more negatively than male infidelity (for summary of means, see Table 6).

Table 6.

Attitudes toward spousal infidelity by participants' caste and gender for Study 3

Brahmins Gounders Thevars
Male infidelity
Men 2.84 (1.41) 2.2 (1.67) 3.15 (1.77)
Women 2.54 (1.54) 3.5 (1.60) 3.04 (1.69)
Female Infidelity
Men 2.57 (1.60) 1.92 (1.55) 2.56 (1.64)
Women 2.4 (1.53) 1.91 (1.49) 2.33 (1.55)

The means for each value are shown on a scale from 1 to 5, with higher scores indicating higher tolerance of infidelity.

Although Brahmin women judged female spousal infidelity more negatively than Brahmin men did, this difference was not significant. There was a main effect for story type among Thevars. They judged female spousal infidelity more negatively than male infidelity [F(1,392)=42.15, p<.001]. For Thevars, the gender difference was not significant. For Gounders, there was a main effect for the gender of the research participant [F(1,401)=14.68, p<.001] and story type [F(1,401)=38.99, p<.001], and the interaction between gender and story type was also significant [F(1,401)=15.29, p<.001]. Gounder men judged female infidelity more strongly than male infidelity. A regression analysis found that the CHS was the strongest predictor of attitude toward female spousal infidelity for Thevar men (b=0.36, p<.01) and Gounder women (b=0.29, p <.01). Both Gounders and Thevars judged female infidelity more strongly than did Brahmins.

4.3. Discussion

Intersections of caste and gender influenced participants' attitudes toward responding to different kinds of insults. Thevar and Gounder women felt that it was appropriate to react violently to personal insults. Brahmins and Gounders endorsed using violence in response to family insult more strongly than Thevars. Men in all caste groups endorsed violent action to caste and family insults than women in all caste groups. There were no gender or caste differences in the endorsement of violent reaction to village insults. The mean scores for judgments for personal insult vignette were lowest for all caste groups (Table 4). It seems that Thevar and Gounder women expected men to actively react to personal insults. This is consistent with Courtright's (1996) observation that men in male-biased sex ratio communities are expected to be aggressive and hypermasculine (also see Hudson & den Boer, 2004). The prestige associated with being a landowning group (Gounders) and the topmost caste group in the caste hierarchy (Brahmins) might have influenced Brahmin and Gounder men's relatively low tolerance of family and caste insults.

A main aim of this study was to establish the prevalence and the strength of honor culture among Thevars. As predicted, CHS was the strongest predictor for attitude toward female infidelity for Thevar men who, as warriors, believe in the need for sons to fight and stand up for their group and family against any insults. Several researchers have noted that in honor cultures, spousal fidelity (particularly female chastity) is highly valued (Giovannini, 1987, Lindisfarne, 1998, Vandello & Cohen, 2003, Weisfeld, 1993). Thevars, particularly Thevar men, viewed spousal infidelity as an affront to personal honor. Another interesting finding was that culture of honor was also a significant predictor for attitude toward female infidelity for Gounder women. Since transferring ancestral land to proper progeny is a major concern for agricultural caste groups, they also seem to value female chastity. This is consistent with ethnographic findings in other agricultural communities (Ortner, 1974). In addition, patriarchal hegemony may also contribute to Gounder women's responses. Gounder women may have internalized legitimizing patriarchal ideologies that prescribe low tolerance for female infidelity (Jost, Banaji, & Nosek, 2004). As expected, all these significant relationships were true only when judging the violent reaction of a male protagonist, not that of female protagonist. Intersections of ecological context, caste, and gender play a critical role in shaping attitudes toward insults and male and female infidelity.

5. General discussion

The major goal of this paper has been to investigate the relationship between cultural ecological correlates of beliefs about gender that contribute to extreme forms of female neglect. The field studies addressed whether caste-specific cultural and ecological factors affect son preference in unique ways among Gounders and Thevars. Findings from the studies described above indicate that agricultural caste groups prefer sons to daughters because of concern that a daughter will take land resources away from her natal family. The warrior castes prefer to have more sons to stand up for them during intergroup conflicts. Considerations of hypergyny are also likely to play a role in the desire not to have too many daughters.

Studies 1 and 2 explored the claim that Gounders' son preference is linked to the preservation of family land. Study 1 found no strong son preference among Thevars and Gounders. Instead, a major portion of them favored an equal distribution of family property. To check whether social desirability might have played a role in the higher number of responses favoring equal distribution of land between sons and daughters, Study 2 asked participants to choose between giving ancestral land to a son with all-male offspring and giving an equal value in cash to a son with all-female offspring. The findings in Study 2 reveal that to inherit ancestral land, it is crucial to have more male offspring than other sons. Attitudes toward gender and extreme forms of son preference are indeed connected to acquiring land resources because acquiring more lands enhances one's social standing and family resources. Interestingly, in both caste groups, a nonsignificant higher proportion of women than men chose to allocate more lands to the son with more male offspring.

It was expected that Gounders, but not Thevars, would exhibit strong son bias in resource allocation tasks. An intriguing revelation from Studies 1 and 2 was that Thevars' attitudes toward resource allocation were similar to that of Gounders'. Why is this the case? Interesting ecological developments in the region where Thevars live may explain this anomaly in the findings. The region Usilampatti, where the data were collected and is predominantly inhabited by Thevars, used to be an arid land. When a Thevar politician became a member of the parliament in the 1970s, he initiated a dam project to irrigate the land. According to the developmental economist Vishwanath (1996), after the successful completion of the dam, the land became more arable and the value of the land in the Usilampatti area went up. This historical development in the region where Thevars live, the building of a dam, and resultant increases in land values and marriage costs could have contributed to this shift in Thevars' attitude and intensified son bias. An inadvertent consequence of the dam project, Thevars' attitudes toward land may have come to mirror that of Gounders'. Perhaps Thevars were copying the practices of Gounders. Such copying of a high-prestige group's behavior has been noted by several anthropologists (Dube, 2001, Henrich & Gil-White, 2001).

To document the existence of a strong honor culture among Thevars, a culture-of-honor task was developed. Individuals' attitude toward insults also strongly predicts their intolerance of spousal infidelity. The relationship was strongest for Thevars, followed by Gounders and Brahmins. Those who were high on honor strongly endorsed the violent response of a protagonist in reaction to a spouse who was cheating. In general, female infidelity was viewed more negatively than male infidelity. Spousal fidelity is a central part of a man's honor, and chastity is a highly valued trait in honor cultures. A woman can be returned to her family if she is found not be a virgin on her wedding night. Although spousal fidelity is a valued trait in all cultures (Buss, 1999), a wife's fidelity takes on an added dimension in honor cultures because a man's honor in the community diminishes if he is known to be a cuckold (Weisfeld, 1993). In many honor cultures, spousal killings for infidelity, called “honor killings,” are still quite common (Weisfeld, 1993). Perpetrators are rarely prosecuted and, even when prosecuted, typically receive a lenient sentence.

Interestingly, the mean scores for the honor task for Gounders were higher than for Thevars. However, CHS was the strongest predictor of approving a violent reaction to spousal infidelity for Thevar men. The mean scores in the honor task were lowest for Brahmins. Although they did not approve of spousal infidelity, their spousal expectations were independent of their views of honor, whereas for Thevars, spousal fidelity is constitutive of their honor culture. Another intriguing finding was that CHS was also the best predictor of Gounder women's attitudes toward female infidelity. Ortner (1974) points that agricultural communities place a high premium on women's chastity and premarital virginity to ensure that family properties are transferred to legitimate heirs. Mahalingam (in preparation) developed a chastity belief scale that measured attitudes toward female chastity. In this study, Thevars and Gounders endorsed chastity beliefs more strongly than Brahmins did.

A combination of paucity of women and the consequent high number of unmarried eligible men with no prospect of finding partners from their caste results in a situation where chances of extramarital affairs are high (Hudson & den Boer, 2004). In such situations, women's chastity is believed to preserve the family honor. Not surprisingly, men have a higher cultural expectation of female fidelity than do women. A confluence of caste and ecological factors augments a universal tendency to hold women's fidelity in higher esteem than men's. Evidence from Thevars' cultural narratives also supports the high value placed on women's chastity among Thevars (Muthulakshmi, 1997).

I predicted that psychological antecedents to female neglect and son preference would be caste specific and that a combination of cultural and ecological factors would affect various beliefs about gender. These field studies suggest that, in their resource allocation tasks, Thevars and Gounders differ from Brahmins significantly in favoring sons, as well as in attitudes toward female fidelity, culture-of-honor beliefs, and social distance measures. The overlap between Thevars and Gounders could be related to ecological changes in the Thevar area. Their son preference, which originally emerged from their honor culture, has also spread to their attitude toward land holdings. Thevar female infants who were at risk due to honor reasons are doubly at risk in an emerging agricultural economy. This further contributes to gender discrimination resulting from resource concerns typical of agricultural caste groups. The evoked culture of these caste groups remains complex and dynamic, yet predictable. Female neglect and a combination of resource constraints and attitudes typical of honor cultures have shaped the cultural psychology of gender in complex ways. An excessive male population contributes to a culture of idealization of masculine and feminine beliefs.

Ecological factors influence the differential treatment of sons of daughters and extreme forms of gender discrimination leading to male-biased sex ratios. There are several consequences. Male-biased sex ratios could contribute to a culture that valorizes masculine and feminine beliefs. Fraternal polyandry may be another option (Goldstein, 1971, Goldstein, 1976). In India, although there are no documentations of the practice of polyandry in communities with male-biased sex ratios, recent newspaper reports suggest that the practice of polyandry is emerging among Jats in Punjab, another landowning caste group with a long history of male-biased sex ratios (Garg, 2005):

[Fraternal polyandry] may be the world's rarest form of marriages in anthropological terms, but in Mansa district of the state, this is a reality that stares you in the face. Visit any village of Boha area—Gandu Kalan, Gandu Khurd, Rendod Khurd, Bhakhrial, Aan-diawali, or Khandkalan—there are families of up to seven brothers married to one woman.

There is considerable debate over whether fraternal polyandry may enhance inclusive fitness (Beall & Goldstein, 1981). Ecological, sociocultural, and political factors can perpetuate mating systems that “entail substantial reproductive sacrifice” (Beall & Goldstein, 1981, p. 11). Cultural beliefs about gender, caste hierarchy, attitudes toward resource allocation, and shifts in caste politics, such as the emergence of Dalits as a political force, may foster a highly aggressive and competitive male culture among Thevars and Gounders. Men have to be successful and competitive in order to find a mate. The socioeconomic mobility of women seems to have a paradoxical effect on the treatment of daughters. Since others in these communities are under tremendous pressure to ensure the financial and academic success of their sons, highly educated women in male-biased sex ratio communities tend to invest more in their sons than in their daughters (Das Gupta & Visaria, 1996). It seems that a combination of economic success and internalization of idealized notions of masculinity may improve the reproductive success of men in these communities. However, such idealized notions of masculinity could also be a source of stress affecting the psychological well being of men (Gilmore, 1990, Mahalingam, 2007, Yim & Mahalingam, 2006). Future research should examine the costs and benefits of the internalization of idealized notions of masculinity and femininity.

6. Conclusions

Extreme systemic female neglect has been reported in India for centuries. A cultural ecological perspective is valuable in identifying the complex relationship among ecology, caste history, and female neglect. An evolutionary psychology explanation of female infanticide should take into account the role of cultural practices and ecological pressures in shaping psychological antecedents to female neglect. This paper has contributed toward building a cultural ecological perspective that synthesizes cultural and ecological models to identify specific psychological correlates that contribute to the perpetuation of extreme forms of gender discrimination.

The desire to preserve land for the immediate family plays a major role in Gounders' desire to have more sons and is an emerging trend among Thevars. Because of the cultural practice of women moving into their husband's family, daughters are viewed as a threat and drain to the natal family resources. For Thevars, son preference is tied to the need to protect family honor, and the dowry cost of marrying daughters to a wealthy family disfavors the desire for more female children (Gaulin & Boster, 1990). In the case of Thevars, the emerging agricultural economy, due to recent ecological changes, also amplifies their existing son biases.

I argue that the relationship between ecology and culture is dynamic, and that macrointerventions, such as building of dams and empowerment of women, result in stricter social control of women's sexuality. Women's preference to marry up (hypogyny) further contributes to intense male–male competition among Thevars and Gounders. The complex interaction between ecology, cultural practices, and attitudes toward gender among these caste groups needs to be understood before meaningful interventions to prevent female infanticide can be designed.


I am extremely thankful for the thoughtful comments of two anonymous reviewers and the editors Martin Daly and Margo Wilson on an earlier version of this paper. I would like to thank Sundari Balan, Bobbi Low, and Viswanathan Ravishankar for their critical comments. I would also like to thank Quadir Ismail for his help with data collection.

Appendix A. Culture-of-honor vignettes

(1) Personal insult

Dorai was waiting for his bus. He was in line, and one person who came late pushed Dorai down, went ahead of him in line, and did not even bother to apologize. Dorai got mad at him and yelled at him.
What do you think of what Dorai did?
1 2 3 4 5
Strongly agree Strongly disagree

(2) Family insult

Malaiappan had a difficult relationship with his neighbor Palanisamy due to dispute over a property. One day, in the middle of an argument, Palanisamy insulted Malaiappan by questioning the fidelity of his mother. Malaiappan got angry, took a hatchet, cut Palanisamy's hand, and injured him badly.
What do you think of what Malaiappan did?
1 2 3 4 5
Strongly agree Strongly disagree

(3) Caste insult

Kumaran borrowed some money from Marudu. He promised to repay in 2 months. Kumaran could not return the money in 2 months because of some unexpected events. He told Marudu that he would return the money on the next month. Marudu was not happy about this and insulted Kumaran that he has shown his “caste buddhi” (meaning caste trait) and is not trustworthy. Kumaran got angry and slapped Marudu for his remark.
What do you think of what Kumaran did?
1 2 3 4 5
Strongly agree Strongly disagree

(4) Village insult

Chinnappan lives in Malaiyur. The neighboring village is Therur. The relationship between the two villages was not good. Chinnappan lends standby generators for village festivals for hire. For the temple festival at Therur, Chinnappan lent his generator. When the power went off, the generator did not work. Chinnappan was trying his best to fix the generator. One of the festival organizers at Malaiyur blamed Chinnappan, saying that he deliberately did this to spoil the spirit of the festival and insulted Chinnappan by saying that “Malaiyurans are untrustworthy.” Chinnappan was upset, and he hit the organizer and broke his nose.
What do you think of what Chinnappan did?
1 2 3 4 5
Strongly agree Strongly disagree

(5) Spousal infidelity

Kanniappan came home 1 day. He found his wife in bed with a friend of his best friend. In a rage of anger, Kanniappan took a knife and stabbed his wife before his friend escaped.
What do you think of what Kanniappan did?
1 2 3 4 5
Strongly agree Strongly disagree


Beall & Goldstein, 1981 1.Beall CM, Goldstein MC. Tibetan fraternal polyandry: A test of sociobiological theory. American Anthropologist. 1981;83(1):5–12.

Bogardus, 1933 2.Bogardus ES. Social distance scale. Sociology and Social Research. 1933;17:265–271.

Boswell, 1990 3.Boswell J. The kindness of strangers: The abandonment of children in Western Europe from late antiquity to the renaissance. New York: Vintage Press; 1990;.

Broude, 1975 4.Broude GJ. Norms of premarital sexual behavior. Ethos. 1975;3:381–402.

Broude, 1980 5.Broude GJ. Extramarital sex norms in cross-cultural perspective. Behavioral Science Research. 1980;15:181–218.

Bugos & McCarthy, 1984 6.Bugos PE, McCarthy LM. Ayoreo infanticide: A case study. In:  Hausfater G,  Hrdy SB editor. Infanticide: Comparative and evolutionary perspectives. New York: Aldine; 1984;p. 503–520.

Buss, 1999 7.Buss DM. Evolutionary psychology: The new science of the mind. Boston: Allyn and Bacon; 1999;.

Buss & Kenrick, 1998 8.Buss DM, Kenrick DT. Evolutionary social psychology. In:  Gilbert DT,  Fiske ST,  Lindzey G editor. Handbook of social psychology. Vol. 2:New York: McGraw-Hill; 1998;p. 982–1026.

Chagnon, 1988 9.Chagnon NA. Life histories, blood revenge, and warfare in a tribal population. Science. 1988;239:985–992.

Charnov, 1982 10.Charnov EL. The theory of sex allocation. Princeton, NJ: Princeton University Press; 1982;.

Clarke & Low, 1992 11.Clarke AL, Low BS. Ecological correlates of human dispersal in 19th century Sweden. Animal Behavior. 1992;44:677–693.

Cohen & Nisbett, 1994a 12.Cohen D, Nisbett RE. Field experiments explaining the culture of honor: Explaining southern violence. Personality and Social Psychology Bulletin. 1994;20:551–567. CrossRef

Cohen & Nisbett, 1994b 13.Cohen D, Nisbett RE. Self-protection and the culture of honor: The role of violence. Journal of Personality and Social Psychology. 1994;20:551–567.

Cohen et al., 1998 14.Cohen D, Vandello J, Rantilla A. The sacred and the social cultures of honor and violence. In:  Gilbert P editors. Shame: Interpersonal behavior, psychopathology, and culture. New York: Oxford University Press; 1998;p. 261–282.

Courtright, 1996 15.Courtright DT. Violent land: Single men and social disorder from the frontier to the innercity. Cambridge, MA: Harvard University Press; 1996;.

Das Gupta & Visaria, 1996 16.Das Gupta M, Visaria L. Son preference and excess female mortality in India's demographic transition. In: Sex preferences for children and gender discrimination in Asia. Seoul, Korea: Korea Institute for Health and Social Affairs, UNFPA; 1996;p. 6–102.

Dickemann, 1979 17.Dickemann M. Female infanticide, reproductive strategies, and social stratification: A preliminary model. In:  Chagnon NA,  Irons W editor. Evolutionary biology and human social behavior: An anthropological perspective. North Scituate, MA: Duxbury Press; 1979;p. 321–368.

Dickemann, 1981 18.Dickemann M. Paternal confidence and dowry competition: A biocultural analysis of Purdah. In:  Alexander RD,  Tinkle DW editor. Natural selection and social behavior. New York: Chiron Press; 1981;p. 417–438.

Divale & Harris, 1976 19.Divale WT, Harris M. Population, warfare and the male supremacist complex. American Anthropologist. 1976;78:521–538.

Dube, 2001 20.Dube L. Anthropological explorations in gender: Intersecting fields. Thousand Oaks: Sage; 2001;.

Ember & Ember, 1994 21.Ember CR, Ember M. War, socialization and interpersonal violence: A cross cultural study. Journal of Conflict Resolution. 1994;38:620–646.

Fisher, 1958 22.Fisher RA. The genetical theory of natural selection. New York: Dover; 1958;.

Frayser, 1985 23.Frayser S. Varieties of sexual experience: An anthropological perspective on human sexuality. New Haven, CT: Human Area Files Press; 1985;.

Fuchs, 1989 24.Fuchs R. Abandoned children: Foundlings and child welfare in nineteenth-century France. Albany: SUNY Press; 1989;.

Garg, 2005 25.Garg B. Draupadis bloom in rural Punjab. Times of India. 2005, July 16;.

Gaulin & Boster, 1990 26.Gaulin SJC, Boster J. Dowry as female competition. American Anthropologist. 1990;92:994–1005.

Gilmore, 1990 27.Gilmore DD. Manhood in the making: Cultural concepts of masculinity. New Haven, CT: Yale University Press; 1990;.

Giovannini, 1987 28.Giovannini MJ. Female chastity codes in the circum-Mediterranean: Comparative perspectives. In:  Gilmore DD editors. Honor and shame and the unity of the Mediterranean. Washington, DC: American Anthropological Association; 1987;p. 61–74Anthropology of the Mediterranean. In U. Cambridge (Ed.), Cambridge University Press..

Goldstein, 1971 29.Goldstein MC. Stratification, polyandry, and family structure in central Tibet. Southwestern Journal of Anthropology. 1971;27(1):64–74.

Goldstein, 1976 30.Goldstein MC. Fraternal polyandry and fertility. Human Ecology. 1976;4(3):223–233.

Harriss-White, 2001 31.Harriss-White B. Gender-cleansing: The paradox of development and deterioration female life chances in Tamilnadu. In:  Sunderrajan R editors. Signposts: Gender issues in post-independence. 2001;[New Brunswick, NJ].

Henrich & Gil-White, 2001 32.Henrich J, Gil-White FJ. The evolution of prestige: Freely conferred deference as a mechanism for enhancing the benefits of cultural transmission. Evolution & Human Behavior. 2001;22(3):165–196.

Hill & Low, 1991 33.Hill E, Low B. Contemporary abortion patterns: A life-history approach. Ethology and Sociobiology. 1991;13:35–48.

Hrdy, 1999 34.Hrdy SB. Mother nature: A history of mothers, infants and natural selection. New York: Pantheon; 1999;.

Hudson & den Boer, 2004 35.Hudson VM, den Boer AM. Bare branches: The security implications of Asia's surplus male population. Cambridge, MA: MIT Press; 2004;.

Hughes, 1988 36.Hughes AL. Evolution and human kinship. Oxford: Oxford University Press; 1988;.

Jeffery & Jeffery, 1997 37.Jeffery R, Jeffery P. Population, gender, and politics. Cambridge: Cambridge University Press; 1997;.

Jost et al., 2004 38.Jost JT, Banaji MR, Nosek BA. A decade of system justification theory: Accumulated evidence of conscious and unconscious bolstering of the status quo. Political Psychology. 2004;25(6):881–991.

Kakar, 1978 39.Kakar S. The inner world: A psycho-analytic study of childhood and society in India. New Delhi, India: Oxford University Press; 1978;.

Keller et al., 2001 40.Keller MC, Nesse RM, Hofferth S. The Trivers–Willard hypothesis of parental investment: No effect in the contemporary United States. Evolution and Human Behavior. 2001;22:343–360.

Kenrick & Luce, 2000 41.Kenrick DT, Luce CL. An evolutionary life-history model of gender differences and similarities. In:  Eckes T editors. The developmental social psychology of gender. Mahwah, NJ: Lawrence Erlbaum; 2000;p. 35–64.

Krishnaswamy, 1988 42.Krishnaswamy S. Female infanticide in contemporary India: A case study of Kallars of Tamilnadu. In:  Ghadially R editors. Women in society. New York: Sage; 1988;p. 186–195.

Levine, 1987 43.Levine N. Differential child care in three Tibetan communities: Beyond son preference. Population and Development Review. 1987;13:281–304.

Levine & Silk, 1997 44.Levine NE, Silk JB. Differential child care in three Tibetan communities: Beyond son preference. Current Anthropology. 1997;38(3):375–398. CrossRef

Lindisfarne, 1998 45.Lindisfarne N. Gender, shame, and culture: An anthropological perspective. In:  Gilbert P editors. Shame: Interpersonal behavior, psychopathology, and culture. New York: Oxford University Press; 1998;p. 246–260.

Low, 1990a 46.Low BS. Sex, power, and resources: Ecological and social correlates of sex differences. Journal of Contemporary Sociology. 1990;27:45–71.

Low, 1990b 47.Low BS. Human responses to environmental extremeness and uncertainty: A cross-cultural perspective. In:  Cashdan E editors. Risk and uncertainty in tribal and peasant economies. Boulder, CO: Westview Press; 1990;p. 229–255.

Low, 1993 48.Low BS. Ecological demography: A synthetic focus in evolutionary anthropology. Evolutionary Anthropology. 1993;1993:106–112.

Low, 2000 49.Low BS. Why sex matters. Princeton University Press; 2000;.

Mahalingam, 2007 50.Mahalingam R. Beliefs about chastity, machismo and caste identity: A cultural psychology perspective. Sex Roles. 2007;56(3–4):239–249.

Miller, 1981 51.Miller BD. The endangered sex: Neglect of female children in rural North India. Ithaca: Cornell University Press; 1981;.

Muthulakshmi, 1997 52.Muthulakshmi R. Female infanticide: Its causes and solutions. New Delhi: Discovery Publishing House; 1997;.

Natarajan, 1997 53.Natarajan S. Watering the neighbor's plant: Media perspectives on female infanticide in Tamilnadu, Monograph no. 6. Chennai, India: M. S. Swaminathan Research Foundation; 1997;.

Nisbett, 1993 54.Nisbett RE. Violence and U.S. regional culture. American Psychologist. 1993;48(4):441–449. CrossRef

Nisbett & Cohen, 1996 55.Nisbett RE, Cohen D. Culture of honor: The psychology of violence in the south. Boulder, CO: Westview Press; 1996;.

Ortner, 1974 56.Ortner S. Is male to female as nature is to culture?. In:  Lamphere MRL editors. Woman, culture, and society. Stanford, CA: Stanford University Press; 1974;p. 67–88.

Ramamoorthy, 2002 57.Ramamoorthy V. Koottaamchoru. Chennai, India: Sandhya Publications; 2002;.

Ransel, 1988 58.Ransel D. Mothers in misery: Child abandonment in Russia. Princeton: Princeton University Press; 1988;.

Rousselle, 2001 59.Rousselle R. “If it is a girl, cast it out.” Infanticide/exposure in Ancient Greece. Journal of Psychohistory. 2001;28(3):303–333. MEDLINE

Sherwood, 1988 60.Sherwood J. Poverty in eighteenth century Spain: Women and children of the inclusa. Toronto: University of Toronto Press; 1988;.

Thurston, 1909 61.Thurston E. Caste and tribes in Southern India. Madras: Government Press; 1909;.

Tooby & Cosmides, 1992 62.Tooby J, Cosmides L. The psychological foundation of culture. In:  Barkon LCJH,  Tooby J editor. The adapted mind: Evolutionary psychology and the generation of culture. New York: Oxford University Press; 1992;p. 19–136.

Torres & Forrest, 1988 63.Torres A, Forrest JD. Why do women have abortions?. Family Planning Perspectives. 1988;20(4):169–176. MEDLINE | CrossRef

Trivers & Willard, 1973 64.Trivers RL, Willard DE. Natural selection of the parental ability to vary the sex ratio of offspring. Science. 1973;179:90–92. MEDLINE

Vandello & Cohen, 2003 65.Vandello JA, Cohen D. Male honor and female fidelity: Implicit cultural scripts that perpetuate domestic violence. Journal of Personality and Social Psychology. 2003;84(5):997–1010. MEDLINE | CrossRef

Venkatachalam & Srinivasan, 1993 66.Venkatachalam R, Srinivasan V. Female infanticide. New Delhi: Har-Anand Publications; 1993;.

Vishwanath, 1996 67.Vishwanath LS. Female Infanticide and the position of women in India. In:  Shah AM,  Baviskar BS,  Ramaswamy EA editor. Social structure and change. Vol. 2:New Delhi: Sage Publications; 1996;.

Weisfeld, 1993 68.Weisfeld GE. Social status and values in traditional Arab cultures. In:  Lee E editors. Comparative biosocial analysis. Vol. 1:Westport, CT: Praeger; 1993;p. 75–97.

Whyte, 1978 69.Whyte MK. Cross-cultural codes dealing with the relative status of women. Ethology. 1978;17:211–237.

Whyte, 1979 70.Whyte MK. The status of women in pre-industrial society. Princeton, NJ: Princeton University Press; 1979;.

Yim & Mahalingam, 2006 71.Yim J, Mahalingam R. Culture, masculinity, and psychological well-being in Punjab, India. Sex Roles. 2006;55(9–10):715–724.

Department of Psychology, University of Michigan, Ann Arbor, MI 48109-1052, USA

Department of Psychology, 3263 East Hall, 512 Church Street, University of Michigan, Ann Arbor, MI 48109-1052, USA

 The research was supported by a faculty seed grant from the Rackham School of Graduate Studies, University of Michigan.

PII: S1090-5138(07)00019-0