Death by family: Honor violence as punishment

The offense

Honor conflicts are traditionally said to arise from violations of female chastity, real or imagined (see, for example, Faqir, 2001: 69). Chastity requires women to present an asexual self in everyday life, to be virgins on the wedding night, and from then out, to have sex solely with their husbands (see, for example, Abu-Odeh, 1996: 150). Beyond chastity, honor extends to relational propriety more generally: in many communities, “social intercourse between unrelated men and women is considered almost the same as sexual intercourse” (Akpinar, 2003: 432). But not all breaches of chastity are even relational. Honor commands women to be subservient to men more broadly (Hasan, 2002: 3–31). Much honor violence punishes autonomy rather than just improper relationships – a woman insisting on completing her education, abandoning an abusive husband, taking a job, or even dressing as she wishes (see, for example, Hasan, 2002: 4; İnce et al., 2009: 539; Pope, 2012: 63). Surveying the honor killing of 230 individuals reported in English-language media around the world, 1989–2009, Chesler (2010) found that 58 percent of victims were slain for being too independent or “western”. What exactly is dishonorable and why it varies across social settings is an important question that will not be pursued here. ⁹ The present analysis holds constant the conduct deemed to be the dishonorable, proposing that the same offense may attract penalties of widely varying severity in different cases.

The honor violated in honor conflicts is that of a group, usually a family. Committed by or on behalf of a group, the collective nature of honor violence differentiates it from the defense of individual honor raised by jealous or jilted men who kill wives or girlfriends in certain Latin American countries, or the individualistic violence of intimate partners in western societies (see, for example, Chesler, 2009; Pimentel et al., 2005). Honor violence is further distinguished from western family violence by being more premeditated (killings are often planned at a family meeting), one-directional (female-on-male incidents are rare), and severe (even a rape leaves a woman vulnerable to punishment) (see, for example, Idriss, 2011: 3; Payton, 2011: 73).

Honor violations have serious social consequences, sharply reducing a family’s social standing – particularly that of its male members. A dishonored family may be excluded from community activities, bear the brunt of mockery and gossip, and experience difficulty finding marriage partners. But lost honor can be regained (see, for example, Sev’er and Yurdakul, 2001: 974). As Pitt-Rivers (1966: 29) stated in a classic essay, “the ultimate vindication of honor lies in physical violence”. Hence, a woman who dishonors her family may be beaten or even killed to restore the family’s status. “‘Now we can walk with our heads held high,’ eighteen-year old Anal said after her sixteen-year old brother had killed their sister” (quoted in Wikan, 2008: 20).

Like most scholarship on violence, the literature on honor violence tends to select on the dependent variable and focus on those cases that did result in assault or homicide. Hence, it often conveys the impression that most honor offenses trigger violence, even death. They do not. As in every system of punishment, sanctions for honor offenses are gradated. The ethnographic record makes clear that families react to dishonor in a variety of ways, including by tolerating the offense, issuing a warning, arranging a marriage, expelling the woman to another location, and seeking compensation from the man’s family (Antoun, 1968; Ginat, 1997 [1987]; Kardam 2005; Van Eck, 2003; Wikan, 1982). The central question then is: what explains the severity of the family’s punishment? In particular, when will the family inflict capital punishment?

Most theories of punishment are of little assistance. Virtually all such theories address aggregate-level patterns: variation in the severity of punishment across groups, such as cities, states, or countries (see Garland, 1990). They do not explain the case-by-case variation that forms the focus of the present inquiry: differences in the probability and severity of punishment across individual conflicts. A notable exception is the work of Donald Black and others, rooted in Black’s distinctive theoretical perspective known as pure sociology (or “Blackian theory”).

Pure sociology

Pure sociology explains social behavior with its position and direction in social space, its social geometry. Social space is multi-dimensional. Interactions between individuals and groups have a direction and distance in each dimension. Theft committed by a poor man against a wealthy man, for instance, has an upward direction in vertical space; the greater the wealth gap between them, the greater the vertical distance of the theft. A directive issued to citizens by an authoritarian government travels farther downward in organizational space than one issued by a democratic government. The sexual assault of a wife or girlfriend spans less relational distance than that of a stranger. And so on.

Variation in social space is associated with variation in conduct. Change the relative elevation (status) of actors or the social distance between them and their behavior changes as well. Beginning with Black’s (1976) seminal work on law, this line of thinking has generated a considerable body of theory (see Campbell, 2011; Senechal de la Roche, 1996). Most of the work has addressed forms of social control, specifying the geometrical conditions under which conflicts attract, for example, avoidance (Baumgartner, 1988), informal therapy (Tucker, 1999), or ethnic hostility (Cooney, 2009a).

Drawing on Black’s (1983) insight that much crime is a form of social control (or the management of conflict), the pure sociology of violence has been a particularly active field. Pure sociologists have isolated the geometrical conditions underlying various types of violent conflict management. These include violence among individuals (suicide and homicide), between groups (feuding), and between groups and individuals (rioting, lynching, vigilantism, terrorism, and genocide) (Black, 2004a, 2004b; Campbell, 2009; Cooney, 1998; Manning, 2012; Phillips, 2003; Senechal de la Roche, 1996).

Honor violence is a species of group-on-individual, or collective, violence. Hence, some of the variables that explain collective violence should also explain honor violence. The type of collective violence closest to honor violence is what Senechal de la Roche (2001: 133–136) calls a communal lynching, in which a highly solidary community assaults or kills an insider, such as a recidivistic thief, liar, killer, or alleged witch. But there are differences: in honor violence, the group inflicting violence is not the community as such but the social unit on which the community is based – the family. A lynching is typically perpetrated by a group of people acting jointly, but honor violence is typically perpetrated by an individual acting alone, though as an agent of a group. ¹⁰ And while a lynching group is typically an ad hoc collection of individuals who come together for a single purpose, honor violence is committed by a more permanent group whose activities extend well beyond social control of its members. For these reasons, honor violence most closely resembles punishment – a group disciplining one of its own (Black, 1990: 78–79). Honor violence bears a particular likeness to legal punishment. Both are triggered by offending conduct (actual or alleged), both target the individual deviant (rather than surrogates), both are inflicted by a group to which the offender owes allegiance, and in both, the most intense form is lethal. From an explanatory perspective, honor killing is thus best seen as a less formal, more communal, type of capital punishment. ¹¹

The pure sociology of punishment finds its most developed statement in Black’s (1993) theory of moralism (the making of enemies). ¹² The theory explains variation in the frequency and severity of punishment with a small number of general principles. Two such principles, both of which can be applied to honor violence, are:

Punishment increases with social distance.

Closeness and mitigation

The occasions on which women could be killed far outweigh those in which they are actually killed. Despite what the members of an honor culture might say about killing automatically following the revelation of an honor offense, qualitative research from settings as diverse as Morocco, Pakistan, Turkey, Israel, Holland, and Sweden reveals otherwise. In reality, families are slow to inflict the ultimate penalty (see, for example, Davis and Davis, 1989: 78). Even among highland Pakistanis – who appear to be among the strictest in matters of honor – discovery of an illicit affair does not necessarily lead to an immediate killing: the woman’s family may wait and see what transpires before deciding to execute her (Knudsen, 2009: 101–103). Families everywhere tend to look first for a solution short of homicide that is consistent with the maintenance of their social ranking. Thus, family members may deny the infraction, refusing to recognize that it has occurred until the evidence becomes overwhelming or it becomes the subject of gossip (see, for example, Onal, 2008: 54–68). They may warn her not to get further involved with a man with whom she has been seen (see, for example, Onal, 2008: 189). They may try to persuade her to break off a relationship that has already started (see, for example, Ginat, 1997 [1987]: 135). They may arrange an abortion if she is pregnant (see, for example, Ginat, 1997 [1987]:155). They may attempt to suppress news of the incident, or lie about what happened (see, for example, Kardam, 2005: 38). They may banish her from the household (see, for example, Davis and Davis, 1989: 78). They may arrange a marriage to the lover or some other man (see, for example, Ginat, 1997 [1987]: 158–159, 175). They may seek compensation from the man’s family (see, for example, Van Eck, 2003: 171–172). Or they may beat or threaten to kill her (see, for example, Akpinar, 2003: 435–436). Some threats are real, but others appear to be just for show. Family members will sometimes bluff, putting on a public display designed to demonstrate their moral resolve even as they seek to avoid killing if at all possible (e.g. Ginat, 1997 [1987]: 163). A case among Israeli Arabs recorded by Ginat (1997 [1987]: 165) provides a rare glimpse of a family’s backstage efforts to find a solution short of murder:

Halima’s mother told me she had told all the family that should the need arise, she herself would kill her daughter. “You are all working, and shouldn’t go to prison. I am an old woman and will only be sentenced to a short term of imprisonment.” The mother explained what she meant by “should the need arise”: “If the family had pressured me to have her killed, I would have thought of a solution. Naturally I intended to do everything to avoid killing her. I thought by announcing that I personally would kill her, it would prevent my family from telling me the family’s honor is soiled.” Then the mother added, with a smile, “I am an old woman who has seen a lot and learned a lot. I have seen girls murdered, and on the other hand can show you girls that have had forbidden sex, committed adultery and yet live unharmed.”

Note, though, that the social distance principle only holds up to a point. If the parties are separated by large expanses of relational and cultural distance, little severity is like to ensue. Where the woman has already moved away and reduced or eliminated contact with her family or where she has long converted to a new culture, then she has, in effect, moved into a different world. As such, she is likely out of reach for sanctions, if indeed not exempt from them altogether. Black (1976: 40–46, 74–78) proposes that the relationship between law and social distance is curvilinear. The ultimate relationship between social distance and punishment for breaches of honor appears to describe the same shape.

Distance and severity

Honor offenses are somewhat more complex structurally than most offenses in that they have two offenders – the man and woman. The social distance between a woman and her family depends directly on her relationship with her family and indirectly on his relationship to her and her family. Thus, a woman who becomes involved with a socially distant man increases the combined distance between the offenders and her family. Such an offense attracts more punishment than an intimate offense. A woman who takes up with an outsider is therefore at greater risk than one who takes up with a man from the same village or town (see, for example, Blum and Blum, 1965: 49–50; Onal, 2008: 34, 236). In one incident in south-eastern Turkey, for example, a woman who had eloped with a man was shot and killed by her father in part because he was a “foreigner” – a resident of a nearby city rather than the village in which the family resided (Pope, 2012: 46). A woman who chooses a man from outside her own clan may similarly be in greater danger (see Scott, 2008: 9). By contrast, pairing off with a relationally close man is less likely to result in execution. In a case in an Arab village in Israel, a woman started a relationship with her husband’s younger brother. Although one of her brothers warned her about her behavior, the conflict was resolved peaceably by the husband divorcing his wife and her marrying his brother (Ginat, 1997 [1987]: 180).

Cultural distance appears to have the same effect. Violations of cultural boundaries – religious, linguistic, racial – are a familiar motif in the back story of honor killings, regardless of where they occur (see, for example, Fisk, 2010; Husseini, 2009: 117–121, 151, 153). A liaison with a culturally distant man increases the cultural distance between the woman and her family and attracts more punishment than a liaison with a culturally close man (see, for example, Chakravarti, 2005: 314–316; Ginat, 1997 [1987]: 188–189). For instance, in Turkey, a Kurdish woman who takes a Turkish lover is at greater risk than one who takes a Kurdish lover (Kardam, 2005: 34–35). Among Dutch Turks, it is more dangerous for an unmarried woman to form a relationship with a Dutch man than a Turkish man (see, for example, Van Eck, 2003: 122–126). And it is more dangerous if she, a Muslim, forms an attachment to a non-Muslim, especially if her family is devout (Van Eck, 2003: 203, 207).

Social distance, then, helps explain why some honor offenses result in homicide but most do not. So too does a second principle of punishment: social status or, in geometrical terms, social inferiority.

Patriarchy

The most glaring form of social inferiority underlying honor killing is patriarchy – the social subordination of women along economic, cultural, functional, relational, legal, and other lines. Despite often considerable local diversity, the literature reveals a consistent tendency in honor communities to separate and stratify the sexes into firm and hierarchical social categories – to equate biology with destiny (see, for example, Baker et al., 1999; Sev’er and Yurkdakul, 2001; Welchman and Hossain, 2005). Thus, societies are generally “patrilineal, patrilateral, and patrilocal: descent is traced through males, the preferred form of marriage is between children of brothers, and preferred residence after marriage is in the household of or nearby the groom’s family” (Davis and Davis, 1989: 84). Male offspring are favored: “the arrival of a baby boy is a cause for lavish celebrations, but female newborns rarely elicit such enthusiasm” (Pope, 2012: 39). Growing up, boys enjoy considerably greater freedom, and are groomed to assume a dominant role in the family and community, that begins with keeping “a sharp eye on [their] older sisters” (Kocturk, 1992: 72). For girls, by contrast, the world contracts as they grow older (Gregg, 2005: 229). Girls often receive less schooling than boys and consequently have higher rates of illiteracy (see, for example, Faqir, 2001: 66; Ikkaracan and WWHR, 1998: 67; Sev’er and Yurdakul, 2001). After puberty, girls are typically required to wear loose clothing and conceal their hair, and may be required to wear a veil. They should “avoid any contact with the opposite sex until marriage” (Eisner and Ghuneim, 2013: 407). Families usually marry off daughters at a young age, particularly in under-developed rural areas (Ikkaracan and WWHR, 1998: 69). On marrying, a woman typically goes to live in her husband’s home, where she is subject to the authority of her husband and, importantly, her mother-in-law (see, for example, Pope, 2012: 50). In adulthood, women and men often remain largely segregated, each inhabiting different physical spaces or the same spaces at different times (e.g. King, 2008: 320–321; Van Eck, 2003: 28–29). Outside the home, in the public sphere – the world of work, politics, law, and religion – men dominate. Women occupy fewer positions than men, especially positions of leadership (Ghanim, 2009: ch. 6). Cross-national data consistently reveal a larger gender gap in countries with higher rates of honor killing than in other regions of the world, particularly in the realms of economic participation and opportunity and political empowerment (Hausmann et al., 2013: 12–13, 18–24).

Women in communities of honor are by no means wholly devoid of influence, however (see, for example, Moghadam, 2007; Sadiqi and Ennaji, 2011). The degree of subordination of women varies considerably by region: male domination is much stronger among some groups (e.g. impoverished rural Afghanis) than among others (e.g. wealthy Jordanian urbanites). It varies by age. Older women often acquire a greater overt say in their own lives and that of the family (see, for example, Gregg, 2005: 339–341). It varies by social class: for example, “male authority decreases in Turkish families where the couple has high levels of income and education” (Kocturk, 1992: 82). But it remains high everywhere compared to the western world. And it appears to continue to survive, carried on by younger people. In a nationally representative 1997 survey of unmarried 16–19-year-olds in Egypt, 92 percent of boys and 88 percent of girls agreed that, “A wife needs her husband’s permission for everything.” Over 80 percent of both boys and girls alike thought the husband alone should be the breadwinner; less than 2 percent of either sex thought men should cook or wash clothes or bathe the children. These results did not vary by socioeconomic status or by region of the country (Mensch et al., 2003). ¹⁶

The extent of the gender gap matters in honor conflicts. The more pronounced gender inequality is, the greater the vulnerability to honor killing: “Women who are unemployed, illiterate and live in impoverished conditions have a higher risk than others of becoming a … victim” (Patel and Gadit, 2008: 688). Women in highland Pakistan or Iraqi Kurdistan are more at risk than women in rural Morocco (Davis and Davis, 1989: 118–125; King, 2008: 319–324; Knudsen, 2009: 98–104, 130–134). Younger women (15–25) everywhere appear to be at greater risk than older women (see, for example, Kulwicki, 2002: 80).

Patriarchy also helps to explain one of the most salient facts about honor killing: men are killed much less often than women. Although the man may be more at fault (as in rape), he is punished less severely. His family generally backs him up; her family is more likely to turn on her (but see Kardam, 2005: 37).

How the woman’s family’s responds to an unauthorized relationship will depend, however, in part on the man’s social status. A downward relationship (from the woman’s family’s perspective) lowers the social elevation of her offense – and becomes more serious. Thus, her family is likely to punish a liaison with a man from a poorer, less influential, family more severely than one with a man from a higher status family (Chakravarti, 2005: 312–314; Kardam, 2005: 34, 43; Kressel, 1981). The consequences of a higher-status liaison are much more favorable for her: “If a girl flees to a rich family, she will live” (Onal, 2008: 255). ¹⁷ But relatively few women can likely escape through high status relationships. As a young Turkish woman lamented: “Where there’s money all our customs are forgotten. When there’s money no one wants to kill the girl. They don’t want a poor husband. But we move in poor circles, where are we supposed to find the rich husbands?” (Onal, 2008: 45–46). On the other hand, a wealthier family may not want to accept a poorer woman, especially one tainted by dishonor. In a Turkish case, for example, negotiations to have a rapist marry the woman he had dishonored ran into difficulties, in part because her family was poorer than his. After the proposed compromise was rejected, the woman was killed by her brother (Kardam, 2005: 39).

Morality

Moral inferiority refers to differences in social actors’ moral status or “respectability” (Black, 1976: 111). Respectability is known by a social actor’s reputation for deviance. A respectable person has an unblemished reputation; an unrespectable person has a record of prior offenses, and the longer and more severe the prior record, the greater the unrespectability. Unrespectability increases the risk and severity of punishment of all types, legal and popular (Black, 1976: 111–117; Cooney, 2009b: ch. 6). The decisions of police, prosecutors, judges, and jurors are invariably influenced by the defendant’s criminal history (see Baumgartner, 1992: 136–141). In US capital cases, for instance, the length and severity of the defendant’s criminal record is one of the strongest predictors of a death sentence over a term of life imprisonment (see, for example, Baldus et al., 1990: 319).

Breaches of honor exhibit the same pattern. A woman from a respectable family who has a bad name herself, especially a reputation for being sexually loose, is more likely to be accused of an honor offense for the same conduct and less likely to be spared, if accused. Among Turkish-Dutch, for example:

A daughter may be killed because her family considers her irredeemably dishonorable … There is simply little to be gained by killing the male transgressor; the girl might simply go on to tarnish her family’s namus (i.e. reputation for chastity). In [one case], the girl had a bad reputation because the divorce from her first husband had been at her instigation. She initially returned home to live with her parents but then left again to move in with her boyfriend. When wedding preparations were almost complete, she left her boyfriend. All of this may have convinced [her family] that there was no hope of her behaving properly. (Van Eck, 2003: 206)

Thus, even if a woman has no prior record of deviance, she may develop one during the course of the conflict. Should she refuse to abide by her family’s solution to the problem, such as marrying an older man, she will commit a further offense likely to be viewed highly unfavorably. In a case in Sweden, for example, a young woman of Kurdish descent had an affair with a man of Iranian descent without her family’s permission. Her family cast her out, forbidding her to return to their town of residence (Wikan, 2008: 112–113). In defying their order,

she was playing with fire … [She] threatened to split the system apart from the inside. She was breaking every rule in the game, not only those concerned with sexual honor and choice of husband. She violated her exile too … [She] had become a multiple offender. (Wikan, 2008: 5)

Shortly afterwards, her father shot and killed her.

The prior deviance need not even be the woman’s own. A woman’s moral reputation can be sullied by the notoriety of a female relative: “if the mother’s reputation is in any way tainted, the girl is tarred with the same brush” (Van Eck, 2003: 204). Thus, an honor offense committed by another female family member can elevate a woman’s own risk of being killed. In an Israeli case, for instance, a girl who lost her virginity at age 12 only narrowly escaped being killed at the hands of her brother. Her younger sister, however, was not so fortunate when she became pregnant at 14: she was strangled and then decapitated by their brother (Kressel, 1981: 149–150).

A woman’s moral reputation can also be tainted by deviance on the part of her close male relatives. Of particular importance is her brother’s or father’s reputation within the extended family group. Public accusations of dishonor – a factor stressed by Ginat (1997 [1987]) – appear to be often made by those who have had prior conflicts with her close male relatives (see, for example, 1997 [1987]: 145).

The lower the relative status of the woman, then, the more vulnerable she is to being killed. But her social inferiority depends not just on how low is her status, it also varies with how high is the status of her accuser: the family.

Organization

An important form of status in honor conflicts is organization – the capacity for collective action (Black, 1976: 85). Groups, by definition, are more organized than individuals. ¹⁸ Among groups, centralization of decision making and corporateness (unity) of action enhances organization. Just as more organized states (e.g. dictatorships) find and punish deviants with greater severity than less organized states (e.g. democracies), so more organized families discipline with greater frequency and strictness than less organized families.

Families in honor communities vary considerably among themselves but, compared to their western counterparts, they tend to be highly organized in several respects. An ideal typical description of the family in honor communities – from which actual families will invariably deviate to one degree or another – would include the following elements. Large, extended, and multi-generational, the family group consists of a man, his brothers, and married sons, together with the wives and children of the men (though not all may live under the one roof). The family has a distinctive identity stretching back in time. Internally, ultimate authority rests with the highest-ranking man. However, for all members “the collective interests of the group tend to take precedence over their personal wishes” (Pope, 2012: 40). The group chooses the marriage partners of its offspring (with varying degrees of input from the young men and women themselves) (see, for example, Delany, 1991: 100–120). Marriage is endogamous (see, for example, Delaney, 1991: 100–110; Kulczycki and Windle, 2011: 1452). Property is shared: land and animals, in particular, are owned by the family rather than by its individual members. Cousin marriage has the effect of keeping property within the family, as the wealth transferred on marriage goes to kin instead of non-kin. Importantly, the family is a corporate unit, acting as a single entity toward the world. In turn, the world treats its members as interchangeable and mutually responsible for each other’s actions. “Strong family unity means that an offense to any member of the family is perceived to be a direct offence to all its members and every family member has a right to intervene in the life of another member” (İnce et al., 2009: 548). Equally, “one family member’s shameful act brings dishonor on the rest of the family” (Abu-Lughod, 1999 [1986]: 65–66). Consequently, “in many cases, women members of the family are the ones who put pressure on male members to kill other female members who are seen to be unchaste” (Faqir, 2001: 72). ¹⁹

Organization, then, tends to increase the strictness with which families define and respond to dishonor, especially that of their lower-ranking members. More individualistic families, by contrast, generally treat honor violations with considerably greater leniency. In an incident in the town of Sohar, Oman, for instance, a woman widely known to be a prostitute was neither condemned by her community nor punished by her husband (Wikan, 1982: ch. 8). Significantly, kinship patterns in the town “approach those found in modern Western industrial societies” in that “patrilineal descent groups… do not in fact emerge as corporate and potentially mobilizable bodies” (Barth, 1983: 112, 131). What matters most, however, is the degree of organization of the family in the particular case, not the organization of families in general. For instance, in a case among Israeli Bedouin, after an unmarried woman became involved in a sexual relationship her life was in jeopardy. However, her father worked for wages in a Jewish settlement and hence “was no longer dependent on his extended family for maintaining a common herd. He was thus able to ignore the decision of the family group and save his daughter” (Ginat, 1997 [1987]: 153).

Cases of this kind are increasingly common. A major reason that 25 of the 27 publicly known cases of dishonor in three rural Arab villages (1988–1995) did not result in an honor killing is the general weakening of the traditional extended co-liable family:

As more individuals of any particular group enter the wage labor market and build up ties and commitments outside the collective responsibility framework of the co-liable group, they will prefer ways of settling disputes that (a) will not put them or their families in danger, and (b) will not disturb their outside working relations by having to return to attend a co-liable group meeting. (Ginat, 1997 [1987]: 192)

More individualistic still are the families of the world’s wealthiest nations. In modern USA and Europe, adults primarily make their own relational, marital, employment, and life-style decisions, acting as autonomous individuals rather than as members of a corporate group. Offspring establish their own families, rather than being part of the husband’s extended family. Family units are smaller, less enveloping, more nuclear, more temporary – and more lenient (Baumgartner, 1988: 21–71). Unmarried pregnancies and births are extremely common, but honor killing is unknown. Over 40 percent of births are to unmarried women in the United States, Netherlands, and Denmark, and over 50 percent in France, Norway, Sweden, and Iceland (Ventura, 2009). The pregnancy of an unmarried woman might still stigmatize her and her family. But even when it does, the woman is rarely beaten or driven away, much less killed by her family.