Don’t Tread on Me

Using both college students and a national sample of adults, the authors report evidence linking the ideology of masculine honor in the U.S. with militant responses to terrorism. In Study 1, individuals’ honor ideology endorsement predicted, among other outcomes, open-ended hostile responses to a fictitious attack on the Statue of Liberty and support for the use of extreme counterterrorism measures (e.g., severe interrogations), controlling for right-wing authoritarianism, social dominance orientation, and other covariates. In Study 2, the authors used a regional classification to distinguish honor state respondents from nonhonor state respondents, as has traditionally been done in the literature, and showed that students attending a southwestern university desired the death of the terrorists responsible for 9/11 more than did their northern counterparts. These studies are the first to show that masculine honor ideology in the U.S. has implications for the intergroup phenomenon of people’s responses to terrorism.

concerns matters of reputation and social standing. Societies structured around the second definition are referred to by social scientists as cultures of honor. Such cultures exist throughout the world, including South America (e.g., Johnson & Lipsett-Rivera, 1998), regions of the Mediterranean (e.g., Peristiany, 1966), and the U.S. South (e.g., Wyatt-Brown, 1982). Scholars who have studied honor in these and related contexts have focused on several dimensions of the construct, including familial status (e.g., having a good family name), female fidelity (e.g., being faithful to one's spouse), and male reputation (e.g., being known as someone who will not tolerate insults). For instance, using Dutch participants (representative of a nonhonor culture) as a comparison group, Rodriguez Mosquera, Manstead, and Fischer (2002) showed that Spanish participants experienced relatively more shame in response to imagining themselves as disgraceful members of their families, and the extent to which they experienced this emotional reaction more than their Dutch counterparts was accounted for by their greater concern with family honor (see also Rodriguez Mosquera, Mastead, & Fischer, 2000). In addition, Vandello, Cohen, Grandon, and Franiuk (2009) found that participants from honor cultures (e.g., Latinos and U.S. southerners), who presumably value female loyalty in relationships, evaluated a woman who remained committed in an abusive relationship more favorably than participants from nonhonor cultures (e.g., U.S. northerners and Canadians), suggesting that cultural concerns with feminine honor could contribute to abusive dynamics in relationships (see also Vandello & Cohen, 2003).
Common to the theoretical arguments for the above findings is the importance of reputation maintenance in honor cultures relative to others. In light of this emphasis, it is not surprising that reputational threats in the form of insults carry special weight in cultures of honor and elicit a variety of negative reactions from members, not the least of which is violence, especially among males. The connection between honor concerns and male violence and aggression has best been demonstrated within the borders of the U.S., comparing so-called honor states in the southern (and sometimes western) U.S. to nonhonor states in the northern part of the country. Although we do not confine ourselves to regional comparisons, it is the connection between honor in the U.S. and aggression that interests us in the present article, but it is a different form of aggression than has previously been considered in the literature. Although prior research has focused on the link between honor and interpersonal aggression, we examine the link between honor and national or collective aggression in the form of militant responses terrorism.

Masculine Honor Ideology in the U.S.
Of the dimensions of honor highlighted above, one is of central importance to the culture of honor in the southern U.S.; specifically, it is a concern with men's reputation for toughness, fearlessness, and aggressiveness in the face of provocations (Cash, 1941;Fischer, 1989;Nisbett, 1993;Wyatt-Brown, 1982). Indeed, Fischer (1989) noted that honor in the historical southern U.S. "meant a pride of manhood in masculine courage, physical strength and warrior virtue" (p. 690). Likewise, Nisbett and Cohen (1996) argued that honor in this region was a reflection of "a man's strength and power to enforce his will on others" (p. 4). The explanation given for why this masculine honor ideology is more prevalent among southerners than northerners comes from historians' observation that Scotch-Irish immigrants, who were themselves immersed in a culture of honor, mostly settled in the southern regions of the country (Fischer, 1989;Leyburn, 1962;Nisbett, 1993;. In the southern U.S., their culture persisted presumably because their families and resources regularly came under threat, and because a lack of organized law enforcement required men to act as the primary defenders of their homes. Succumbing to sufficient provocations in these communities marked one as an easy target for exploitation, so violence was often used as social proof that one's family and resources could not be taken advantage of without serious, perhaps lethal, retaliation following (Fischer, 1989;Wyatt-Brown, 1982; for an alternative view on the origins of honor cultures, see Henry, 2009).
Researchers have identified several social influences that have propelled this cultural concern with masculine reputation into the present, especially among southern Whites (e.g., Brown & Osterman, 2012;Cohen & Nisbett, 1997;Vandello, Cohen, & Ransom, 2008). Regional differences in argumentrelated homicides (Nisbett, Polly, & Lang, 1995), favorable attitudes toward "honorable" violence (Cohen & Nisbett, 1994), aggressive responses to personal slights , and school violence (Brown et al., 2009) have been cited as evidence of its persistent effects. To date, however, scholars have almost exclusively emphasized the implications this mentality has for responses to provocations at the personal level. We move beyond this focus by considering the link between the ideology of masculine honor in the U.S. and militant responses to provocations experienced at the national level in the form of terrorism.

Masculine Honor Ideology and Militant Responses to Terrorism
Masculine honor ideology in the U.S. could contribute to militant responses to terrorism for at least two reasons. First, protection of family and possessions is closely associated with masculine honor. Therefore, endorsing actions intended to safeguard one's homeland from threats would be expected of people who value the honor ethic among men. Second, it has been well established that U.S. men who are concerned with masculine honor tend to respond to personal insults with aggression (e.g., Cohen et al., 1996;Daly & Wilson, 1988). Construing acts of terrorism as national insults (Slocum, 2007) suggests that people influenced by this ideology will respond similarly to terrorist threats because doing so reflects an unwillingness to be disrespected or intimidated, whether at the personal or national level. Likewise, to the extent that honor cultures coincide with cultural collectivism (for examples of such covariation, see Osterman & Brown, 2011;Peristiany, 1966), and the merging of the personal and the collective within the individual self-concept, national insults might be experienced like the personal insults that are so inflammatory to culture-of-honor men .
Although no research to date has directly linked the ideology of masculine honor in the U.S. with responses to terrorism, some work has touched on related topics. Ducat (2005), for instance, has argued for a connection between honorrelated concerns and conservative politics, of which national defense is an important component, and Slocum (2007) has proposed that the post-9/11 policies of George W. Bush's administration reflect the honor concerns associated with southern U.S. culture (see also McCullough, 2008). More direct evidence related to our hypothesis comes from an archival study by Cohen (1998) showing that Congresspersons from honor states were more supportive of "hawkish" national security policies (e.g., using military force in the Gulf War) than their counterparts in nonhonor states. Thus, it seems plausible to hypothesize a link between masculine honor ideology in the U.S. and militant responses to terrorism. In Study 1, we test this idea using an individual difference measure of masculine honor ideology endorsement, and in Study 2, we test this idea by distinguishing between respondents from an honor state and respondents from a nonhonor state, as has traditionally been done in the literature.
Importantly, we also include women in Study 2. Theorists have posited that women play an important role in sustaining and perpetuating the culture of honor in the U.S. Wyatt-Brown, 1982). For instance, Sam Houston's mother awarded him a musket when he joined the army with the admonishment to "never disgrace it" by fleeing combat to save his life (Wyatt-Brown, 1982, p. 51). Likewise, Andrew Jackson's mother instructed him to never take legal action against someone who slandered or harassed him, but to instead settle such matters himself (Nisbett & Cohen, 1996, p. 86). Although this evidence is anecdotal, it seems reasonable to predict from it that women shaped by the culture of honor might support militant responses to terrorism. Such support would not likely be motivated by women's interest in maintaining their own reputations for strength and toughness, nor would women who endorse militant responses necessarily engage personally in the militant defending. Nonetheless, women in honor cultures might pronounce such responses as right and good and call upon men to defend their country's good name from the insult of an attack. Finding support for this hypothesis would add to the growing body of evidence that certain honor-related phenomena emerge among women as well as men (Barnes et al., 2012;Osterman & Brown, 2011).

Overview of Studies
The goal of our first study was to establish the predicted association between masculine honor ideology in the U.S. and militant responses to terrorism, including (a) openended responses to a hypothetical terrorist attack on the Statue of Liberty, (b) defensive vigilance in situations involving potential terrorists, (c) support for using severe interrogations with suspected terrorists, and (d) support for national defense actions-for example, the "war on terror," controlling for a broad range of covariates shown to predict similar outcomes (e.g., right-wing authoritarianism [RWA], social dominance orientation [SDO]). Because of the theoretical focus on White males from the southern and northern U.S. in previous culture-of-honor research, we followed this pattern in the pilot study and in Study 1 to establish the validity of our new measure of masculine honor ideology endorsement. In Study 2, we used the traditional approach to operationalizing the culture of honor in the U.S. by comparing White respondents from an honor state to White respondents from a nonhonor state on their support for lethal retaliation against the terrorists responsible for 9/11. We also included women in Study 2 to explore whether they exhibit the militant reactions we expected to observe among men.

Pilot Study: A Measure of Masculine Honor Ideology in the U.S.
Research on the culture of honor in the U.S. has relied heavily on Census Bureau divisions as an indirect indicator of masculine honor ideology endorsement-that is, states and persons belonging to the southern (and western, with the exception of Alaska and Hawaii) U.S. are assumed to be more concerned with masculine reputation, than those belonging to the northern U.S. (e.g., Cohen, 1998). However, because individuals' acceptance of this ideology does not perfectly follow regional boundaries (Cohen & Vandello, 2001;Leung & Cohen, 2011), and because these divisions are always used to represent differences in masculine honor ideology endorsement among persons, we sought to measure this endorsement directly, at the individual level, much as other scholars have done with related dimensions of honor among Spanish and Dutch participants (Rodriguez Mosquera et al., 2002;see also IJzerman, van Dijk, & Gallucci, 2007) and among Chileans and Canadians (Vandello et al., 2009).
Our interest in constructing a new measure of masculine honor ideology endorsement is three pronged. First, only one item on Rodriguez Mosquera et al. (2002) scale addresses the importance of a man defending himself against an insult; the remaining statements concern issues of sexual prowess and familial dynamics. Given the importance of aggressive retaliation in response to provocations in research on the U.S. culture of honor, we felt this theme should be better represented in a masculine honor ideology measure. Indeed, the importance of retaliation to the ideology of honor can be seen in a recent study by Leung and Cohen (2011) who assessed participants' internalization of honor norms by having them rate their support for actors' use of violence in a series of filmed provocations (see also Henry, 2009, Study 4). Second, although Vandello et al.'s (2009) scale better reflects the U.S. culture of honor described by scholars (e.g., Cash, 1941;Fischer, 1989;Wyatt-Brown, 1982), it blends items about masculine and feminine honor together, and we wished to measure beliefs about masculine honor independently. Finally, and importantly, none of the above measures have been distinguished from other constructs that might be expected to covary with the ideology of honor. We created the Honor Ideology for Manhood Scale (HIM), in response to these issues.
The HIM includes eight statements about the contexts in which men have the prerogative to use physical aggression for personal and reputational defense (e.g., "A man has the right to use physical aggression against another man who insults his mother"). These statements were adapted from items examined by Cohen and Nisbett (1994), who found that White men from honor states endorse the use of physical aggression for the purposes of protecting their property, family, and personal reputations more than do White men from nonhonor states. The HIM also includes eight statements about the defining qualities of "real men," such as selfsufficiency, pugnacity, and physical toughness (e.g., "A real man will never back down from a fight"). All 16 items included on the HIM are presented in the appendix and were responded to by participants on a scale ranging from 1 (strongly disagree) to 9 (strongly agree).
The pilot study, part of a larger investigation, was designed in part to test for regional differences in the HIM independent of constructs such as religiosity, conservatism, education, and income, all of which vary regionally and could account for any differences we observe in honor ideology endorsement among White U.S. men (e.g., Brown et al., 2009). Data were collected through Knowledge Networks (KN), an Internetbased polling agency that maintains a national probability sample of U.S. citizens and conducts online survey research for the government, universities, and corporations. KN obtains its respondents via random digit dialing and addressbased sampling. The former method randomly samples from the universe of U.S. landline telephone numbers, and the latter randomly selects mailing addresses from the U.S. Postal Service Sequence File. Computers and Internet access are provided to panel members who lack these resources, thereby allowing KN to reach individuals within the population who are not current Internet users and do not own a personal computer. Participants who are not compensated with computers and Internet coverage are included in modest incentive programs (e.g., raffles and monetary giveaways), which serve as compensation for completing surveys.
Consistent with prior research on the U.S. culture of honor, we obtained a sample of 328 White males from the southern (i.e., South Census Region: South Atlantic, South Central, and West South Central) and northern (i.e., Northeast Census Region: New England and Middle Atlantic; Midwest Census Region: East North Central, and West North Central) U.S. from KN's nationally representative response panel, with the former group representing honor states and the latter representing nonhonor states (Cohen, 1998). The age of these respondents ranged from 18 to 40 (M = 29.47, SD = 6.64). Because of missing data on more than two of the HIM items, the final sample was reduced to 323 (southern: n = 165; northern: n = 158). In addition, because of a large number of missing responses to the single item concerning the frequency of religious service attendance (n = 54), we imputed scores for these missing values using the average frequency of attendance for each region in the sample (M = 3.17 and 3.46 for north and south regions, respectively). In the analyses reported below, all data were adjusted with a demographic weighting parameter derived from the U.S. Census Bureau's Current Population Survey for the purpose of generalizing study findings to the population. 1 We submitted the HIM's 16 items to an exploratory factor analysis using principle axis factoring as the extraction method. One dominant factor emerged, accounting for 48.8% of the variance in HIM ratings; the remaining two factors accounted for much smaller amounts of variance (8.7% and 5.9%, respectively) and appeared to reflect formatting factors based on differences in wording between the two types of statements on the scale (i.e., "A real man . . . " and "A man has the right to use physical aggression . . . "). Importantly, all items loaded positively and strongly on the first factor, with a median, unrotated factor loading of .70 (range = .47-.83). We, therefore, combined these items to form our measure of honor ideology endorsement. The HIM's internal reliability was excellent in this sample (α = .94).
Further support for the HIM comes from a regression analysis that substituted Gastil's (1971) continuous statelevel measure of southern cultural influence for the dichotomous, regional variable used above. We assigned participants values on Gastil index using their state of residence. Although this measure has been criticized (e.g., Loftin & Hill, 1974), it currently is the only continuous proxy of the culture of honor available for U.S. states, and it has been used successfully in prior culture-of-honor research (e.g., Barnes et al., 2012;. As with the dichotomous regional index, Gastil's measure also predicted HIM scores, β = .19, t(317) = 3.55, p < .001, controlling for religious service attendance, conservatism, income, and education. Thus, we have added confidence in the HIM's ability to capture endorsement of one of the central features of the cultural ideology of honor in the U.S.-specifically, defense of masculine reputation for strength and toughness.

Study 1
Because protecting hearth and home from threats is important to male honor, and because insults are not tolerated by honor-minded men, we expected a positive association between the HIM and militant responses to terrorism to emerge. In Study 1, we sought not only to establish this association but also to determine whether it remained when we controlled for a host of potentially related covariates (e.g., RWA, SDO, religious fundamentalism), some of which have been linked in prior research to hostile reactions to outgroups, including terrorists (Altemeyer & Hunsberger, 1992;Cohrs, Moschner, Maes, & Kielmann, 2005;Crowson, 2009;Pratto, Sidanius, Stallworth, & Bertram, 1994;Sidanius & Liu, 1992;Wylie & Forest, 1992).

Method
Participants. Participants were 191 young White males from a national, online panel maintained by Market Tools Incorporated (MTI), an Internet-based marketing research firm. Members of MTI's respondent panel are recruited via online advertisements and are compensated for their participation with credit usable in an online store. The open nature of the panel allows MTI access to individuals fitting a wide range of demographic characteristics but does not ensure that it is representative of the U.S. population (www.zoomerang. com). A nontrivial number of participants failed to respond to a subset of the scale items in the survey (n = 59). To recover a portion of these data, scale totals were computed for all participants who were missing responses on no more than one item for a given instrument. This left a sample of 177 respondents, of whom 75 were from the southern U.S. (i.e., South Atlantic, South Central, and West South Central) and 101 were from the northern U.S. (i.e., Middle Atlantic and New England; region data were missing for one person). It should be noted that some of the remaining participants failed to answer the open-ended question about the hypothetical terrorist attack (n = 27), so analyses involving this outcome have slightly fewer respondents and, therefore, different degrees of freedom from the other analyses.
The age of participants ranged from 18 to 55 (M = 31.95, SD = 6.77). Regarding education, 19.8% had completed only high school, 26.6% finished some college-level work, 32.8% completed college, 9.0% had some graduate or professionallevel training, 11.3% had completed graduate or professional school, and 0.6% (one) failed to provide this information. No information on participants' income levels was available.
Materials and Procedure. All materials for the study were administered online in the order described below. Items comprising the first four measures were all created for this study.
General aggressiveness (α = .89). Seven items tapped participants' general aggressive tendencies. Three items ("It's easy for me to get angry," "I have a short fuse," and "It doesn't take much to set me off") were statements to which respondents indicated their level of agreement on a 1 (strongly disagree) to 7 (strongly agree) scale, and four items were one-word adjectives (aggressive, peaceful [reversed], belligerent, and quarrelsome) that respondents identified as being either self-descriptive (7 = [This adjective] describes me very well) or not (1 = [This adjective] does not describe me well). These items were included to show that the HIM is not isomorphic with dispositional aggressiveness, despite the emphasis on aggressive retaliation in some of the HIM's items.
Patriotism (α = .89). Participants rated their level of agreement with the following three statements on a 1 (strongly disagree) to 7 (strongly agree) scale: "Being a U.S. citizen is an important part of my identity," "I have a lot of American pride," and "I think America is the greatest country in the world." 2 Religiosity (α = .94). Four items assessed respondents' level of religiousness ("I am very devoted to my religious faith," "I think of myself as a very religious person," "I am not at all religious" [reversed], and "My religious faith is a very important part of my life"). These items were responded to on the same 7-point scale used for patriotism.
Conservatism. Using a scale ranging from 1 (very liberal) to 9 (very conservative), with the midpoint labeled "moderate," participants answered the following question: "How would you describe your political orientation?" Honor ideology for manhood (α = .92). Participants rated their level of agreement with the 16 HIM items using a scale ranging from 1 (strongly disagree) to 9 (strongly agree).
Religious fundamentalism (RF; α = .93). Altemeyer and Hunsberger's (2004) 12-item RF scale measures defensive adherence to religious beliefs irrespective of the sectarian origins of those beliefs. Sample items include, "When you get right down to it, there are only two kinds of people in the world: the Righteous, who will be rewarded by God, and the rest who will not" and "The fundamentals of God's religion should never be tampered with or compromised with others' beliefs." Agreement with each statement was registered by participants on a scale ranging from 1 (strongly disagree) to 7 (strongly agree).
RWA (α = .87). Zakrisson's (2005) 15-item version of Altemeyer and Hunsberger's (1992) RWA scale was administered. Respondents rated their agreement with each statement on the same scale used for the RF items. Sample statements include "Our country desperately needs a mighty leader who will do what has to be done to destroy the radical new ways and sinfulness that are ruining us" and "Our country will be great if we honor the ways of our forefathers, do what the authorities tell us to do, and get rid of the "rotten apples" who are ruining everything." SDO (α = .88). Participants completed Pratto et al.'s (1994) SDO scale by rating how positively they felt toward statements such as "Some people are just more worthy than others" and "It is important that we treat other countries as equals" (reverse scored). Ratings were made on a 1 (very negative) to 7 (very positive) scale.
The above scales were treated as predictor variables in our analyses. The criterion variables are presented next in the order they were completed by participants.
Statue of Liberty attack and terrorist-directed hostility. Participants read about a hypothetical attack on the Statue of Liberty. The attack, supposedly orchestrated by Afghani terrorists, resulted in the destruction of the monument's head and the death of 250 persons. The scenario was accompanied by the image of a headless Statue of Liberty on the front page of a fictitious issue of the New York Times, with a headline that read, "Afghani Terrorists Strike Lady Liberty: 250 Dead, Dozens Seriously Injured." Participants were instructed to write about "the thoughts and feelings [they] would have if the Statue of Liberty were actually attacked." Responses were coded by two raters for expressed hostility (i.e., themes of hatred, aggression, and vitriol) toward the responsible terrorists. Ratings were made on 7-point scales, with higher values representing more hostility. Raters' judgments were highly correlated, r = .88, p < .001, and were averaged together to create a composite measure of terrorist-directed hostility.
Defensive vigilance (α = .98). Participants were presented with four ambiguous situations in which the key actor in each might or might not be a terrorist. For example, You are standing in line at the post office when a darkskinned man dressed in Middle Eastern garb walks in carrying a large, seemingly unmarked package. He is breathing hard, sweating profusely, and keeps looking at his watch.
Participants rated how "suspicious," "threatening," "hostile," and "distressing" the target's behavior seemed to them on a scale from 1 (not at all) to 9 (extremely). Next, participants were presented with the following statements: "How vigilant or alert to danger would you be in the above situation," "Would this man's behavior suggest to you that something bad would happen if action were not immediately taken to prevent it," and "Would you think it appropriate to report this man's behavior to an authority (e.g., the postmaster or police)?" Responses to these items were made on 9-point scales with appropriate pole labels (e.g., "not at all alert" vs. "very alert" and "definitely not" vs. "definitely yes"). All responses were averaged together to form an index of defensive vigilance across all four ambiguous threat scenarios. Support for severe interrogations (α = .96). Four statements captured the extent to which respondents supported the use of severe interrogations with suspected terrorists (e.g., "Interrogators should do everything in their power to draw information out of suspected terrorists, even if it means using methods that cause those persons lasting physical or psychological problems"). Ratings were made on a 1 (strongly disagree) to 9 (strongly agree) scale. Support for military action (α = .80). Finally, participants rated on a scale from 1 (strongly disagree) to 9 (strongly disagree) their level of agreement with six statements about their support for using aggressive military action in the U.S.'s campaign against terrorism, including "I am in favor of the U.S.'s war against terrorism," "It is entirely appropriate to engage in preemptive attacks on countries that are suspected of harboring or supporting terrorists," and "If necessary, the U.S. should use nuclear [and, in a separate question, chemical] weapons to defend its interests against those of terrorists." Four of these items were borrowed from Pyszczynski et al. (2006).

Results and Discussion
Intercorrelations and descriptive statistics for all Study 1 variables appear in Table 1, and detailed regression results can be seen in Table 2. The goal of Study 1 was to determine whether an individual difference measure capturing endorsement of one of the central features of the U.S. culture of honor (i.e., concerns with the defense of masculine reputation) might predict militant responses to terrorism, controlling for conceptually related measures, such as RWA, SDO,and RF. 3 To test this possibility, we examined the HIM's ability to predict each of our outcome measures, including participants' open-ended responses to the Statue of Liberty attack (i.e., terrorist-directed hostility), controlling for the covariates of RF, RWA, SDO, patriotism, and aggressiveness. Regressing terrorist-directed hostility onto the variables listed above revealed the HIM as a significant predictor of this construct, β = .31, t(143) = 3.81, p < .001, indicating that the HIM predicts participant-generated hostile responses to terrorism. Importantly, the HIM also predicted responses to our Likert-scale survey items tapping defensive vigilance, β = .23, t(170) = 3.24, p = .001, support for severe interrogations, β = .19, t(170) = 2.79, p = .001, and support for military action, β = .15, t(170) = 2.11, p = .036. Taken together, these findings demonstrate the unique predictive utility of the HIM for responses to terrorism.
In our initial regression analyses, we omitted religiosity and conservatism as covariates because of their strong conceptual overlap with RF and RWA, respectively; however, we recognize the potential contribution these variables could make to our results (e.g., Crowson, Thoma, & Hestevold, 2005), so we reran our equations adding religiosity and conservatism to our original set of covariates. Even with these additions, the HIM predicted terrorist-directed hostility, β = .32, t(141) = 3.82, p < .001; defensive vigilance, β = .22, t(168) = 3.02, p = .003; and support for severe interrogations, β = .17, t(168) = 2.42, p = .016; its association with support for military action was marginally significant, β = .14, t(168) = 1.90, p = .059. We believe the persistent ability of the HIM to  predict the outcomes assessed in this study, even in the presence of such important covariates, highlights the HIM as a unique index of a cultural ideological construct in the U.S. not captured by these other individual difference measures. These results also show, for the first time, the relevance of this ideology to responses to terrorism, including hypervigilance to terrorism-related cues and support for extreme military actions (e.g., nuclear and chemical bombs).

Study 2
Study 1 showed that individual endorsement of masculine honor values predicted hostile and militant responses to a hypothetical terrorist attack. This is one of only a few recent studies that have attempted to capture the culture of honor at an individual level (e.g., Barnes et al., 2012;Leung & Cohen, 2011). To better demonstrate that the honormilitancy association in the U.S. has a collective/cultural manifestation, our final study shifts from the individual level of analysis to the regional level used in most pervious research. Specifically, Study 2 involved a reanalysis of U.S. students' reactions to the terrorists attacks of 9/11 (see Barnes & Brown, 2010;Brown, Wohl, & Exline, 2008). We compared reactions of White students at two universities, one located in the southwestern region of the U.S. (Oklahoma) and the other located in the northeastern region of the U.S. (Pennsylvania), with respect to their desire for lethal retaliation against those responsible for the attacks. Based on the treatment of these states in prior research, we reasoned that participants at the school in the southwestern U.S. would exhibit the culture of honor, whereas participants at the school in the northeastern U.S. would not (e.g., Cohen, 1998). In addition, because women may help sustain and perpetuate the culture of honor in the U.S., even if they do not personally display the same violent behaviors that this culture rewards among men Vandello et al., 2009;Wyatt-Brown, 1982), we investigated whether White women from an honor state would exhibit an aggressive response to a national-level provocation alongside their male counterparts. The dependent variable of interest in Study 2 was the extent to which participants wanted those responsible for 9/11 to be killed, and we controlled for levels of patriotism and religiosity to ensure that any regional difference obtained was not simply a function of these potential confounds.

Method
Participants. A collection of questionnaires was administered during the 2 weeks following 9/11 to 442 White students (268 females, 173 males) at two prominent universities in the U.S.: The University of Oklahoma (OU: n = 193) and Pennsylvania State University (PSU: n = 249). According to prior research, OU is located in a state characterized by a culture of honor, whereas PSU is not (Cohen, 1998). An additional 111 students completed the questionnaires, but their data were excluded from this study either because their ethnicity could not be determined (n = 18) or because their non-White demographic is not the focus of culture-of-honor theory and research in the U.S. Importantly, no other racial or ethnic group had sufficiently large numbers at both campuses to serve as an adequate comparison group. Course credit was granted to students in exchange for participation. An additional three participants were lost from the sample because of missing responses on one or more key variables. This left a final sample of 439 students (OU: n = 191: 70 male; PSU: n = 248: 103 male). Specific religious affiliation was unknown for most participants. Of the remaining participants (n = 153), all identified themselves as Christian (17% Catholic).
Measures and Procedure. The week following 9/11, researchers at OU and PSU compiled a series of questionnaires designed to tap different aspects of participants' psychological reactions to the event. 4 Patriotism was measured by taking the average of participants' rated agreement with three statements, which were created for this study and were similar to those used in Study 1: "I consider myself to be patriotic," "I have strong feelings of patriotism about the USA," and "I feel very proud to be a U.S. citizen." Ratings were made on a 7-point scale, with higher ratings indicating stronger agreement (α = .88). Religiosity was measured with five questions, also created for this study: "I consider myself to be a religious person," "I consider myself to be a spiritual person," "I am not religious" (reverse scored), "I am a member of an organized religion," and "I regularly attend religious services." These ratings were also made on a 7-point scale with higher ratings indicating stronger agreement (α = .89). To tap desires for lethal retaliation against the 9/11 terrorists, participants responded to the question, "To what extent do you feel right now that you want those responsible for the attacks to be killed," on a 9-point scale ranging from 1 (not at all) to 9 (completely).

Results and Discussion
Regressing the lethal retaliation item simultaneously on university affiliation, patriotism, and religiosity revealed a significant association with university affiliation, β = .17, t(435) = 3.57, p < .001, d = .35. Specifically, participants from OU reported that they wanted those responsible for 9/11 to be killed (M = 6.80, SE = 0.20) more than did participants from PSU (M = 5.86, SE = 0.17). 5 In addition, patriotism was a significant covariate in this analysis, β = .21, t(435) = 4.50, p < .001; religiosity was marginally significant, β = −.09, t(435) = −1.80, p = .073. Importantly, endorsement of lethal retaliation differed by university among males, β = .17, t(169) = 2.28, p = .024, d = .36 (OU: M = 6.90, SE = 0.32; PSU: M = 5.94; SE = 0.27), and females, β = .16, t(262) = 2.66, p = .008, d =.34 (OU: M = 6.73, SE = 0.25; PSU: M = 5.82; SE = 0.23). From these results, it is possible to conclude that residing in an honor state where an ideology of honor dominates among Whites is associated with advocating the death of the terrorists responsible for 9/11, and this association cannot be attributed to regional differences in patriotism or religiosity. Furthermore, Study 2 extends the results of Study 1 to show that the culture of honor could lead to lethal attitudes in response to terrorism among women as well as men-one of the first such demonstrations in the literature on the U.S. culture of honor and violence (for other honor-related outcomes among women, see Barnes et al., 2012 andBrown, 2011).

General Discussion
The present studies are the first to provide evidence that the cultural ideology of honor implicated in interpersonal violence (e.g., Cohen et al., 1996; has implications for the intergroup phenomenon of people's responses to terrorism. In the pilot study, we reported evidence that endorsement of masculine honor ideology in the U.S. can be successfully tapped at the individual level using the HIM (cf. IJzerman et al., 2007;Leung & Cohen, 2011;Rodriguez Mosquera et al., 2002;Vandello et al., 2009). We then proceeded to use this instrument in a study using a national sample of White males (Study 1). Participants in this study were asked to respond to four indicators of militant responses to terrorism. Consistent with expectations, scores on the HIM significantly predicted (a) terroristdirected hostility, coded from open-ended responses to a hypothetical terrorist attack; (b) defensive vigilance across multiple, ambiguous threat scenarios; (c) support for severe interrogations with suspected terrorists; and (d) support for military action in the U.S.'s campaign against terrorism. It is noteworthy that scores on the HIM were predictive of these responses controlling for a host of related measures, such as RWA, SDO, patriotism, and religious fundamentalism. These results provide further support for the value of measuring masculine honor ideology at the individual level (e.g., Barnes et al., 2012;Osterman & Brown, 2011;Rodriguez Mosquera et al., 2002;Vandello et al., 2009) and point to the contribution this ideology makes to issues of national and global importance.
Study 2 used the more common approach social psychologists have taken to capture the culture of honor in the U.S.: regional differences. In this study, we examined college students' desires for retaliation against the terrorists who attacked the World Trade Center and Pentagon on September 11, 2001. As expected, White students from a school in an honor state in the southwestern U.S. reported a greater desire for lethal retaliation during the 2 weeks following the 9/11 attacks than did their counterparts from a school in a nonhonor state in the northeastern U.S. This finding extends the responses to the hypothetical measures of Study 1 to real responses to an actual terrorist attack. Also in Study 2, we examined whether any gender differences occurred, and none were found-a noteworthy departure from the bulk of literature on aggression and violence in the U.S. culture of honor, which shows that with respect to interpersonal violence, women from honor states do not exhibit the same elevated rates of homicide and assault that males from honor states do . Recent research, however, reveals that women living in honor states do exhibit some of the same honor-related behaviors that men do, including excessive risk taking (Barnes et al., 2012) and suicide (Osterman & Brown, 2011). We suspect that the reason for this difference in gendered behavior patterns concerns the fact that the endorsement of lethal retaliation is not the same as personally engaging in homicide or assault. Indeed, homicidal violence is something that women rarely engage in, and when they do, it is not typically for the same reasons that men do (Daly & Wilson, 1988). Even so, the present results underscore the interesting possibility that women who live in honor cultures, or who embrace the beliefs and values of an honor culture, might display a number of important motives and behaviors that are similar to those of honorculture men.

Limitations and Future Directions
Absent from this research is evidence that masculine honor concerns in the U.S. shape behavioral responses to terrorism. We do not wish to downplay the importance of such evidence, but we feel it is noteworthy that when it comes to a nation responding to terrorism, most people never have the opportunity to interrogate a terrorist suspect or fire artillery into a terrorist-harboring country. People do, however, have ample opportunities to express their attitudes about terrorism through debates with peers, opinion polls, voting, and letters to congresspersons. Consequently, we believe that the collection of dependent variables examined here constitutes a good starting point for behavioral research in this area.
The hypothetical nature of the Statue of Liberty attack and defensive vigilance items is another limitation of this research, as numerous scholars have noted that forecasted responses in hypothetical contexts are often incongruent with responses to real situations (e.g., Ajzen, Brown, & Carvajal, 2004;Armor & Sackett, 2006;Barnes & Brown, 2010). This limitation does not apply, however, to our interrogation and military action items, which assessed participants' position on these issues in general, not with respect to hypothetical scenarios. Likewise, this limitation does not apply to Study 2, which evaluated individuals' reactions to the attacks of 9/11 within 2 weeks of the event. Thus, the data from this latter investigation nicely complement the hypothetical stimuli used in Study 1 and supports the validity of Study 1's conclusions.
Our focus in these studies was on reactions to foreign terrorist attacks as a function of the ideology of masculine honor in the U.S., and we have argued that this construct influences the form these reactions take. If we turned the tables, however, we could consider how this ideology might function as an impetus for committing acts of terrorism. Whether a perceived injustice is perpetrated by a government against its citizens, or one country against another, the ideology we have described might motivate individuals to defend their convictions with violence or, at least, to judge less harshly those who share their convictions and use violence to defend them. Indeed, researchers have implicated honor-related motives-in particular, the desire to restore lost honor or to prevent future losses to honor-as a common feature underlying many terrorist attacks, including the rather extreme instances of suicide bombings (e.g., Atran, 2003;Bloom, 2005;Crenshaw, 2007;Kruglanski, Chen, Dechesne, Fishman, & Orehek, 2009). As Kruglanski et al. (2009) noted, "In some cases, [one's cultural] norms and ideologies may identify the suicide mission against one's enemies as a most honorable act, lending one a sense of immense veneration and significance" (p. 338). Although suicide bombings might seem extreme and "exotic" examples of behavior, Osterman and Brown (2011) have recently found that more "mundane" instances of suicide in the U.S. can be linked to a culture of honor. Given this parallel finding, it might very well be that even acts of domestic terrorism in the U.S. are driven, in part, by honor motives. This argument raises the interesting possibility that the promotion and defense of honor might motivate the same people both to respond with militant aggression to acts of terrorism and to engage in terrorist acts, under certain circumstances.

Conclusion
In his State of the Union address in 2001, President George W. Bush issued this ultimatum: "Every nation in every region now has a decision to make. Either you are with us, or you are with the terrorists." Insofar as Bush's warning reflects his cultural upbringing (McCullough, 2008;Slocum, 2007), his words suggest to us that the dynamics of the U.S. culture of honor might be played out on the national level as well as the individual one (Slocum, 2007). In today's complex world, the bifurcation of humanity into those who are "with us" and those who are "with the terrorists" might impede otherwise healthy international relations (see also, Kruglanski, Crenshaw, Post, & Victoroff, 2007). In the end, military conflict might not be averted, but insofar as honor-based motives to aggress against terrorists can be critically evaluated, perhaps decisions about international confrontations that cast an eye toward the long-term costs and consequences of such conflicts can be made.
2. It could be that hostile responses to terrorism are isolated to or are particularly pronounced among those who are patriotic and identify with the cultural ideology of honor, but no evidence was found for this hypothesis in either Study 1 or Study 2, both of which assessed patriotism. 3. Although we tested for regional differences in the HIM and our outcome measures in Study 1, we found no significant differences by region, with the exception of support for military reactions. It should be noted that although the differences were nonsignificant, southerners endorsed the HIM more than northerners did in Study 1, as in the pilot study, and the regional differences on the remaining three dependent variables in this study were consistently in the predicted direction. We suspect that the lack of significant regional differences is due to the nonrepresentative nature of the sample in Study 1 compared with the sample used in the pilot study, although we cannot be certain of this. However, our failure to find consistent regional differences in Study 1 led us to focus more attention on this matter in Study 2. 4. We gratefully acknowledge the help of Jennifer Bosson and Elizabeth Pinel in gathering data used in this study. 5. Not surprisingly, most participants in our sample expressed a desire for lethal retaliation. This resulted in a negatively skewed distribution for this item. Squaring the data adequately corrected this problem. The results obtained by analyzing the transformed lethal retaliation item were the same as those reported in the text.