In pre-colonial Papua New Guinea, a collection of small stateless warrior societies, payback (or retributive homicide) was deemed to be a legitimate sanction. Under custom (or perhaps more appropriate customary law) a homicide, whether intentional or deliberate, could be sanctioned with a payback killing, which could be targeted toward the wrongdoer or a member of his or her kin (Reference TrompfTrompf 1994). Now a payback killing is an offence itself under Section 299 of Papua New Guinea's Criminal Code Act 2013 labeled Wilful Murder.
Despite being a crime, which can and does attract the highest possible prison sentences, payback is still practiced in many parts of Papua New Guinea including the New Guinea Islands. While there are no data on its prevalence, some believe that it is more widely accepted and practiced in rural areas, where the population is considered less sophisticated and the reach of the post-colonial state is particularly weak.Footnote 1 This view is articulated by Papua New Guinea's national newspaper, the Post-Courier (2010), which stated that payback killers should “leave their terrible ‘customs or culture’ behind in their villages.” Such a statement is consistent with Reference MerryMerry's (2003: 964) observation that some post-colonial urban elites tend to portray themselves as progressive while characterising rural communities as being “steeped in old cultural practices … that they are not willing to abandon.”
While the Papua New Guinean courts once attempted to distinguish customary motivations from more malevolent ones, they effectively no longer do. Indeed, the state seems to have taken aim at customary practices that are contrary to current Western norms, with the purpose of eradicating them through criminal law. In addition to the increasingly tough stance taken by the courts toward payback (see below), parliament has recently increased the penalties for those convicted of killing alleged sorcerers. In 2013, parliament passed the Criminal Code (Amendment) Act 2013 that applies the death penalty to the new crime of Wilful Murder of a Person on Account of Accusation of Sorcery (Section 229A) and repealed the Sorcery Act 2013.Footnote 2 These amendments were partly in response to some highly publicized killings and pressure from human rights groups to deter them (Amnesty International 2013). While the killing of alleged sorcerers differs from payback (which is akin to retributive homicide),Footnote 3 these recent moves underscore the state's increasing tendency to use criminal law to eradicate violent customary practices and to transform society.
States elsewhere have had various degrees of success in transforming society through criminal law: for instance, the practice of sati was largely eradicated following its criminalization by British colonial administrations (see Reference SharmaSharma 1988)Footnote 4; however this outcome can be contrasted by Reference MassellMassell's (1968) account of the Soviet state's failure at using the law as an instrument of revolutionary change in Soviet Central Asia in the 1920s; while Reference Shell-DuncanShell-Duncan et al (2013) have found mixed results in relation to the criminalization of female genital mutilation/cutting in Senegal.
The focus of this article is to shed light on how internalized beliefs (or internalized legal norms) are formed using payback as a case study, with a particular focus on the role that criminal law may play in forming internalized norms. The remainder of this article is as follows. First, there is a description of payback contextualized within the broader practice of custom and the state's response. Then a framework for conceptualising the internalization of norms concerning the legitimacy of payback is developed. This is followed by an empirical investigation into attitudes toward payback in the New Guinea Islands, including sampling and data, descriptive statistics, estimation, and key findings. The article concludes with a discussion of the results in terms of the theoretical framework and a conclusion.
Contextualising Payback
The “Traditional” Customary Practice of Payback
What follows is a description of what Reference TrompfTrompf (1994: XV) describes as one “of the more remarkable features of Melanesian life: payback killing.”Footnote 5 It should be seen as a Weberian ideal type, given that Papua New Guinea is extremely ethnically and culturally diverse. It should also be noted that some such as Reference ChanockChanock (1998) contest the very concept of customary law, suggesting it to be a colonial phenomenon invented by local informants for various purposes, including self-gain and the protection of the sacred. Finally, just like state law, custom can also vary over time (see the next section for a discussion the potential drivers of change). Therefore, the focus of this description is on the “traditional” customary practice of payback in Melanesia within the broader practice of customary law (henceforth custom).
The principles of reciprocity and kinship are the cornerstones of custom in Papua New Guinea and play an important, and interrelated, role in the type and enforcement of sanctions for wrongdoing.Footnote 6 Across much of Melanesia enforcement of customary sanctions was the primary responsibility of the person wronged and his or her kinship group, commonly referred to as self-help.Footnote 7 Kin were obliged to protect their group and help enforce customary sanctions, if they wished to have the same support in return. In the absence of a pre-colonial state, self-help was an important, if not necessary, mode of redress and social control (Reference Narokobi, Crocombe and RocheNarokobi 1996).
The principle of reciprocity requires that returns must be made for an act, and this applies to “helpful and beneficial actions as well as harmful ones” (Reference Strathern and ScaglionStrathern 1981: 6). In relation to harmful ones, and specifically in cases of homicide, persons must effectively be “paid for” (Reference Strathern and ScaglionStrathern 1981: 13). At one end of the spectrum was the payment of compensation and peaceful agreement while at the other end was “self-help and reprisal through force” (Reference StrathernStrathern 1995: 229).Footnote 8 On the latter, Reference DinnenDinnen (1988: 30) states that:
The concept of ‘payback’ or retaliation is an important social principle in dealing with ‘wrongs’ in traditional Papua New Guinean societies. The form of ‘self-help’ resorted to depends as much upon the social context within which the wrongful act occurs as upon the nature of the act itself. ‘Payback’ or retaliation may express itself in at least two different ways. In its simplest and most obvious form, retaliation might take the character of direct physical violence, eg. the ‘payback’ killing. A more indirect form of retaliation would involve sorcery, the threat of which might act as a sanction for proper conduct.
As highlighted by Dinnen, the social context of payback is critical. Group loyalty and interdependence is of utmost importance in Papua New Guinean society. In this regard one must be careful not to assume that Melanesian society was (or is) one made up of “self-contained” Western-style individuals (Reference Li PumaLi Puma 2001).Footnote 9 The importance of kinship (or group loyalty more generally) in relation to wrongdoing has been labeled the sliding scale of justice by Reference Lawrence and BrownLawrence (1969): this sees lighter and mostly nonviolent punishments imposed on those who have closer social proximity, while heavier and often violent sanctions imposed on those with greater social distance. That is, despite payback being acknowledged as a legitimate traditional sanction, it was usually reserved for outsiders and for settling scores in tribal fighting.Footnote 10 It was also common that the victim of a payback killing was not the culprit but a member of the culprit's family or clan, a principle approaching joint or group liability (or indiscriminate retribution).Footnote 11 Furthermore, where someone has been killed, retribution or compensation demands will be forthcoming, regardless of intention or level of care taken, and therefore, the standard of responsibility “is far more akin to one of absolute liability than to one based on fault” (Reference Gabi, Zorn and BayneGabi 1975: 183). With the advent of roads and motor vehicles, if a car driver killed someone regardless of care taken, they were generally considered responsible for taking a life and they could expect retribution or compensation demands to be directed toward them (Reference O'Neill, Zorn and BayneO'Neill 1975).Footnote 12
Not only does the literature suggest that was payback a legitimate traditional sanction across Melanesia, some suggest there could be moral and social obligations to carry out a payback killing. For instance, Reference TrompfTrompf (1994: 27) states that “ostracism” could await those unwilling to perform an “atoning payback” for lost kin in some parts, while Reference ReedReed (2003) documents people imprisoned for carrying out payback motivated by custom and group loyalty. For men, the practice of retributive violence was also seen as a mechanism for asserting one's “manliness” and increasing one's prestige within the group (Reference Strathern and ScaglionStrathern 1981; Reference TrompfTrompf 1994).
Finally, given the importance of the group, in contrast to individual aspects of personhood, some discussion is necessary of what it represents, both “traditionally” and in more recent times. Reference TrompfTrompf (1994: 112) considers that the traditional “Melanesian security circle” consisted of the extended family, a cluster of more distant relatives within one's tribal area and other trusted families where special-exchange partnerships existed. However, with increased urbanization and migration wontokism (literally “one talk” in Pisin) is a prominent in phenomenon Papua New Guinean society, especially in urban environments. In addition to traditional clan based security circles, wontokism sees people who speak the same language cooperate and support one another in urban environments (sometimes even if they are traditional enemies back home). Variants of wontokism include provincialism and regionalism, that see a greater sense of obligation toward people from the same province or region. In urban environments, particularly in Port Moresby, there are also a number of criminal raskol gangs, which are also usually populated along language or regional lines. These mutations of the “group” have led Reference TrompfTrompf (1994: 347) to speak of the development of a “neo tribal culture” in urban environments.Footnote 13
The State's Response to Payback
During the colonial period the vast majority of the population, apart from those living in or immediately around towns, continued to use custom with little interference from the colonial authorities. These arrangements, described by Reference Weisbrot, Weisbrot, Paliwala and SwayerrWeisbrot (1982: 66) as a situation where the state criminal law “was effectively reserved for European residents,” continued up until the mid-1960s. Those indigenous Papua New Guineans who were tried in colonial courts were usually sentenced leniently when motivated by customary considerations (including payback) as it was acknowledged they were “not criminals in the true sense of the word” (Reference GoreGore 1965: 88). To justify lenient sentences, which normally ranged from the rising of the court to 3 to 4 years imprisonment for homicide, the courts applied a broad interpretation of the partial defence of provocation based on concept of sophistication, which saw a defendant's level of education and exposure to the transplanted law as important factors in sentencing (Reference GoreGore 1965).
However, in response to a report by Reference DerhamDerham (1960), a professor of jurisprudence at the University of Melbourne, there was a change in policy and there was an attempt to roll out the transplanted criminal law across the entire population, primarily in an effort to achieve convergence between custom and the state criminal law.Footnote 14 Indeed, Professor Derham (1963) made it unmistakably clear that he saw criminal law as a powerful tool for social change. Over time, the courts have taken an increasingly tough stance toward payback in an effort it eradicate it. Now payback cases attract the highest possible sentences. In The State v. Monodoao,Footnote 15 National Court Judge Ellis sentenced a man convicted of wilful murder to life imprisonment for a payback killing. In his judgement he told the prisoner he should have been given the death penalty to serve as a deterrent, and put future payback killers on notice. In conclusion Justice Ellis said (reported by Reference KelolaKelola 2010):
The incidence of payback killings and deaths from tribal fighting has become a cause for concern. Right-thinking people regard such conduct as offensive. However, as there [are] still some in this country who regard payback killings as an age old custom, it is necessary to change such attitudes and to change the conduct of those who commit payback killings, regardless of whether they occur in a village, a rural town or the national capital.
On face value, using criminal law to eradicate payback and transform society would seem fit for purpose. Deterrence theory presumes that an individual will commit a crime if the expected punishment is less than the expected benefit. While there are sound reasons to be sceptical of this most basic approach to regulating human behavior, Reference Levitt, Miles, Polinsky and ShavellLevitt and Miles (2007) who reviewed the vast empirical literature found that differences in the expected punishment, either through the level of enforcement or magnitude of sanction, does seem to be related to the prevalence of crime, all other things being equal. In reaching this conclusion, they did report two important qualifications; the difficulty in separating deterrence from incapacitation effects and that a marginal increase in already high sanctions (e.g., the death penalty versus a lengthy prison term) may have little effect of deterrence.Footnote 16 Importantly, however, the expected punishment attached to a given behavior is only one factor that can affect its prevalence; indeed all things are not equal. Other factors, such as psychology, biology, emotions, society, and gender can also play an important, and perhaps a larger, role in criminality and criminal activity.Footnote 17
From the introduction, we know that states have had various degrees of success in transforming society through criminal law. If one were to accept the empirical validity of deterrence theory, one reason for this mixed success could simply be due to variations in enforcement activity. That is, while high magnitude headline sentences might be attached to a given activity, punishments may be rarely enforced. The expected punishment in such a case (the magnitude of the penalty multiplied by the probability of enforcement) could actually be low, resulting in a low level of deterrence, despite having high headline sentences for those few who get caught.
In Papua New Guinea, despite the courts imposing high magnitude sentences for payback killings, law enforcement levels are low. This sees many law breakers avoid state punishments, whether their crimes are motivated by payback or not. On average Papua New Guinea has one police officer per 1,084 people compared to the global average of 400 seeing criminal law enforcement relatively weak and geographically confined.Footnote 18 The police mainly confine themselves to the towns and their surrounds and rarely venture to the more remote areas, in part due to a lack of funding for transportation. The Papua New Guinea Law and Justice Sector Secretariat (2007: 11–39) reports a national arrest rate for reported murders at 27 percent and a prison escape rate of 7 percent per year. Therefore, while there are high magnitude sanctions for payback killings there are low levels of enforcement, especially in remote and rural areas, thereby limiting the deterrence effect of criminal law in eradicating payback.
The seminal work of Reference HartHart (1997) suggests that legal compliance is achieved through both external penalties (even if someone does not accept the law as it stands) and internalization of the law (where the law is accepted and used as a guide for conduct). That is, in addition to any external (deterrence and incapacitation) effects that law may generate, internalized beliefs (or internalized legal norms) may also play an important role in the prevalence of payback. Shedding light on how internalized beliefs are formed is the focus of this article: in particular the role that criminal law may play in forming internalized norms of individuals in a society that has been subjected to a legal transplant very different from its own pre-existing legal order. Therefore, the next section discusses the various influences on the internalization of legal norms and how they might be transmitted.
Conceptualising the Internalization of Legal Norms
To better understand and test how individual attitudes toward payback may be formed, a conceptual framework for norm internalization in a transplant society is developed. This framework draws on Reference Bisin and VerdierBisin and Verdier's (2010) evolutionary biology model which suggests that norms are transmitted through both vertical and horizontal mechanisms and highlights the importance of the intensity of exposure to these norms and transmission efforts in the probability of their adoption. In terms of internalized norms associated with the practice of payback, those who consider it to be a legitimate sanction are labeled “traditionalist” while those who do not consider it to be a legitimate sanction are labeled “neoteric.”Footnote 19
There are reasons to believe that criminal law should also play a role in the formation of internalized beliefs through its expressive role or in what Reference MerryMerry (2003: 973) describes as the “cultural production of norms.” Numerous authors have highlighted the expressive function of law and have suggested that the state is able to create, perpetuate and dismantle norms among its citizenry through its ability to make statements that signal good, bad or neutral behavior that are then internalized among its populace.Footnote 20 Similarly, some suggest that customary norms that run counter to the state law can be changed through increased access to courts and formal law enforcement. Reference AldashevAldashev et al. (2012: 798) stress that even if formal state courts are not used, their mere presence “can actually pull custom in its direction” through the threat of one of the disputants resorting to it—which they label the magnet effect. Indeed, Reference MerryMerry (2003: 943), in noting that only a small portion disputes make it to court concludes that “compliance depends largely on individual consciousness of law.” However, others such as Reference BicchieriBicchieri (2006) have highlighted the resilience of norms, while Reference Carbonara, Parisi and Von WangenheimCarbonara, Parisi, and Wangenheim (2012) even suggest that increases in the strictness of laws that run counter to social norms may actually lead to countervailing effects that can reinforce them.
Criminal law is not the only tool available to a state wishing to eradicate “bad” norms. In Place and Time Reference Bentham, Schofield and EngelmannBentham (2011: 174)Footnote 21 makes a distinction between “indirect legislation” and “direct legislation,” for a state wishing to transform society. While he did not provide precise definitions, indirect legislation refers to educative measures (e.g., schooling, instruction, and public education) while direct legislation refers to the (criminal) law itself. In this regard, Reference Bentham, Schofield and EngelmannBentham (2011) and others such as Reference Posner and RasmusenPosner and Rasmusen (1999: 382) suggest that non-legal measures may be more effective in eradicating “bad” norms by making such behavior “dishonourable.” In an effort to reduce definitional confusion going forward, those channels associated with Reference Bentham, Schofield and EngelmannBentham's (2011) notion of indirect legislation will be referred to as “non-legal.”
In addition to the state, other sources are also likely to play an important role in the formation of internalized legal norms. In Papua New Guinea, internalized beliefs will almost certainly depend heavily on the practice (and notions) of local custom, handed down from one generation to another, perhaps in a manner similar to how language is transmitted. In handing down local custom we can expect parents, customary leaders, and others within local communities to play an important role (Reference Bisin and VerdierBisin and Verdier 2010; Reference HenrichHenrich 2001). Other non-state influences on the internalization of legal norms may include missionaries, non-government organizations, aid agencies, expatriate businesses, and the media.
Importantly, custom itself may mutate, either endogenously or through the influence of exogenous forces. For instance, Reference WoodmanWoodman (1986) argues that the arrival of the sweet potato to the island of New Guinea centuries ago (and well before European colonization) revolutionized agriculture and society. In addition to the state influencing custom, Reference TrompfTrompf (1994) and Reference Wiessner and PupuWiessner and Pupu (2012) have highlighted the importance of Christian missionaries on Papua New Guinean society and the practice of custom. Reference GoddardGoddard (2009) even suggests that Australia's former overseas aid agency (AusAID) may have influenced custom by emphasising its restorative elements and downplaying its more violent and retributive elements through its various education and extension programmes. In summary, in addition to directly influencing internalized norms of individuals, governments and non-state groupings may also influence custom itself and therefore have an indirect effect on norm internalization.
It would seem that there are numerous sources of influence and transmission channels that can affect the internalization of legal norms. Sources of influence particularly relevant to Papua New Guinea would seem to be the state, missionaries, aid agencies and advocacy groups, expatriate businesses, and of course, custom itself. In terms of transmission channels, the state can use direct legislative efforts (e.g., criminal law) and/or non-legal channels (e.g., schooling and public education programs) while non-state groups are assumed to transmit via the latter. Furthermore, it would seem that all of these sources have the potential to influence the internalized norms of an individual either directly (e.g., inculcation through schooling or through socialization effects at work) or through altering the practice (or notions) of local custom itself. Given the potential for custom to mutate, a distinction is made between “traditional custom” and “current custom”—while acknowledging that each, and what is contained in each, is likely to be contested.Footnote 22
As can be seen from viewing Figure 1 below, traditional customary norms are represented as being transmitted vertically, while norms associated with the transplanted legal order (from a variety of sources as outlined above) are represented as being transmitted horizontally; noting that these horizontal influences can not only directly affect an individual's internalized beliefs but also indirectly by potentially altering the practice and/or perceptions of custom. Consistent, with Reference Bisin and VerdierBisin and Verdier (2010), the probability of an individual adopting the “traditionalist” view versus a “neoteric” view concerning payback will depend on the intensity of exposure and on the transmission effort from the various sources and channels. We can also expect the adoption of an internalized stance toward the practice of payback to depend on individual personal characteristics (e.g., gender). The scheme represented in Figure 1 does not attempt to provide an exhaustive account of sources or channels of legal norm internalization, but rather provide some conceptual clarity for empirical investigation.Footnote 23
Figure 1. Formation of Internalized Norms Concerning Payback in Papua New Guinea.
Before embarking on the empirics, it is worth taking stock of what has been established thus far: first, payback killings were once a legitimate customary sanction across Papua New Guinea while now they are deemed to be grave crimes that attract high magnitude state sanctions; second, internalized attitudes toward the legitimacy payback should depend on the intensity of the vertical transmission of custom (noting that custom itself may be mutable) and the intensity of the horizontal sources, including the state's direct legislative efforts (criminal law) and non-legal sources, from both the state and non-state sources.
Research Method
The primary focus of the empirical investigation is to shed light on the role that horizontal transmission channels play in the internalization of norms. Specifically, the aim is determine whether adopting a traditionalist stance toward payback has a statistical relationship with a measure for exposure to direct legislative efforts (proximity to state criminal law enforcement) and non-legal measures (quantity of education and the ability to speak English).Footnote 24 Before outlining the survey method and results, a short description of the New Guinea Islands, including a brief outline of crime and punishment is provided.
Crime and Punishment in the New Guinea Islands
The provinces of East New Britain, West New Britain, and the Autonomous Region of Bougainville (along with New Ireland and Manus Island) form the New Guinea Islands which lie off the north east coast of the New Guinea mainland. As in the rest of Papua New Guinea, the majority of inhabitants continue to live a semi-subsistence lifestyle tending small gardens.Footnote 25
The Islands have a common colonial history. They were first colonized by Germany in the late nineteenth century; taken over by the Australian military at the beginning of the First World War; under Japanese military control during much of the Second World War; then again under Australian military rule, followed by a period of Australian civilian rule. Since 1975 they have been provinces within the Independent State of Papua New Guinea.Footnote 26
Unlike some parts of Papua New Guinea which are reported as having some of the highest crime rates in the world, crime rates in the Islands, both in urban and rural communities are relatively low.Footnote 27 Like the rest of Papua New Guinea, police numbers are low compared to world averages and police mainly confine themselves to urban centres (see Papua New Guinea Law and Justice Sector Secretariat 2007). Despite this, it would be a mistake to conclude that wrongdoers are not punished. In both urban and rural areas, non-state enforcement continues to provide an important role in punishing criminal (and noncriminal) behavior.
An account of customary sanctions in the New Guinea Islands is presented below. This account is based interviews with community leaders (and community members) in each of the provinces visited.Footnote 28 A common view expressed was that for a payback killing to be permissible under “traditional” custom it must occur on the same day that the initial wrong was committed.Footnote 29 If payback cannot be enforced on the same day, compensation should be demanded instead.Footnote 30 The only exception to this rule is when a compensation demand is not met. Under these circumstances a payback killing may again be permissible, and in this sense it can be seen an enforcement mechanism for compensation demands (which are usually seen as a substitute for payback). In cases of homicide, compensation payments ranged from Kina (K) 20,000 (US$8,000) to more than 100,000 K (US$40,000) depending on the place and circumstances. In a country where average annual income was approximately US$1,300 per year at the time of the survey, even US$8,000 is a considerable sum. Given the magnitude of compensation demands, help from an extended kinship network is necessary for payment.Footnote 31 Usually, the families of the wronged and wrongdoer will come together and agree on a settlement, but sometimes a sum is simply demanded. Whether the wrong has been punished through payback or a compensation payment, it is usual for both parties to come together in a ceremony to signal the restoration of peaceful relations.
The customary leaders interviewed expressed a preference for compensation payments and discouraged payback. This suggests that payback is perhaps no-longer a legitimate sanction under “current custom,” but only under “traditional custom” and raises a number of questions over what custom is, how it can be identified and who, if anyone, has the ability or authority to articulate it. Some customary leaders and survey respondents on the Island of New Britain also expressed the view that violent payback was not their “traditional custom,” but rather that it has been imported by outsiders. While these statements cannot be completely ruled out, it is almost certainly not the case. Early accounts of Reference HahlHahl (1897) and those compiled by Reference Sack and EpsteinSack (1974) suggest that while compensation payments and sorcery were widely used to sanction wrongs in New Britain, violent indiscriminate payback also took place. As discussed earlier, the ubiquity and necessity of self-help across Papua New Guinea, due to the statelessness nature of the societies and a warrior culture has also been highlighted by Reference EpsteinEpstein (1974); while Reference TrompfTrompf (1994) also suggests that the practice of violent payback extended throughout Melanesia at the time of European contact—and also explicitly cites cases in the Islands of New Britain and Bougainville.Footnote 32
Survey
The survey took place in six remote villages and in each of the three urban capitals, and two former capitals of East New Britain, West New Britain, and the Autonomous Region of Bougainville. Data were collected via a survey questionnaire in both English and Pisin during October to December in 2010. In total, 200 questionnaire responses were collected; 23 from the Autonomous Region of Bougainville, 115 from East New Britain and 62 from West New Britain. Papua New Guinea is a difficult place to conduct this kind of research and all possible efforts were taken to avoid sampling bias. The sampling method used in both urban areas and remote villages was aimed at gaining a cross section of the community, given a lack of residency records and the difficulties of a unrecognized visitor entering some communities uninvited (both urban and rural). In towns and villages, the aim was to have a respondent from every third residence encountered, however in some villages and urban communities we were guided by local officials or relied on kinship networks. While the need for a random sample was stressed, and appreciated, the sampling method was not strictly randomized. In an effort to gain a more representative cross section of communities, household sampling was supplemented by sampling at marketplace and village stores due some residential “no go” areas (for outsiders). For consistency with the market samples, there were no follow-ups with the household samples. Of those asked to participate in the survey, approximately 36 percent declined.Footnote 33 As can be seen from Table 3 below, the respondents were on average relatively well educated and with higher levels of income compared to the Papua New Guinean population as a whole (World Bank 2013). This can partly be explained by the Islands Region having a reputation for being well educated and prosperous but it also suggests a relatively “middle class” sample. From Table 3, it can also be seen that bulk of respondents were drawn from East New Britain, the largest province in the New Guinea Islands in terms of population and that there are a higher proportion of male respondentsFootnote 34 Based on the conceptual framework developed, we can expect that higher levels of education, income, and formal employment would bias down the number of respondents who agree with the use of payback, while (based on the literature) higher levels of male respondents should increase it. Given the stated importance of Christian missionaries to Papua New Guinean society, it should be noted that Islands Region, and in particular all of the areas surveyed, were predominately Catholic suggesting that any effects they generate in relation to the internalization of norms concerning payback should be broadly uniform across the sample areas.
The survey focused on collecting data on internalized attitudes toward the legitimacy payback. The data were gathered using third person vignettes involving a homicide (and the potential for payback) which were read out to respondents—one accidental and the other deliberate. These were chosen over direct questioning to reduce reporting biases, especially social desirability bias, which can be particularly prominent in matters of ethics and criminal conduct (Reference Chung and MonroeChung and Monroe 2003). The vignettes and potential responses were adapted from the payback literature and two court cases in 2010.Footnote 35
Then in randomized order, the respondents were presented with various potential responses of the deceased's family and asked if they agreed or disagreed with each of them. Respondents were also given the option of registering neither agree nor disagree and respondents were not precluded from agreeing with multiple (mutually exclusive) responses. The potential responses included a payback killing, going to the police, seeking a compensation payment, and doing nothing. In this way, the vignettes were consistent with Reference GerberGerber's (2000: 962–63) approach of using “the concept of decision as the focal point of analysis” in gaining an understanding of foreign laws. The two vignettes are presented below.
Vignette: Scenario 1
Imagine the following scenario. It is very late on a Friday night and Joseph and his friend have been drinking alcohol and are walking back to their village. Given their level of drunkenness they sit down on the side of the road and fall asleep. Not long after, a van comes past and runs over Joseph and his friend. The driver stops but both Joseph and his friend are dead. The police interview the van driver and conclude that he was driving within the speed limit, and that the death of Joseph and his friend was an unfortunate accident. Indicate how you feel about the actions taken by Joseph's family.
They demand compensation from the van driver or his family.
They attack the van driver and kill him.
They attack a member of the van driver's family and kill him.
They go to the police and demand the van driver be charged with reckless driving causing death.
They do nothing as the death was an accident.
They throw rocks at the van and damage it.
Vignette: Scenario 2
Imagine the following scenario. Two men from the same village, John and Anthony got into a heated argument, and things got out of hand. Anthony attacked John with a bush knife, chopped his arm off and killed him. Indicate how you feel about the following actions of John's brother.
He kills Anthony's brother and burns down his house.
He meets with Anthony to work out a compensation payment.
He goes to the police and demands that he is charged with murder.
He kills Anthony.
He kills Anthony's brother.
He demands compensation from Anthony.
The aim of using the vignettes was to identify the traditionalist internalized viewpoint, defined as those who consider payback to be a legitimate sanction. Given that each respondent was free to agree with more than one response in each of the scenarios, the survey was designed to draw out those who agree with the use of payback, not whether they have a preference for it over other potential responses to a homicide. For instance, as can be seen in Table 1 below, some respondents agreed both with payback and going to the police.Footnote 36 In this sense the vignettes do not provide information on whether the respondents agree more with the use of payback than with the use imprisonment (or compensation payments) or vice versa, but only if they agree with the use of payback in the scenarios presented.Footnote 37 While there is no reason to believe that respondents systematically misrepresented their views this cannot be ruled out, as in any stated preference revelation exercise.Footnote 38 Table 2 below provides a list of the variables collected and their definitions.
Table 1. Stated Responses to the Vignettes
Results
As can be seen, only 12 percent of respondents agreed that the family should accept the decision of the police and do nothing more for Scenario 1 (accidental killing), while 69 percent agreed with the family demanding compensation from the driver, and 69 percent agreed with the family going back to the police and demanding the driver be charged. In relation to Scenario 2 (deliberate killing) 81 percent agreed with going to the police and jailing the offender while 56 percent agreed with demanding a compensation payment.Footnote 39
Overall, 36 percent of respondents agreed with the a payback killings in one or both of the scenarios. For Scenario 1 (accidental killing), 15 percent of respondents agreed with the use of payback while 29 percent agreed in the case of Scenario 2 (deliberate killing). Note that the aggregate figure highlights the fact that some respondents agreed with the use of a payback killing for the accidental death involving a motor vehicle and not in the deliberate killing following a heated argument among two men from the same village, and vice versa. It is also noteworthy that of those who agreed with the use of payback, slightly more agreed with the killing a member of the culprit's family rather than the actual culprit. These responses highlight the fact that the “traditional” customary norms of strict liability and group liability still hold some currency. The summary statistics for the variables used in the econometric analysis are presented in Table 3 below.
Table 2. List and Description of Variables
Table 3. Summary Statistics
Econometric Analysis
To test for the influence of horizontal norm transmission channels, the statistical relationship between those who have revealed themselves as having internalized a traditionalist viewpoint in relation to payback and the explanatory variables is estimated using a maximum likelihood probit multiple regression; where stated agreement with the use of payback is represented by 1 (traditionalist) and disagreeing is represented by 0.Footnote 40 Using this approach, it is possible to test for the presence of direct legislative channels (proximity and access to the state criminal law) and non-legal channels (measures for education quantity and quality and socialization effects).Footnote 41 Importantly, the use of multiple regression analysis enables the relationship to be estimated while controlling for a range of other factors that could also influence the response. The control variables include province, sex, income, age, type of employment, and whether the respondent had been a victim of a crime or is local to the province.Footnote 42
Conscious of the potential for multicollinearity, eight equations are presented in Table 4. The correlation coefficient between the variables Urban and Distance is −0.88, which is not surprising given that the police in Papua New Guinea are based in the town centres and usually require those who seek their assistance to go to town. In addition, the two pairs of variables Crime Victim and Violent Crime, and Income and Formal are also highly correlated, with respective coefficients of 0.72 and 0.67. Finally, there is also a reasonably strong correlation between English and Schooling with a coefficient of 0.52. Therefore, each variable is omitted in at least one of the equations below, with the variable Urban being included more often due to the higher number of recorded observations.Footnote 43
Table 4. Agree with the Use of Payback
Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1.
First, it can be seen that male respondents are more likely to agree with the use of payback suggesting they may have more (retributive) violent tendencies. This result is consistent with the literature suggesting payback was seen as a mechanism for asserting “manliness.” It can also be seen that the provincial dummies are significant in many of the equations, underscoring the heterogeneity of Papua New Guinea, and even within the Islands.
In terms of norm transmission channels, the results suggest that horizontal non-legal channels may be playing an important role. The variable Formal (representing formal employment) is significant and negative in the first three estimated equations, suggesting that those who work in formal employment are also less likely to agree with payback killings; remembering that the bulk of formal employment in the survey area is either with the government or expatriate firms. However, given the loss of significance with the exclusion of Income (and a change in the sign) it suggests the results in terms of formal employment are unstable. However, in all of the estimations the variable English is significant and negative, suggesting that those who speak English are less likely to agree with payback. The results also show that the variable Schooling (years of schooling) is not statistically significant. Conscious of the potential for multicollinearity given the degree of correlation between English and Schooling both were estimated with the absence of the other (and other highly correlated variables). As can be seen in Table 4, English remains statistically significant across each estimation, while Schooling remains insignificant even with the absence of English, Tertiary (education), and Formal (employment). These results suggest that the channel does not seem to be the quantity of education. While these results are discussed later, the significance of English could be capturing the quality of education and/or more general (non-legal) socialization effects.
The other significant result is that those respondents that have greater exposure and access to the state criminal law were more likely to agree with the use of payback. As can be seen, the variables measuring proximity to state criminal law enforcement Urban (whether the respondent resides in or near a town) and Distance (motorized time to police station) are both highly significant with signs that suggest that residents that have greater access to criminal law and are subject to greater enforcement levels are more likely to agree with the use of payback. In Equation 1 all observed variables are included in the equation and Distance is not statistically significant. However, when Urban is removed from the estimated equation, Distance is highly significant and the coefficient has changed sign, suggesting that multicollinearity was present in Equation 1. In Equation 2, the variable Distance is both significant and negative suggesting that the further the respondent is from a police station the less likely they are to agree with the use of payback. Therefore, overall these results suggest that there is no positive relationship between proximity of the state criminal law and the internalization of state legal norms. Indeed, they suggest that respondents are more likely to agree with payback when they are closer to the state criminal law. These results suggest that when controlling for a range of observed variables that may influence internalized norms, those in closer proximity to the state law (and with greater access to it) are more likely to adopt a traditionalist stance and agree with the use of payback.
Internalized norms toward the legitimacy of payback may also be gained from views on the criminal penalty for such activity. Respondents were also asked their view on the appropriate state sanction for a payback killing: 42 percent of respondents agreed with life imprisonment (the standard headline sentence for wilful homicide in Papua New Guinea), 38 percent agreed with a reduced sentence, and 20 percent agreed that there should be no prison sentence at all. It could be argued that respondents who consider that payback should be punished by life imprisonment have fully adopted the norms associated with the transplanted state legal order, that is, they have adopted the neoteric viewpoint.Footnote 44 In this respect, the data can provide a test of robustness of the previous results. However, it must be stressed that agreeing or disagreeing with payback is very different from agreeing or disagreeing with state penalties for such activity. For instance, one may disagree with cocaine use but also disagree with state prison terms for such activity. Alternatively, one may even compartmentalise their views within the two legal orders, and for instance, agree with use of payback under custom but simultaneously also agree with the imprisonment under state criminal law.Footnote 45 Despite these caveats, the regression results are broadly similar to those of the vignettes. They do not provide any evidence of direct legislative transmission into internalized norms, while they do suggest transmission through quantity of education. While the estimations are not presented owing to space limitations, the variables capturing proximity to the criminal justice system are statistically insignificant, suggesting that once controlling for other factors, respondents are equally as likely to agree with life imprisonment for the perpetrator of a payback killing irrespective of whether they live in an urban or rural environment, again, providing no evidence of internalized norm transmission based on proximity to the state criminal justice system. In terms of non-legal norm transmission, in this case, the variable Schooling is both significant and positive, suggesting those with more schooling are more likely to agree with current criminal punishments for payback.
Discussion
The survey results provide no evidence of the horizontal transmission of legal norms through geographical proximity to the state criminal system. However, there is evidence of transmission through non-legal channels: specifically the ability to speak English is consistently significant using the vignette data. Despite English being one of Papua New Guinea's three official languages, it is most people's third language, behind their local language, and Pisin, the lingua franca of the New Guinea Islands and much of Papua New Guinea. Apart from some rare instances, English is only learned at school and only spoken in formal education and for some government and business activities (Reference Franken and AugustFranken and August 2011). Therefore, this variable could be capturing quality of education and/or familiarity with Western norms through education, employment, or some other channel. In terms of educational quality, Reference SongySongy (2007) concludes that it varies considerably in Papua New Guinea, as does English proficiency (even with the same level of formal schooling).Footnote 46 While these results suggest the existence of non-legal horizontal norm transmission, as English is almost only learned through horizontal transmission channels, it is not possible to pin-point the exact channel due to the inter-relatedness of the variables. Furthermore there may be some degree of reverse causality, that is, those who have an internalized disposition toward Western legal norms are perhaps more likely to learn and speak English as a result.
On the whole, these results would seem to provide support for Reference Bentham, Schofield and EngelmannBentham's (2011) preference for non-legal measures in transplant societies. Contained in his “rules” for colonial states contemplating the eradication “repugnant” customs, Reference Bentham, Schofield and EngelmannBentham (2011: 174) concludes that: “indirect legislation should be preferred to direct: gentle means to violent: example, instruction, and exhortation should precede, or follow, or if possible stand in the place of, law.” Given Bentham's status as one of the founding fathers of deterrence theory, his preference for non-legal measures over legislative measures in transplant societies seems to be uncannily astute.
While these results do not find any evidence of norm transmission through proximity to the criminal law (i.e., the criminal law acting as a magnet in forming internalized beliefs) this channel still cannot be ruled out, as the results could simply be capturing the reality of living an urbanized existence in Papua New Guinea. For instance, such an existence may be one characterized by higher levels of raskolism (criminal gangs), murder, suspicion and harsher, and more depersonalized interactions (Reference Koczberski, Curry and ConnellKoczberski, Curry, and Connell 2001; Reference ReedReed 2003). There is also reason to believe that some of the negative aspects of urban life are exacerbated by group loyalty, which could explain the urban and rural divide in terms of the vignette data. As would be expected, the survey also found urban areas to be more ethnically heterogeneous than rural areas. This could be capturing a consequence of the phenomenon of the sliding scale of justice based on kinship ties as outlined by Reference Lawrence and BrownLawrence (1969). That is, despite payback being acknowledged as a legitimate traditional customary punishment, it was (and is still) usually reserved for outsiders of the group. Therefore, the imposition of heavier and more violent customary punishments in ethnically fractionalised settings, that is urban areas, would be expected. In this regard, Reference Strathern and ScaglionStrathern (1981: 13) also suggests that disputes in ethnically diverse environments are more likely to lead to “violence or threats of violence.” If this were the case horizontal norm transmission could be taking place through proximity and access to the state legal system, but it is outweighed by other factors. Of course, another explanation is that the criminal law produces countervailing effects as suggested by Reference Carbonara, Parisi and Von WangenheimCarbonara, Parisi, and Wangenheim (2012).
It is hoped that these results will spur further research in three broad areas. The first is gaining a better grasp of the exact nature of non-legal channels and sources associated with norm internalization. This would involve further untangling the different, but closely related non-legal channels, and determining the relative importance of state and non-state sources within this process. Also, while this study has established statistical correlation, further empirical research is necessary to claim causality, especially given the possibility of omitted variable bias and reverse causality. The second area involves better understanding, and testing, the reasons why urban residents are more likely to adopt a traditionalist stance toward payback than their rural counterparts. While there is an explanation put forward, it is acknowledged that this is little more than informed speculation, and there are other possibilities that require further investigation and testing. The third area for further research concerns the mutation of custom (and notions of what custom is and was), the sources and channels of these mutations, and the effect that this may have on internalized norms. It is noteworthy how closely these findings accord with descriptions of custom written some time ago, including during the colonial period, suggesting that certain conceptions and principles of what custom was or is are quite persistent. But on the other hand, perceptions of what “traditional” custom was seemed to be different on the Island of New Britain. Further investigation into these phenomena and the indirect influence of horizontal sources of norm formation through changed perceptions of custom would seem to be a particularly rich field for further research.
Conclusion
For over one third of survey respondents, payback remains a legitimate sanction despite it being illegal for more than 130 years. This highlights the persistence of customary norms even when they are in direct opposition to the state's criminal law. Alternatively, these results could be seen as a success story in social engineering; in less than 130 years payback has gone from being a legitimate sanction to being considered a wrong itself by almost two thirds of the survey respondents. Perhaps this success is all the more impressive when we remember that any real attempt to roll out the transplanted criminal law beyond the urban centres only began in the 1960s.
Either way, the results show no evidence of a change in internalized beliefs through direct legislative channels. That is, there is no evidence of a complementary relationship between a respondent's proximity to state law enforcement and a rejection of the customary norm of payback. In fact, there is evidence of an opposite relationship; those respondents living in urban areas are more likely to be traditionalists and agree with the use of payback than those respondents in rural areas. However, it must be remembered that attitudes toward payback and proximity to the state criminal justice system are likely to be endogenous. Even with this caveat, what can be said with some certainty is that if there is a magnet effect between proximity to the criminal law, it is a relatively weak force. In contrast, there is evidence of norm transmission through non-legal channels. While there is a close association between the non-legal horizontal variables used in the econometric analysis, making the exact transmission channels difficult to identify, the significance of the ability to speak English in terms of disagreeing with the practice of payback is a notable result.
