Home | Databases | WorldLII | Search | Feedback Papua New Guinea Law Reform Commission

Domestic Violence in Urban Papua New Guinea, Occasional Paper 19 [1986] PGLawRComm 1 (1 January 1986)

LAW REFORM COMMISSION OF PAPUA NEW GUINEA

DOMESTIC VIOLENCE
IN
URBAN PAPUA NEW GUINEA

edited by
Susan Toft

OCCASIONAL PAPER NO. 19, 1986

Copyright 1985 by the Law Reform Commission of Papua New Guinea

Printed at the Government Printing Office, Papua New Guinea and published by the Law
Reform Commission of Papua New Guinea.

LAW REFORM COMMISSION OF PAPUA NEW GUINEA

Commissioners

Joseph Faupugu Aisa, Acting Chairman
Roy Tiden
Lepani Watson
Imnaculata Kereku
Robert Lak
Allan Marat

Secretary

Josepha N Kanawi

Research Staff

Sao Gabi
Susan Toff

Administrative Staff

Luke Nalo
Ari Heai
Theresa Imamaeva
Jack Uke-e

The Commission is situated in the 4 Mile Government Offices.

Postal Address: P0 Box 3439
BOROKO
Telephone: 258755

NOTE

This publication is one in a group to be produced by the Papua New Guinea Law
Reform Commission during 1985 and 1986 on the topics of Marriage and/or
Domestic Violence. The others are:

• Occasional Paper No. 18

MARRIAGE AND DOMESTIC VIOLENCE IN RURAL PAPUA NEW GUINEA
results of a rural survey.

• Monograph No. 3

DOMESTIC VIOLENCE IN PAPUA NEW GUINEA
a collection of articles by professional anthropologists.

• Monograph No. 4

MARRIAGE IN PAPUA NEW GUINEA
a companion to Occasional Paper No. 18 ethnographic information from the
survey villages.

• Report No. l4

DOMESTIC VIOLENCE
summarising and concluding the recent research into domestic violence, with
recommendations.

ACKNOWLEDGEMENTS

Thanks go to all the people who have contributed to the studies presented in this publication either through participation or administrative support.

The main urban survey, that of low income earners, was only possible because of the co-operation of the University of Papua New Guinea Geography Department, through Stephen Ranck, and the personnel departments of Burns Philip (Papua New Guinea) Ltd., Steamships Ltd., the Electricity Commission and the Defence Force. The survey of elites was done through the support of the Public Services Commission and those senior public servants who responded to the questionnaire.

Rosa Au conducted the urban settlement survey as part of her degree studies and her assistance was very much appreciated. Dr. Alec Ekeroma is thanked for kindly giving us permission to publish the results of his hospital survey. Thanks go to Vagi Raula who drew the graphs, and to Mathew Mobutuna of the National Mapping Bureau for the map. Ari Heai typed the manuscript of this publication, presenting the printer with camera-ready tables and text. She is commended and thanked for her outstanding work and for her patience and perseverance.

LAW REFORM COMMISSION
TERMS OF REFERENCE
DOMESTIC VIOLENCE

I, TONY BAIS, M.P., Minister for Justice, by virtue of the power conferred on me by Section 9 of the Law Reform Commission Act, 1975, and all other powers me enabling, refer the following matter to the Law Reform Commission for enquiry and report –

Because -

1. domestic violence is contrary to the principles of our Constitution; and

2. the law does not enable the police and courts effectively t o protect women from domestic violence.

Enquire into and report to me on -

1. the nature and extent of domestic violence as a social problem; and

2. the legal remedies available for complaints of domestic violence; and

3. any changes to the law which may be necessary or desirable to achieve the protection of women from domestic violence; and

4. the steps which should be taken to bring the problem of domestic violence to the public notice.

In undertaking this reference you will –

1. consult with such bodies or people as you consider appropriate, including the National Council of Women; and

2. consider whether 'domestic assault' should be a specific offence in criminal law so that the police are obliged to prosecute even when the victim will not proceed with charges; and

3. examine the law of evidence and the defence of provocation with particular reference to domestic assault; and

4. consider any other relevant aspect of the topic as may be revealed during the enquiry.

When making your reports on the matters the subject of this reference you will attach drafts of any legislation required to give effect to any of the recommendations in those reports.

Dated this 18th day of August, 1982.

ANTHONY BAIS, M.P..
Minister for Justice.

---------------------------------------

TABLE OF CONTENTS

INTRODUCTION

On 18 August 1982 the Papua New Guinea Minister for Justice, the Honourable A.L. Bais, MP, passed to the Law Reform Commission a Reference on Domestic Violence. The Reference was in response to a request from the National Council of Women (Appendix I). A resolution passed at their Fifth Convention 1981 stated that crimes against women, particularly domestic violence and rape, seemed to be increasing and that there were inadequate controls to protect women.

In formulating an approach towards this as a research topic it was decided that the nature and extent of domestic violence should be investigated through a series of studies. The initial study was rural. A survey was carried out at the end of 1982 and results are published in Law Reform Commission of Papua New Guinea Occasional Paper No. 18 and Monograph No. 4. This present paper deals with domestic violence in the modern urban environment of Papua New Guinea and includes five approaches to the topic: a survey of urban elites; a survey of urban low income earners; a survey of two urban settlements; a hospital survey; and two case studies of battered urban women.

The urban elite and urban low income earner surveys are analysed together, with comparative data from the previous rural survey, but otherwise the studies are presented independently. They all point to the fact that domestic violence is a phenomenon of married life for many Papua New Guinean couples. The conclusions to each chapter explain the general situation. The findings will be collated in Law Reform Commission of Papua New Guinea Report No. 14 which will serve as a synopsis for the research on domestic violence conducted by the Commission.

The articles in this present publication are innovative efforts to produce base-line data on domestic violence in Papua New Guinea and they reveal a high rate and tolerance of domestic violence. This could be a partial explanation for the present highly publicised problems of sexual harassment and rape of women.

If violence is occurring between the sexes within the family, then beyond the family even less respect to women may be shown. The rape of women from an enemy group or in a 'pay-back' conflict between men was traditional in many areas. Today, the violent mistreatment of women, much of it apparently exercised on strangers who are random victims, has brought strong public reaction and criticism. This concern prompted the following studies. Both the concern and the studies point to changes in Papua New Guinean attitudes towards marital violence. The country's decision makers need to be aware of these changed attitudes when they plan future development initiatives.

Only 12%¹ of Papua New Guinea's total population is urban, but it is urban dwellers who first publicly expressed concern about the occurrence of domestic violence, mainly wife-beating. The urban population is drawn from a rural pool of nearly three million people divided among over seven hundred linguistically discrete ethnic groups. Thus people with diverse cultural backgrounds come together in an urban environment which has developed almost entirely since the Second World War. People are drawn from rural to urban areas by many factors, important among which are employment opportunities, the availability of services such as health and education, the sophistication of the urban life-style and the presence of wantoks² to provide hospitality and support. Nearly all urban dwellers retain personal rights to the use of land in a rural location, but once having adapted to the town environment few people, especially the more educated ones, find village life congenial for more than brief visits. Coupled with that, in many rural areas, someone who returns permanently to the village is seen as a drop-out, a failure, and there is consequently great social pressure on the urbanite to struggle remorselessly on, not only to survive in the town, but to make a success of it there. Port Moresby, with a population, of over 124,000 is approximately twice the size of the next largest city, Lae. Its role as national capital contributes to making it both the major administrative and commercial centre and it is thus attractive to the aspiring rural person.

Notes

1. Estimates of population are based on the National Statistical Office (January 1982) Papua New Guinea 1980 National Census Pre-Release: Summary of Final Figures. Government Printer, Port Moresby.

2. Wantok is a pidgin word meaning ‘one language’, it is used to express the close relationship between people who share the same language and are therefore from the same cultural group. Sometimes used generally to mean ‘friend’.

CHAPTER ONE
DOMESTIC VIOLENCE IN AN URBAN CONTEXT
WITH RURAL COMPARISONS

by Stephen Ranck and Susan Toft

INTRODUCTION

The aim of this chapter is to measure the incidence, frequency and intensity of domestic violence amongst urban low income workers and urban elites. The primary data come from questionnaire surveys. The first survey covered low income earners in Port Moresby. It was carried out in 1984. The second survey, of urban elites, was done in 1985. The results of these two urban surveys are compared and contrasted with a similar survey, carried out in 1982 (Toft & Bonnell, 1985), covering rural sections of Papua New Guinea society.

The questionnaire survey technique was used for its speed and wide coverage. The internal consistency within samples, and the patterns revealed among the different samples, point to the basic validity of the technique in this context. In other words, it was the most appropriate data gathering tool given the aims, resources and time for this project. Questionnaires of this sort cannot probe too deeply and will not reveal all the underlying mechanisms triggering domestic violence. What they do very well is to set the minimum parameters for violence and bring the issues clearly into focus for any ameliorative initiatives which may follow or for any future in-depth research. Any bias will generally be in the direction of under-reporting. Therefore, the trends shown in this survey must be taken very seriously as they show the situation in its best light.

Urban low income earner survey

A survey of the low income segment of the urban population was organised in conjunction with the Geography Department at the University of Papua New Guinea. Thirty-six students enrolled in a course 'Cultural Geography' were trained in survey methods and techniques over a period of four weeks. They were then taken out to specific urban locations in Port Moresby with questionnaire papers which they used for formal interviews. Supervision of the students was tightly controlled. The work was an integral part of their course for which they received credits towards a Bachelor of Arts degree: this provided an incentive for the students to procure materials of good quality and ensured a high standard of performance.

As marriage and domestic violence are topics which concern people's personal lives, they could be sensitive to frank discussion in a survey. During the previous urban settlement and rural surveys, however, it was found that respondents were open, co-operative and interested. They appreciated that the family is at the core of their social life and were very willing to contribute to an enquiry seeking to identify problems which threaten marital and family stability. It was felt that a similar approach would work in the urban low income earner survey, but interviewers were cautioned to be sensitive and respect respondents' rights to privacy. Fortunately, respondents were generally co-operative, as observed and as opined by the student researchers in post survey de-briefing. Consistency between male and female results corroborates the quality of the data.

During the rural survey (Toft and Bonnell, 1985) students lived for a month in villages where they developed a working relationship within a homogenous community. This was not possible in the urban situation for various practical reasons including finance, accommodation and other demands on student time. It was felt that in face to face interviews with strangers, people would be most likely to be honest if interviewed in private, and certainly not in the presence of their spouse. As the majority of formal sector low income earners (mainly manual workers) in Port Moresby are men, it was decided to interview the men at their places of work and, also during normal working hours, to visit housing areas identified as low-cost (National Statistical Office 1982) where women could be expected to be at home attending to domestic responsibilities. In principle, female students were to interview women and male students to interview men. This was difficult in practice because, of the 36 students involved in the survey only nine were women, and their task of going around on a door to door survey was time consuming. Consequently, three male students who were mature and capable were selected to interview women employed in low income jobs at their places of work; they, with two women (one of the authors and a senior member of Steamships Personnel Department) interviewed 35.9% of the female respondents in this way.

The target population of low income earners had to be carefully identified. Small business enterprises which had been approached for assistance were unwilling to release employees individually for up to 30 minutes per interview. Two large companies, however, gave the study full support: Burns Philip (Papua New Guinea) Limited and Steamships Trading Company Limited. Without the help of their personnel departments and the managers of various subsidiary businesses the task would have been daunting and very difficult to monitor and control. As it was, easy access was provided to employees in bottling plants, a Laundry, steel works, supermarkets, hotels, a quarry, dock yards, a timber yard, garages, and wholesale storage yards. The total labouring work-force in each place was interviewed. In addition to Burns Philip and Steamships, assistance was given by the Electricity Commission, which made employees available for interview and gave authorization for a door to door survey of the Commission's housing complex, and the Defence Force gave permission for a door to door survey of houses occupied by families of lower ranks at Murray Barracks. Low cost housing areas in the suburb of Gerehu were visited door to door and all houses were covered in each housing area (Gerehu, the Electricity Commission and Murray Barracks)¹. Students were taken by bus to pre-arranged places, handed questionnaire papers which were collected when the bus returned to pick up the students, and supervisors did spot-checks during the sessions. Each student was expected to complete a minimum of twenty questionnaires, an average of five during each of four field sessions. Some students needed five sessions to complete their quota and some took an extra session to complete extra questionnaires. A total of 666 people were interviewed of whom 368 (55.3%) were men and 298 (44.7%) were women (Table 1). This is out of an estimated Papua New Guinean urban population in Port Moresby of 131,525, in September 1984². Presently, there are no data to indicate with any precision what proportion of the Port Moresby population would be within the formal low income sector. The respondents represent a wide diversity of residential areas (Table 2).

The previous rural survey served as a preparatory study both for this urban survey of formal sector low income earners and for the separate survey of urban elites. Appropriate questions in the rural questionnaire which related to marriage problems and domestic violence were extracted and used to compose urban questionnaires (Appendices 2 to 4), and on the basis of the rural results, coded answers were formulated. Thus, direct comparisons can be made between the results of the three surveys: rural, urban low income and urban elite. These are presented in this paper.

Urban elite survey

Accessibility to elite respondents was much more difficult than to the lower income groups and elite sensitivities were more pronounced. Responses in the urban elite section may consequently carry a wider margin of error than in either of the other two surveys. However, there is a consistency in the pattern of results which is obvious and which suggests reliability. Further, any bias should be that of under-reporting which means that the urban elite survey shows minimum figures for domestic violence rather than any exaggerations. This principle enhances the value of the survey.

The Public Services Commission of Papua New Guinea lent its support to this survey and directed that all ‘married’ (see definitions below) public servants of high rank, Clerk Class 10 and above, should receive a postal questionnaire.

TABLE 1: CHARACTERISTICS OF SAMPLES ACCORDING TO SEX,
ACE AND MARITAL STATUS AT TIME OF SURVEY

*The term ‘living together’, as opposed to ‘married’, was recorded as accurately as possible during the low income survey by interview, but it was clear from responses to the postal elite survey that some respondents had not differentiated between the terms ‘married’ and ‘living together’. In cases where both categories were marked, to indicate that the respondent was married living with a spouse, the ‘married’ category only was coded. But there may be respondents who marked ‘living together’ when they were in fact married.

Departments supplied lists of staff who fell into that category and officers were sent separate questionnaires for themselves and their spouse (see Appendices 5 and 6). This was organised in such a way that each person had an addressed envelope in which to seal and return the anonymous questionnaire through the Public Service internal mail system. Of 530 officers circulated (ie. 1060 questionnaires distributed), 277 papers were returned: 178 out of 530 from men, which is a 33.6% response rate, and 99 out of 530 from women, a response rate of 18.7%. Only a handful of senior public servants are women (the official figure was 27 at the time of survey) and most of the women who received questionnaires did so through their husbands (Table 1). It is not possible to know how many of the 178 male respondents in fact took home the questionnaires to their spouses, but it is clear from papers returned that many of the women were of a lower educational
standard than the husbands and had difficulty in completing the form; several questionnaires in excess of 99 were received but had to be discounted because respondents had clearly not understood all the questions, or had spoiled their paper in some way.

TABLE 2 : RESIDENTIAL AREAS OF LOW INCOME EARNER RESPONDENTS

Definitions and background

For the purpose of this study, marriage is defined as ‘....a union between a man and woman such that children born by the woman are recognised as the legitimate offspring of both partners’ (Committee of the Royal Anthropological Institute of Great Britain and Northern Ireland, 1974:71).

There are three forms of marriage recognised and practised in Papua New Guinea today: marriage according to custom, marriage according to church law and marriage according to state law which is often referred to as general law marriage (Marriage Act, Chapter No. 280, Revised Laws of Papua New Guinea 1980). The three types of marriage are not exclusive. For instance church marriages often include the registration of marriage according to the Marriage Act. In urban areas a fourth form, the de facto marriage, is an increasing phenomenon, largely for the convenience of the couple involved. This type of liaison is discussed in more detail in Chapter 4. For the purpose of the urban surveys de facto relationships are not distinguished from other forms of marriage because many urban Papua New Guineans today do not readily recognise the distinction.

The marriage union is arranged, solemnised and maintained according to varying practices throughout society, and involves partners in a number of reciprocal rights and duties. These rights and responsibilities are vested in the social roles held by individuals according to their status within a particular society. Husbands and wives have many roles within the context of marriage, the extended family and society at large. The expectations that members of a society have of individuals in their roles are determined by social values and norms, and the failure of individuals to fulfil these expectations exposes them to criticism. The problem of expectations spouses have of each other and failure on the part of one of them to fulfil obligations became a focal point of interest during the rural survey. It featured importantly in the results of that survey and needs some explanation.

The survey term ‘failure to meet obligations’ covers a wide variety of unfulfilled cultural expectations concerning the behaviour of husbands and wives in marriage. In the sample of villages covered, the prime concern was failure of the wife in traditional obligations, including household duties. The wives themselves agreed that such behaviour gave legitimate grounds for criticism, and although there is clearly the concept of a lazy wife, it was also clear during interview that women may deliberately withhold services as a sign of dissatisfaction with their husbands.

‘Failure to meet obligations’ as stated by both male and female respondents was usually referring to the wife’s rather than the husband’s failings. Examples of traditional services which rural wives are obliged to perform include cultivating the garden, preparing food, looking after pigs and other wealth, looking after and feeding relatives, giving the husband sexual access, contributing to general family welfare and performing ritual functions when necessary. One of the researchers in an ethnographic description of Uma village, Southern Highlands, wrote:

From my own observations, the women do more work than the

men. Wives show a lot of respect to their husbands by performing

their duties properly in order to please their husband.

On the other hand the men show little respect to the wives.

They let the wives do a lot of work but do not let them

participate in decision making. When sharing goods and

possessions, often the husbands do not make a fair distribution

and they keep the money after the sale of goods in the

market. They also hit their wife when she fails to do her

work properly without realising that wives are loaded with

a lot of work. Beside this, the husbands often take a new

wife, causing problems within the families (Warus n.d.)

In the male-dominated societies which are prevalent in rural (neo-traditional) Papua New Guinea women are under greater daily pressure than men to perform duties. Men also have specific obligations but their traditional tasks tend to be less related to the immediate needs of the family and seem to have been more influenced by change than those of women. For instance, a man whose father would have been mainly engaged in territorial defence may himself be occupied primarily in the production of cash crops. His wife, however, still the provider of subsistence crops and the housekeeper, holds roles which are in principle unchanged. Men's obligatory duties are sometimes more arduous than women's but are irregular, and in general men have more options concerning their own activities on a day to day basis.

Women’s failure to meet obligations emerged as a major cause of problems and violence in rural areas. Therefore it was surmised that, in the urban environment where the pattern of life is different and roles are less clearly prescribed, even more serious marital problems would arise because the expectations spouses might have of each other, based on traditional concepts of roles and standards, would be impossible to fulfil.

There is a second term used in the survey which needs clarification: ‘sexual jealousy’. Jealousy seems to be an accepted justifiable emotion and a genuine cause for grievance in Papua New Guinea, whether relating to sexual suspicion or envy of success in others. Traditionally, a jealous person was one who could be expected to attack the object of his or her feelings through magic. So it was in the past, and is still today, considered inadvisable to arouse jealous feelings in a person.

Jealousy relates to sex through the possessiveness which a person may feel for a sexual or marital partner. Both behaviour which arouses jealousy and jealous behaviour are often culturally determined. General ethnographic information on Papua New Guinea tells of numerous and varied traditional ways in which sexual segregation is upheld. To ethnocentric Western eyes, noting outward appearances only, Papua New Guineans may seem to be immodest. To many Westerners, for instance, exposure of a woman's torso by the scantiness of traditional styles of dress has a sexual connotation. In fact, in Papua New Guinea strong and often subtle social constraints distance men from women, prescribing social behaviour and thereby formalising relationships. In some traditional Papua New Guinea societies the level of segregation between sexes could be compared with that found in Muslim countries where the separation is overtly expressed and displayed.

The word adultery is not always used in Papua New Guinea in its literal sense: sexual intercourse between two people when at least one of them is married. Thus, when people give adultery as a questionnaire response the meaning may well not include the physical relationship which it would in the West; rather it indicates offended, suspicious or jealous emotions which have been provoked by the behaviour of a spouse in relating to a member of the opposite sex. Further confusion is due to the fact that the word adultery does not always have a precise equivalent in translation. It is also due to the fact that in some Papua New Guinean societies, to be seen in circumstances which traditionally would be considered compromising can be as serious as if sexual intercourse had occurred. Such circumstances might be talking to someone who custom dictates should be at a social distance or to someone with whom normal sexual relations would not be taboo. For example by Question 5, ‘What do you think causes problems in marriage?’, on the questionnaire of a male respondent in the village of Vavua, West New Britain, the interviewer wrote: “[A problem is caused] if a married man gives betel nut, or other things, to another woman who is not a relative, and the wife finds out and suspects him
of having an affair”.

Jealous feelings may be particularly exacerbated when the social distance which men and women are customarily expected to observe starts to break down, as is occurring in the changing rural and urban spheres of Papua New Guinean society. It is better to talk of sexual suspicion or sexual jealousy rather than adultery as a source of marital problems. The questionnaire category ‘sexual jealousy’ includes adultery, real or suspected, and behaviour on the part of a spouse which is considered loose or tantamount to adultery according to custom.

Sexual jealousy featured as a major problem and cause of violence in the rural survey. Therefore, it was expected that problems due to sexual suspicion and jealousy would increase in the anonymity of the urban situation, where spouses can move around more independently of each other and of relatives or of co-villagers who would know their activities in the rural context.

There is a relationship between a wife’s failure to meet her obligations and some other categories of questionnaire response, especially sexual jealousy, which is difficult to assess. If a wife is displeased with her husband, and is possibly jealous, she may express her resentment by withholding services, by not cooking for example, as a result of which her husband may hit her. The stated problem may be that she failed in her obligations: he hit her because she did not cook. But the original, hidden cause of the problem is that she was jealous, whether justifiably so or not.

SURVEY RESULTS

Below, the findings from the three surveys are presented, compared and contrasted. Short titles have been given to each group surveyed. ‘Rural’ is used to designate the rural survey. ‘Urban Low Income’ is used to designate the low income earners who are married and working in the urban formal sector, and married women whose husbands may or may not be formally employed (93.3% of the female respondents had employed husbands; 18.5% of the male respondents had employed wives). ‘Urban Elite’ designates high ranking government employees who returned their postal surveys (50.3% of them had wives who were employed).

Causes of problems in marriage

The questionnaire respondents were all asked to name what in their opinion were the three greatest sources of marital problems. Figure 1 shows the responses. It illustrates a consistency between male and female responses which is reflected in most of the survey results. Emphasis on problems may differ but rankings of importance generally show consistency between the sexes.

As mentioned in the Introduction, the two main causes of marriage problems in rural areas are sexual jealousy and a wife’s failure to meet marital obligations, and it was supposed that both these problems would be more severe in the urban environment. This is not the case. They are still important problems, but they have rivals in alcohol and money. Access to alcohol is much easier in the urban environment and money is required for food and other essentials.

Seventy-one per cent of the urban low income women (70.8%) place alcohol at the top of their scale. It is understandable, as men are the primary consumers of alcohol, that fewer of them (40%) accept that this is a problem. Elite men (60.7%) and women (65.7%). however agree with one another in placing it at the top of their list. Urban elite men are perhaps more objective in their judgement; it is unlikely that our respondents were mainly non-drinkers because alcohol features elsewhere in elite results.

Sexual jealousy is seen as the second most common cause of problems by both the urban groups of women: 61.1% of the low income group and 52.5% of elite group. Unfortunately, there is nothing to show whether the problems stem more from male or female provocation. The fact that the men in all three surveys saw sexual jealousy as slightly less of a problem than the women did could be connected with the fact that the majority of men feel they have proprietary rights over their wives, in which case male possessiveness is not seen as a problem. It is likely that women generally feel more threatened by the freedom men have, even the traditional freedom to take more wives, and are more vulnerable as victims of both male jealousy and male promiscuity.

Rural

Urban Low Income

Urban Elite

The men from the urban low income group see money problems as the biggest obstacle to marital peace. Slightly more women in that group (57.4%, compared with 56.0% of the men) also think that money causes problems, but it is third on their list after alcohol and sexual jealousy. It is to be expected that, for this group of people, money, or lack of it, would present problems. The urban elites also see money as a problem but both the men and women, like the low income women, place it after alcohol and sexual jealousy. The relationship between money problems and alcohol consumption is obvious, and it is impossible to know whether the fact that alcohol features so strikingly as a major problem is connected more to the effect it has on a household budget or to a husband's drunken behaviour.

The elites presumably have more money at their disposal than other urbanites, but the pressures on them to spend are just as demanding as on other groups. Not least, their urban low income relatives and wantoks³ expect their help. Apart from providing a certain ‘know how’ when required, the provision of money is the basic way in which urbanites can help their rural cousins. The demands are varied - air fares, school fees, medicine, bride-price contributions, pocket-money etc. - but money is the solution. Problems from relatives and wantoks feature much more in the elite survey than in the others and this is likely to be strongly connected with financial demands from a wide circle of people. Troubles over bride-price feature as a minor problem in all three surveys. When respondents stated that this was a problem it was more with reference to pressure being put upon the couple by parents and less to do with lack of cash. The wife's family want to extract bride-price payments in cash or kind from the man or his family and the main way they do this is through the couple. This causes friction between their daughter and her husband or between the couple and either family.

The category ‘failure to meet obligations’ is markedly different in the three sets of survey results. This is particularly important because it indicates a change in attitudes and concepts of roles. In the rural survey men saw a wife's failure to meet obligations as the major cause of problems. Women also felt this was a big problem, but neither sex complained much about a husband failing in his obligations. In the urban low income survey we see indications of women changing their stand and slightly more of them named a husband's failure to meet his obligations as a cause of problems than did those who named a wife's failure: 17.4% as opposed to 15.1%. The elite women complained almost as much about men failing in their obligations (46.5%) as men complained about women (48.9%). The following extract from the data in Figure 1 makes the picture clear (Table 3).

TABLE 3: ATTITUDES TOWARDS MARITAL OBLIGATIONS

The urban male is apparently under greater domestic pressure than is the rural male. Urban male roles tend to follow those of the Western world, where the man is openly seen as the principal provider for the family. The rural male has had many of his traditional functions as protector and hunter for the family severely eroded, leaving him under much less pressure to perform on a day to day basis regarding subsistence and survival. Table 3 shows clearly that some urban men accept that husbands not meeting obligations cause problems in marriage. Amongst the urban low income group, both men and women feel that husbands are not meeting obligations, which points to an increasing sensitivity amongst men about the changing urban male domestic role. This is most evident amongst the urban elite, where the highest response rates regarding husbands failing to meet obligations are shown, but in this instance it is the women who are relatively most concerned. The two groups, men and women, are behaving identically. They both see there are problems regarding the fulfilment of obligations but tend to point an accusing finger at the other group. Many men and women now acknowledge that in the urban environment men have 'domestic' obligations towards their wives which they should fulfil. So two changes have occurred: the role of the male within the family and the attitudes of both sexes towards male and female roles.

Urban women are less prepared than rural women to admit fault in failing to meet obligations. This is a new attitude which indicates that many of them no longer accept that they are servant to man. They have reassessed their role and rejected their previous subordinate position. The urban elite women do not show this new attitude the same way as the women in the low income group, but it is evident from Table 3 that the attitudes of male and female urban elites follow a similar trend. Many of them recognise they are not meeting obligations and each sex sees the other as failing even more in this respect than itself. This is indicative of the strain which urban elite marriage partners must be under. They realise the new requirements entailed by modern marriage and approximately 30% of husbands and wives noted themselves failing in obligations, whereas around 47% felt their spouses were more at fault. It is understandable that people should more readily find fault in a partner than in themselves when under pressure in an entirely new socio-economic and material environment.

In the ethnically heterogeneous urban environment, unlike rural areas, mixed marriages (marriages between Papua New Guineans from different ethnic groups) are not unusual. Expectations people have in the roles of husband and wife are based largely on their own socialisation and experience, so if two people from different backgrounds with different role expectations marry, there may be conflict over the performance of one or both of the partners in their role as spouse. Another source of conflict in these marriages can be the two sets of relatives with their two differing sets of expectations. Mixed marriages are seen in Figure 1 as a minor new cause of problems. It must also be noted that mixed marriages can have benefits which the data do not touch upon. For example, mixed marriages sometimes provide a means of escape from village obligations and arranged marriages.

TABLE 4: ARE RESPONDENTS AND SPOUSES FROM THE SAME
LANGUAGE CROUP?

The extent to which mixed marriages exist in urban areas can be gauged by responses to a question asking if a respondent and his or her spouse are from the same language group. Table 4 shows that approximately two-thirds of the low income group are from the same cultural group and one third have mixed marriages. For the elites, it is closer to half of the group which has mixed marriages, and the higher number of elite people who named mixed marriages as causing marriage problems is consistent with this higher proportion of mixed marriages. The difference between elite male and female results reflects the degree to which returned questionnaires were not in fact from couples; this is because sometimes only one of a couple, either male or female, responded and also because of the much higher number of male respondents.

On the positive side for mixed marriages, there is a reduction in marital problems due to simple dislike of the marriage partner. In the rural areas, where people are often pressured by families into making particular marriages, dislike of partner was thought to be a problem by over one third of the respondents (35.1% men, 33.9% women). In both the low income and elite surveys this category is reduced to less than 10% of the respondents. There could be several factors contributing to this urban situation, such as the fact that people are more free to choose their own marriage partners, the need for couples to co-operate more closely in organising their lives coupled with the independence of the nuclear family in the urban context and better opportunities to leave disliked partners.

Children, in as much as they cause problems in marriage, did not feature in any specific way in questionnaire responses. Presumably there must be some problems between spouses in raising children but these were not deemed serious enough to mention.

Lack of accommodation features only in the low income earner survey and as a minor problem. Housing in Port Moresby i s expensive and in short supply. Doubtless over-crowding occurs and the extended family system is used to off-set the shortage. Respondents may well accept the bad housing and may, therefore, not have complained as much as their living conditions might warrant. Further, the low income sample was taken from individuals generally better housed than many others because their employers often provided housing for male employees and their families. This bias may mask housing as a source of marital problems amongst the informal sector and other income earners.

The main focus of the questionnaire was domestic violence and this will be discussed in detail later in this paper. But the category ‘husband hitting wife’ as a cause of problems in marriage is interestingly low. Domestic violence is not seen as a major problem although women expressed more concern about it than men. The situation is that, in general, domestic violence does not threaten the stability of a marriage. It is accepted as a part of married life and not seen as a problem unless it becomes too severe and too frequent.

A final glance at Figure 1 shows that there seem to be more categories of problems which the urban dweller identifies. Life appears more complex in the urban environment and it may be that as new sources of problems arise the old problems assume less significance, rather than disappear.

The people to whom marriage problems are taken

Having identified causes of marital problems respondents were asked if they ever discuss their own problems with anyone. Table 5 shows that the majority of people do discuss

TABLE 5: IF PEOPLE HAVE MARRIAGE PROBLEMS, DO THEY
DISCUSS THEM WITH ANYONE?

their problems and Figure 2 identifies the kind of people to whom the respondents go for discussion. The urban dwellers have more options than the villagers regarding the categories of people available, but relatives are the ones to whom people turn most, except for the urban elite women. The role of friends becomes more important in the town and particularly the elite women depend on friends in this way. The elites also try to sort out the problems themselves more than the other two groups, though the low income town people also do this to some extent. After relatives and friends, the church plays a significant role, and it is clear that formal services provided by welfare agencies, the ‘Lifeline’ independent counselling service, the police and the courts are not well used, except perhaps in the case of the rural village court. This latter is anyway in effect operated by relatives, friends and wantoks of the parties concerned. Marriage problems are basically seen as private affairs, and not the concern of the state.

Figure 2 : The people to whom marriage problems are taken

Rural

Urban low income

Urban elite

People most likely to help a beaten wife

Figure 3 looks at a different group of people, those most likely to come to the aid of a beaten wife. The answers are in many ways predictable. Again, relatives are the people
most involved in help, in all three survey groups, with a variety of other categories of helpers emerging in the urban context. The extent to which friends are thought to be involved in the urban setting shows a new trend. It is commonly held that people will not interfere in what is seen to be a private matter between a husband and wife, but it seems as if friends and neighbours may in fact be assuming the role that relatives held in the rural situation. The responses to this survey question were based on supposition and opinion, not facts.

Figure 3: People most likely to help a beaten wife

Rural

Urban low income

Urban elite

Measuring domestic violence

A way of trying to gain an impression of the extent to which domestic violence exists was through a coded question asking in the rural questionnaire. ‘How many husbands hit their wives in this village?’ and in the low income and elite questionnaires, ‘How many husbands hit their wives in your neighbourhood or among your friends?’. Table 6 shows the responses.

TABLE 6: HOW MANY HUSBANDS HIT THEIR WIVES?

The highest response rate for no husbands hitting their wives (the category ‘none’) is amongst the urban low income, which is consistent with this group showing the least violence. Amongst the rural group nearly everyone knew of violence, which is hardly unusual given the open communal nature of village life.

Gauging the extent of violence becomes more difficult in distinguishing between the knowledge of ‘some’ and ‘many’ husbands hitting their wives amongst friends and neighbours. If the two categories are combined, the pattern for ‘none’ as reference repeats itself, pointing to less violence amongst the urban low income group. As separate entities, the 'some' and 'many' categories show a slightly different pattern which may reflect fact or a reticence on the part of the elite to nominate many friends as demonstrably violent. Or, the elite home setting, often among strangers in a high-cost housing area, may isolate people to some extent from knowledge of their friends' activities and thus account for the pattern of elite responses.

The category ‘all’ is highest amongst the rural sector and lowest amongst the elite with the urban low income between. This final category must make up a special sub-set of particularly violent friends and neighbours.

A subsequent question given only in the urban and elite surveys asked, ‘Does a next door neighbour at home hit his wife?’ Table 7 shows a noteworthy consistency in replies to this factual question (although the concept of which neighbours are actually ‘next door’ is possibly ambiguous).

TABLE 7: THE NUMBER OF NEXT DOOR NEIGHBOURS WHO HIT WIVES

Approximately two-thirds of the respondents said they live next door to a man who hits his wife. It now seems appropriate to look at the number of respondents who admit to having hit their spouse (Table 8) and those who claim to have been hit (Table 9).

TABLE 8: THE NUMBER OF PEOPLE WHO HAVE HIT THEIR SPOUSE

The figures are very consistent in that Table 8 shows 66% of the rural men say they have hit and in Table 9, 67% of the women say they have been hit. Similarly, 55.5% of the male urban low income earners say they have hit, and 56.1% of the women say they have been hit; and 62.5% of the elite men say they have hit whilst 62.2% of the women have been hit. Readers are reminded that, except to some extent in the elite survey, respondents were not married couples.

TABLE 9: THE NUMBER OF SPOUSES WHO HAVE BEEN HIT

There is less consistency in the matter of wives hitting husbands: 33.4% of the rural women claim to have hit and 29.9% of the men admit to having been hit; 24.3% of the urban low income women say they have hit, but 37.3% of the men claim to have been hit; 49% of the elite women have hit and 50% of the men have been hit.

The figures point to the likelihood that well over half the married male population on a national level hit their wives and roughly half that number of women hit their husbands. We are careful not to say that women retaliate, because there were some female respondents who admitted to having hit husbands who had never hit them (4.5% of the low income urban males said they have been hit but have not themselves hit). The women are inconsistent, however, and it looks as if elite women are asserting themselves unlike the others. This is relevant to the next aspect of this paper which is an examination of attitudes towards domestic violence.

Attitudes compared with practices

Respondents were asked if, in their opinion, it i s acceptable for a man to hit his wife. The answers are recorded in Table 10. Urbanisation and education/income show a negative correlation to thinking that hitting is alright. The majority of urban low income men and even more urban elite men do not think that hitting a wife is acceptable, whereas the majority of rural men sampled do think it is alright. New attitudes are arising in the new environment.

TABLE 10: IS IT ACCEPTABLE FOR A MAN TO HIT HIS WIFE?

Although men's attitudes are mellowing this is not altogether matched by their behaviour: 21.3% of the elite men (but only 8.8% of the low income men) who admit they have hit, had previously stated that hitting is unacceptable behaviour. Women's attitudes are changing and there is a difference between the rural and urban environments in the gap between those married women who are hit and those married women who accept a husband's right to hit (regardless of their own particular case).

TABLE 11: COMPARISON OF ATTITUDES AND PRACTICES:
HUSBANDS HITTING WIVES

*Compare with Table 12.

TABLE 12: COMPARISON OF ATTITUDES AND PRACTICES:
WIVES HITTING HUSBANDS

Compare with Table 11.

In the rural survey there was a 11.9% gap (67% who have been hit minus 55.1% who accept a husband's right to hit) between these two categories of married women, but in the low income survey the gap increased to 30.8% (56.1% minus 25.3%) and in the elite survey it was 25.8% (62.2% minus 36.4%). It is not surprising that in the context of wife-beating women's attitudes are changing faster than those of men. Women (except the very few who suffer from psychosis) do not like being beaten even if they accept a man's right to hit them. It is possible that a new sense of identity already expressed by urban women in their attitude towards male and female obligations in a marriage (see page 15) is causing them to object strongly to beatings. This in itself may influence the men enough to reduce the incidence of violence, although some men's attitudes will also be influenced in this respect by external factors such as education.

With regard to the acceptability of wives hitting husbands (Table 13), attitudes are again inclined to move away from supporting hitting, between the rural neo-traditional environment and the urban environment; but a comparison of attitudes and practices is again informative (Table 12). According to the men, the incidence of violence against them is actually increasing, regardless of attitudes. Is it more likely that the urban men are exaggerating wifely violence or that urban women are not always admitting that they hit? Probably the latter, or there may be some sample bias as couples were not interviewed, except partially among the urban elite, which shows the closest correlation.

TABLE 13: IS IT ACCEPTABLE FOR A WIFE TO HIT HER HUSBAND?

An outstanding feature of the figures is that although only 25.3% of the low income women think it is alright for a husband to hit his wife (Table 11), 33.2% of them think it is alright for a wife to hit her husband (Table 12). The justification is, of course, self-defence. This attitude provides a further example of the move women are making away from their subordination to men.

Reasons for domestic violence

Figures 4 and 5 show attitudes regarding what, in the opinions of respondents, justifies a man or woman hitting his or her spouse. In all three surveys, with the exception of the low income urban women, men and women agree that a wife's failure to fulfil her obligations is the main justification for her husband to hit her. However, the difference between the surveys in the percentage of respondents who say this is important (Table 14).

TABLE 14: JUSTIFICATION FOR WIFE-BEATING

Adaptations to the urban environment are taking place. In addition to being less in favour of violence, the urban dweller has to some extent accepted that a woman's role is different in the urban environment. There is a small group of town people who feel that a wife's job, in cases where she has one outside the home, can interfere with her domestic obligations, and this is seen as a reason to hit. There is a small group of men and women who are prepared to state quite bluntly that a husband has a right to hit his wife. In Figures 4 to 7 the rural category ‘domestic argument’ is hard to assess, but the implication is that if a woman argues with her husband he has the right to hit her. He is expected to assert his control in an argument, whatever the cause. The altercation may have stemmed from her failure to please him in some way, or may have had a number of other causes, but physical dominance resolves the issue.

Sexual jealousy features in all three surveys as a provocation to violence. It is rated below failure to meet obligations but is seen by the surveyed groups as the second main justification for a man to hit his wife. There are two exceptions: one, prominence given by the rural respondents to the category ‘domestic argument’ which, as just mentioned above, clouds the issue and which probably includes arguments triggered by sexual jealousy; the second is the elite urban men who rank self-defence above sexual jealousy as a justification for hitting their wife, probably because, as has already been established, they are under attack more than the two other groups of men.

Figure 4: Opinion on what justifies a husband hitting his wife

Rural

Urban low income

Urban elite

*The category ‘domestic argument’ was used in the rural survey but was replaced by ‘money problems’ in the two urban surveys.

Figure 5 :Opinion on what justifies a wife hitting her husband

Rural

Urban low income

Urban elite

*The category ‘domestic argument’ was used in the rural survey but was replaced by ‘money problems’ in the two urban surveys.

In addition to being seen as major causes of marital problems, money problems and alcohol were also seen as causes of violence. It is noteworthy, however, that a small group of town people, both low income and elite, accept that when a husband is drunk he cannot be held responsible for his behaviour - it is acceptable, or excusable, if he hits when drunk.

From Figure 5, it seems that opinion on what justifies a wife hitting her husband is in many ways similar to when husbands hit wives. A husband’s failure to meet obligations is there but to a lesser degree. Self-defence is the most obvious difference between the two graphs. They indicate that there is some acknowledgement that women have a right to defend themselves. A new element in the urban responses is that women in towns sometimes become drunk – an unusual occurrence in Papua New Guinea. Apart from a small number of elite men, few people said that a wife has an automatic right to hit her husband: there must be a valid reason.

Figure 6 looks at reasons given by respondents for why they have hit. In all three surveys the majority of responses from women name self-defence. The majority of responses from men cite a wife’s failure to meet obligations. Figure 7, which gives an alternative perspective, the victims’ reasons for why he or she was hit, confirms this. Readers are again reminded that, except partially in the elite survey, respondents were not couples. The low income urban men report that they were hit for failing to fulfil obligations more than the women said that this was the reason for hitting. Once more the general category of ‘domestic argument’ can confuse the issue. The most common reason low income urban women gave for being hit was that the husband was drunk. The men however, did not perceive things this way. A similar situation exists amongst the urban elite women who name alcohol as a major problem (though not the greatest one) whereas the elite urban males make no mention of alcohol as a problem.

In other words, it appears that when a husband is drunk he finds fault with his wife (often, either for not fulfilling obligations or for making him jealous) and hits her. She identifies the alcohol as triggering the attack, whereas he blames her behaviour.

Figure 6 : Reasons why respondents hit spouse

Rural

Urban low income

Urban elite

Figure 7 : Reasons why respondents have been hit by spouse

Rural

Urban low income

Urban elite

*The category ‘domestic argument’ was used in the rural and urban low income surveys but was replaced by ‘money problems’ in the elite urban survey.

Some incidents are completely alcohol related: the man is drunk, the wife reacts with non-submissive behaviour, perhaps refusing to prepare him food (i.e. she fails to fulfil an obligation) in the middle of the night or perhaps nagging him for his drunkenness and the money spent on drink, and a fight erupts. This is very different from the rural areas where hitting was generally not associated with alcohol. There appears everywhere to be, however, a scenario of the wife who is hit because she does not somehow come up to the husband's expectations and a proportion of beaten wives who strike out in self-defence.

The frequency of marital violence

Respondents were asked how often violent fights erupt and Figures 8 and 9 show the responses. The main pattern which the figures display is that the majority of couples have violent fights once a year or more, although not as often as every month. The women say with consistency, both as attackers and victims, that there is more violence than the men say there is. The difference is particularly pronounced among the urban elites and this is notable because responses from men and women in this group to other questions have shown general agreement. The men say clearly that violent fights are less frequent than the women perceive them to be. Maybe this is because many of these men, perhaps disapproving of domestic violence in theory, are tending to minimise the problem (21% of them who said it is not acceptable for a man to hit his wife still said they themselves hit).

Another noticeable feature in the data is the higher frequency of violence in the urban low income group among the small number of people who hit often: every month, every fortnight or every week. Although overall there are fewer people who hit in the lower income survey group than in the rural or urban elite groups, it seems that when violence does exist there are more cases in this group where it is a constant problem. That Figures 8 and 9 indicate violent fights occur with less frequency in rural rather than urban areas is not considered significant: rural life has a timelessness which makes it difficult for respondents to be accurate in such estimates.

Objects used in domestic violence and serious injuries sustained

In the rural survey several questions were asked about the use of weapons in fights between spouses (Occasional Paper No. 18 which contains the rural survey results shows that the main justification for rural men to use a weapon against wives is sexual jealousy and the main justification for rural women to hit husbands with a weapon is self-defence).

Figure 8 : The frequency with which people have
hit their spouse

Rural

Urban low income

Urban elite

Figure 9: The frequency with which people
have been hit by their spouse
Rural

Urban low income

Urban elite

In order to keep the urban questionnaire as brief as possible, these were reduced to two questions asking factually which weapons had been used. It was an attempt to see i f town women, like rural women, use weapons more than men, and also to try to assess the extent to which weapons are used.

Figures 10 and 11 tally reasonably well in comparing what the beaters and the beaten have said. It is necessary to look at the precise figures, which appear in Tables 15 and 16, to see the full picture. These tables may be interpreted in one of two ways. Either the victims of violence are over-estimating the use of weapons or the perpetrators of attacks are underestimating the use of weapons. It seems much more likely that the attackers, or hitters, of Table 15 are under-estimating their use of weapons. Note that this behaviour is common to both sexes.

In the case of hitting with hands, the picture changes somewhat. The correspondence is almost perfect between the males of Table l5 who say they have hit with their hand and the females of Table 16 who say they have been hit by a hand. However, reversing all other trends of weapons, more females from all three samples state in Table 15 that they have hit their spouse with the hand than all males in Table 16 say they have been hit. One explanation for this may be that the males consider the incidents too trivial or unmanly to mention. Simply forgetting the incident is also possible and points to an unconcerned attitude.

Weapons used vary between the urban and rural environments. Sticks are much more common items in rural areas and they are used more, whereas stones and belts are obviously handier to urbanites. Knives must be considered the most dangerous weapon of those specifically recorded, with males facing the brunt of knife attacks. Except in the urban elite group, the responses between those attacking with a knife and those being attacked differ by approximately 5%, which again indicates under-estimation on the part of attackers.

Both the rural and urban low income females are attacking with weapons other than hands at about the same rate, if Table 16 is accepted as the true picture. Table l5 shows urban low income females stating a much lower general use of weapons than their male counterparts in Table 16. In Table 16, the urban low income males report more stone and knife attacks against them than the urban elite and rural men.

Figure 10 : Weapons which respondents have used to hit spouses

Rural

Urban low income

Urban elite

Figure 11 : Weapons used to hit respondents

Rural

Urban low income

Urban elite

TABLE 15: WEAPONS USED BY RESPONDENTS
Multiple responses given by people who have hit (See Figure 10)

TABLE 16 : WEAPONS USED TO HIT RESPONDENTS
Multiple responses given by people who have been hit (See Figure 11)

The minority of urban low income women who attack their husbands may be using weapons slightly less than rural women. Certainly, this is their claim in Table 15, but it is not backed up by males in Table 16. Further, a higher percentage of those women who hit, than those men who hit, use weapons. This same pattern is evident amongst the urban elite, although there is a decline in weapon use amongst the elites of both sexes. Women use weapons more often than men for the number of times they hit. In other words, proportionally more women than men use weapons, and, in proportion to the number of times they hit, women use weapons more frequently than men.

This is consistent with the fact that women are often hitting in self-defence and are usually physically weaker than men. It cannot always be assumed that a fight including weapons causes worse injuries than one involving fists. A severe beating by fists alone is enough to hospitalize an individual. In rare cases, it may even prove fatal. However, as a general rule, the authors assume that fists are less harmful than most weapons, and that any rise in the use of particularly dangerous arms such as knives must be viewed with concern. Thus there must be rising concern for the presently small subset of urban low income women who are frequently involved in violent incidents and who are becoming more prone to using dangerous weapons, generally in retaliation for attacks on themselves.

In order to have some idea about the degrees of violence - how serious injuries might be - respondents in the urban environment who said they had been hit were asked if they had ever had to go to the hospital for treatment after being involved in a violent marital dispute. Table 17 displays a consistency between replies from both the low income and elite groups: 29% of the women and 12% of the men in both groups who have been hit have receive a hospital treatment at least once.

TABLE 17: PEOPLE WHO HAVE RECEIVED HOSPITAL TREATMENT
AFTER A MARITAL DISPUTE

Reporting domestic disputes to the police

The same respondents were asked if they had ever reported a marital dispute to the police. This would not necessarily reflect degrees of injuries but could do so: there is a general tendency to regard domestic disputes as private and if the police are called in it implies a serious escalation of the dispute with possible ramifications of a broken marriage and disputes over child custody. The calling in of police removes the action from the private domain of the family to the public domain of the state. Table 18 shows that presently the police are rarely involved in assaults which involve only marriage partners. The question then arises as to what policies the state should pursue with regard to what is criminal assault in a legal sense but which carries a whole set of other socially accepted connotations as outlined in the findings of this paper.

TABLE 18: PEOPLE WHO HAVE REPORTED MARITAL DISPUTES
TO THE POLICE (URBAN ONLY)

Changing attitudes towards domestic violence

One of the most striking characteristics of the survey data is the consistency with which the urban low income women are both against violence and apparently less violent than the rural or urban elite women. The urban low income men follow this trend but to a lesser extent.

One possible reason for this relates to ethnic background. In the rural survey differences between the regions were noted (with caution, because the four regions are administrative not cultural divisions, although cultural similarities often occur between geographically adjacent or related groups). It was found that there was less domestic violence in the Papuan region than in the Highlands or Momase⁴ (Toft and Bonnell, 1985). The low income urban survey was conducted in Port Moresby, which is situated in the Papuan region, and Tables 19 and 20 show how the population sample was heavily weighted with Papuans: whereas in the rural survey only 21.1% of the married respondents were from Papuan provinces, this number was approximately doubled in the urban elite survey and tripled in the urban low income survey.

TABLE 19: PROVINCE OF ORIGIN OF RESPONDENTS

RURAL SURVEY: 21.1% of the married respondents were from Papuan provinces; 19.8% of the whole sample were Papuan.

Note: No response was given by 2 males and 2 females in the low income group, and by 14 males and 19 females in the elite group. It is thought that some of the elite respondents did not state their province in the interests of complete anonymity.

TABLE 20: PAPUAN RESPONDENTS

TABLE 21: RESPONSES TO QUESTIONS ON THE ACCEPTABILITY OF
HITTING CROSS-TABULATED BY REGIONS OF PAPUA NEW GUINEA*

Is it acceptable for a man to hit his wife?

N.B. 17 cases missing.

Is it acceptable for a wife to hit her husband?

N.B. 16 cases missing.
*Because of differences in missing case numbers due to the newly introduced variables, the percentages are not exactly the same as in the original tables but they are certainly comparable.

TABLE 22: RESPONSES TO QUESTIONS ON THE PRACTICE OF
HITTING, CROSS-TABULATED BY REGIONS OF PAPUA NEW GUINEA*

Respondents who have hit their spouse

N.B. 23 missing cases.

Respondents who have been hit by their spouse

N.B. 23 missing cases.
*Because of differences in missing case numbers due to the newly introduced variables, the percentages are not exactly the same as in the original tables but they are certainly comparable.

In order to test the possibility of regional bias, responses from the low income urban survey were grouped according to regional origin of respondents and Tables 21 and 22 show the situation. The averages have been weighted in order to allow comparison between the four regions and the weighting is equivalent to per cent of the total for each region. The Islands region male category has too few cases for any serious consideration. It is only the other 3 regional groups of men and all 4 regional categories of women which can seriously be compared. In this instance Table 22 indicates that the large Papuan element in the sample is not the cause for the lower tolerance and practice of marital violence amongst urban low income earners.

Ethnic background has influenced other aspects of the results, and a number of interesting lines of investigation appear in the Tables:

• There appears to be more actual violence among lower income Highlands couples, but the women condone this violence less than the Papuan and Momase sample groups.
• It is also notable that by region the rates of acceptability for a man to hit his wife are generally much less than for a wife to hit her husband.
• Hitters are underestimating marital violence in all regions. This is demonstrated in the comparison between those respondents who have hit and those who have been hit.
• Finally, women from the Islands, Papua and Momase are involved in approximately the same amount of violence, whereas Highlands men and women experience relatively more.
  

A second possible reason for the lower level of violence in the urban low income survey could be associated with the number of women who work. Money matters were rated highly as causes of problems in urban marriages and it is possible that this could be less so among couples where both partners are employed. Data were extracted to see if those cases where both spouses were working differed from the cases where only one spouse worked. No patterns emerged which are indicative of any important differences. This can be seen in Table 23 which shows the two groups are very similar to the mean answer for both groups and to each other. Thus whether one or both spouses are employed does not seem to be an important variable in any prediction of domestic violence. (It may be worth pursuing cases where only the wife has formal employment, but the above samples did not have enough cases to make any judgement in this regard).

TABLE 23: RESPONSES TO QUESTIONS ON THE ACCEPTABILITY AND
PRACTICE OF HITTING CROSS-TABULATED BY ONE INCOME AND
TWO INCOME FAMILIES*

*Because of differences in missing case numbers due to the newly introduced variables, the percentages are not exactly the same as in the original tables but they are certainly comparable.

A third explanation for the different levels of acceptance and practice of domestic violence is age differentials. The elite women are almost all under 30 years of age and it was posited that their youth could account for the differences between them and the urban low income women.

TABLE 24: COMPARISONS OF THE RESPONSES TO QUESTIONS ON THE
ACCEPTABILITY AND PRACTICE OF VIOLENCE BETWEEN THE URBAN
ELITE AND URBAN LOW INCOME CROSS TABULATED BY AGE*

**NA indicates "not applicable" due to lack of sufficient cases.
*Because of differences in missing case numbers due to the newly introduced variables, the percentages are not exactly the same as in the original tables but they are certainly comparable.

The two urban samples can be compared in terms of age and this is done in Table 24 which sub-divides by sex and by two age groups: (1) 30 years old and under, and (2) over thirty years old. In this table the 30 years old and under group of women in the urban elite sample may be compared with the same group in the urban low income sample for the same four measures as was done for the regional breakdowns in Tables 21 and 22, and for Table 23 showing one and two income families. This exercise clearly shows that the large relatively youthful female component of the urban elite sample has not biased or lead to a difference in the results compared with the relatively youthful group of urban low income women. The latter group still reports a lesser propensity to approve of or accept domestic violence, except in the last category of actually being hit, where the 30 and under urban low income women are receiving marginally more assault than the same age group of elite women. To summarise, it seems that neither regional affiliation, dual family incomes, nor age are the key explanative factors for the variations in domestic violence reported in the samples studied.

A fourth explanation for the lower rate and tolerance of violence among the urban low income women is a marked change in female attitudes and a lesser change in male attitudes. This is suggested by an unexpected factor which emerges in Table 24. What appears is that amongst both the urban low income and elite groups of the 30 and under males, there is a greater acceptance and practice of violence. It is possible that the less submissive stance of the urban women is causing the younger men to attempt to assert their authority in the marital relationship. The younger men feel threatened and are responding violently.

TABLE 25: INDICES OF THE CHANGING ATTITUDES TOWARDS
DOMESTIC VIOLENCE
PEOPLE WHO ANSWERED YES TO THE FOLLOWING QUESTIONS

Table 25 (which converts 'Yes' responses in Tables 8, 9, 10 and I 3 to indices based on 100 as the rural values) clearly demonstrates these changed attitudes. In urban as compared with rural areas, a greater proportion of men are being hit and a lower proportion of women. Also, both sexes in the urban environment demonstrate considerably less acceptance of hitting. These changing attitudes to marital violence must be related to education and experience. This, however, would also account for changes among the elite. It then remains to explain why the urban low income group have a lower incidence of violence than the elite group.

A difference between the two groups is that the elite women and their husbands are under more social and professional pressure than the low income earners. Pressures come from wantoks, who have endless expectations of expenses and expertise; also from the desire or need to maintain a standard of living commensurate with their elite position (including the payment of school fees for all their children); and from t h e need to perform a t work.

The data suggest that among the low income earners, who are, to a large extent, making a success of town life and making adaptations to modernity, domestic violence is less common than among rural people, where old attitudes prevail, and also less than among elite urban people, where particularly the men's behaviour does not conform to the men's own changing attitudes.

CONCLUSION

There is considerable violence within many marriages in Papua New Guinea. The data point to a long rural heritage of domestic violence which is changing in the urban context and possibly lessening as a function of education although the data are not extensive nor clear enough to make any absolute claim regarding trends.

Marriage is basically a civil contract engaged in by the partners (and, according to custom, their extended families) with a complex series of reciprocal rights and obligations. The state has a clear interest in the promotion of harmonious marriage and the lessening of any of its citizens' exposure to violence. However, it faces a most difficult task in the case of marital violence. A strictly legalistic approach cannot solve the problem and runs a risk of worsening the situation.

Marriage is a partnership. The partners depend on each other for psychological, social and economic benefits. In Papua New Guinea, as elsewhere in the world, certain levels of violence are tolerated (or occasionally even expected) in a certain proportion of marriage partnerships. Were the state to attempt to legislate against all marital violence, the result must be to destroy the marriage partnership. When a spouse depends on economic support from the marital partner who is incarcerated for domestic assault the victim of assault must suffer twice: first the assault, secondly the imprisonment of the breadwinner for assault. Needless to say, any children in the family must also suffer.

This is the essential paradox or conundrum faced by the state. Obviously, legislation alone cannot work. Laws for criminal assault are already on the books but the average spouse will not use them to charge the marital partner. Only when the marriage appears to be dissolving and all its various roles are disintegrating may such steps be taken. Otherwise the relationship continues as a partnership, with possible threats of legal action, calls to the police in extreme cases but actual court hearings rarely taking place. The police are often seen in the vague role as mediator because they know, once the dispute is calmed, the partnership will continue.

Whilst doing all it can to facilitate the handling of marital violence through existing channels in the courts and other agencies, the state must turn from dependency on legislation t o other means in any genuine attempt t o reduce domestic violence. Where there is irredeemable violent breakdown in a marriage the state works to protect the victims. This is legislated for already in Papua New Guinea. The effectiveness of the legislation may be questioned because the attitudes of the general population may not clearly support the legislation.

Much domestic violence is the result of accepted attitudes. Any solution to the lessening of domestic violence must attack this core problem. Simply stated, if one's attitude is such that frustrations may be released by hitting a spouse when some perceived fault arises, domestic violence will continue. If the attitude is changed to a belief that it is basically wrong to hit a spouse in the face of frustration then other avenues to vent disappointment and anger will be utilised. It would seem that the state should pursue a course of changing attitudes against the acceptability of domestic violence rather than further punitive legalistic measures which can only weaken the family, a consequence against the interests of the state.

Attitudes and practices do not correlate perfectly in the data but they show a strong trend in the same direction away from marital violence. Attitudes may be held at a variety of levels. Publicly espoused attitudes may not be fully practised in private. Obviously the depth of commitment one has to an attitude will determine how often one acts upon it. Therefore, anyone working to lessen domestic violence faces the difficult task of changing attitudes at very personal and deep levels of practice.

Attitudes amongst some groups do appear to be changing, away from the acceptance of violence. The state should act to further such trends. A positive step would be to sponsor and act upon in-depth studies of families in which violence occurs and in those where it does not. The aim would be first to learn how some couples avoid violence and why others do not. Then work could be undertaken t o promote the type of behaviour and conflict resolution which the non-violent families are using. Work may also be done to diminish the use of weapons, kicking and punching with closed fists, on an assumption that some violence will always be present in some marriages but that its most physically damaging effects may be lessened.

Finally, special attention should be paid to the small group of urbanites whose families are affected by violence which repeats itself at short intervals and shows a high incidence of the use of dangerous weapons. It will be critical to future planning to know if this is a growing trend.

Notes
1 Every household was approached but only those with married women present were surveyed.
2 The National Statistical Office population figure for national citizens in Port Moresby in September 1980 is 112,429, and a compound 4% annual growth rate has been added to reach a figure for September 1984.
3 See definition on page 2.
4 Momase is the current abreviation for the New Guinea mainland coastal
provinces comprising Morobe, Madang, East Sepik and West Sepik.

References
COMMITTEE of the Royal Anthropological Institute of Great Britain and Northern
Ireland, 1974 (Reprinted, Sixth Edition), Notes and Queries on Anthropology.
London: Routledge, Kegan and Paul.

NATIONAL STATISTICAL OFFICE, January 1982. Papua New Guinea 1980 National Population Census Pre-Release: Summary of Final Figures. Port Moresby: Government Printer.

TOFT, S and BONNELL, S. 1985 Marriage and Domestic Violence in Rural Papua
New Guinea, Occasional Paper No.18, Law Reform Commission of Papua New Guinea, Port Moresby.

WARUS,J. n.d Ethnographic Data on Marriage in Uma, Kagua. (Highlands Region).
In Toft, S. (ed) Marriage in Papua New Guinea. Monograph No.&, Law Reform Commission of Papua New Guinea, Port Moresby.

CHAPTER TWO
MARRIAGE AND DOMESTIC VIOLENCE
IN TWO URBAN SETTLEMENTS

by Rosa Au

INTRODUCTION

This paper presents one of a series of surveys of domestic violence conducted by the Law Reform Commission of Papua New Guinea. Its purpose is to provide a link between a survey of culturally homogenous rural communities and surveys of culturally heterogeneous urban communities. The aims were first, to try to discover to what extent domestic violence exists; secondly, to examine attitudes towards marriage and domestic violence, including support systems and ways of coping with the problem; and thirdly, to try to discover the causes of domestic violence.

The population of Port Moresby, the capital city of Papua New Guinea, was 122,761, at the time of the 1980 National Census, of whom 75% lived in squatter settlements and 10% in urban villages². Urban villagers are people who, generally speaking, live on traditionally owned land around which the city of Port Moresby has expanded. Squatters are migrants from elsewhere who settle on vacant land they do not own. Some squatter settlements are ethnically mixed, but some contain people who come from the same home area in the country and from an ethnically homogenous group within the city. The latter type was the focus for this study on marriage and domestic violence, so that findings from homogenous rural communities could be compared with small-scale ethnically homogenous groups who had transplanted themselves into the urban environment.

Choosing the Sample

Two squatter settlements were chosen:

Ragamuka: near Six-Mile Dump, which is the home of about 700 people from Gumine district of Chimbu Province in the Highlands; a specific section of the settlement with a population of 363 was used for the survey.

Vada Vada: in the East Boroko area of Port Moresby, where two groups of Goilala from Central Province, numbering 333 at the time of survey, live. (The official Vada Vada settlement area includes a separate group of Mekeo people, Central Province neighbours of the Goilala, but they were not included in the survey).

In selecting the sample and location for the study, the following criteria were used:

l. Ethnic Homogeneity: a main objective of this study was to make comparisons with results from the rural study which was also based on homogeneous groups.

2. Workable Size: because of time constraints it was decided that a settlement population of 300-500 would be appropriate, and a minimum of 10% of the populations of each of the two settlements (Ragamuka 763 and Vada Vada 333) was the target number for interview. In my study I interviewed 25 males and 25 females from each of these two settlements, which is a 14% and 15% sample size respectively.

3. Previously Established Contacts: also due to time constraints, and because of the sensitive nature of the study, it was considered important to operate through renewed personal contacts in settlements I had previously worked in.

Methods of collecting data

I attended orientation classes with the Diploma in Social Development students at the Administrative College during November 1982. These classes were on methods of social research in preparation for the rural study and included the preparation of the questionnaire and other methods of data collection.

An interview questionnaire for settlement respondents was adapted from the individual questionnaire for villages designed for the rural survey and a structured interview schedule for persons knowledgeable about the areas for study. (See Appendices 2, 7 and 8).

CHARACTERISTICS OF THE RESPONDENTS

Personal Details

Ragamuka: Of the population of 363 in Ragamuka settlement in December 1982 50 people were interviewed, 25 men and 25 women. Out of the 50, 16 males and 23 females were married (78%). Of the 8 single male respondents, 6 were young men between the ages of 17 and 19 years who were unemployed and living with relatives. There were l8 mature males, of whom 2 were single and 16 married, 24 mature females, of whom 2 were single, and there was one old married female.

Vada Vada Settlement: Out of a population of 333, 50 respondents (25 men and 25 women) were interviewed. Out of the 50 people, 17 males and 20 females were, married (74%). 2 male and 2 female respondents were widowed (8%). Altogether 82% of the interviewees in this settlement were seen as mature or old.

TABLE I: PERSONAL DETAILS OF SAMPLE

From both settlements then, 19% were single and 81% were either married, widowed or divorced; only 15% of the sample were classed as young and, therefore, less experienced in matters relating to the questionnaire than older people.

In both settlements, we were able at times to interview both husband and wife, that is couples, but the majority were respondents whose spouses (if married) were not available for interview.

Employment Details (Table 2)

Ragamuka: It was stated that 16 members of respondents' households, (whether the respondents themselves or otherwise), worked for the government and 24 for private companies. 16 were self-employed on a full-time basis and 41 worked casually: 18 in gardening/selling garden produce, 6 in firewood collecting/selling, 15 in collecting bottles/selling, 2 in other casual activities (e.g. selling betel nut) and 5 in other forms of income-earning capacities. From talking to respondents, it appeared that the black-marketing of beer and prostitution were known to be taking place. However, no respondent acknowledged these forms of activities in the households they were representing. Only 8 respondents stated that both they and members of their household were completely unemployed.

Vada Vada: Responses showed that 9 members of respondents' households worked for the government and 14 worked for private companies; 2 were fully self-employed, 34 were in casual employment: 15 in gardening/selling produce. 12 in firewood collecting/selling, 6 in bottle collecting/selling, one in other casual employ and 2 in other forms of income earning employment. A total of 12 people said they were unemployed in one form or another. Like the Ragamuka group, individual respondents stated black-marketing of beer and prostitution took place, but no respondent acknowledged this for his/her household.

TABLE 2: OCCUPATIONS OF MEMBERS OF RESPONDENTS'
HOUSEHOLDS

In both Ragamuka and Vada Vada Settlements unemployment was said to be very high. However, I was unable to obtain exact unemployment figures for these two settlements. Because of the way the question was phrased, 'Are you, or is anyone else living in this house, employed?', in an attempt to discover how many people had access to regular financial support, it is certain that the same employee was cited by more than one respondent. It is also possible that individuals had more than one kind of employment. For this reason the figures do not represent actual numbers of people employed. But it can be said that, of the respondents, 63% of them (80% from Ragamuka and 50% from Vada Vada) lived in households where there was one wage earner or more, working either for the government or private employers, and 20% of them (8% from Ragamuka and 12% from Vada Vada came from households without any specific form of income.

MARRIAGE PROBLEMS

There were a number of stated causes of problems in marriage. Sexual jealousy, a category which includes jealousy, suspicions about adultery and husbands taking new wives, was said to be the highest cause of problems between married 3 couples; 68% of all respondents thought this (see Table 3). Next on the list came what people called a 'lazy wife'. Some wives probably are lazy, but it really means that for some reason a wife does not perform duties or fulfil the obligations expected by her husband, and 59% of the respondents named this as a cause of problems. It should perhaps also be looked at in connection with the category 'lack of respect' because it was men (48%) more than women (22%) who thought this a problem. Maybe when wives fail in their duties some men feel that this shows a lack of respect.

Third on the list, between the categories 'wife's failure to meet obligations' and 'lack of respect', comes 'alcohol'. More women said this is a problem than men, but altogether it was named by 50% of the respondents. Alcohol was seen as more of a problem than gambling, which was placed next by 32% of the respondents, and lack of money, named by 30%, but the three categories must all be linked because money for food is sometimes wasted on alcohol or gambling. In general, men are the drinkers and women the gamblers. Domestic violence was only named as a problem by 26% of the respondents but this, as will be shown later, does not mean that only 26% of the respondents are involved in domestic violence; it is just not seen as a problem in all cases.

Problems over bride-price, children and relatives are at the bottom of the list, and it is interesting to see how low a position is given to relatives in these ethnically homogenous settlements. Perhaps people just accept demands made by relatives, or perhaps it is these low income level people who cause the problems for wealthier relatives elsewhere in the urban setting.

TABLE 3: PROBLEMS IN MARRIAGE

Note: Respondents were asked to name three causes if possible.

People were asked whether they discussed their marital problems with anybody. Table 4 shows the answers to this question. Over twice as many people said they do discuss these problems as people who said they do not.

TABLE 4: DO PEOPLE DISCUSS THEIR MARRIAGE PROBLEMS WITH ANYONE?

Table 5 shows that, when asked who they turned to for advice or discussion, the majority of those who responded stated that help was sought from relatives: mostly their own relatives and not their in-laws. Only 17% of those who responded stated that they sought help from outside their family systems, mostly from village court officials and settlement elders. Only one person claimed to have sought advice from a welfare officer. Failure to seek welfare assistance could be due either to a lack of welfare services, or, if available, the services not being within easy reach. Generally then, when faced with marriage problems, respondents from these two settlements sought advice or assistance from within their own family.

TABLE 5: THE PEOPLE RESPONDENTS TURN TO FOR ADVICE OR
ASSISTANCE WHEN FACED WITH MARITAL PROBLEMS

Note: More than one response was sometimes given.

Sexual jealousy featured as the supposed biggest source of marriage problems (Table 3): Questions relating specifically to adultery were asked and Tables 6 and 7 show attitudes towards extra-marital sex. When asked whether it was permissible for a married man to have sexual relations with a single woman, 65% of the respondents (60% of males/70% of females) stated 'No' whilst 35% (40% of males/30% of females) said 'Yes'. The majority of the respondents, 82% (80% of males/84% of females) stated that a relationship between a married man and married woman is wrong. With regard to a married man having sex with a prostitute, only 64% of the respondents said 'No' (62% of men/66% of women).

Men and women seem to share broadly similar views on adultery. The most notable feature of the figures is that men and women in Ragamuka showed less concern than the Vada Vada people about a man having sex with a single woman or prostitute. Their stronger attitude regarding his having sex with a married woman may reflect attitudes towards women who commit adultery rather than men: Table 7 shows very strong attitudes in both settlements against women committing adultery.

TABLE 6: IS IT ACCEPTABLE FOR A MARRIED MAN TO COMMIT ADULTERY?

TABLE 7: IS IT ACCEPTABLE FOR A MARRIED WOMAN TO COMMIT ADULTERY?

Divorce

Various factors were given as having contributed to divorce in these two settlements and the pattern is not very different from the reasons given for marital problems. More people thought sexual jealousy contributes above anything else towards divorce - it was mentioned by 58% of the respondents. A wife's failure to meet obligations is thought to be the second most important causal factor, named by 37% of the respondents. Alcohol was third but i t is interesting that only 26% of the respondents felt it contributed towards divorce, whilst 50% felt it caused problems in marriage, though to some extent this will be due to the fact that fewer responses were given to the divorce question: people were asked to name at least three causes of problems, whereas they need only have given one cause of divorce and on average gave only two. Gambling seems to be more of a problem in Ragamuka than Vada Vada and domestic violence featured slightly more strongly in that settlement, as both Tables 3 and 8 show.

TABLE 8: CAUSES OF DIVORCE (OPINION)

Note: More than one response was sometimes given.

Table 9 shows opinions were divided as to whether divorce is more frequent now than it was in the past. Of those people who stated an opinion that divorce is increasing, the highest reason, given by 20% of the male respondents, was Westernisation and breakdown of traditional practices. Female respondents from both settlements disagreed with the men. This was because the male respondents generally felt that women were getting big-headed and not performing duties as required by custom and tradition. Of course, most female respondents did not regard this as a problem. Interestingly, although Westernisation polled highest when respondents were asked why there was an increase in divorce today, when asked for reasons for divorce, Westernisation and breakdown of tradition polled very low and came in the 'other' category.

TABLE 9: IS THERE MORE DIVORCE NOW THAN IN THE PAST?

Of the 29% of the respondents who stated they have been divorced at one time or another (Table 10), the majority were male; the highest rate was amongst the Ragamuka men, of whom 44% have been divorced. If single respondents are omitted from the calculations, the overall number of divorcees increases to 36% a very high divorce rate. This situation is possibly being created by the nature of the relationship the settlement has with the traditional home area: defacto marriages involving people who are migrating to and from the rural area may be short-lived. Of divorced male respondents, the largest number gave the wife not fulfilling her obligations as their reason for divorce (59%) (Table 11), while for divorced female respondents, the most frequently given reason for divorcing was sexual jealousy (see definition in Chapter One). Table 11 gives a break up of those respondents who had been divorced at one time or another.

TABLE 10: RESPONDENTS WHO HAVE BEEN DIVORCED

TABLE 11: REASONS FOR HAVING DIVORCED

Note: Percentages are based on the numbers of divorced persons of each sex (17 males and 12 females). More than one response was sometimes given.

DOMESTIC VIOLENCE

Generally, it was stated that men are entitled to hit their wives, as indicated by 80% of the respondents: 82% of the male respondents and 78% of the female respondents (Table 12A).

TABLE 12: IS IT ACCEPTABLE FOR A HUSBAND TO HIT HIS WIFE?

A: Husband permitted to hit wife (with hands)