CAIN Project: Issues: Marriage - Robinson, G., Cross-Community Marriage in Northern Ireland

CAIN Web Service

Cross-Community Marriage in Northern Ireland - Section 2

[Key_Events] Key_Issues] [Conflict_Background]

Text: Gillian Robinson ... Page Design: Fionnuala McKenna
This page is currently under development


2.1 Methodology

There are difficulties involved in studying a potentially vulnerable group within a conflict situation. Those who are being studied may perceive risks in being researched. This may not be entirely without justification. Since they are capable of being defined as deviant, perhaps of being judged as having betrayed their community by marrying beyond its borders, couples may feel themselves to be potentially vulnerable to intimidation or worse.

The advisory group agreed that an in-depth interview would be the preferable method of obtaining the information necessary to create a 'biography' as was the research objective. As we have seen the possible areas of interest in looking at cross-community marriage are many: however, it was decided to limit this study to look at the social supports and social constraints on couples. A sample of up to 100 cases was targeted.

As no list exists of people in a mixed marriage it was considered impractical and beyond the resources of this study to attempt a random sample. At the initial talks one of the possible problems raised with doing this type of research was the difficulties in obtaining a sample of any size, as previous researchers (for example, Lee, 1981) had experienced difficulties. However the Relate and CMAC contacts felt that they had a wide enough connection of counsellors Province-wide to enable sampling using a snowballing technique.

So the research procedure was used whereby the researcher made an initial contact with those who had attended the initial meetings. They then made further contacts through their organisations and friends, colleagues and so on. Following this, an initial contact was made to a couple by an intermediary known to, and trusted by, the couple. If the couple were agreeable, the intermediary passed on their names and address to the researcher who then made further contact with the couple giving them more information about the project. At this time, further details on anonymity and confidentiality were given to the couple. If one or other partner (or both) was then agreeable to take part, an interview appointment was set up.

Because of the range of people working and connected to these organisations it was felt that a wide variety of couples could be found from all sections of life in Northern Ireland. It seemed important to ensure that couples were found at all stages in the family cycle as this would enable us to see if there were any changes over time in the experience of those in a cross-community marriage. Likewise, it seemed important to see people in both rural and urban areas as it appeared from previous research that the situation might differ between such areas.

While there was some evidence (Lee, 1981) to imply that Roman Catholic women were more likely to marry Protestant men, than Protestant women were to marry Catholic men it was felt important that not only should we attempt to interview equal numbers of Catholics and Protestants, but also equal cases where the man was Catholic and the man was Protestant. This is of particular relevance, as it is traditional in Ireland for a couple to marry in the bride's church.

However, in practice, we interviewed all couples who were identified to us and who agreed to take part in the project within the time constraints.

The interview schedule (see Appendix 1) was developed in consultation with the advisory group, and covered in chronological order, the person's own family background; schooling; area of residence; importance of religion in the family; previous experience of mixed marriage; work and leisure pursuits; and likewise for their partner. It then went on to examine how the couple met; development of the relationship; reaction of family; friends; church and community. From there it moved to the decision to marry and associated decisions such as where to marry and how to rear any children they might have. It looked at the reaction of family and friends at this time as well as the support or otherwise offered by the Churches in arranging their wedding. Moving on from the wedding their married life was examined, and in particular the baptism and schooling of any children. Finally, people were asked if they thought being in a mixed marriage had had any effect on their relationship, and also about any thoughts they had on how support to such families might be improved if they saw that as being necessary.

The interviews were to be conducted at the convenience of the respondents, either in their home or another venue. It was hoped to tape record the interviews with the respondent's permission. Obviously in the course of the interview, respondents might mention names of family members, clergy or others; likewise they might mention workplace or other factors which might identify them. To ensure confidentiality, after each interview the researcher changed all names and in some cases, for example, rural areas changed, or deleted place names. The interviews were then transcribed by experienced typists. The researcher checked the transcripts for accuracy and confidentiality and the final drafts were printed.

Analysis was then carried out looking at the biographies in chronological order.

The final transcripts will be lodged with Relate who may make them available, at their discretion to bona-fide researchers.

It was decided to avoid any publicity during the course of this study to further ensure confidentiality and to obtain the trust of respondents.

2.2 Interviewee recruitment and co-operation rate

From the names that were given to the researcher only one person refused to co-operate when contacted. She said that her husband was not happy for her to take part.

But how difficult it was for those trying to get couples to agree to have their name referred on to the researcher in the first instance is not known. However one minister did write

'Due to heightened local sensitivities surrounding this issue I would prefer not to become involved in your study at this time' (Presbyterian minister in Co. Down November 1991)
Telephone calls from other intermediaries often went along the lines
'I have one name for you now and I may have another, I think they will agree to take part'
Figure 1 outlines the points of contact


The time during which the interviews were carried out was one of the most violent periods in the recent history of Northern Ireland with several major atrocities, for example, Teebane where Protestant workmen were murdered as they returned from work and the Ormeau Road betting shop murders where Catholics were murdered by Protestant paramilitaries. In fact one Belfast woman and her son were murdered in November 1991 allegedly because she had converted to Catholicism when she had married twenty years earlier.

A total of 79 interviews were completed, all except one between 1st October 1991 and 31st January 1992. A further possible 23 interviews might have been completed had we been able to extend the interviewing time.

2.3 The characteristics of the couples represented

The 79 interviews represented some 96 individuals or 64 families as Table 1 shows

Table 1 Details on interviews completed

Total number of interviews completed79
… Both partners interviewed separately 15
… Joint interview17
Total number of individuals represented 96
Total number of families represented64

The interviews were conducted at a time convenient for the respondents and this meant that in a majority of cases, the interviews were carried out in the evening when people had returned from work and had children in bed. Some interviews were conducted at weekends. In the majority of cases the interviews were conducted in the respondent's home, however, in three cases, the respondent preferred to be interviewed in The Queen's University of Belfast; in one case, the interview was conducted in a restaurant and in two cases the interview was conducted at the respondents place of work.

Every effort was made to conduct the interviews with each partner individually, however in 17 cases the couple preferred to be interviewed together. These interviews while worthwhile were more difficult for the researcher and in some cases it is possible that the same depth of information was not achieved as in the individual interviews. This opinion is based on the fact that in the individual interviews several times a respondent would say something on the lines of 'I don't know if this is how John would see it' or 'I don't know if Sheila really realised how difficult things were at home at that stage, I would certainly have tried to hide it from her a little'. Many times the couple would query the benefits in both partners being interviewed separately. In the 15 cases where this did happen the transcripts provide hard evidence of the benefits of hearing both sides of the story.

Interviews ranged in length from 45 minutes to 120 minutes, however this is solely the time that was taped. In many cases the researcher was with the respondent(s) for much longer than the actual interview time. In one case arriving to a home at 7:30pm and not leaving until 11:00pm.

At the outset it was hoped to have approximately equal numbers of participants from both the Roman Catholic and Protestant communities and in fact this hope was fulfilled. As it happened couples were taken as referred and there was no attempt made to ensure equal numbers. Table 2 shows the breakdown by religion and sex.

Table 2 Breakdown of respondents by religion and sex

Roman Catholic

Religion is taken as religion before marriage. In point of fact very few converts were encountered in the course of this project and in almost all of the cases where conversion did occur the partner who converted did not practise. One woman whose husband had converted said

'...but to me I would still maintain Samuel is Protestant to this day...' (Roman Catholic woman who married almost 20 years ago).
Another woman had made the decision to convert to Catholicism and had been taking instruction before meeting her husband. She said that she would not really consider herself a 'mixed marriage' but for the purposes of this study the fact that she had been reared in a Protestant home until her late teens would include her.

The Protestants represented in the study came from the following denominations: Presbyterian, Church of Ireland, Quaker, Methodist, while some attended small house-groups or gospel halls.

It will be of no surprise that more women took part in the study than men, however it was considered very satisfactory that over a third of the people represented in the study were men.

Respondents were not asked directly about their work or income however as most interviews took place in the respondents home it was possible to estimate their social class. The majority of respondents would have been middle class, but there were some couples from a more working class background. Respondents were much more likely to have had third level education than the population in general.

Moving on to look at family characteristics. In 32 of the families the man was the Catholic partner and in other 32 families the man was Protestant. Of the 64 families some 48 had children while 16 did not. This is a higher representation of childless families than in the population generally (approximately 10%). This can be partly explained when the length of time married is examined as Table 3 shows. (Two couples were expecting their first child)

Table 3 Length of time married

0-5 years
6-10 years
Over 10 years
Divorced/ Separated
Not yet married

Nineteen couples had been married for less than 6 years and in fact many of these were married only one or two years. In Section 3.5 of the report the couples who are deliberately childless will be examined in more detail. Respondents were not asked directly about their age although in some cases this information has been offered in the course of the interview. From length of time married we can infer that the respondents cover a broad age span. The youngest person taking part was in her early twenties, while the oldest was a woman in her sixties.

It is likely that the sample is over-representative of couples in a healthy relationship given the nature of the sample selection whereby couples who were experiencing difficulties may not have been approached by intermediaries or may not have agreed to take part. The people who were divorced or separated (4 in total) while not big in number do shed some light on the possible effects of being in a mixed marriage on the breakdown of that marriage and their opinions on this will be examined in some detail later in Section 3.7. The fact that our respondents are fairly evenly spread over the length of time married allows for some examination of changes over time, for example in the role of the Churches. Obviously those most recently married will have clearer memories of the problems or lack of them encountered in actually trying to get married. While the years may have dulled these memories for people married for longer time periods they will have experience of a different kind to impart for example the effects if any on their children of coming from a mixed marriage.

For the 48 families with children the church of baptism is outlined in Table 4.

Table 4 Religious baptism of the children

No religion (not baptised)
Roman Catholic
Undecided (baby)
Joint / Alternate
* 1 family had first two children baptised Roman Catholic and the remainder were not baptised

In over half the families with children they had been baptised in a Roman Catholic service. Baptism and religious upbringing will be examined in detail in Section 3.5.

However there would appear to be a slight tendency for the children to be more often brought up in the religion of the mother than of the father as Table 5 demonstrates. This is particularly true when the mother is Roman Catholic.

Table 5 Religious baptism of the children by mother's religion

Mother Protestant
Mother Roman Catholic
Roman Catholic
Not baptised

Another characteristic of respondents which was considered important in the initial discussions about the study was area of residence. We have already seen that more mixed marriages are conducted in the Down and Connor diocese. It was felt that there might be differences between city and country. The numbers in this study are too small to be broken down into more than two groups for confidentiality reasons. Therefore we have some 35 families residing in the Belfast area with the remaining 29 families being outside that area, some in rural areas with others in small towns. Interviews were conducted in Ballynahinch, Londonderry, Comber, Newry, Omagh, Coleraine, Dungannon, Enniskillen, Ballymena, Ballymoney, Irvinestown, their hinterland and some other small rural townlands throughout Northern Ireland as Map 1 illustrates

Return to Publication Contents

Picture of Help Icon Picture of Help Icon Picture of Help Icon

If you have any questions or comments regarding the CAIN service please send an email message. Or you can contact the CAIN Project Manager directly at the following email address:

© 1996-1998 CAIN Project

Last Modified by Dr Martin Melaugh :

Back to the top of this page