TABLE 3.PERCEIVED CAUSES OF BREAKDOWN 16,323,p <.000).Friction with relatives (p AND BREAKDOWN'S TIME OF ONSET .03),housing problems (p =.000),husband's Time of Onset drinking (p =.016),and husband's cruelty (p (Years of Marriage) .004)all were most commonly reported by those Perceived Causes 0-12-5 6-10 11+ who married when younger than 21 years.Women Sexual who married in their 30s or above also were rather 1.53 1.56 1.47 1.81 Other woman 1.77 170 1.63 155 likely to nominate the husband's drinking,an out- Children 1.82 1.68 1.82 1.93 come that is partly due to the fact that these later Husband's drinking 1.61 1.86 1.83 1.69 marriages include a number of remarriages (see Husband's cruelty 1.61 1.84 1.92 1.93 below).Financial problems are most common Note:Only significant causes are included.Causes are among the later (over 30s)and earlier (under 21) scored dichotomously (yes 1,n0=2). marriages.The interaction between age at mar- riage and sex of respondent is not significant. contribute to this effect.Where the spouse has no religious affiliation,women are likely to nominate Marital Status sexual incompatibility as a cause,whereas men There was no overall difference between di- with a religous spouse (of any kind)are more like- vorced and separated respondents,but two of the ly to make this attribution.(This parallels the in- univariate analyses show a trend towards signifi- teraction reported above between respondent's cance.The separated were more likely to nomi- own religious status and complaints of sexual in- nate sexual incompatibility as a cause (p =.025), compatibility and partly reflects a fairly high level the divorced to nominate the husband's affair of religious homogamy in the sample.)A similar with another woman (p =.041).This latter asso- pattern occurs for complaints about an "other ciation,though not particularly strong,seems man."Women with nonreligious husbands were logical,since it would be expected that in a certain more likely to refer to their own affairs with other number of such cases the man would wish to sub- men,whereas husbands of religious wives (Roman sequently marry the other woman involved.There Catholic or Protestant)were more likely to see was no interaction with sex.When respondents in their wives'infidelity as a cause of the break- their first and later marriages were compared,no down.The reverse picture emerges for the other multivariate difference emerged,but two univari- two significant causes,friction with relatives and ate analyses were significant.Complaints of the husband's cruelty.Here it is the wives of religious, husband's drinking (p =.039)and gambling and in particular of Roman Catholic,husbands (p =.009)were much more common among sec- who most commonly report both these com- ond marriages. plaints,whereas it is the men with nonreligious wives who do so.Since only four men mentioned Parental Approval of the Marriage their own cruelty as a cause (none of them Roman The attitudes that respondents perceived their Catholics),not too much weight can be placed on mothers and fathers to hold towards the marriage the cruelty outcome;but the data with regard to were very similar.Some 27%of both mothers and problems with relatives are more interesting.Only fathers disapproved or strongly disapproved,25 2 of the 32 women with nonreligious husbands- were neutral,and 40%%approved or strongly ap- constituting 6.25 of this group-made this com- proved.A woman's choice,however,was much plaint,in contrast to 23.9%of the Protestant more likely to be disapproved by both her mother wives and 26.8%of the Roman Catholic wives. (32%compared,with 16%0 for men)and by her Husbands of nonreligious wives,however,report father (320%compared with 14%for men).The much friction(42.9%),as do husbands of Protes- association of parental attitude with perceived tant wives (31.4%),but not those of Catholic causes was significant (F 1.32,df 16,193, wives (11.8%).If the inference is made that"fric- p=.048),with three causes making the major tion with relatives"usually means the spouse's contribution.Husband's cruelty (p =.006)and relatives,it appears that nonreligious husbands housing problems (p =.014)were both associated either have very agreeable relatives or else have lit- with greater disapproval,while lack of communi- tle or no contact with their families of origin, cation (p =.009)was associated both with strong whereas the families of nonreligious wives seem disapproval and with parental neutrality (a cate- particularly likely to become involved. gory that included lack of knowledge of the Age at Marriage parental attitudes).Subsequent friction with rela- tives,a complaint that might be expected to be Age at marriage was also a highly significant associated with parental disapproval,showed a predictor of perceived causes (F 1.76,df nonsignificant trend in this direction (p =.061). August 1984 JOURNAL OF MARRIAGE AND THE FAMILY 557 This content downloaded from 211.80.94.134 on Mon,19 Dec 2016 05:27:55 UTC All use subject to http://about.istor.org/termsTABLE 3. PERCEIVED CAUSES OF BREAKDOWN AND BREAKDOWN'S TIME OF ONSET Time of Onset (Years of Marriage) Perceived Causes 0-1 2-5 6-10 11 + Sexual 1.53 1.56 1.47 1.81 Other woman 1.77 1.70 1.63 1.55 Children 1.82 1.68 1.82 1.93 Husband's drinking 1.61 1.86 1.83 1.69 Husband's cruelty 1.61 1.84 1.92 1.93 Note: Only significant causes are included. Causes are scored dichotomously (yes = 1, no = 2). contribute to this effect. Where the spouse has no religious affiliation, women are likely to nominate sexual incompatibility as a cause, whereas men with a religous spouse (of any kind) are more like- ly to make this attribution. (This parallels the in- teraction reported above between respondent's own religious status and complaints of sexual in- compatibility and partly reflects a fairly high level of religious homogamy in the sample.) A similar pattern occurs for complaints about an "other man." Women with nonreligious husbands were more likely to refer to their own affairs with other men, whereas husbands of religious wives (Roman Catholic or Protestant) were more likely to see their wives' infidelity as a cause of the break- down. The reverse picture emerges for the other two significant causes, friction with relatives and husband's cruelty. Here it is the wives of religious, and in particular of Roman Catholic, husbands who most commonly report both these com- plaints, whereas it is the men with nonreligious wives who do so. Since only four men mentioned their own cruelty as a cause (none of them Roman Catholics), not too much weight can be placed on the cruelty outcome; but the data with regard to problems with relatives are more interesting. Only 2 of the 32 women with nonreligious husbands- constituting 6.25%7o of this group-made this com- plaint, in contrast to 23.9% of the Protestant wives and 26.8% of the Roman Catholic wives. Husbands of nonreligious wives, however, report much friction (42.9%), as do husbands of Protes- tant wives (31.4%), but not those of Catholic wives (11.807). If the inference is made that "fric- tion with relatives" usually means the spouse's relatives, it appears that nonreligious husbands either have very agreeable relatives or else have lit- tle or no contact with their families of origin, whereas the families of nonreligious wives seem particularly likely to become involved. Age at Marriage Age at marriage was also a highly significant predictor of perceived causes (F = 1.76, df = 16,323, p < .000). Friction with relatives (p = .03), housing problems (p = .000), husband's drinking (p = .016), and husband's cruelty (p = .004) all were most commonly reported by those who married when younger than 21 years. Women who married in their 30s or above also were rather likely to nominate the husband's drinking, an out- come that is partly due to the fact that these later marriages include a number of remarriages (see below). Financial problems are most common among the later (over 30s) and earlier (under 21) marriages. The interaction between age at mar- riage and sex of respondent is not significant. Marital Status There was no overall difference between di- vorced and separated respondents, but two of the univariate analyses show a trend towards signifi- cance. The separated were more likely to nomi- nate sexual incompatibility as a cause (p = .025), the divorced to nominate the husband's affair with another woman (p = .041). This latter asso- ciation, though not particularly strong, seems logical, since it would be expected that in a certain number of such cases the man would wish to sub- sequently marry the other woman involved. There was no interaction with sex. When respondents in their first and later marriages were compared, no multivariate difference emerged, but two univari- ate analyses were significant. Complaints of the husband's drinking (p = .039) and gambling (p = .009) were much more common among sec- ond marriages. Parental Approval of the Marriage The attitudes that respondents perceived their mothers and fathers to hold towards the marriage were very similar. Some 27% of both mothers and fathers disapproved or strongly disapproved, 25% were neutral, and 40% approved or strongly ap- proved. A woman's choice, however, was much more likely to be disapproved by both her mother (32% compared, with 16% for men) and by her father (32% compared with 14%7o for men). The association of parental attitude with perceived causes was significant (F = 1.32, df = 16,193, p = .048), with three causes making the major contribution. Husband's cruelty (p = .006) and housing problems (p = .014) were both associated with greater disapproval, while lack of communi- cation (p = .009) was associated both with strong disapproval and with parental neutrality (a cate- gory that included lack of knowledge of the parental attitudes). Subsequent friction with rela- tives, a complaint that might be expected to be associated with parental disapproval, showed a nonsignificant trend in this direction (p = .061). August 1984 JOURNAL OF MARRIAGE AND THE FAMILY 557 This content downloaded from 211.80.94.134 on Mon, 19 Dec 2016 05:27:55 UTC All use subject to http://about.jstor.org/terms