Copyright Sociological Research Online, 1998

 

Pilcher, J. (1998) 'Gender Matters? Three Cohorts of Women Talking about Role Reversal'
Sociological Research Online, vol. 3, no. 1, <http://www.socresonline.org.uk/3/1/10.html>

To cite articles published in Sociological Research Online, please reference the above information and include paragraph numbers if necessary

Received: 20/10/97      Accepted: 19/2/98      Published: 31/3/98

Abstract

Cohort is an important predictor of gender-role attitudes, as a number of surveys have shown. In this article, I undertake a comparison between cohorts of women on the issue of role reversal, with a primary focus on the qualitative differences in what was said and by whom, rather than in how many said what. It is my argument that a qualitative analysis is revealing of the way in which cohort acts to influence the very language used to report 'agreement' or 'disagreement' on matters of gender. Via an analysis of responses to an interview question on role reversal, it is shown that historical location via cohort operates to make permissible and/or available, some ways of talking rather than others. Consequently, on the issue of role reversal, gender featured as a more relevant category in the talk of the oldest cohort than in the talk of the younger cohorts.

Keywords:
Cohort, Gender Issues, Role Reversal, Qualitative Research, Vocabularies

Introduction

1.1
The relationship between cohort and socio-political orientation has long intrigued social scientists, principally because of the role cohort processes are understood to play in historical change. The concept of cohort sensitises us to the fact that an individual's location in historical time, via date of birth and death, exposes them to certain experiences, crises and events and excludes them from others. Following Mannheim (1952), youth is identified as a key period of exposure to experiences, crises and events, and is argued to have lasting ideological effects throughout a person's subsequent life. As a result of differential exposure and exclusion due to cohort location in historical time, people of different ages are likely to experience the same social and cultural events differently. Following from this, cohort replacement, via the birth, ageing and eventual death of all members of a cohort, is regarded as a key element in the production of historical change. In Mannheim's words, the 'fresh contact' of new cohorts with the already existing cultural and social heritage always means a 'changed relationship of distance' and a 'novel approach in assimilating, using and developing the proffered material' (Mannheim, 1952: p. 293). In sum, cohort is recognised as a key engine of social change, via its shaping of socio-political orientations over the life course. Debate continues, however, as to the precise ways cohort interrelates with period and stage in life course effects in shaping socio-political orientations (see Alwin et al, 1991 for review).

1.2
A subject of some interest to social scientists, especially in the United States, has been the status of cohort as a predictor of gender role attitudes amongst women and, accordingly, the part cohort replacement has played in changes in the gender order (for example, Roper and Labeff, 1977; Slevin and Wingrove, 1983; Thornton et al, 1983; Bell and Schwede, 1985; Witherspoon, 1985; Harding, 1988; Misra and Panigrahi, 1995; Scott et al, 1996). This evidence on gender role attitudes suggests that younger cohorts of women are more liberal and egalitarian in their attitudes to sex roles than are older women. For example, the 1985 British Social Attitudes survey made use of the following statement to tap attitudes to gender roles and the traditional pattern of the domestic division of labour: 'A husband's job is to earn the money; a wife's job is to look after the home and the family'. Patterns of 'agreement' and 'disagreement' with this statement showed marked differences by age group. Thus, whilst 21% of women aged 55 and over 'agreed strongly' with the statement, 14% of women aged 35-54 and 8% of women aged 18-24 did so. In contrast, 24% of women aged 18-24 'disagreed strongly' with the statement, compared to 25% of the 35-54 year olds and 12% of those aged 55 and over (Witherspoon, 1985: p. 93, table 3.13).

1.3
Studies on cohort variation in attitudes to gender roles have generally been conducted via surveys, and hence with a quantitative research design. The survey approach has a number of distinct advantages as a method of examining the significance of cohort. It allows claims of representativeness and generalisability of findings, and often includes the examination of the correlation between cohort age and other variables, including marital and employment status. The survey approach also facilitates cross-national comparisons (for example, Scott et al, 1996) and thereby the development of explanations of changes in gender attitudes and behaviour over time and between cultures. Whilst surveys make these important contributions through providing, say, data on the proportions of each cohort 'agreeing' or 'disagreeing' with a particular questionnaire item (as in the British Social Attitudes example, above), they are not designed to explore the equally interesting question of differences in the ways in which agreement or disagreement may be expressed by each cohort. Surveys give data on the 'final product' (the 'attitude') but can tell us little about how that product was formulated. Arguably, pre-coded responses 'do violence' to the richness and variety of what might otherwise be said on a particular issue. In contrast, qualitative studies, through allowing interviewees to answer in their own words, do allow a detailed examination of the vocabulary used to report responses to gender roles and an exploration of the ways in which this may vary by cohort. Furthermore, as I have argued elsewhere (Pilcher, 1994a), if the sociology of cohort influences on socio-political orientation is a sub-field of the sociology of knowledge, then a methodological approach suited to the interpretation of meaning suggests itself as the most appropriate method of investigation, with a particular focus on the words, vocabularies or discourses employed (Mannheim, 1960; Mills, 1967a, 1967b; Dant, 1991). Yet, cohort studies of women's gender role attitudes using a qualitative research design remain rare. Consequently, although we know from surveys that patterns of agreement and disagreement on the issue of traditional gender roles show marked differences by cohort, we have little sense of the 'reasoning', particularly in relation to gender, that results in the reported disagreement or agreement.

1.4
This article aims to fill some of the gaps in knowledge about cohort differences amongst women on gender role issues, via an analysis of responses to role reversal. The article addresses the following research questions. First, in a study with a qualitative research design, are cohort differences as apparent as they are in surveys? Second, when allowed to answer in their own words, do women of different cohorts report their agreement or disagreement in the same or in contrasting ways? Third, within their talk about role reversal, are there differences between the cohorts of women in the acknowledgement given to gender as a category?

The Study

2.1
Data presented in this article were collected as part of a study comparing the accounts three cohorts of women gave on a range of gender issues and the women's movement. Data were collected in 1989, via semi-structured interviews with women who were members of the same family (mother, daughter and adult granddaughter). In other words, an age-structured sample was achieved via multi-generational family groupings. Due to their differing dates of birth and the lineage gap, parent-offspring generations always belong to different cohorts. Using this strategy means a neat and ready made cohort division and avoids the recognised problem of where to locate cohort boundaries. Using related samples, however, does introduce the complicating variable of family socialisation (for a full discussion of the difficulties associated with the empirical investigation of cohort, see Alwin et al, 1991; Pilcher, 1995). In total, nineteen families of three generations were interviewed (fifty seven women), representing three broad cohorts. The oldest cohort ranged from 62-87 years, the middle from 38-56 years and the youngest from 17-29 years. The sampling strategy employed for the study was an 'opportunistic' one. The families of women were contacted through meeting older women at pensioners' clubs or day centres for elderly people, via personal contacts and via snowballing from both of these sources. The women lived in urban areas of South Wales, the majority having been born and having lived most of their lives in this same area. All the women were white. Using the Registrar General's Classification of Occupations (Office of Population Census and Surveys, 1980) to categorise the women's own occupations, over half of the women were categorised as 'middle class', with most falling into the intermediate (Class IIIn) category. A majority of the male partners (or fathers in the case of the non-cohabiting youngest generation) of the women were categorised as 'working class', via their work in skilled, semi-skilled or unskilled manual occupations. Most of these men worked in skilled manual occupations (Class IIIm). Whether determined by the women's own occupations, or the occupations of their male partner, the single largest grouping in the sample was Class III (49% under both measures) and thus the sample was bunched around the middle of the social class scale. Interviews were held with the women on an individual basis, were tape recorded and fully transcribed, since having a full record of the women's responses was central to the research task of focusing on words and vocabularies. The topics or issues I raised in the interviews, whilst not aiming to define 'gender issues', were intended to reflect concerns which have been central to the feminist agenda, both recently and historically. These included issues relating to gender in the domestic context, in paid work, in public life and politics, in sexuality and reproduction, and in culture (full details of the sample and methods employed can be found in Pilcher, forthcoming). Analysis of the interview data centred around the intensive comparison of words and phrases and from this, sets of vocabularies were identified (for studies which have taken a similar approach on different topics, see, for example, Mansfield and Collard 1988; Brewer, 1990; Duncan and Edwards, 1997). In this article, quotations from the women who were interviewed are accompanied by a pseudonym, by their age and by an abbreviated reference to indicate their generational position (G1 for the oldest, G2 for the middle and G3 for the youngest).

Accounts Of Role Reversal

3.1
A 'web of constraining influences' (Morris, 1987), including gender role and family ideology, the gendering of paid work and the policies and practices of the agencies of the welfare state, militate against any fundamental and widespread change in gendered divisions of labour. Consequently, despite probable increases in recent years, households operating on a system of 'role reversal', with the woman as the full time paid worker and breadwinner, and the man as the unpaid domestic worker in the home, remain very much a rarity. Morris (for example, 1985) has shown that even in the context of men's redundancy and subsequent unemployment, strong social pressures remain which hinder the renegotiation of the domestic division of labour. Role reversal households are, then, unconventional and remarkable. On a symbolic level, role reversal households represent a challenge to the various components of traditional gender role ideology, including ideas about motherhood and children's 'needs', dependent femininity, and men as providers and heads of households. For some couples, reversing roles may be a conscious choice, precisely because it puts gender 'out of place' through its radical alteration of conventional gender and generational relationships at the level of the household. For other couples, a reversal of roles may be forced upon them and reflect changes in opportunities faced by some women and men in the labour market. Whether through choice or force of circumstance, role reversal disturbs the conventional patterns of responsibilities and power relations between women and men which are central to the performance of masculinity and femininity. For this reason, role reversal presents itself as an ideal issue through which to examine accounts of gender roles and how these differ qualitatively between cohorts.

3.2
In wording the interview question on role reversal, I deliberately did not use the term 'role reversal'. I anticipated that some of the women might have been unfamiliar with the term and further, that if individual women were familiar with it, the term itself might be associated with pre-conceived notions. The question was therefore phrased as follows: 'What would you think of a marriage or a relationship where the man stayed at home and did the housework and looked after the children, if there were any; and the woman was the one who went out to work full time?'. The accounts that the three cohorts of women gave in response to this question were broadly categorised into 'hostile' and 'sympathetic'. The majority (12 out of 18) of the oldest cohort gave accounts hostile to role reversal, whilst the majority of the middle (13 out of 18) and the youngest cohorts (13 out of 19) gave sympathetic accounts. This finding is in line with patterns of cohort difference on gender roles revealed by, for example, the 1985 and 1994 British Social Attitudes surveys (see Scott et al, 1996 for discussion). The qualitative analysis of the accounts of role reversal focused on the contrasting types of vocabulary used by the women in response to the interview question. Four types of vocabularies were present in the women's accounts of role reversal and the core issue of the gendered responsibility for domestic work that it tapped in to. These vocabularies were labelled as traditional, individualism, equality and structural constraints. As illustrated below, the vocabularies were used to differing extents and in different ways according to cohort.

The Oldest Cohort

4.1
In the hostile accounts given by the oldest cohort (12 out of 18), role reversal was something they said they did not like or, in two accounts, had difficulty even imagining. A traditional vocabulary was used to convey their disapproval. For example, in reflecting upon role reversal, the women of this cohort identified the unusual circumstances which would bring men in to the domestic environment, such as their unemployment or ill health.

...Unless it was that the husband couldn't find a job and the wife was able to work.
Nancy Caswell, aged 69 (G1)

...If it was a case of the husband not being able to go out to work, if there was something wrong with him, well, all right. That's fair enough...Somebody's got to be the breadwinner.
Irene Harvey, aged 62 (G1)

4.2
Six women made reference to such forces of circumstance, under which role reversal might be countenanced. Even so, in some of the women's accounts, a man's incapacity to be a breadwinner did not automatically mean that he should then become completely responsible for domestic work. For Nancy Caswell, Agnes Baker and Dora Griffiths, men could 'help' their wives if they themselves were not working but, 'not completely look after the family and the house' (Agnes Baker, aged 78, G1). Moreover, two women of this cohort expressed a concern over the effect of role reversal on the man's feelings and status. For example,

I don't think I would like to go out [to work] and leave my husband at home doing work.// Well, it makes the husband think he's not looking after you properly, really.
Irene Harvey, aged 62 (G1)

4.3
The conditions under which role reversal might be acceptable (a man's unemployment or inability to work, his 'helping' rather than having complete responsibility for domestic work and, in both cases, 'if the man was prepared to do it' (Doris Ascote, aged 72, G1), are core characteristics of a vocabulary of hostility which presented role reversal as an infringement of traditional gender roles. This is clearest in the account's of five women which contained the phrases 'a man's place' and/or 'a woman's place'. For example,

Actually, I don't agree with that. I got to be truthful...a man's place is in work, provide and a woman's work is to have her babies and look after them and look after the home.
Sybil Richards, aged 71 (G1)

No, I wouldn't like that. I think a man's place is in work. No, I don't think I'd like that at all.
Dora Griffiths, aged 68 (G1)

Well, I'm not at all in favour...I think a woman's place is in the home. Rose Jessop, aged 73 (G1)

4.4
Apart from Sybil Richards and Dora Griffiths, the women specified the 'place' of the woman, particularly as a mother to children, rather than the 'place' of the man. Five other accounts given in response to role reversal, did not contain the references to women's and men's 'place', but nevertheless made explicit reference to traditional gender roles. Here, the role of men as breadwinners featured more strongly:

I don't think it's right.//...I think you get married and your husband will keep you or work for you.
Ivy Keating, aged 87 (G1)

Well, I just wouldn't like it.//...when the husband brings the money home, that's great, ain't it.[1]
Irene Harvey, aged 62 (G1)

We have been brought up with the man being the breadwinner, primarily. That was the norm when I was young.
Nancy Caswell, aged 69 (G1)

4.5
Ivy Keating, quoted above, was the only woman of the oldest cohort who gave an indication that her vocabulary of the 'proper' roles of men and women was perhaps out of place in contemporary society. After saying about husbands 'keeping' their wives, she added that this was probably 'an old fashioned idea, I suppose'.

4.6
Six women of the oldest cohort gave sympathetic accounts of role reversal, often specifying the sorts of circumstances which might lead to this unusual arrangement. Two women, although favourable to the idea of role reversal, used a traditional vocabulary, with the man as the main focus of their concern.

If there was no work about for the man, yes...As long as the man wasn't demoted, do you understand me, because of that, you know.
Doreen Owens, aged 70 (G1)

Well, to me, I think that's a good idea. If he wants to do that.
Elsie Farrall, aged 74 (G1)

4.7
In contrast, the other four sympathetic accounts given by women of this cohort had the position and status of the woman as a primary focus. In two accounts, circumstances where the woman partner had a 'good job', earning more money than the man, were specified. For example:

I don't see as it makes any difference really, if the wife has a good job...
Yvonne Daniels, aged 75 (G1)

4.8
Three women conveyed their views on role reversal via telling 'success stories' of couples known to them who had reversed roles, either because the woman partner had 'a better job', 'earned more' or had 'a very important job' compared to the man. Direct knowledge of successes therefore seemed an important factor influencing the sympathetic construction of role reversal by the oldest cohort of women. In stressing the practicality of the arrangement according to the circumstances faced by the couple involved, the women drew upon a vocabulary of individualism.

The Middle Cohort

5.1
A minority of accounts (five out of 18) given about role reversal by the middle cohort were hostile in content, and all drew upon a traditional vocabulary. For example:

I wouldn't like it. I wasn't brought up that way...the man went out to work and the woman stayed at home with the children.
Susan Griffiths, aged 44 (G2)

Wouldn't like it.//...it's a man's place to go out to work.
Valerie Harvey, aged 40, (G2)

5.2
Two women were only prepared to countenance the idea of role reversal under circumstances which forced the man to be in a domestic environment, that is, if he was unemployed or unable to work due to ill health. Two others explained their antipathy to role reversal on the grounds that men are incompetent at housework, a feature not present in the oldest cohort's accounts: 'Some men, they still don't do it the way you do it, you know' (Miriam Powell, aged 43, G2). Valerie Harvey (aged 40, G2) also doubted whether her husband would do the housework to the same standard as she did, if he had responsibility for it. When I then suggested to her that he might soon learn, she replied that he should not. She was suggesting that it would not be proper for him to gain a high level of competency in housework, because this would violate traditional gender roles. As in the hostile accounts of the oldest cohort, these women's hostile accounts were constructed around a focus on men's roles and status.

5.3
The middle cohort of women were twice as supportive in their accounts of role reversal than were the oldest cohort. They too specified the circumstances under which role reversal might occur, thereby suggesting that it is an unusual way of dividing responsibilities between women and men. In five of the thirteen supportive accounts, approval of the idea of role reversal was accompanied by reference to the relative earning power of the couple. For example:

Well, if the woman has got a better job and is bringing more money into the house. And if they are agreeable, I don't see anything wrong.
Angela Farrall, aged 44 (G2)

That's fine, it's sensible. I think the one who has the greatest earning power ought to do the work.
Rosemary Thomas, aged 46 (G2)

Well, I certainly don't see anything wrong with it. I mean, if the woman could earn more money than the man. Well then, that is fair enough.
Carol Mitchell, aged 38 (G2)

5.4
In these examples, the practicality of the arrangement is stressed according to the paid work circumstances a couple might face. This vocabulary of individualism, where the main feature is the suitability of the arrangement, according to the circumstances, preferences and/or character, of the particular couple concerned, was a prominent one in the sympathetic accounts of this cohort. For example:

Fine, if it suits them and they are happy...
Judith Ascote, aged 48 (G2)

That's fine, if they are happy doing that.
Janice Caswell, aged 44 (G2)

I wouldn't mind. I think my husband's got more patience than I have...I think he would be quite willing to stay at home.
Rita Parry, aged 42 (G2)

5.5
Unlike the oldest cohort, no women of this cohort gave a sympathetic account of role reversal via a traditional vocabulary. In other words, none expressed concerns about the feelings and status of men in a situation of role reversal (although Angela Farrall (aged 44, G2) did say that she was unsure how men in general would feel about having a domestic role), nor were the circumstances of the man's unemployment or inability to work specified. Of the middle cohort, Janet Morgan (aged 47, G2) was alone in expressing her positive assessment of the idea of role reversal in terms of the principle of equality, saying '...Yes, it's equality, isn't it really. Yes, I have got no objections to that'. With these exceptions, women of this cohort constructed their sympathetic accounts of role reversal with a vocabulary of individualism.

The Youngest Cohort

6.1
The six hostile accounts of the youngest cohort were comprised of a traditional vocabulary. None of these women had children themselves and only one was cohabiting or married. A particular feature, present in four accounts, was the referencing of women's roles as mothers. For example:

No. I wouldn't like that. I'd like to be there with the children.
Rhian Keating, aged 24 (G3)

...(I think) the mother should look after the child...the child needs the mother more really.
Sharon Owens, aged 18 (G3)

6.2
The role and status of men also featured in the hostile accounts given by the youngest cohort. One woman objected to role reversal on the grounds of men's incompetency in domestic work. She cited media representations of role reversal as her evidence:

I don't know about that [role reversal]. I think the woman should stay at home and look after the children and do the housework...I don't think the men should stay at home because I don't think they could cope. Like, I saw this film on T.V., 'Mr. Mum', and the man stayed at home and looked after the children and that, and it didn't work out.
Mandy Mitchell, aged 17 (G3)

6.3
In two other hostile accounts, reference was made to the feelings and status of men who had a domestic role, given that this would be an infringement of their traditional roles as men:

I think it would be humiliating for a man...Like, if I was a man, I wouldn't like to be kept.
Elaine Griffiths, aged 19 (G3)

I don't think the man would like it all that much, would he. Men like to go out do the work, rather than stay home and do the housework.
Lorraine Morgan, aged 19 (G3)

6.4
The vocabulary thus presents role reversal as a threat to masculine identity, whilst women who stay at home do not have to 'swallow their pride' or feel humiliated at 'being kept'.

6.5
There were two features present in the hostile accounts of the youngest cohort which were not found in other hostile accounts. Wendy Caswell acknowledged that her antipathy to role reversal was not in line with a feminist stress on equality:

For me personally, I wouldn't like it...you know, women's lib and all that kind of stuff but I would want to be with the children.
Wendy Caswell, aged 22 (G3)

6.6
The second feature unique to the youngest cohort's hostile accounts of role reversal was an awareness, in two cases, of the gendered structure of the labour market which militates against role reversal. Lorraine Morgan (aged 19, G3) said that she would not be able to 'earn enough money' in her line of work (office worker) to enable her to reverse roles. Similarly, Wendy Caswell (aged 22, G3) said that 'It will always be that men have got better jobs', so 'not a lot' stay home and look after the children. Here, rather than expressing their hostility to role reversal in ideological terms, the women are identifying structural constraints which inhibit role reversal, and make it an impractical option for many households.

6.7
The thirteen sympathetic accounts of role reversal given by the youngest cohort specified the sorts of circumstances which would lead to role reversal. In six accounts, sympathy for role reversal was expressed through referencing the relative earning power and employability of the couple. For example:

That's fair, I mean, if she can earn more. If he can't get a job and she finds one, I don't see the difference.
Rebecca Daniels, aged 29 (G3)

I think it's all right...if the wife have [sic] got a good job and the man hasn't got a job, then there is no point in them both being home.
Denise Nicholl, aged 27 (G3)

Fine. It depends, you know, normally the man can earn a better wage. But if the woman can earn a good wage, why not?
Bethan Parry, aged 24 (G3)

6.8
In these accounts, it was especially the status of the woman partner's job that was cited as important (six examples), with the enforced circumstances of men's unemployment cited only twice. In Bethan Parry's account (above) the structural constraints which militate against role reversal are identified (a vocabulary of structural constraints). Rather than reflecting on the circumstances that might bring men in to the domestic environment (a feature of the oldest cohort's accounts), for these women of the youngest cohort, the focus was the circumstances which encourage women to continue in paid work. Concern for damaged masculinity was expressed by one woman, who said that a man in a role reversal situation would have to 'swallow his pride' but that if he did reverse roles, this would be 'to his credit' (Lindsay Farrall, aged 20, G3). Otherwise, the status and feelings of men did not feature. Instead, there was a concern with the happiness and needs of the couple:

Oh, I think that was great...if you've got that kind of relationship.
Karen Lestor, aged 29 (G3)

Well, if they are happy, that's fine. If it suited them, that's fine.
Isabel Ascote, aged 21 (G3)

6.9
In speaking of the practicalities of role reversal, according to the circumstances, character and preferences of the couple involved, these women employed a vocabulary of individualism.

6.10
A distinctive feature of this cohort's accounts was an explicit interpretation of role reversal as a challenge to traditional gender roles or as an example of gender equality in practice. For example,

Good.//...Well, long enough that women have stayed in the home.
Ruth Richards, aged 17 (G3)

I agree that women should have equality.
Kirsty Harvey, aged 19 (G3)

6.11
One other woman of this cohort employed a similar vocabulary in her account of role reversal. Hayley Baker (aged 23, G3) explained her view that role reversal was a good idea 'on the surface' because it is something that men do not usually do and that it would therefore be good for society as a whole. These are examples of a vocabulary of equality, where role reversal is presented as 'good thing' on ideological grounds.

Discussion

7.1
The data show that the women in the study responded in contrasting ways to role reversal, along the lines suggested by cohort studies with a quantitative survey design. The broad age ranges within each cohort (62-87, 38-56 and 17-29) grouped together women who had lived through different historical circumstances. Yet, within each grouping, women seemed to have enough common socio-historical location to influence their constructions of role reversal along cohort lines. Women in the oldest cohort tended to give hostile accounts, mostly portraying role reversal as a violation of traditional gender roles. Women in the younger cohorts tended to give sympathetic accounts, where role reversal was mostly presented as an acceptable arrangement, according to the particular circumstances faced by a couple. Unlike in cohort surveys, however, no marked difference was found in frequency of support for traditional gender roles between the middle and the youngest cohorts. In other words, the division was between the oldest cohort and the younger cohorts, rather than a three-fold division, with the middle cohort in-between, as commonly reported in cohort surveys.

7.2
The proportions of the middle and youngest cohorts giving hostile or sympathetic accounts of role reversal was the same, but qualitative analysis showed the influence of cohort location on the vocabularies employed within these accounts. Four types of vocabularies were identified within the women's accounts of role reversal and the core issue of the gendered responsibility for domestic work that it tapped in to. These vocabularies were labelled as traditional, individualism, equality and structural constraints. Of these, the traditional and the individualism types were the most frequently employed. A traditional vocabulary was predominantly used in hostile accounts, and was especially made use of by women of the oldest cohort. It presented role reversal as a violation of traditional gender roles, responsibilities and power relationships. It contained descriptions of traditional and conventional gender roles and responsibilities, and included the prescriptive use of the phrases 'a man's place' and 'a woman's place'. References were made to forces of circumstance which might place men in a domestic context, such as their unemployment or ill health, and concern was expressed over the detrimental effects on masculinity. Doubts as to men's competency in domestic work also formed part of this vocabulary. In essence, then, a traditional vocabulary portrayed role reversal as an infringement of traditional, conventional and 'proper' gendered roles and responsibilities, particularly from a man's point of view. Within this vocabulary, gender was presented as a legitimate basis on which to determine roles and responsibilities, and in a prescriptive and inflexible manner. A vocabulary of individualism was predominantly used in accounts sympathetic to role reversal, and especially by the middle and youngest cohorts. Its main feature was the emphasis on the practicality of role reversal according to the paid work circumstances, preferences and characters of the couple concerned. Within this vocabulary, role reversal was presented as a choice that should be made on practical grounds, according to the employment status and prospects of each partner in a couple, including their earnings potential. Within this vocabulary, ideological justifications and the gender of each partner seemed irrelevant; rather, role reversal was a matter of what 'suited' the particular couple and their particular circumstances. The equality vocabulary presented role reversal as an example of gender equality or feminist principles in practice. It was only used by women in the younger cohorts and mostly in their sympathetic accounts of role reversal. Here, role reversal was supported on ideological grounds, and portrayed as representing a valid challenge to traditional and conventional divisions of labour, which disadvantage women. Within this vocabulary, therefore, gender was suggested as a significant, if illegitimate, basis on which experiences and opportunities are shaped. Where an equality vocabulary was used by women in their hostile accounts, the status of role reversal as a challenge to gender inequality was recognised but not felt to be appropriate to their own preferences or circumstances. A vocabulary of structural constraints was only found in hostile and sympathetic accounts given by the youngest cohort. This vocabulary identified the structural pressures which make role reversal very difficult to achieve. In particular, the gendered nature of paid work was said to make it unlikely that a woman would have the sort of job to enable her to be a 'breadwinner', rather than her partner. This type of vocabulary suggests a recognition of wider structures of gender inequality which constrain choices and options available to couples at the level of the household. Within this vocabulary, gender is invoked as an important determinant of opportunity and as highly relevant to the household strategy a couple might develop.

7.3
Two types of vocabulary, traditional and individualism, were made use of by all three cohorts. However, the traditional vocabulary was employed mostly by the oldest cohort and the vocabulary of individualism mostly by the middle and youngest cohorts. Vocabularies of equality and, especially, structural constraints were not used at all by the oldest cohort, but especially by the youngest cohort. Within sociology, vocabularies are argued to reflect the context of their use (Mills, 1967a; 1967b), including the wider context of socio-historical location, via life course and cohort membership. For the oldest cohort, therefore, using a vocabulary which constructed role reversal as a violation of traditional gender roles, responsibilities and relationships reflected their socialisation experiences and life courses lived within a particular gendered socio- historical structure. The historical context of these women's life courses was one where men only undertook limited household work, married women had a limited involvement in paid work and where a marked gendered division of labour was the norm. According to the women themselves, their own earlier involvement in paid work was intermittent, broken (if not ended) by marriage or children. They also reported on their long held responsibility for domestic work, which often continued after their husbands' retirement (see also Mason, 1987). Data on these women's accounts of men's involvement in housework, published elsewhere, showed that they were concerned to specify the circumstances and extent of men's participation. Their talk showed them to be resistant to the substantial involvement of men in the domestic sphere and to hold low expectations about men's domestic work activities (Pilcher, 1994b). In the context of these women's socio- historical experiences, therefore, talking about 'a man's place' and 'a woman's place', noting the force of circumstances that would bring a man in to the domestic context and expressing concern about the effects this would have on his masculinity, were all appropriate, valid and acceptable ways of talking about role reversal. For the same reasons, talking about role reversal in terms of individualism was not an acceptable or, in the case of equality or structural constraints, possibly even available (Emmison and Western, 1990; Frazer, 1988), vocabulary, for most women of this cohort. An important contingency affecting the sympathetic construction of role reversal for the oldest cohort was direct knowledge of role reversal 'success': three out of six women giving sympathetic accounts of role reversal told 'success stories', employing a vocabulary of individualism in the process.

7.4
In the accounts of the two younger cohorts, a traditional vocabulary was less evident, both in terms of frequency and explicitness. Conversely, an individualistic vocabulary (citing practicality of circumstances, personal needs and happiness) was more often employed. These cohort differences in use of vocabulary also allow an interpretation along the lines of discursive acceptability and availability arising from cohort location in historical 'gendered opportunity structures' (Walby, 1997). The two younger cohorts experienced a different socio-historical context to that faced by the oldest cohort, including in terms of changed expectations about married women and paid work, with knock-on effects for expectations about men and domestic work. Married women have increasingly participated in the formal labour market in greater numbers since the 1950s and, as Martin and Roberts (1984) have argued, we should therefore expect to find differences between older and younger women on gender roles which reflect these different employment-home relationships. Changes over time in women's participation in paid work, coupled with an erosion of men's former security in the role of worker, have contributed to the breaking down of the association of women wholly with the domestic sphere and men wholly with the sphere of paid work. Even where women of the middle and youngest cohort gave accounts hostile to role reversal, their vocabulary contrasted with that of the oldest cohort. Hence, rather than framing their objections via referencing men's circumstances, feelings and status, some women of the middle cohort cited men's incompetence in the performance of domestic work, whilst women of the youngest cohort referenced the needs of children or structural constraints. As suggested by Scott et al (1996), for younger women, opposition to role reversal may be more acceptable or permissible if expressed in ways which do not reference men's needs, as in the older cohort's accounts.

7.5
Data discussed elsewhere (Pilcher, 1994b) on the middle cohort of women's accounts of men's participation in housework showed they were considerably more accepting of men doing housework than were the oldest cohort. Data on these women's own domestic arrangements at the time of the interviews showed them retaining responsibility for domestic work, despite significant levels of engagement in paid work. Some women, consequently, spoke of their often inequitable domestic arrangements as a 'bone of contention' between themselves and their husbands, and there were reports of efforts to increase their husbands' input (see also Brannen and Moss, 1991). Although many of them had not yet reached that stage of their life course, accounts of men doing housework given by the youngest cohort showed that they were emphatically supportive of the involvement of men in housework. Frequently, it was men not doing housework that was an issue and they held high expectations of men's participation in the domestic sphere (Pilcher, 1994b). For the younger cohorts of women, therefore, gender did not excuse the participation of men in housework. Nor did gender preclude the participation of women in paid work, including as 'breadwinner'. Therefore, in their accounts, role reversal was largely constructed as an issue of practicality: the gender of who does what was immaterial. Although women of the younger cohorts responded in these ways to the idea of role reversal, none had actively experienced it themselves. Younger women may be more likely to construct role reversal as an issue of practical choice according to circumstance, but given the 'web of constraining influences' (Morris, 1987) on gender roles and responsibilities, this is unlikely to translate into 'breadwinning wives' and 'househusbands' on a large scale.

7.6
Whilst over time, women's changed relationship to paid work may have contributed to the erosion of the relevance and acceptability of traditional vocabularies of gender roles and responsibilities, other 'vocabularies of motive' (Mills, 1967a; 1967b) have gained greater hold. Individualist ideologies, comprising (amongst other elements) a belief in autonomy and the free action of individuals, have become an increasingly prominent feature of Western industrial society and an increasingly acceptable 'vocabulary of motive' (Lukes, 1973; Hutson and Jenkins, 1989). Indeed, it was this vocabulary with which most accounts sympathetic to role reversal were constructed, especially by women of the younger cohorts. Equality and structural constraints vocabularies were rarely employed in the accounts of role reversal, but where they were, it was mostly by women of the youngest cohort. Arguably, this is a reflection of their greater exposure to egalitarian and feminist ideologies which have made systemic and systematic gender inequality a focus of concern. In contrast to the emphasis on individual choice contained within individualism, the equality and structural constraints vocabularies suggest a recognition of restricted opportunities collectively faced by people, especially women, on the basis of gender. However, the findings of this study indicated that, on the issue of role reversal, individualism is a more accessible and appropriate vocabulary of motive than feminist influenced equality and structural constraints vocabularies. Whilst both ideologies can be viewed as progressive, the prevalence of individualism over feminist influenced vocabularies means that the collective oppression of women on the basis of their gender loses prominence, along with recognition of systemic patriarchal structures and systematic practices of sexism. Contrasting experiences of and exposure to traditional gender roles and practices, and to the ideologies of individualism and feminism are suggested, then, as aspects of the differing gendered opportunity structures faced by the three cohorts of women, which are reflected in the vocabulary they used to talk about role reversal. In their accounts of this gender issue, it is possible to identify (in Mannheim's terms) the 'fresh contact' first the middle and then the youngest cohort has had with the pre-existing gender heritage, and their consequent 'novel and distinctive approaches' to gender divisions. The pattern of findings suggest that future cohorts of older women will experience their later life informed by progressive 'vocabularies of motive' on gender matters, thereby accelerating and propelling further change in the gender order.

Conclusion

8.1
In offering such explanations of why the women used different vocabularies to talk about role reversal, I am concurring with the type of explanation put forward by those reporting survey data on cohort differences. Whether qualitative or quantitative, both types of research design are, after all, investigating the same phenomena and both hold as axiomatic the principle that findings are indexical to the social context . However, the research issues addressed by surveys and by small scale qualitative studies are different ones. They each have different purposes (Hammersley, 1996), give data on different things, and may produce findings which are similar or distinct (Mason, 1994). It is my argument that findings from qualitative studies, such as that reported on here, are an important supplement to surveys of cohort differences in women's gender role attitudes. They indicate that women's responses to gender roles vary by cohort in much more complex ways than is suggested by survey data on proportions 'agreeing' or 'disagreeing' with a particular questionnaire item. Whilst surveys can claim representativeness and other advantages, qualitative studies are better placed to reveal the varied and complex ways in which women of different cohorts construct gender issues and the influence cohort has on the very language they use to do so.

Note

1 In transcribing the taped interviews, I was concerned to reflect the way in which responses were spoken. Here, for example, this response was not put to me as a question, and therefore is not punctuated by a question mark.

Acknowledgements

The research on which this article is based was supported by a University of Wales Postgraduate Studentship.

References

ALWIN, D., COHEN, R., AND NEWCOMB, T. (1991). Political Attitudes Over The Life Span: The Bennington Women After Fifty Years. Madison, Wis.: University of Wisconsin Press.

BELL, M. and SCHWEDE, K. (1985) 'Roles, Feminist Attitudes and Older Women', Women in Politics, vol. 5, no. 1, pp. 5 - 22.

BRANNEN, J. and MOSS, P. (1991) Managing Mothers. Dual Earner Households After Maternity Leave. London: Unwin Hyman.

BREWER, J. (1990) 'Talking About Danger: The RUC and the Paramilitary Threat', Sociology, vol. 24, no. 4, pp 657 - 674.

DANT, T. (1991). Knowledge, Ideology And Discourse. A Sociological Perspective. London: Routledge.

DUNCAN, S. and EDWARDS, R. (1997) 'Single Mothers in Britain: Unsupported Workers or Single Mothers?' in Duncan, S. and Edwards, R. (editors) Single Mothers In An International Context. London: UCL Press.

EMMISON, M. and WESTERN, M. (1990) 'Social Class and Social Identity: A Comment on Marshall et al', Sociology, vol. 24, no. 2, pp. 241 - 253.

FRAZER, E. (1988) 'Teenage Girls Talking About Class', Sociology, vol. 22, no. 3, pp. 343 - 358.

HAMMERSLEY, M. (1996) 'The Relationship Between Qualitative and Quantitative Research' in J. Richardson (editor) Handbook of Qualitative Research Methods. Leicester: BPS Books.

HARDING, S. (1988) 'Trends in Permissiveness' in R. Jowell, S. Witherspoon and L. Brook (editors) British Social Attitudes. Aldershot: Gower.

HUTSON, S. and JENKINS, R. (1989) Taking the Strain. Milton Keynes: Open University Press.

LUKES, S. (1973) Individualism, Oxford: Basil Blackwell.

MANNHEIM, K. (1952). 'The Problem of Generations' in K. Mannheim, Essays On The Sociology Of Knowledge. London: RKP (first published 1923).

MANNHEIM, K. (1960). Ideology And Utopia. London: Routledge (first published 1936).

MANSFIELD, P. and COLLARD, J. (1988) The Beginning of the Rest of Your Life? A Portrait of Newly-Wed Marriage. London: Macmillan.

MARTIN, J. and ROBERTS, C. (1984) Women and Employment. A Lifetime Perspective. London: HMSO.

MASON, J. (1987) 'A Bed of Roses? Women, Marriage and Inequality in Later Life' in P. Allat, T. Keil, A. Bryman and B. Bytheway (editors) Women and the Life Cycle. London: Macmillan

MASON, J. (1994) 'Linking Qualitative and Quantitative Data Analysis' in A. Bryman and R. Burgess (editors) Analyzing Qualitative Data. London: Routledge.

MILLS, C.W. (1967a). 'Situated Actions and Vocabularies of Motives' in I. Horowitz (editor) Power, Politics And People. The Collected Essays of C. Wright Mills. Oxford: OUP (first published 1940).

MILLS, C.W. (1967b). 'Language, Logic and Culture' in I. Horowitz (editor) Power, Politics And People. The Collected Essays Of C. Wright Mills. Oxford: OUP (first published 1939).

MISRA, R. and PANIGRAHI, B. (1995) 'Change in Attitudes Toward Working Women: A Cohort Analysis', International Journal of Sociology and Social Policy, vol. 15, no. 6, pp. 1 - 20.

MORRIS, L. (1985) 'Renegotiation of the Domestic Division of Labour in the Context of Male Redundancy' in H. Newby, J. Burjra, P. Littlewood, G. Rees and T. Rees (editors) Restructuring Capital. London: Macmillan.

MORRIS, L. (1987) 'Constraints on Gender', Work, Employment and Society, vol. 1, no 1, pp. 87 - 106.

OFFICE OF POPULATION CENSUS AND SURVEYS (1980) Classification of Occupations. London: HMSO.

PILCHER, J. (1994a) 'Mannheim's Sociology of Generations: An Undervalued Legacy', British Journal of Sociology, vol. 45, pp. 481 - 95.

PILCHER, J. (1994b) 'Who Should do the Dishes?' in J. Aaron, T. Rees, S. Betts and M. Vincentelli (editors) Our Sisters' Land. The Changing Identities of Women in Wales. Cardiff: University of Wales Press.

PILCHER, J. (1995) Age and Generation in Modern Britain. Oxford: Oxford University Press.

PILCHER, J. (forthcoming) Women Of Their Time. Generations, Gender Issues and Feminism, Aldershot: Ashgate.

ROPER, B. and LABEFF, E (1977) 'Sex Roles and Feminism Revisited. An Intergenerational Attitude Comparision', Journal of Marriage and the Family, vol. 39, no. 1, pp. 113 - 119.

SCOTT, J., ALWIN, D. and BROWN, M. (1996) 'Generational Changes in Gender Role Attitudes', Sociology, vol. 30, no. 3, pp. 471 - 492.

SLEVIN, K. and WINGROVE, C. Ray (1983) 'Similarities and Differences among Three Generations of Women in Attitudes Toward the Female Role in Contemporary Society', Sex Roles, vol. 9, pp. 609 - 624.

THORNTON, A., ALWIN, D and CAMBURN, D. (1983) 'Causes and Consequences of Sex Role Attitudes and Attitude Change', American Sociological Review, vol. 48, pp. 211 - 227.

WALBY, S. (1997). Gender Transformations. London: Routledge.

WITHERSPOON, S. (1985) 'Sex Roles and Gender Issues' in R. Jowell and S. Witherspoon (editors) British Social Attitudes. The Fifth Report. Aldershot: Gower.

Copyright Sociological Research Online, 1998