The social definition of women's smoking behaviour

Sot. Sci. Med. Vol. 20, No. 12, pp. 1269-1278, Printed in Great Britain

THE SOCIAL

1985

0277-9536/85 $3.00 + 0.00 Pergamon Press Ltd

DEFINITION OF WOMEN’S BEHAVIOUR

SMOKING

ANDREA KNOPF ELKIND Department of Epidemiology and Social Research, University Hospital of South Manchester, Kinnaird Road, Manchester M20 9QL, England Abstract-The history of women’s smoking behaviour is one of changing normative definitions. Recent trends have been explained in terms of the symbolic value of smoking, representing for women freedom and independence. This view is emphasised by advertising. However, other evidence suggests the continued existence of an older, more negative cultural stereotype. A two-part study of young women undergoing professional training for nursing and teaching throws some light on the way in which female smoking behaviour is currently socially interpreted. The first phase indicated that among the minority of parents who had expressed their attitudes towards their daughter’s smoking in relation to sex-role norms, smoking was presented as unacceptable for women. More than half the sample perceived a negative cultural stereotype to be operating in contemporary society and two-thirds recognised its existence in the past. This stereotype presents smoking as a male behaviour and hence inappropriate for women. Women who do smoke are liable to be labelled as having unfeminine or degrading attributes. The stereotype operated more strongly in the general social background rather than in reference to personal relationships and hence its influence on contemporary behaviour is likely to be limited. It was rejected as out-dated or a male belief by some but nevertheless it represented the personal opinion of others. In terms of a more favourable definition the female smoker was perceived in terms of an elegant/sophisticated dimension and in relation-to an extrovert personality. The view of sample members that the growing acceptability of women’s smoking was related to social change indirectly supported the view that sees smoking as symbolic of independence. Those who saw smoking in neutral terms, i.e. as not having sex-role attributes, perceived smoking in this sense as normal social behaviour for men and women alike. The second phase suggested that smokers and non-smokers have divergent views about the image of the female smoker. The non-smoker’s image was based on the older cultural stereotype (‘unladylike’), whereas the smokers were more likely to take a view corresponding to the perspective that sees women’s smoking as symbolic of social change and greater independence (‘liberated’). The non-smokers had a clear and positive image of ‘girls who don’t smoke’ (‘feminine’), whereas for smokers the female non-smoker lacked a distinctive identity. The study thus suggests that traditional concepts of appropriate female behaviour continue to inhibit smoking among some women, whereas others perceive it as an aspect of independent behaviour. The roles of advertising and health education are discussed.

INTRODUCTION Smoking has undergone a long history of changing normative definintions [l] and this is especially true in relation to women. It has been hypothesised that smoking may be just one behaviour that was formerly suppressed by social norms prescribing appropriate behaviour for women and is now disinhibited [2]. In this paper the way female smoking behaviour is socially interpreted will be examined. Cigarettes were introduced to Britain in the midnineteenth century but it was only the daring or bohemian woman who smoked in public [3]. By the turn of the century a lady might smoke, but only in private, alone or with friends [4]. During the 19141918 war smoking in public became acceptable for women [5] and in the inter-war period women’s emancipation expressed itself in many ways including more open indulgence in tobacco [6]. By the early twenties cigarette smoking was a society vogue [7,8], but not everyone approved of the new behaviour. In 1919 the New Statesman reported that a young women in a restaurant had a cigarette knocked from her mouth by an irate waiter [9].

in 1921 less than 1% of tobacco sold was consumed by women, by 1930 it was 5% and by 1939 11% [lo]. In the mid-thirties smoking was seen as a normal means of relaxation for the busy business woman [l 11.Nevertheless smoking remained a minority habit among women and many did not take it up until well into adult life [12]. In the case of men, among whom smoking had become firmly established in the First World War [7], the onset of smoking occurred in adolescence [ 121. The tobacco companies recognised the growing market among women. Wills, for example, used flapper girl and Mary Pickford images in their advertising and later featured figures from the world of entertainment [7]. For the generations for whom cinema-going was a major pastime [6] the association of smoking with romance, glamour, independence, sophistication and social success was reinforced by the Hollywood stars of the thirties and forties [13]. The views of those who disapproved of smoking are less easy to discover, but are hinted at by suggestions that at various times girls have smoked as a gesture of progressive defiance, bohemian daring, or

1269

1270

ANDREAKNOPF

as a way of asserting one was uninhibited, emancipated or even fast [3]. Whatever the state of opinion, however, it was World War Two that transformed behaviour. The changes in women’s role in society were rapidly accelerated [5]. The smoking habit among women was boosted to the point of take-off and became accepted as normal behaviour for adult women at least in certain situations and contexts [12]. In the post-war period among successive generations of women the trend was towards earlier experimentation and earlier take-up of regular smoking [12]. Smoking became ubiquitous. The term ‘nonsmoker’ acquired a pejorative inflection, while smoking was as natural as eating or drinking. In advertisements cigarettes were portrayed as indispensable to relaxation, sophistication, confidence, virility and femininity [3]. However, in 1962 a new element was introduced with the publication of Smoking and Health by the Royal College of Physicians [14], which inaugurated the health-conscious era of smoking behaviour. The prevalence of smoking among men had already begun to decline before this and continued to do so afterwards, so that between 1948 and 1982 the proportion of male cigarette smokers fell from 65 to 38% [lo, 15,161. Among women, however, of whom some 41% were smokers in 1948 [lo] prevalence remained fairly stable until in the mid-1970s when some decline became evident for the first time and by 1982 33% of women were smokers [16]. This convergence in the level of smoking among men and women has been most marked among the younger age groups [lo, 161. Although a smaller proportion of younger women are now taking up smoking the age of onset has continued to grow younger. By the mid-sixties the onset of smoking occurred during adolescence in men and women alike [12] but the schoolgirl smoker was sufficiently unusual for a national study of children’s behaviour to be confined to boys [17,18]. In the eighties boys still begin to experiment earlier but girls catch up during the secondary school years so that between the ages of 11 and 16 there is little difference between the sexes in relation to prevalence [18]. Among the arguments put forward to explain these post-war trends are those which emphasise changes in social values generally and in women’s role in society in particular. Clark [19] has discussed the impact of the new values of the sixties on the teenage subculture. Other American commentators have suggested that smoking symbolises for women emancipation, independence and freedom [2&22] and female smoking patterns have been related to the concept of women’s liberation [23]. This viewpoint is epitomised by the American advertising slogan of the sixties for ‘Virginia Slims’ cigarettes, ‘You’ve come a long way, baby’. However, smokers and non-smokers are equally likely to identify with the women’s movement [19] and the trends towards the increasing development of smoking were established during World War Two [12], well before ‘liberation’ became a popular concept. Jacobson [24] argues that while social change may have influenced women to take up smoking, such ideas are irrelevant when considering why they continue once the habit is established. In this view, far from being a symbol of liberation, the cigarette is a

ELKIND

sign of dependence and oppression consequent on women’s unequal position in society. Nevertheless, smoking has always been a gender-related phenomenon [25] and underlying women’s adoption of smoking as a social phenomenon is the more general trend towards self-determination of social behaviour among women [2]. In developed countries the ideal image of women has shifted radically. Attractiveness, sexuality and inter-personal facility remain valued but to these have been added independence, assertion and career achievement [2]. It is to establish and reinforce the connection with such attributes that advertising aimed at women is at least in part directed. In the early seventies cigarette advertisements were still promising sex, sophistication and elegance [24] but in 1977 the Advertising Standard Authority amended its code to preclude suggestions that ‘female smokers are more glamorous or independent than non-smokers’ or that ‘smoking enhances feminine charm’ [26]. A great weakness of such regulation is that it deals only with explicit claims; the advertisers simply respond by emphasising implicit claims [27]. The images promoted represent solutions to experienced problems of identity [28], for example, women’s tennis, representing exactly the qualities of independence, assertiveness and success, has been used to promote the women’s brand ‘Kim’. However, set against this is evidence for the continuing existence of an older cultural stereotype. Within the teenage sub-culture, for example, working class adolescents attribute important meanings to smoking, including the signifying of identity, the negotiation of independence from adults and the strengthening of peer group solidarity. However, for girls such meanings can be confused. A girl who smokes to establish an adult sophisticated image may be misinterpreted as promiscuous, while one who smokes to attain equality with her boyfriend may find this clashes with his view of appropriate female behaviour [29]. A double standard may operate so that for girls smoking always signifies low social status to peers, whereas for boys it does not necessarily do so [30]. Although girl smokers are not judged more severely than boys [31] it is possible that different normative standards still apply to male and female smoking behaviour. The image of the smoker remains male-oriented, despite changes in prevalence [32], and traditional ideas of femininity still inhibit smoking for some American women [25]. A study designed primarily to investigate the smoking behaviour of student teachers and nurses during training provided some further clues to the way in which women’s smoking behaviour is currently socially defined and to the types of images that smokers and non-smokers are perceived as projecting. THE STUDY

The data presented here are derived from two separate phases of a single longitudinal study conducted among groups of trainee nurses and teachers in the early stages of their professional training. Its purpose was to consider the extent to which the smoking behaviour of young women undertaking a

The social definition of women’s smoking behaviour

professional training can be explained by reference to women’s smoking behaviour in general and to what extent it relates to factors inherent in their training and future occupation. The study therefore included some consideration of smoking with reference to women’s role in society, a topic covered in the first phase described here. Because the findings were somewhat unexpected the second phase was developed to explore the issues further. The sample is small and because it is limited to two occupational groups cannot be said to represent the views of young women generally. The study should therefore be interpreted within these confines. Phase one: method The sample consisted of female student teachers enrolled for a B.Ed. degree at a college of higher education and female nurses (pupils, general students and paediatric students) in training at a school of nursing. Sixty-nine nurses (100% response) and 38 student teachers (88%), 107 individuals in all, completed a questionnaire at their entry to training and all the nurses and 36 student teachers were subsequently interviewed during their introductory course in nursing school or their first term at college. The data were analysed by the x2 test. At their entry to training 27% of the sample as a whole were smokers (occasional and regular consumers). Thirty-three percent of the nurses were smokers but only 16% of the student teachers, and there were also marked differences between the sets of nurses, 46% of the pupil nurses, 37% of the general student nurses and 11% of the paediatric students being smokers. These variations reflect occupational patterns [33,34]. The age range of the sample was between 17 and 40 although 80% were aged 17-19. Fifty-three percent were of middle class social origin (Registrar General’s [35] social classes I, II and III non-manual) and 47% of working class origin (III manual, IV and V), based on parental occupation. Phase one: jindings Parental expectations. Children’s initial concept of the attributes of the smokers and the normative definition of smoking is most likely to be derived from their family experience. The relationship between parental example and attitudes and the bahaviour of the child is not a simple one and is likely to involve a complex matrix of the adolescent’s attitudes towards his or her parents, including both conformity to and deviation from their expectations [36]. These expectations may be different for girls as opposed to boys, as indeed is the experience of adolescence as a whole [37]. In the sample those who had a brother were asked “Was your parents’ attitude the same for you as it was for your brother(s)?” Only 13 sample members (18% of those who had a brother) felt their parents had taken a stricter line with them and this was mainly based on the belief that smoking is less socially acceptable for women.

TO USshe’d say it wasn’t right for ladies to smoke but it’s o.k. for men or boys.

1271

To a lesser extent it was also seen as an example of parental protectiveness. In general greater parental strictness towards a girl was exhibited by working class parents (P c 0.05) and in those families where the father was a smoker (P c 0.1). All interviewees were asked “Do you think the fact that you are a girl made any difference to how your parents felt about your smoking?’ The majority of parents had not expressed their attitudes towards their daughter’s behaviour in terms of sex-role norms. Whatever sex I was, their principles would be the same. However, some 30% of the sample thought their parents had taken a more negative view because they were girls. This was mainly related to the social acceptability of women smoking and to the perceived image of the female smoker. yes, because it comes from this idea of it being unfeminine and I think it’s more acceptable when a lad smokes. They said it’s not ladylike, not socially done. They played hell the first time they saw me in the street. Again, negative attitudes towards female smokers were found more frequently where the father was a smoker (P < 0.05) and this was irrespective of the mother’s smoking behaviour. My father, if you go out and see a woman smoking, it doesn’t look very good. He doesn’t feel it looks nice on my mother or sister. The social perception of the female smoker. Similar themes emerged when interviewees were asked “Does the fact that a girl smokes make any difference to the sort of impression other people have of her?“, “Did it make a difference in the past?’ and “Why do you think things have changed?‘. When these questions were devised it was the researcher’s expectation that respondents would be aware of the older cultural stereotype but that their own view would correspond more to the positive image presented in advertising. When it became clear at the pilot stage that a largely unfavourable view was being described an extra question was added for those respondents who initially replied in completely negative terms, i.e. “Do you think smoking can ever give a favourable impression of a girl?“. Forty-three per cent of the sample thought that the type of impression given by a girl who smokes is a negative one, while a further 18% thought there were both positive and negative aspects to the image. Only one person answered the question in entirely positive terms. Thirty-five percent though it made no difference and 5% answered ‘don’t know’. In their replies sample members suggested that smoking was perceived socially as a masculine behaviour. Therefore the female smoker was regarded as unfeminine or unladylike and had various degrading attributes ascribed to her. A lot of people think it looks far worse when a girl smokes. If I see a girl walking around with a cigarette I think it looks worse It’s not a feminine thing to do, it’s the sort of thing you expect men to do. It’s like a lot of things we do, it’s not so socially acceptable for a woman to do it. It’s dirty for a woman to do it. It just looks worse on women, it looks common. It looks as if the

1272

ANDREA KNOPF ELKIND

person’s cheap, a bit like Hilda Ogden. You can imagine them with their curlers cleaning the step with a fag hanging out of their mouth. Also involved in this social aesthetic aspect of smoking.

unacceptability

is the

Women put over an impression of beauty, health and cleanliness. Smoking seems dirty and doesn’t fall into the right category for a woman. A younger version ‘tough girl’.

of the negative

image

was the

When I was at school it was always the extrovert type who smoked. The ones that wanted to be older. It gave them extra confidence. It fitted in with their image, tough.. . I thought girls who smoked were rough, but now I meet girls who smoke and I see it’s not so. I meet a wider circle of friends. You get to know them as people. How the female smoker is seen might thus be dependent on her age, but also on the onlooker’s age or smoking habits. Sample members who mentioned these views did not necessarily subscribe to them themselves, rather they were aware they were current in society. Some regarded them as old-fashioned, the views of an older generation, or felt that such an image represented a male point of view.

. . It’s just an attitude you get among your relatives. It matters less now. To me it doesn’t but to my boyfriend, if he saw a girl who smokes it would turn him right off. Some lads, they think it looks common for a girl to smoke, but to me it doesn’t make any difference. I wouldn’t alter my views of a person just because they smoked. However, no clear consensus emerged as to whether boys found the female smoker attractive or not. Q. Can smoking give a favourable image? A. To the right person.. . If they think that’s the type of person they’d like to go out with. Q. What type is that? A. If a boy smoked and a girl smoked he’d think they were the same type of person. If you don’t smoke he might think you’re a bit of a stick-in-the-mud. At school it was like that. The clever ones didn’t smoke and the not-so-clever ones did, so it was in two groups. Among those who saw the female smoker in a more positive light were some who saw it in terms of a more extrovert personality.

. . . it can be read into it a person’s more outgoing, it seems to relate to shyness and quietness with a non-smoker, the little goody girl image, not smoking. Q. What about the image smoking gives? A. Adventurous, outgoing, modern. For some it meant ‘being cool’ while others attributed elegance and sophistication to the smoker. A. I think it can be ladylike. It can be elegant. Women can smoke in an elegant way. Q. How do you mean, elegant? A. If YOUsee these Hollywood stars, Bette Davis, she could smokk elegantly. I asked my friends’ mothers when did you start smoking and they say when Bette Davis smoked. Those who felt that whether or not a girl smoked made no difference to the impression people would have of her tended to reject the idea of labelling smokers in this way or felt that with equality of the sexes or the extent of smoking among women such

views were irrelevant, smoking is simply a social habit. You’re equal these days, it doesn’t make any difference what they do. Quite a few of my friends smoke, it doesn’t affect me, it doesn’t stand out, it’s just something they do. Expressing either positive or negative views about the female smoker was not related to personal smoking behaviour, although among those who replied in both positive and negative terms the non-smokers tended to emphasize the negative. What of the past? Two-thirds of the sample thought that in the past a female smoker would have been seen in more negative terms than now, 11% though she would have been regarded more positively, 2% felt her image would have had both positive or negative elements and 21% thought it made no difference or did not know. The negative cultural stereotype of the female smoker was perceived by sample members to have been more in evidence in the past, and for some this was regarded as an aspect of a society that was male dominated. Because of the way women were treated in the past, and a lot of these ideas are still passed down. Only now women are feeling less inhibited. This idea I have of not liking to see women smoke hasn’t come from myself, it’s from outside. The idea that ladies have to behave just so is a male idea. It’s always been accepted that men smoke. There was a stigma attached to girls who smoked. Q. What stigma? A. It just wasn’t accepted that a conventional woman would smoke, a woman who should be at home looking after a family . . Change in attitudes towards women smoking, in terms of greater social acceptability, was related by sample members to greater equality between men and women, changes in women’s roles and attitudes and to more liberal social attitudes generally. Things have changed because of equal rights and women’s lib. Everything’s changed. Men and women are more equal. It’s accepted now. Everybody does what they want to do. It’s a lot less important what people think of you. Some sample glamour in the had been more feel this change of smoking as

members connected smoking with past. In general, those who thought it acceptable then than now tended to was related to the current perception anti-social or unhealthy.

I think less so in the past because more people used to smoke then, girls and boys. It was the accepted thing then. Thinking when my mum and dad were about sixteen they said everybody smoked then. They thought it was good for your health then, it was very popular, kept your weight

down. Q. Why do you think things have changed? A. People realise the dangers, less young people are smoking.

Smokers were more likely than non-smokers to think the female smokers had been seen more negatively in the past than now (P < 0.1). Phase one: discussion These findings, given the limitations of a small and unrepresentative sample, seem to demonstrate a

The social definition of women’s smoking behaviour specifically female dimension to the social definition of smoking behaviour and indicate that a negative image of the female smoker is more in evidence than might be expected. They suggest that a double standard still applies to this aspect of women’s behaviour. When the role of parental expectations of appropriate female behaviour was examined it was noted that the majority of parents had not expressed their attitudes towards their daughter’s smoking behaviour in terms of sex-role norms. Among those who had however this was largely with reference to the cultural stereotype which sees smoking as male behaviour and hence unacceptable for women. In terms of this stereotype women who smoke are liable to be labelled as having unfeminine or degrading attributes. Such attitudes appeared to occur most strongly in families where the father was a smoker, suggesting that it is in such households, irrespective of the mother’s behaviour, that smoking is most likely to be indentified as a male behaviour. There was also some indication that such views were class related, being stronger in working class families. In broader terms more than half the sample perceived the negative stereotypic view to be operating in society in general and two-thirds recognised that it had existed in the past. How important is it in influencing the contemporary behaviour of young women? As the questions moved away from personal relationships-parents’ expectations in relation to siblings, parents’ general expectations of female behaviour, the contemporary view of the female smoker, the perceptions of the past-so negative references increased. This suggests then that the stereotype operates more strongly at the general background level rather than among the more direct influences on behaviour. Certainly, for many members of the sample it was clear that such views were regarded as out-dated, the opinion of an older generation and not those relevant at the peer group level. In general there was little evidence that in the sample as a whole awareness of the stereotypic view was related to behaviour. Nevertheless while some distanced themselves from the stereotype even though acknowledging its existence, it was clear from the interviews that for some non-smokers the negative social perception of the female smoker represented their personal opinion and reinforced their own behaviour. The aesthetic aspect of smoking is one element in the view that it is unacceptable for women. Being found attractive by others is especially valued by young women although most smokers are unconvinced that giving up smoking would affect their attractiveness [32]. What represents attractiveness is in the eye of the beholder and sample members came to no clear conclusion about whether boys found the girl smoker attractive or not. A recurrent theme in the discussion of the perception of smoking was the way in which women’s behaviour is defined in relation to men. Thus in the past in a more male-dominated society smoking was socially unacceptable for women. Even in contemporary society however smoking continues to be defined as a male-oriented behaviour, acceptable and even expected for men; unfeminine, unladylike or worse for women. Moreover, this definition is inter-

1273

preted as a male point of view and such a view is perceived as operating at the peer group level. Although the study did not investigate the issue it may be thought unlikely that men would define their smoking behaviour in relation to women in this way. Despite the pejorative interpretation that is still attached to smoking by women, the behaviour is clearly much less likely to attract social disapproval than in the past. This move in social perception away from unacceptability in the past to greater acceptability in the present was largely interpreted by the sample in terms of greater equality between men and women, change in women’s roles in society, change in attitudes of and towards women or more liberal social attitudes generally. Such a view indirectly supports the definition of female smoking in terms of the ‘independent woman’ often suggested by advertising and social commentators. It is of interest that the increasing acceptability of smoking for women has occurred in a period in which smoking itself, as a social behaviour, has become less acceptable, increasingly stigmatised and regulated [l, 231. Certainly it is no longer ‘natural’ [3] although whether or not it is perceived as deviant is more open to question [23,32]. Those sample members who saw women’s smoking as less acceptable now than in the past did not interpret this in relation to sex-role norms but in terms of issues such as health and social nuisance. What then of the more favourable definition of female smoking behaviour? When the study was devised the expectation was that a positive image of smoking among women would emerge and in particular that this would reflect the messages of advertising. American teenage girl smokers perceive the kinds of people identified in cigarette advertising as attractive, well-dressed, enjoying themselves, sexy, young and healthy [19]. In Britain where greater controls operate the advertiser must rely on implicit messages [27] particularly evident in the life-style approach. Thus brands aimed at women, such as ‘Kim’ or ‘More’, are often seen featured with cocktails, suggesting elegance and sophistication. Such a dimension did emerge from the positive view of female smoking presented by the sample as did evidence that smoking was seen to reflect an extrovert personality. Moreover, implicit in the idea that smoking has become acceptable for women because of social change is the view that smoking symbolizes independence and liberation. Nevertheless, this more positive view was not widely expressed, although it was given greater emphasis by smokers. Non-smokers tended to acknowledge it in passing whereas smokers described it in greater detail. However, in a sense those who rejected altogether the concept of a specific image for the female smoker can also be put on the ‘favourable’ side of the account for underlying their opinion was the view that smoking is equally acceptable for men and women and that it was a normal behaviour for women. Phase two: method

It is possible, however, that the tendency to depict the female smoker in unfavourable terms may have been related to the context of the questions, which

ANDREA KNOPF ELKIND

1214

into a later stage of the project. This was a questionnaire completed by 66 nurses and 28 student teachers about 8-10 months after they began their training. Under the headings ‘girls who smoke’ and ‘girls who don’t smoke’ participants were given the identical list of 34 descriptive words in alphabetical order and asked to tick as many as they thought applied in each case. Most of these had been used or implied during the interviews to describe female smokers and non-smokers, both favourably and unfavourably. In addition a few terms were added for balance and six non-related words were included. The full list can be found in Table 1. Underlying the selection of the 34 words was not a wish to devise a list of opposites but an attempt to make use of the diversity suggested by the interviews.

Table 1. Percentage of sample selecting words to describe girls who smoke and girls who do not smoke Girls who smoke (%)

Girls who do not smoke

(%)

Words selected by at least 30% of respondents

Unladylike common Liberated Rough Cheap

Feminine Intelligent Attractive Confident Well-behaved

51 42 33 32 31

Words selected by less than 30% of respondents

ScIUffY Tough Grown up Confident Outgoing Adventurous Cool Self-righteous Sophisticated Ugly Attractive Plain Intelligent Studious Elegant Quiet Well-behaved Pretty Short Stick-in-the-mud Tall Old-fashioned Shy Feminine Neat Nice sweet Beautiful Prim

Sample size Mean selections per person Mean selections per word

55 51 31 31 30

27 Nice 23 Neat 21 Elegant 20 Pretty 20 Sophisticated 16 Grown up 16 Quiet 13 Beautiful 12 Cool 11 Sweet 9 Old-fashioned 9 Prim 7 Shy 7 Liberated 6 Studious 6 Outgoing 6 Adventurous 5 Plain 5 Self-righteous 5 Cheap 5 Stick-in-the-mud 4 Common 4 Short 3 Unladylike 3 Rough 3 Scruffy 3 Tall 2 Tough 2 Ugly n = 94 n

25 23 22 22 21 19 17 16 15 15 14 12 12 11 11 10 9 9 7 5 5 4 4 4 3 3 3 3 3 = 94

4.7

5.2

13.1

14.3

Phase two: jindings

The findings of the second phase confirm and develop some aspects of the first phase but fail to substantiate other elements. The frequency with which each word was chosen is shown in Table 1. When an arbitrary line is drawn at words selected by 30% or more of respondents it is clear that two quite distinct sets of characteristics are delineated for ‘girls who smoke’ and ‘girls who don’t smoke’. ‘Girls who smoke’ were most frequently described as ‘unladylike’ (55x), ‘common’ (51%), ‘rough’ (31x), ‘liberated’ (31%) and ‘cheap’ (30%). Of these, only ‘liberated’ (11%) was selected by more than 5% of the respondents to describe ‘girls who don’t smoke’. They in turn were most often described as ‘feminine’ (5 1%) ‘intelligent’ (42%), ‘attractive’ (33x), ‘confident’ (32%) and ‘well-behaved’ (31%). Of these only ‘confident’ (20%) is used with any frequency to describe smokers. The expected ‘elegant/sophisticated’ dimension of the perception of smokers failed to appear, and in fact these words were used more often to describe non-smokers. When the sample is divided into those who were smokers or not at the time of this phase of the survey (Table 2) a rather more complex picture emerges, the x2 test being used to compare the groups. Negative characteristics were attributed to the girl who smokes significantly more frequently by those who were not themselves smokers. They selected ‘unladylike’ (62%) and ‘common’ (59%) much more often than smokers (36% and 28% respectively), and they chose ‘rough’

came immediately after those about parental attitudes, or to the sample members’ beliefs about the interviewer’s own expectations. Therefore the next phase of the enquiry was intended to measure the extent to which these negative images were held in the sample, to explore further the possibility of a more positive image and to examine the way the female non-smoker was perceived. Because the larger study of nurses and student teachers was longitudinal in design it was possible to incorporate the next phase

Table 2. Words most freauentlv selected bv smokers and non-smokers Smokers (n = 25)

Not smokers (n = 69)

Girls who smoke

Liberated* Unladylike

(56%) (36%)

Unladylike? Comment Rough? Cheapt Scruffy

(62%) (59%) (38%) (36%) (30%)

Girls who do not smoke

Attractive

(32%)

Feminine? Intelligent Attractive Confident Well-behaved

(59%) (46%) (33%) (33%) \ (33%) ,“,

*Selected significantly more frequently by smokers than non-smokers, using corrected x2 test, P < 0.05. tSelected significantiy more frequently by non-smokers than smokers. Other words may have been selected to a significantly different extent by sub-groups of the sample but this series of tables is restricted to those selected by at least 30% of a sub-group.

1275

The social definition of women’s smoking behaviour

during the interviews, that labelling smokers in relation to sex-role norms or their personal attributes had no relevance.

and ‘cheap’ three times more frequently. Smokers, by contrast, saw the primary characteristic of the ‘girl who smokes’ as ‘liberated’, some 56% making this choice compared with 22% of non-smokers. Smokers did not select ‘elegant’ or ‘sophisticated’ to describe the female smoker any more frequently than nonsmokers nor were they any more likely to choose words suggesting extroversion, such as ‘adventurous’, ‘outgoing’ or ‘confident’. ‘Girls who don’t smoke’ were described more often as ‘feminine’ by non-smokers (59%) than by smokers (28%) and the former also selected ‘intelligent’ more frequently than the latter (46% compared with 28%) although this difference was not significant. Among smokers only one word was used by more than 30% of the sample to describe ‘girls who don’t smoke’-‘attractive’ (32%). Table 3 shows that sample members of middle class and working class origin tended to agree about the negative attributes of ‘girls who smoke’ but working class individuals were more likely to see the female smoker as ‘liberated’ (44%) than those with a middle class background (18%). About four in ten of both groups saw ‘girls who don’t smoke’ as ‘intelligent’ but for the working class sample members the most distinctive attribute was ‘feminine’ (64%) which they selected significantly more often than middle class respondents (39%). When both smoking and class are taken into account these contrasts become even sharper, although the numbers in each sub-group are very small. Of the 14 working class smokers 79% selected ‘liberated as a characteristic of ‘girls who smoke’, compared with only 28% of middle class smokers (n = 1 1), 29% of working class non-smokers (n = 31) and 15% of middle class non-smokers (n = 38). Middle-class smokers were the least likely to subscribe to the view that the girl who smokes is ‘common’ or ‘unladylike’. In relation to ‘girls who don’t smoke’, 74% of working class non-smokers selected ‘feminine’ compared with 47% of middleclass non-smokers, 43% of working class smokers and 9% of middle class smokers. This last group were also the least likely to subscribe to the view that ‘girls who don’t smoke’ are ‘intelligent’. Finally it should be noted that 17% of the sample ticked no words at all in relation to ‘girls who smoke’ and 20% failed to do so in relation to ‘girls who don’t smoke’. This response is probably best interpreted in terms of the view expressed by some sample members

Phase two: discussion

These findings suggest that smokers and nonsmokers have divergent views about the image of the female smoker. The non-smoker’s view is based on the older cultural stereotype which sees smoking as a male-oriented behaviour and hence unacceptable for women, whereas the smoker’s view corresponds to the perspective that sees women’s smoking as symbolic of social change and greater independence. Moreover, while the non-smoker has a clear and positive image of the female non-smoker as feminine, for the smoker she lacks a distinctive identity. The contrast between the smoker’s image of the female smoker as ‘liberated’ and the non-smoker’s image of the female no+smoker as ‘feminine’ was very marked among respondents with a working class social origin, reflecting the association between class and the interpretation of behaviour noted in the first phase. Smoking is more prevalent among women in the manual rather than non-manual socio-economic groups [ 161and the connection between smoking and class probably represents the effects of a whole complex of sub-cultural differences in behaviour and outlook [12]. This contrast in the interpretation of smoking in relation to a concept of traditional femininity on the one hand and of a definition of the independent woman on the other may express a fundamental split in social attitudes relevant to other areas of behaviour. Although the study did not include the participants’ self-image these findings may be interpreted in relation to the view that smoking is adopted in part because it confirms an already existing set of self-relevant beliefs (although those beliefs are not necessarily always positive ones) [381. The question then arisesdo such beliefs also sustain the behaviour once adopted? Jacobson [24] has rejected the view that smoking symbolizes emancipation for women but this study suggests that for some female smokers at least smoking is associated with the concept of the independent woman. Smoking may become an integral part of the smoker’s self-concept and for some become part of their role-defining behaviour [39]. Hence the meaning smoking has for young women may be relevant both

3. Most frequently

of middle class and working Middle class (n = 49)

Girls who smoke

Unladylike Common Liberated* Cheap ROUgh

(58%)

Scruffy Rough

(53%) (51%) (33%) (31%)

Intelligent Feminine

(43%) (39%)

Feminine* Confident Intelligent Well-behaved Attractive Neat

(64%) (4%) (4%) (40%) (38%) (33%)

Unladylike

Common

Girls who do not smoke

*Selected significantly

Working class (n = 45)

more frequently

by working

class respondents.

K-E (3104,) (31%)

1276

ANDREA KNOPF ELKIND

to the initiation and maintenance of the smoking habit. This phase of the study failed to confnm the definitions developed from the first in terms of elegance and sophistication or an extrovert personality, although such images have been noted elsewhere [40,41]. Clearly a study of this nature is dependent on the particular words selected for inclusion, different images will emerge if different words are used. However, the majority of the terms incorporated here were based on the opinions expressed by sample members. In retrospect it might have been useful to include reference to those concepts of women’s smoking behaviour that relate to aesthetics, stress and weight control [24]. Another issue arises with reference to the use of the term ‘girls’: would a different set of responses have been obtained if the term ‘women’ had been used? The first phase suggested the possibility that the younger female smoker appeared to be more negatively perceived and it5would be useful to examine the role of age differences>!in this respect. Studies of children and adolescents [3 1,38,42] do not suggest any markedt,differences between the sexes in their image of the smoker. Indeed, it has been suggested that the social impact of smoking on the adolescent’s social image is the same for both sexes [43]. This lack of differentiation between the sexes may have been related to the age of the groups studied so that they tended to perceive smoking among their contemporaries as a male phenomenon, or to the type of image available for selection. This study however suggests that distinctively female characteristics are ascribed to smokers and non-smokers. The most frequently selected words-‘liberated’, ‘unladylike’, ‘feminine’, all belong in the female dimension. However, given the limitations of the study, these findings and those of the first phase, can be interpreted only as a preliminary foray into this field, needing to be confirmed before any generalisations can be made. What is required now is a larger study of a sample more generally representative of young women, and one which incorporates a multivariate approach taking account of dimensions such as smoking behaviour, parental attitudes, social origin, occupations and other issues which underly both smoking behaviour itself and the development of social attitudes. CONCLUDING

REMARKS

In relation to smoking the medical and commercial interests compete in their attempts to convince smokers and potential smokers of the validity of their contrasting arguments. How have they responded to the changes in values and roles that have affected women? In Chapman and Eggar’s [44] analysis of the decoding of advertising they suggest that advertisements address themselves to the negative qualities of the product and its users by promising positive aspects which will be made salient in relation to a cultural myth. Among the many contradictions about smoking that advertising seeks to resolve, this study suggests that it is not only promising positive gains to the female image in terms of cultural myths such

as the ‘independent woman’, it is also addressing the negative cultural stereotype that condemns smoking as unacceptable for women. One of the explanations suggested for the much later and slower rate of decline of smoking among women compared with men in the post-war period [lo, 161 is the failure of health education to respond to women’s needs. In their analysis of behaviour for 1972 the authors of the General Household Survey [45] suggested that health education campaigns up till that time had been aimed primarily at men. Subsequently when women were featured the approach was essentially sexist, women being exhorted not to smoke, not for their own sake, but for the sake of others. Thus women were appealed to as girl-friends, wives and mothers and the messages often depended on generating guilt and anxiety [24]. Proponents of this view suggest that strategies against smoking had evolved as a consequence of the ‘Male Domino Effect’. The early medical research was on men and, in terms of numbers, lung cancer was then primarily a male problem. Health education was devised by men for men and failed to convince women [46]. It should be noted, however, that it has also been argued that health education has been successful in interrupting the diffusion of the habit among women [47]. What implications do the divergent definitions of smoking behaviour suggested by this study have for health education? Should, for example, the message to women be designed to emphasise the negative cultural stereotype? Such negative images may reinforce non-smokers’ behaviour but the evidence presented here suggests such a view would be rejected by smokers as out-dated, representing male values, or irrelevant at best, offensive at worst. Insofar as these issues have been addressed in health education the approach has been to emphasise other negative qualities of the smoker-the ‘nobody wants to kiss a girl who tastes like an ashtray’ type of message. However, studies suggest that while female smokers are aware of the aesthetic drawbacks they see these in less negative terms than non-smokers and they are not deterred because of them [48,49]. Moreover, Chapman and Eggar [44] argue that health education cannot succeed by attacking images perpetrated by cigarette promotors. Rather, what is needed is the development of a positive identity for the non-smoker by the establishme,nt of an image structure not presently in evidence. 1 Figure 1 outlines the most salient features of the image structure this study suggests. Non-smokers do have an image of girls who do not smoke, seeing them as primarily feminine and to a lesser extent intelligent. The smoker by contrast fails to identify any clear image in relation to the non-smoker. The task both for future research and for health education is to address themselves to filling in the missing segment

Girls who smoke Girls who do not smoke

Fig. 1. The female dimension of smoking.

The social definition of women’s smoking behaviour

and thereby developing a positive image of nonsmoking to which both the female smoker and nonsmoker could respond. Such an approach meshes well with thinking in health education seeking the promotion of a positive image of non-smoking in a more general sense [50]. Despite the major changes in smoking behaviour that have occurred during the twentieth century this study supports the view [25] that traditional concepts of femininity continue to inhibit smoking among some women. For other women, however, smoking is perceived in terms of the self-determination of behaviour [2]. Underlying this contrast is a wider question about the determinants of the values relating to the perception of appropriate female behaviour in general. Acknowledgements-My thanks are due to the staff of the institutions involved in the study and especially to the nurses and student teachers who took part. The study was funded by the DHSS.

REFERENCES 1. Nuehring E. and Markel G. E. Nicotine and norms: the re-emergence of a deviant behaviour. Social Problems 21, 513-526,

1974.

2. Grim E. R. The female smoker: research and intervention targets. In Psychological Aspects of Cancer (Edited bv Cohen J.. Cullen J. W. and Martin C. R.). Raven Press, New York, 1982. 3. Jeger L. The social implications. In Common Sense about Smoking (Edited by Fletcher C. M., Cole H., Jeger L and Wood C.). Penguin Books, London, 1963. 4. Mullen C. Cigarette Pack Art. Hamlyn, London, 1979. 5. Marwick A. Britain in the Century of Total War: War, Peace and Social Change 19#-1967. Bodley Head, London, 1968. 6. Thomson D. England in the Twentieth Century (1914-63). Penguin Books, Harmonsworth, 1965. 7. Alford B. W. E. W. D. & H. 0. Wills and the Development of the U.K. Tobacco Industry 17861965. Methuen, London, 1973. 8. Charlton A. Galsworthy’s image of smoking in the Forsyte Chronicles. Sot. Sci. Med. 15A, 633-638, 1981. 9. Graves R. and Hodge A. The Long Weekend: a Social History of Great Britain 1918-1939. Norton, New York, 1963. 10. Lee P. N. Statistics of Smoking in the United Kingdom. Research Paper 1, 7th Edition, Tobacco Research Council, London, 1976. 11. Bradstock L. and Condon J. The Modern Woman: Beauty, Physical Culture, Hygiene. Associated Newspapers Ltd, 1936. 12. McKennell A. C. and Thomas R. K. Adults’ and Adolescents’ Smoking Habits and Attitudes. HMSO, London, 1967. 13. Jacobson B. Warning: smoking chokes your anger. Spare Rib Issue 109, August, 1981. 14. Royal College of Physicians. Smoking and Health. Pitman Medical, London, 1962. 15. Cape11 P. J. Trends in cigarette smoking in the United Kingdom. Health Tren& 10, 49-54, 1978. 16. office of Population Censuses and Surveys (1983) OPCS Monitor, GHS 8313. 17. Bynner J. M. The Young Smoker. HMSO, London, 1969. 18. Dobbs J. and Marsh A. Smoking among Secondary School Children. HMSO, London, 1983.

1277

19. Clark R. Cigarette smoking among teenage girls and young women: summary of the findings of a survey conducted for the American Cancer Society. In Public Education about Cancer, UICC Technical Report Series Vol. 24 (Edited by Wakefield J.). International Union Against Cancer, Geneva, 1976. 20. Zagona S. V. and Zurcher L. A. An analysis of some psychosocial variables associated with cigarette smoking in a college sample. Psycho/. Rep. 17,967-978, 1965. 21. Garret N. Smoking: now and then. Can. Nurse. 22-26, 1973. 22. Soffer A. You’ve regressed a long way, baby. Archs intern. Med. 138, 1779, 1978. 23. Markle G. E. and Troyer R. J. Smoke gets in your eyes: cigarette smoking as deviant behaviour. Social Problems 26, 611625, 1979. 24. Jacobson B. The Ladykillers: Issue. Pluto Press, London,

Why Smoking is a Feminist

1981. 25. Dicken C. Sex-role orientation and smoking. Psychol. Rep. 51, 483-489, 1982. 26. Advertising Standards Authority. The British Code of Advertising Practice, Appendix M, 1977.

27. Hoffman W. E. The impact of tobacco advertising and promotion on cigarette consumption. In The Smoking Epidemic: A Matter of World- Wide Concern (Edited by Ramstrom L. M.). Proceedings of the 4th World Conference on Smoking and Health. Stockholm. 1980. 28. Chapman S. and Fitzgerald B. ‘Brand preference and advertising recall in adolescent smokers: some implications for health promotion. Am. J. Publ. Hlth 72, 491494, 29. Murray

1982.

M. The social context of smoking during adolescence. Paper presented at the 5th World Conference on Smoking and Health, Winnipeg, Canada, 1983. 30. Newman I. M. Status configurations and cigarette smoking in a junior high school. J. Schl Hith 40,28-31,

1970. 31. Aitken P. P. Ten- to Fourteen-Year-Ola!s

and Alcohol,

Vol. III. HMSO, Edinburgh, 1980. 32. Marsh A. and Matheson J. Smoking Attitudes and Behaviour. HMSO, London, 1983. 33. OPCS. Smoking and Professional People. Office of Population Censuses and Surveys, London, 1977. 34. Hawkins L., White M. and Morris L. Smoking, stress and nurses. Nurs. Mirror, 155, No. 15, 18-22, 1982. 35. Registrar General. Classification of Occupations, 1980. 36. Wohlford P. and Giammona S. T. Personalitv and social variables related to the initiation of smoking cigarettes. J. Schl Hlth 39, 544-552, 1969. 37. Surgeon General. The Health Consequences of Smoking for Women. U.S. Department of Health and Human Services, Washington DC, 1980. 38. Chassin L., Presson C. C., Sherman S. J., Corty E. and Olshavsky R. W. Self-images and cigarette smoking in adolescence. Pers. Sot. Psychol. Bull. 7, 670-671, 1981. 39. Mausner B. and Platt E. S. Smoking: A Behavioural Analysis. Pergamon Press, Oxford, 1971. 40. Mausner B. Reoort on a smoking clinic. Am. Psvchol. 21, 251-255, 1666. 41. Newman I. M. Capturing the energy of peer pressure: insights from a longitudinal study of adolescent cigarette smoking. J. Scht HIth 54, 146-148, 1984. 42. Bewley B. R. and Bland M. The child’s image of a young smoker. Hlth Educ. J. 37, 236-241, 1978. 43. Barton J., Chassin L. and Presson C. Social image factors as motivations of smoking initiation in early and middle adolescence. Child Dev. 53, 1499-1511, 1982. 44. Chapman S. and Eggar G. Forging an identity for the non-smoker: the use of myth in health promotion. Int. J. Hlth Educ. 23, Suppl., I-16, 1980. 45. Office of Population Censuses and Surveys. The General Household Survey 1972. HMSO, London, 1975. 46. Bostock Y., Jacobson B., White P. and Seymour L.

1278

ANDREAKNOPF ELKIND

Ad-man’s bull’s_eye-researcher’s blank. The Health Services 9 July, 1982. 47. Gritz E. The challenge to public education: a multi-national perspective. Paper presented at the 5th World Conference on Smoking and Health, Winnipeg, Canada, 1983. 48. Fishbein M. Social psychological analysis of smoking

behaviour. In Social Psychology and Behavioural Meditine (Edited by Eiser J. R.) Wiley, New York, 1982. 49. Loken B. Heavy smokers’, light smokers’ and nonsmokers’ beliefs about cigarette smoking. J. appl. Psychol. 67, 616-622, 1982. 50. Leathar D. S. Images in health education advertising. Hlth Educ. J. 39, 123-128, 1980.