‘Please don’t put the whole dang thing out there!’: A discursive analysis of internet discussions around infant feeding

The promotion of breastfeeding is an important focus of intervention for professionals working to improve infant health outcomes. Literature in this area focuses largely on ‘choices’ and ‘barriers to breastfeeding’. It is our argument, however, that women's cultural context plays a key role in infant feeding ‘choices’. In this article, we explore contested representations of infant feeding and infant feeding choices in public debates conducted on a large British parenting website. To sample dominant representations of infant feeding circulating in UK culture, two threads were chosen from the debating board of a busy online parenting community (105 and 99 individual posts, respectively). Participants on the threads were largely women. A feminist informed Foucauldian discourse analysis was used to deconstruct the intersecting constructions of gender, childhood and motherhood implicit in public discussions about infant feeding choices. We identify dominant constructions of women who breastfeed or bottle feed, social representations of both forms of infant feeding, and explore the relationship between constructions of infant feeding choices and constructions of ‘good’ or ‘bad’ motherhood. This analysis functions to trouble the individualist assumptions underpinning the notion of infant feeding ‘choices’, considering the cultural context within which British mothers ‘choose’ how to feed their babies.


Introduction
McBride Henry (2010) suggests that the culture within which women are located will have a significant influence on their choices around infant feeding, and their experiences of breastfeeding. Despite this important insight, public discourse around infant feeding operates within an individualist frame of reference, focusing predominantly on health promotion (e.g. focused on infant feeding 'choices' and on 'barriers' to breastfeeding). Within such discourses, responsibility for breastfeeding is located primarily with the mother, and midwives, health visitors and other health professionals are exhorted to convince individual women that 'breast is best'. In this article we want to trouble the individualist assumptions underpinning infant feeding, exploring how these apparent 'choices' are represented and constituted socially, and considering the cultural context within which British mothers 'choose' how to feed their babies. We explore the social representations of breastfeeding within the British cultural context, arguing that this will give a more nuanced understanding of the complex and subtle pressures which shape women's experiences of, and 'choices' around, breastfeeding. We consider how social representations of infant feeding are constituted in a more contested and public domain. To this end, we selected breastfeeding debates extracted from the community discussion section of a popular UK parenting website.

Representations of infant feeding choices in academic and health promotion literature
The health benefits of breastfeeding for both mother and child have been well established (e.g. Fewtrell, 2004;Kramer et al., 2001;Labbok, 2001;Wilson et al., 1998). Current World Health Organisation guidance recommends exclusive breastfeeding until the infant is six months old, followed by provision of breast milk until at least the age of two (WHO, 2003), while in the UK, the National Service Framework for Children, Young People and Maternity Services (2004) actively prioritises the promotion of breastfeeding and suggests 'Breastfeeding is the best nutrition for infants and is associated with better health outcomes for the mother and her infant' (p. 35). These medical and official expert discourses position breastfeeding as the 'optimal choice', lending the practice of breastfeeding a sense of legitimacy as an aspect of being a 'good mother' (Marshall, Godfrey, & Renfrew, 2007;Murphy, 1999).
Much literature around how parents decide to feed their infants tends to draw on a liberal humanist discourse of feeding 'choices'. The deployment of the notion of 'choice' here bears clear echoes of the concerns expressed by McRobbie (2008) that 'the individual is compelled to be the kind of subject who can make the right choices ' (p. 19). Parents are seen as agents, making (informed or less informed; constrained or relatively free) choices to breastfeed, bottle feed or mix feed their children. Academic breastfeeding literature tends not to problematise this emphasis on choice and agency: it may explore factors that influence these choices, or perhaps explore the difficulties parents experience in keeping to the choice that they have made (e.g. problems experienced in breastfeeding). However, when exploring the factors that influence choice, the literature focuses largely on how to increase the uptake of breastfeeding as a healthier option for mothers and for infants (e.g. Witters-Green, 2003) Much of this literature focuses on the application of social cognitive models to an understanding of how women make decisions around infant feeding (e.g. Giles et al., 2007;Hamilton, Daniels, White, Murray, & Walsh, 2011;Swanson & Power, 2009). For example, McMillan et al. (2009) used the theory of planned behaviour to examine women's 'feeding intentions'. They explored the breastfeeding beliefs that underpinned women's feeding intentions and decisions, with the explicit aim of intervening to increase breastfeeding among women, particularly in socio-economically deprived areas. Their paper identifies three categories of women: 'inclined actors', who intended to breastfeed and did, 'inclined abstainers', who intended to breastfeed but did not, and disinclined abstainers, who did not intend to breastfeed and who did not. The authors concluded that interventions targeted at the underpinning beliefs that appeared to produce these 'categories of woman' could be effective in increasing uptake of breastfeeding. Interesting in this particular paper is the identification of women as 'actors' versus 'abstainers', and as 'inclined' versus 'disinclined'. This language explicitly frames women as 'choosing' actively to breastfeed, but not as making a parallel choice to formula feed. The woman who bottle feeds is seen here as a passive abstainer. This construction ties in with notions of rational agency, that explicitly associate being an 'active chooser' as more positive and healthier than being a passive person who, in the language of 'abstention', fails to choose. In this manner, psychological literature around breastfeeding, with its focus on manipulating both attitudes and practice, reproduces representations of breastfeeding as a 'right choice' for 'good mothers'. The language around infant feeding buys into the rhetoric of women as consumers, 'choosing' how they feed their infants, but nonetheless references to actors and abstainers, to 'increasing breastfeeding uptake' implicitly constrain supposed freedom to choose for women, through the identification of certain choices as implicitly wrong.
This article is typical of social cognitive approaches to an understanding of infant feeding decisions, clearly problematic in their location of 'decision making' as an individual, agentic activity. Social factors may be taken into account superficially as a part of the individual decision making process, but they do not engage sufficiently with the complex nexus of discursive positionings (of motherhood, good mothering, femininity, working womanhood, etc) within which women form their decisions and experiences around breastfeeding.
In contrast to this highly individualised social cognitive framework, we suggest breastfeeding as a practice is probably best understood as one that is historically and culturally embedded: breastfeeding advice and practices vary cross culturally (e.g. Maher, 1992;Vincent, 1999) and historically (Henschel & Inch, 1996). In building this argument we draw on a feminist informed analysis of breastfeeding literature, rather than relying on the more social cognitive orientation of most health psychological work. Marshall et al. (2007) suggest that breastfeeding is a part of the whole relationship between a mother and baby, couched within the transition to motherhood, and that it cannot be understood in isolation from the person's life and social context. Women must negotiate representations that conflate breastfeeding with 'good mothering' (Carter, 1995;Marshall et al., 2007;Shaw, 2004) -a construction that invokes a 'moral minefield' drawing on professional and social discourses of what it means to be a good mother. Women navigate multiple and contradictory socio-cultural and discursive practices to negotiate decisions around and experiences of infant feeding. For example, breastfeeding is often represented as 'natural', a construction that implicitly suggests that, in normal circumstances, it should simply happen as an expression of good, natural mothering (Wall, 2001). Further, Bowes and Domokos (1998), in their analysis of the experiences of Pakistani and white women suggest that breastfeeding is best understood as a 'negotiated project', heavily influenced by the power hierarchy of the hospital, by the social isolation experienced by many in the home environment during early experiences of motherhood, and shaped by class and ethnic concerns. To breastfeed 'successfully', women must deal with complex representations of the public and the private, negotiating the demand that good mothers 'should' breastfeed with the public taboo that continues to operate around breastfeeding in public (Johnson, Williamson, Lyttle, & Leeming, 2009;Stearns, 1999). Sheehan, Schmied, and Barclay (2009) conducted a constructionist grounded theory study, exploring women's perspectives on how they made infant feeding decisions, and found that few women transparently accepted the view that 'breast was best'. Their account suggested that women actively deconstructed the idea of what 'best' was. While acknowledging medical orthodoxy that breast is best, they suggested that it is not the only option available to women, and that 'everybody's best is different'. They suggested that women's experiences of feeding decisions have to be understood within the context of their post-natal experiences and socio-cultural context and are not made in isolation by individual agents. Similarly, Bartlett (2002) suggests women's breastfeeding experiences are not homogenous, and that the intersection of socio-cultural and physiological factors shape these lived embodied experiences. Kelleher (2006) suggests that this embodied experience can only be understood as an aspect of broader social experiences of embodiment and the social construction of the feminine body. To maintain the identity of 'good mothering', the breastfeeding body must be maintained as asexual. This is clearly linked to the social positioning of mothers as relatively asexual, constituted within the traditional 'Madonna whore' dichotomy, a social representation of western femininities that positions women as either good mothers, or as sexual beings, but never (or at least not comfortably) both (e.g. Bryant & Schofield, 2007).
While this broader socio-cultural context has been unpacked in some feminist literature in relation to the experiences of women making feeding choices, the focus has largely been very directly on breastfeeding itself (with alternate forms of feeding very clearly cast as 'Other' choices), and less has been said about the more general cultural milieu surrounding feeding practices. Bartlett (2000) notes that while pregnancy and images of pregnant women have become more culturally mainstreamed and even eroticised (for example, Demi Moore's naked photo shoot while pregnant), that representations of breastfeeding remain largely sequestered from cultural accounts of sexuality. Henderson's (1999) analysis of Australian news reporting on breastfeeding, suggested that the notion that 'breast is best' is a dominant ideological position, but that it is rarely acknowledged in popular representations as a pragmatic one.
Drawing on and extending this feminist literature, we explore how breastfeeding functions at the intersection of social and cultural representations of motherhood, femininity, and feminine embodiment. We suggest it is important to explore the broad range of discursive practices within which decisions around and experiences of infant feeding are constituted, particularly exploring how it is nested within moralising representations of what mothers and women should do and be. We also suggest that representations of breastfeeding must always be understood in terms of the ways such representations both produce and are produced by notions of other forms of infant feeding, like bottle and mixed feeding.

Method Aim
Our aim was to consider the discursive milieu within which women's choices around infant feeding are constituted. We wanted to explore dominant and alternate representations of breastfeeding and bottle feeding, considering how these images of infant feeding are couched within other discursive formations (for example, ideas around mothering and femininity).
Talking about breastfeeding: Online debating forums Mitra and Watts (2002) suggest that cyberspace offers a discursive realm within which people and institutions voice themselves. Tanner (2006) argues that participants in online debating forums generate public opinion and collective memories in their discussion around particular topics. In other words, debating forums offer a clear space for representations of social phenomena to be produced. They offer an ideal arena to explore 'everyday' talk (or at least, talk not explicitly generated for research purposes) about issues like infant feeding, enabling the expression of multiple, fraught and contradictory ideas and representations of the issue.
The internet parenting community has emerged in the last decade or so as an important source of support and self expression for parents, particularly new parents. As Mungham and Lazard (2009) point out, this is not always a straightforward supportive environment, however. These forums often function as a site, for example, in which contested definitions of good and bad parenting are negotiated. In this article, we chose to sample threads on breastfeeding from a busy UK based parenting website, as a context where constructions of infant feeding and representations of good and bad parenting are produced, reproduced and contested. This site has numerous discussion boards, including general discussion boards, support forums for parenting issues like breastfeeding, co-sleeping, weaning, etc., and a general debating board, in which members are able to discuss politics, social issues, parenting, etc.
To ensure an ethical sampling of internet discussion around breastfeeding, we chose our research context fairly carefully. We decided to use a debating board rather than, say, a 'help and advice' board for several reasons. We felt it was likely that more contested views around breastfeeding would be expressed on a debating board (it was anticipated that help and advice boards would be more oriented to support than to discussing the pros and cons of different methods of infant feeding). We also felt that it was more ethically sound to choose public domain threads on a board that was explicitly not about help-seeking, since help-seeking threads had been posted for support and potentially from a position of some vulnerability, while debate oriented threads more explicitly engaged in a public space with the politics of breastfeeding -and in that sense that their posting was far more in the spirit of this article.
When selecting threads for analysis, we looked for threads on which there had been substantial discussion and debate, and in which multiple views around infant feeding were expressed. We obtained the consent of board administrators to use two threads, and also asked their permission to write to individual members of the web forum whose data we wished to use, explaining who we were, detailing the quotes we wished to use and the nature of the research. Participants were largely supportive, and several participants asked us to send copies of the research when published. (For those members of the site who indicated that they did not wish to be included, quotes were removed from the analysis.) Participants on the threads largely self-identified as women (there was one selfidentified male poster, on one of the threads). The two threads selected included a total of 105 and 99 individual posts respectively. We have not attempted to analyse the responses of individual posters in this analysis, preferring instead to focus on the threads as a whole, exploring how social representations circulate within the exchanges on the thread. Quotes from the threads are used exactly as they were posted on the internet forum, and have not been adjusted for grammar, spelling, or to remove commonly used internet phrases. For readers unfamiliar with commonly used internet phrases, we have added them, and any other explanatory information needed, in double brackets.

Analysing the data
The selected debating threads were analysed using a Foucauldian informed Discourse Analysis (e.g. Graham, 2005;Parker, 2005). We systematically unpack the discursive subject and object positions within the text, exploring the connections and tensions between them, the contradictions and coherences in the text, exploring how webs of discursive formations function to constitute infant feeding, and to enrol the reader in particular versions of social reality as they read (Parker, 1999). Within the analysis, we explicitly resist the tendency to reproduce dominant representations of breast and bottle feeding through talk about women's feeding 'choices', focusing instead on how infant feeding choices are couched within discourses around femininity, 'natural' and 'nurturing' motherhood, and notions of childhood (particularly discourses like the best interest of the child.) As such we aspire to 'dissect, disrupt and render (. . .)strange' (Graham, 2005) the familiar terrain of typical public debate around infant feeding with its powerful 'breast is best' discourse, by exploring how parents are enrolled in dominant representations of feeding children, of mothering, of femininity and sexuality, how they play with, are bound within and resist such representations. Our focus is on the productive effects of various ways of talking about breastfeeding: how it enables and constrains aspects of parenting practice. As critical researchers our approach is 'to engage in struggle, to reveal and undermine what is most invisible and insidious in prevailing practices.' (Ball, 1995, p. 267)

Analysis and discussion
From the analysis, five intersecting discursive constructions are located. In Breastfeeding as 'primary' and 'natural', and the related construction, Bottle feeding as 'second best', naturalising discourses of breastfeeding are evident, conflating natural breastfeeding, womanhood and motherhood. We consider the implications of this for the social positioning of women who bottle-feed. Bottle feeding is often represented within the social representations of infant feeding as a second best choice for mothers and children. Feeding choices are seen as being constituted within class and other kinds of social positionings. In Sexualisation vs. natural functions, we explore how this construction of feeding as the natural choice for women and for mothers constructs a tension between the role of breasts as a part of natural motherhood, and their other apparent function as sex objects. We were interested in the way that sexualisation and natural bodily functions were set as being in opposition in the way women's breasts were described. Participants often suggested that breastfeeding is a natural act for women, and that breastfeeding is the primary function of breasts (not their sexual functions). The sexualisation of women's breasts, together with the idea of the 'natural feminised beauty of breastfeeding has consequences for women's choices around infant feeding and around feeding in public, as we explore in the sections Breastfeeding as a provocative act and Discretion: Mothering as public/private nexus. Participants debated quite fiercely whether breastfeeding can and should be practiced in a discrete way, with ideas around this being strongly framed by the suggestion that breasts are, after all, sexual objects, and that 'reasonable women' take steps to ensure that their breasts are not 'exposed'.

Breastfeeding as 'primary' and 'natural'
Participants in the online debates we analysed often represented breastfeeding in a taken for granted way as a natural feeding method (and by extension, as the 'best' way to feed infants). Within these online discussions around infant feeding, the 'naturalness' of breastfeeding is advanced as an indisputable argument for its adoption, with this method of infant feeding represented as the 'obvious' choice within the debates. For example, as one of the participants suggests: 'that is the reason women have breasts in the first place'. In this quote, breastfeeding is clearly represented as a natural choice for women, with the breast positioned here as a symbolic expression of natural womanhood and natural childcare. The purpose of the breast is to feed babies.
However, breastfeeding is also represented in these debates as a natural choice for infants: 'My friend sent me a picture of her new baby. This tiny little baby, with a bottle stuck in her gob. It just looked so wrong. Made me want to cry.' The participant here clearly presumes that the milk in the bottle is formula. Bottle feeding is frequently conflated with formula feeding in participants' discussion, in a manner that obscures the possibility of bottle feeding expressed milk. Having made this assumption, the participant draws a clear distinction between breast milk as the natural food substance for children, and formula milk as artificial, synthetic. This idea of artificiality functions here to produce formula feeding as a necessarily wrong choice for women. Interesting is that in both instances, what is seen as either natural or unnatural here is not the substance itself -not the milk -but the means by which the milk is transmitted to the child: the bottle or the breast. The rights and wrongs of breastfeeding are not understood as mere nutritional issues. There is something in the specific naturalness or artificiality of the breast or bottle that is seen as important. In the second quote above, the language used of the bottle 'stuck in her gob', positions the bottle itself as intrusive, disrupting the apparently natural relationship between mother and child, mediated by her breasts, by the 'maternal body'. This emphasises the breast as primarily an object for infant feeding (not a sexual object) and identifies woman as the proper nurturer of children, by virtue of having breasts.
The notion of breastfeeding as 'natural' extends into discussions about breastfeeding in public and private spaces. Barriers to breastfeeding were identified by several participants, and were understood as a failure within the cultural context to understand breastfeeding as natural. For example, as we will demonstrate later in this article, negative reactions from others when breastfeeding in public were seen as problematising breastfeeding as somehow 'wrong' for women. Such negativity is framed as a fundamental misunderstanding of the purpose of women's breasts as instruments of infant feeding. Familiarity with and exposure to breastfeeding is often suggested as a means to both challenge the stigmatisation of public feeding, and increase women's uptake of breastfeeding: 'I can totally understand why people don't breastfeed. Before I had my first child, I only remember seeing someone breastfeeding once ever . . . When I had my first child there were no babies in my extended family, I was the first of my generation to have a child. I was also the first among my friends . . . I did bf ((breastfeed)), but I was sure before my babe was born that I wouldn't like it, and asked the midwife how long I would ''have to'' breastfeed for. Breastfeeding does seem natural and wonderful once you have done and enjoyed it, but if you have little or no experience of seeing it done, it can seem rather odd and alien'.
An interesting feature of this particular extract is its focus on a discourse of 'experience and exposure'. This discourse suggests we need to be exposed to breastfeeding in order to find it normal and acceptable. It is the lack of public representations of breastfeeding, the lack of available role models and of regular public practice of breastfeeding that makes it seem 'odd and alien'. Breastfeeding needs to be lived, and lived with, in order to understand its apparently obvious naturalness and wonderfulness.
Some participants indicated that women and children had a 'right' to public breastfeeding, and used the notion of the naturalness of breastfeeding as a justification for this position, e.g. 'Breastfeeding in public should be a completely natural and unchallenged thing'. While we would in no way object to women's right to feed their infants as they wish, nonetheless, we suggest that this warranting of public breastfeeding on the grounds of its 'naturalness' functions to entrench less empowering representations of women.
One problem with this focus on the 'naturalness' of breastfeeding is its inevitable implied other term: the 'unnatural' bottle feeder. The natural, primary status of breastfeeding is built up in talk about formula feeding. Mothers who 'chose' not to breastfeed were represented in debaters talk as unfathomable to breastfeeding mothers. For example, one participant says: 'I'm more intrigued as to why some women never even attempt to breastfeed and actually decide not to prior to their first baby's birth'. In these kinds of representations, the decision not to breastfeed is positioned as a choice that is difficult to understand. By talking about deliberately not breastfeeding as 'intriguing', bottle feeding is constructed as non-normative, as something that requires explanation, not a 'natural choice'. It is clear in this extract that 'choosing' not to do the natural thing is seen as wrong choice -an unnatural choice that is constituted as antithetical to ideas of natural, nurturing motherhood. This choice it represented as necessarily deviant -so much so as to be incomprehensible to 'natural mothers'.

Bottle feeding as 'second best'
By constituting breastfeeding as the natural and primary feeding method, bottle feeding was often constituted as a second and poorer option: 'I have 2 close friends who have never attempted to breastfeed any of their children. I have never felt that I can ask them why as I'm concerned they may feel it implies criticism by asking. It did make me feel sad to see my friend's newborn son rooting at his mum and him being given a bottle of formula. Yes he is a happy healthy baby (and is very very bonny!) but I felt a pang for him and did wonder why she chose not to try breastfeeding as her first choice' The expression of sadness problematises the choice to formula feed, and the feeling of a 'pang for him' functions as an implicit commentary on the inappropriateness of that choice. In positioning the loss as the child's, the implicit sanction is that these mothers have made choices that are not 'in the best interest of the child', that they have chosen a second best, rather than a best method of feeding, and that as a consequence, their children are disadvantaged, and to be pitied.
Mothers who had chosen bottle feeding as first choice often justified their choice in the context of the constraints of their own lives and/or appeared to be moved to defend their decisions in a manner that breastfeeding mothers did not. In the following extract, we see interesting tensions in a bottle feeding mother's account of her infant feeding decisions 'I couldnt just give up all my life to breastfeed my baby it would have been stupid and unhealthy both physically and mentally. When i was pregnant with ds2 i decided from the start i wasnt breastfeeding and he was formula fed from day 1, as was ds3. I have no regrets, i did what i wanted to do with my children and its worked out ok for me. I dont mind people asking why i didnt breastfeed cos i dont feel i have to justify anything i do with my children' The participant begins with a justification of her choices (she could not breastfeed for health reasons), then goes on to describe how she never feels she has to justify how she fed her baby. The contradictory statement that she does not have to justify anything she does with her children is an interesting one, since it clearly invokes the logic of the mother as expert on her own children, functioning to protect her against the expert health discourse that suggests that 'breast is best'. Murphy (2000) suggests that bottle feeding mothers often resist being positioned as 'bad mothers' by emphasising their other, responsible actions in their care of their children. This justificatory statement emphasises the regulative social power of the 'breast is best discourse'. While it may not have pressured this mother to breastfeed, it nonetheless functions in a censorious manner that demands she justify this 'wrong choice'. Marshall et al. (2007) suggested that women do not passively receive health information about infant feeding choices. She notes that many women who start off breastfeeding and then switch to bottle feeding later position themselves as 'good mothers' by using other discursive resources -in the extract above, the participant draws on the notion of good mother as expert on her own children, using a 'mother knows best' discourse. She also explicitly defends her choice to bottle feed on health grounds, pointing out that it would be both physically and psychologically damaging for her to try. These justifications stand in contrast to the other position she holds, that 'I did what I wanted to do with my children and it's worked out ok for me'. It is clear in this extract that to simply position herself as not wanting to feed would have located her as a 'selfish' mother, making 'second best choices' for her children. Such a position would be inconsistent with dominant discourses of good, nurturing mothers who sacrifice all in the best interests of their children. Despite protestations that she does not need to justify her choices, it is nonetheless obvious that, within dominant discourses around mothering and infant feeding, she has to do a great deal of positioning work to represent herself simultaneously as a good bottle feeding mother.
Justifications for formula feeding can be found throughout these debates, often accompanied by the 'I don't have to justify my choices' tag. Defences for bottle feeding include illness, midwife's advice, and pressure from family. Mothers in this context do not routinely position themselves as 'choosing' formula feeding. Rather they locate the causes of bottle feeding as beyond their control. Bottle feeding is represented not as a poor choice, but as a non-choice. In the context of these debates around infant feeding, mothers worked hard to demonstrate unusual circumstances that are structured to defend them against the criticism of having made an 'unnatural' choice. In contrast, breastfeeding mothers were not called on to justify their choice of feeding method, unless their practices were non-normative (e.g. extended breastfeeding). (This pattern echoes several other instances of discursive practice around normative notions of femininity, e.g. Reynolds and Wetherell (2003) note that women often are asked to account for their single status in a manner that married women rarely are).
In these debates, clear distinctions were drawn around breast and bottle feeding that explicitly positioned formula feeding as a young mother's choice. In contrast, women who identified themselves as more adult, as more mature, either said they had breastfed or positioned themselves as having tried to breastfeed in the first instance: 'i know loads of women well girls mainly who have bottle fed through choice in fact i am the only person i know who has breastfed or even tried, but most of them wont because they think its embarrassing and don't want people looking at them, these are the young mums groups who are all aged between 15 and 25 (makes me laugh though most of them wouldnt think twice of flashing them about after a couple of vodkas)' Here formula feeding is delegitimised through its association with representations of the 'irresponsible' younger mother. The link between formula feeding and younger women also serves to stigmatise and further problematise the younger mother for not making the 'right' feeding method choice. Young women's discomfort with exposing breasts is mocked as illegitimate, because such young mothers are, in this extract, located as sexualised young women, ladettes out for a bit of fun, happy to expose themselves while drinking, but not to use their breasts for the more natural and feminine purpose of breastfeeding. Sexualised exposure is framed as necessarily wrong, an expression of deviant femininity, while the proper activity of breastfeeding is represented as appropriately feminine and nurturing.
Age and class interweave in these debates in complicated and exclusionary ways. Across forum discussions, class was often invoked subtlety or implicitly. For example, in the statement 'This tiny little baby, with a bottle stuck in her gob. It just looked so wrong', bottle feeding is constituted as 'common' through the use of the word 'gob'. Class is raised more explicitly in some strands of debate. For instance, 'Formula feeding is the done thing. Most mothers on our estate choose formula. A lot of them are also very young'. The term 'estate' (referring to a neighbourhood with social housing) emphasises the class location of the women described. Formula feeding is represented as a class linked fashion choice, not carefully considered, but simply done because that's how things are on the estate in question. Young mothers are positioned as not reflecting on their parenting choices, not consciously 'choosing' as middle class and mature women apparently do. Implicit in many of the contributions to the debate (including the one above) is the image of the young working class mother from a 'council estate', making a range of poor parenting choices. For example: 'you just know that she'll be the same one, fag in hand, pushing a pushchair, and the baby's got a bottle full of coke'. The association between bottle feeding, young mothering, class, and poor parenting are locked in these representations in ways that position formula feeding as the method of bad parents. 'I saw this ''woman'' walking her very small baby (3 or 4 mos) in one of those crappy lightweight ''umbrella'' pushchairs with a bottle propped up and formula just running down each side of the poor little girl's mouth (I mean thick runnels of the stuff!) straight into her neck whilst her Mother absently windowshopped. Just made me CRINGE! OMGs! ((Oh My God!)) (Makes me sicky now thinking of it!)' The placing of 'woman' in inverted commas is rhetorically interesting. The mother is described as an inadequate, disengaged, and by virtue of this, her womanhood is called into question. This is coherent with feminist accounts which have suggested that 'good motherhood' and 'good femininity' have become so conflated that being a good mother without conforming to ideas of nice femininity is rendered discursively impossible. As Gillespie (2000) suggests 'Motherhood has predominantly been perceived as natural for women, the desire for it inevitable, unquestioned and central to the constructions of normal femininity' (p. 223).
Historically, feeding 'choices' have always been shaped and constrained by class positions: this is not a new phenomenon, but the form class influence takes has shifted over time. One woman notes 'I think that goes back to my mother 's generation (1960s/1970s) where formula was seen as ''posh'' because it was scientific, while breastfeeding was what poor working class women did'. Another suggests 'Formula was also a status symbol -it meant that you had money and could afford not to breastfeed'. Expert notions of good parenting shift historically to reflect social and cultural trends, as much as real medical expertise (Burman, 2008). In the debates we have analysed, both expert and everyday representations of 'good parenting' are seen as having clear implications for 'good' infant feeding choices. The uptake of particular feeding practices are linked to ideas about class, status, and good mothering more generally. In the extracts we have considered, representations of class and age (particularly the idea of young working class mothers) function to underscore the idea of formula feeding as a second best choice.

Sexualisation versus natural functions
In online debates around breastfeeding, sexualisation of breasts is frequently represented as a barrier to breastfeeding. The sexualisation of women's breasts is implicitly problematised: if the natural function of breasts is to feed infants, then sexualisation of women's breasts must therefore be something else, something additional to their natural purpose. The sexualisation of breasts is rendered as cultural, and extraneous to the natural function of breasts, for breastfeeding. Given the construction of mothering as sexless womanhood (the Madonna whore polarisation so frequently a feature of western notions of femininity, e.g. Bryant & Schofield, 2007), the idea of the breast as both sexual and maternal is difficult to sustain. 'How do you feel about page 3 girls in the tabloids? It amuses me that in the UK we have pictures of ''celebs'' in various states of undress all over the place. Even Konnie Huq from Blue Peter ((a very wholesome children's television programme)) more or less had her norks out at the spiderman premier. But a b-feeding ((breastfeeding)) woman, who to be honest with you is hardly showing any norkage, gets funny looks and comments!' The use of the term 'norkage' (breasts) positions the sexualisation of breasts in popular culture as something humorous, rather silly, and perhaps a bit 'common'. The image of the 'page 3 girl in tabloids' is one that is simultaneously quintessentially British, but also references a particular kind of Britishness. (The 'page 3 girl' refers to the regular presence of a picture of a topless young woman, on page 3 of The Sun tabloid newspaper.) Class overtones are again in evidence in relation to the sexualisation of breasts, which is positioned here alongside the 'council estate' images referred to in the previous section. Like the earlier representation of the housing estate mother who, with her 'cheap umbrella pushchair' and her baby with its formula dripping down its neck offended the debater's sense of aesthetics and good taste, sexualised exposure of the breast is seen here as something a little bit tacky (tabloid), not really very nice. Using the term 'norkage' aligns it with ladette culture 'getting them out when you've had a few vodkas', whilst the introduction of the image of the 'page 3 girl' at the same time evokes a flavour of the Benny Hill or the Carry On Films (quintessentially British comedic TV programmes and films, characterised by a fondness for slapstick comedy, and the sexualisation of women in apparently humorous situations). Through these associations, sexualisation of breasts is positioned as somewhat outside the range of 'nice femininity', and by extension, breastfeeding is represented as proper, appropriately feminine, and rather more discreet.

Breastfeeding as a provocative act
Discretion is often represented as the key to successful public breastfeeding. For example, one debate participant suggested 'I do understand the ''please don't put the WHOLE DANG THING out there'' mindset too, though'. Necessary partial exposure that is discrete and maternal is represented as acceptable. However, the phrasing of this injunction suggests that more than this is again an aesthetic problem, and a breach of good taste. Getting the 'whole dang thing' out is positioned as provocative. We suggest this is because such 'indiscrete breastfeeding exposure' crosses boundaries between acceptable maternal public breasts, and sexualised, tacky and inappropriate ones. This distinction between exposure of the breastfeeding and the sexualised breast is maintained by other posters: 'TBH ((to be honest)) if I did then so what? It's not like it's gratuitous flashing of flesh, there's a good and valid reason for it and I'm pretty damn sure than whenever I breastfed in public then I was baring a lot less than a lot of the girls wandering round wearing their normal summer clothes . . . ' A contrast is drawn here between the 'virtuous' nudity of breastfeeding, and the more 'inappropriate' but to a degree socially accepted, mainstreamed sexualised nudity. Exploring links between breastfeeding practices and sexuality, Bartlett (2005) suggests breastfeeding and sexuality are constituted in social practices as antithetical social arenas, and that maternal sexuality is rendered inexpressable by these kinds of discursive practices, and that the idea of breastfeeding as a pleasurable activity is rendered illegitimate this way. Boundary crossing between the asexual maternal breast and the sexualised exposed breast is implicitly positioned in the debates as unacceptable to this parenting community: 'I recall recently coming home from work when this woman whipped out her breasts in front of all and sundry to feed her sleepy child. She was pretty blatant about it although no-one said anything, looking at the child sitting next to her, he looked very uncomfortable. As much as I would welcome law to not restrict breastfeeding in public, I would also welcome part of that law that states that the mother should be as modest as possible'.
The notion of 'blatant' breastfeeding is an interesting one. As Dykes (2005) suggests, the lack of visible (or 'blatant'?) breastfeeding is implicated in women's failure to learn breastfeeding within a communal context and can contribute to women's loss of confidence in their own bodies. In the extract above 'modesty' is represented as a more natural feminine position for mothers to take up. Typically the term 'blatant' is associated with negative and antisocial acts (for example, being a 'blatant liar'), so to see it associated with an apparently positive and nurturing feminine activity perhaps signals some of the ambivalence that still surrounds both the sexualised and the nurturing role of the female breast. The unexpressed discomfort of those around the breastfeeding mother is seen here as an understandable reaction, with the act of breastfeeding implicitly framed as a bit embarrassing. Interesting too here are the absent traces of masculinity and the role of the masculine observer -represented here by the uncomfortable male child. The potential discomfort of others (particularly young male others) is prioritised over the needs of mother and child. Also, while breastfeeding has repeatedly been represented as a natural choice, and as associated with good mothering, it is interesting that in this extract it is presented as problematic in public because of the risk of exposing a male child to breastfeeding. Clearly this representation blurs the boundary significantly between the sexual and nurturing aspects of breasts, and equates the act of breastfeeding to an act of self-exposure in front of a minor. The final sentence of the extract -the idea that the law should intervene to make mothers modest is particularly interesting, as it appears to equate modesty and mothering, and casts 'blatant' acts of breastfeeding in a more prurient light. In suggesting that idea of needing to legislate modest feminine behaviour, the debate participant appears to invoke ideas of the modest (implicitly middle class) feminine to regulate the expression of dangerous female sexualities. The suggestion is that by relaxing the ban on female public nudity, protective pro-breastfeeding legislation enables 'blatant' (i.e. sexualised) exposure of the breasts, a risk which needs to be controlled through further legislation. Clearly, in the view of this participant, the public (and particularly boy children) need to be protected from this kind of 'exposure'.
The male observer functions as an unarticulated presence throughout these accounts. This observation is never explicitly discussed but implicitly present in both the discussions about the sexualised breast, and the need for discretion. Women's breasts are implicitly constituted as the object of the male gaze, through their association with sexualisation, and through the sense that they should be guarded from casual observation by feminine 'discretion'. Women need to be 'discreet' in their feeding practices, because to do otherwise is to risk exposing themselves as sex objects.

Discretion: Mothering as public/private nexus
The preoccupation with the idea of modesty and discretion is found in the contributions of several participants. Some of the accounts here the tension identified by Johnson et al. (2009) between the representation breastfeeding as a moral duty for good mothers, and the discomfort and frequent censure that attend public breastfeeding. One participant suggests 'I'm going to put my neck on the line here and say the following: I absolutely hate seeing women breastfeeding in public. If they can be discreet about it, that's fine. But don't make it obvious to others'.
Breastfeeding is framed in this account as a private act, only acceptable if this shroud of privacy can be sustained. Discretion is essential. This is a particularly problematic positioning, since, as Hoddinott and Pill (1999) have suggested, lack of exposure to public breastfeeding might result in mothers lacking an embodied understanding of breastfeeding. This idea of discreet acts of infant-feeding is extended by another poster: 'Restaurants and other similar public places should have a small area for those who have taken sick, fallen unconscious or for bfding ((breastfeeding)), be it a corner in the staffroom or managers' office. I don't think much mothers would disagree with going somewhere like that especially in the situation where someone was voicing a complaint or felt offended'.
The idea of a sequestered area for those who are 'sick, unconscious or breastfeeding' is presented in this account as a possible solution to the problem of public breastfeeding. This rhetorical conflation of breastfeeding with being 'sick' or 'unconscious' implicitly pathologises breastfeeding, positioning the breastfeeding mother as someone in need of special consideration and treatment. Interesting in this extract is the sense that a complaint or a feeling of offence on the part of a member of the public is represented as perfectly legitimate here: nobody could object to going off to conduct an apparently private act in private. The sense that 'mothers wouldn't disagree' is implicitly gendered, drawing on ideas of femininity as 'being nice', not wanting to impose, and being considerate of others. The account leaves the complaint itself unchallenged, and the member of the public's outrage is seen as requiring no justification -the participant presumes consensus amongst women here ('I don't think much mothers would disagree'). Compliance is presented here as the reasonable response. This sense of women being 'reasonable' and 'appropriate' is echoed by other participants: 'That said though, I've seen women covering their child with their jacket while they bf ((breastfeed)), so the majority do use commonsense. It's the very few who literally want to over-advertise it to everyone that makes it bad for everyone else'.
The appeal to commonsense in this passage functions to suggest that 'discretion' is the norm for all sensible women. Being 'discreet' about public breastfeeding is seen as the reasonable feminine option. No 'reasonable' woman would want to be indiscreet, preferring instead to show appropriate propriety in relation to certain forms of female nudity. In this extract 'being blatant' is seen as spoiling things for 'sensible women'. However, other participants offer other foils to this argument. For example, one participant, recounting an experience where she was asked by restaurant staff to go to a quiet room, suggests that being asked to feed 'discreetly' is problematic, as it forces her to justify her choices. She says: 'I wasn't in the mood to reply to the staff members discussing it at the time but why the hell should I have to anyway?' The demand for justification for her feeding practices is seen here as violating her privacy and her rights as a mother. By forcing her to account for herself through a challenge to her right to feed her child in a restaurant, the apparently private choice to breastfeed is rendered public again, and this sense of a violation of her right to privately choose how she feeds her child is clearly expressed in her outraged statement 'why the hell should I have to (justify myself) anyway?'.
Another participant defends the right to public breastfeeding by drawing on alternate representations of nurturing femininity. Resisting the idea of 'blatant breastfeeding', the participant suggests: 'I have never seen an indiscreet feeder and contrary to some I think the sight of a woman breastfeeding is serene and peaceful.' The participant uses the image of the serene Madonna, the acceptable public face of the breastfeeding woman. The terms 'serene and peaceful' underscore the idea of the naturalness of the act of feeding a child -to see a woman this way is perfectly in keeping with the ideals of nurturing femininity. However, perhaps more intriguing in this statement is the idea that, had a woman been 'indiscreet' in her exposure of her breasts, objections to that would, of course, have been perfectly reasonable and legitimate. The aesthetically pleasing, serene and discreet breastfeeding mother is an acceptable public sight, the appropriately de-sexed Madonna. The blatant breastfeeder is not.
The tension between public and private acts, and public and private choices, threads through these accounts lending layers of complexity to the act of breastfeeding. On the one hand, how you feed your child is positioned as a private choice, and as a parent, that choice functions in the private domestic sphere, and should not be questioned. At the same time, exercising that choice in public leads to far more complex and contradictory ideas about what is and what is not acceptable. This tension is captured particularly strongly in this extract: 'I couldn't actually give a damn how others feed their babies, and am very aware that just because someone is bottle feeding their baby doesn't mean they didn't/don't breastfeed/desperatly want to. Unfortunately i've also spoken to women who feel harrassed for breastfeeding not working out despite their best efforts. All i really believe is that we should all be left alone to feed our babies as we wish and where we wish, no matter how we do it'.
A mother's feeding choice is represented here as a private one: how we choose to feed our infants is our own business and nobody else's. The participant flags up that when we see any act of feeding in a public space, we are not aware of the private back story that might inform that act. This notion of feeding as a private choice is somewhat belied by the other aspects of this statement though. The back story is presented as justificatory -essentially, she suggests that observers do not know whether mothers have tried and failed to breast, or whether they might have wanted to breastfeed and not been able to. This is implicitly framed here as a proper reason for not breastfeeding. Breastfeeding is still positioned here as the proper choice, but the participant appeals to people to respect that that choice is also a private one. At the same time, the idea of breastfeeding as the right feeding choice is, of course, produced within a context of extensive public campaigns that suggest that 'breast is best', and that 'good mothers' will of course choose what is best for their children. The interface of these public and private choices is seen in one woman's account of the protective role of the law in breastfeeding practice: 'I think it's needed, although it's a sad reflection on our society that we need legislation to make it unacceptable to stop a woman nursing wherever she is. I've breastfed each of my 3 children in lots of different public places, but never more comfortably and confidently than in Scotland, knowing that legislation had been passed there, so nobody could dare challenge me'.
The law, it is suggested, protects a woman's rights to nurse as she chooses. This public intervention (the statute) secures the public enactment of a private and personal choice: this public intervention enables women to feed without private sequestration. However, there is a palpable tension in these accounts between the need to draw on a public rights discourse, and a private choice to protect the space of breastfeeding. Mothers can legitimately breastfeed in public, but they should do so privately (i.e. they should be discreet). Mothers can draw on public resources (the law) to secure the privacy of breastfeeding, enabling her to practice this private act of breastfeeding unassailed by public expression of disapproval. This tension between public and private in social representations of breastfeeding seems to us to reflect and reproduce a particular complexity around representations of mothering itself. On the one hand, mothering is seen as something intimate and personal, an aspect of the private domestic sphere. However, mothering also always occurs in the public eye, under scrutiny -we are expected to be able to account for our private mothering decisions to a watching public.

Conclusion
The discursive formations identified in our analysis of these debates around breastfeeding are united by what we have referred to as an 'exposure discourse': the exposure of the breast, exposure of the maternal self, visible expressions of maternal impulses and of care -united in the universal-private image of the breastfeeding mother. This flags up the complex and contradictory pressures that are brought to be on women as they negotiate decision making around infant feeding. On the one hand breastfeeding is positioned as the right moral choice for good mothers through the discursive imperative that 'breast is best' (Murphy, 2000), but on the other hand public breastfeeding is often problematised in the UK (Nicoll, Thayaparan, Newell, & Rundall, 2002). The sexualisation of breasts, and the idea that breastfeeding should be both discreet and private constructs a discursive tension that positions breastfeeding as somehow morally wrong, conflating the sexiness of breasts with the de-sexed imagery of motherhood. The suggestion that breastfeeding is both the best choice, but also something that should be done in private creates a moral tension for mothers to navigate.
Debates around breastfeeding are heavily shaped by liberal humanist notions of choice, and by traditional notions that parenting operates within a private domainour choices around how we feed should be regarded as private choices. However, as we have seen in this analysis, infant feeding choices are anything but straightforwardly private. They are open to public scrutiny -through health promotion, through legislation, through representations of women and sexuality, and women as mothers. UK women's choices around infant feeding are constituted within the kinds of social representations of infant feeding, of class, of good and bad mothering, of women and sexuality, and of breasts and breastfeeding, that circulated within the debates under consideration. In particular, it is important to note that discourses of femininity, mothering, femininine sexualities etc function to enable and constrain particular parenting practices -all choices are not equal within debates around infant feeding and parenting. Choices are constituted within this complex nexus of social representations and the implications of this need to be carefully considered in public health campaigns around infant feeding. For example, understanding the way that infant feeding choices might be used as a further tool to derogate young mothers and poorer mothers might be an important insight for health professionals working with women in these circumstances. Simply reiterating their choice is not enoughtheir choices need to be understood against the backdrop of the social context within which choices are constituted.
Within these online debates, the construction of the good maternal body (Johnson et al., 2009;Stearns, 1999) and the way that it functions in relation to women's feeding practices are clearly in evidence. However to fully comprehend this social construction, it is necessary to engage with the parallel construction of its denigrated 'other'. The breastfeeding mother is constituted as good against a range of classed, sexualised and de-naturalised alternative others. To fully engage with discursive practices around breastfeeding requires that we think through in complex ways the manner in which it is constituted at the expense of other ways of performing motherhood.