Using liminality to understand mothers’ experiences of long-term breastfeeding: ‘Betwixt and between’, and ‘matter out of place’

Breastmilk is widely considered as the optimum nutrition source for babies and an important factor in both improving public health and reducing health inequalities. Current international/national policy supports long-term breastfeeding. UK breastfeeding initiation rates are high but rapidly decline, and the numbers breastfeeding in the second year and beyond are unknown. This study used the concept of liminality to explore the experiences of a group of women breastfeeding long-term in the United Kingdom, building on Mahon-Daly and Andrews. Over 80 breastfeeding women were included within the study, which used micro-ethnographic methods (participant observation in breastfeeding support groups, face-to-face interviews and online asynchronous interviews via email). Findings about women’s experiences are congruent with the existing literature, although it is mostly dated and from outside the United Kingdom. Liminality was found to be useful in providing insight into women’s experiences of long-term breastfeeding in relation to both time and place. Understanding women’s experience of breastfeeding beyond current usual norms can be used to inform work with breastfeeding mothers and to encourage more women to breastfeed for longer.


Introduction and background
Breastmilk is widely considered as the optimum nutrition source for babies; good quality evidence has shown that it is an important factor in both improving public health and reducing health inequalities in both developed and developing countries (Horta and Victora, 2013;Ip et al., 2007;Renfrew et al., 2012). Breastfeeding promotion is a UK policy priority (Oakley et al., 2013;Public Health England, 2014). Some have argued that encouraging more women to initiate and sustain breastfeeding also brings potential economic benefits to society by reducing healthcare costs through improvements in the health of women and children (Pokhrel et al., 2015;Renfrew et al., 2012).
UK policy and guidance builds on World Health Organization (WHO) recommendations that all babies are exclusively breastfed for the first 6 months of life. WHO (2003) also suggests that breastfeeding should continue after 6 months alongside weaning foods 'up to two years of age or beyond ' (pp. 7-8). This second recommendation is not explicitly supported by the UK government unlike other countries, for example, the United States and Australia, which are more prescriptive (American Academy of Pediatrics, 2012;National Health and Medical Research Council, 2013).
UK breastfeeding initiation rates are high but rapidly decline, and only 34 per cent of women breastfeed at 6 months, only 1 per cent exclusively (McAndrew et al., 2012). The numbers breastfeeding in the second year and beyond are unknown. Long-term breastfeeding (also referred to as 'extended', 'full-term' or 'on-going', by breastfeeding women and in the literature) is uncommon in post-industrial countries; women who continue are at best misunderstood (Dowling and Brown, 2013) and at worst vilified, with any publicity about long-term breastfeeding usually generating negative commentary (Bekiempis, 2012;Lau, 2012;Time Magazine, 2012).
This study examines the experiences of a group of women breastfeeding long-term in the United Kingdom (the majority of whom were support group attendees), drawing on Mahon-Daly and Andrews' (2002) work on liminality and breastfeeding. Liminality has been applied to a range of health issues (including breastfeeding), but not breastfeeding long-term. A small amount of scholarship examines the relationship between women's accounts of their ability to continue breastfeeding and issues of space and place (Boyer, 2012;Groleau et al., 2013;Pain et al., 2001;Smyth, 2008). This article focuses on the ideas of 'betwixt and between' and 'matter out of place' (Douglas, [1966(Douglas, [ ] 2002Turner, [1969] 2009). It shows how liminality can be used to think about breastfeeding experience and focuses on liminality as 'a term denoting a time and a space' (Czarniawska and Mazza, 2003: 269). This article's contribution is to consider the experience of women who have breastfed beyond 'usual' limits, examining this through a specific conceptual lens, and thinking about what can be learned from this to encourage breastfeeding for longer.

Experiences of long-term breastfeeding
A small body of research examines women's experiences of long-term breastfeeding. Most studies are from North America and Australia; with recent findings very similar to those in an earlier body of work (including Hills-Bonczyk et al., 1994;Morse and Harrison, 1987;Reamer and Sugarman, 1987;Sugarman and Kendall-Tackett, 1995;Wrigley and Hutchinson, 1990). Women breastfeeding long-term in the 1980s and 1990s talked about the difficulties faced carrying out what is perceived and felt as a socially unacceptable and stigmatised practice. They experienced gradual withdrawal of support from 6 to 8 months, received comments such as 'Are you still nursing?' at 9-10 months, with increasing pressure/coercion to wean if still breastfeeding at 12 months and beyond. Support from organisations such as La Leche League 1 (LLL) was important. 'Support' is usually written about in terms of normalising the experience and receiving advice and encouragement. Participants spoke of the emotional benefits, bonding and close relationships with their children; the perceived 'naturalness' of long-term breastfeeding; and the importance of child-led weaning.
More recent work finds that women still manage long-term breastfeeding within unsupportive cultural and social contexts (Stearns, 2011). Studies report social stigma, the influence of societal attitudes and family and wider support, as well as the way in which long-term breastfeeding is often hidden, taking place within the family or in secret (Buckley, 2001;Gribble, 2007Gribble, , 2008Rempel, 2004). Strong social pressure to wean continues (Gribble, 2008;Stearns, 2011), and breastfeeding beyond a few months is 'extend(ing) beyond the boundaries of appropriate public behavior' (Stearns, 2011: 525). Long-term breastfeeders still emphasise the importance of closeness and relationship, and connect this with 'natural mothering' (Rempel, 2004: 307) and nutrition and nurture (Gribble, 2008). LLL meetings are experienced as normalising spaces (Stearns, 2011), with mothers seeking support from groups and 'sub-cultures within which breastfeeding beyond infancy … [is] normal and expected' (Gribble, 2008: 12).
Women who continue breastfeeding in the United Kingdom experience similar issues (Britton, 2000). Faircloth (2010bFaircloth ( , 2011 relates long-term breastfeeding to maternal identity and risk by examining the 'accountability strategies' of long-term breastfeeding women and the relationship between attachment parenting (Schön and Silvén, 2007;Sears and Sears, 2001) and long-term breastfeeding (2010a). Her participants reason that what they do 'feels right' (Faircloth, 2011), and she emphasises the importance of LLL in validating women's decisions and recognising an alternative way of being (Faircloth, 2010a).
Interviews for a UK-based health experiences project (HealthTalkOnline, 2011(HealthTalkOnline, , interviews carried out in 2005(HealthTalkOnline, -2006 include eight that address long-term breastfeeding. In these, women talk about breastfeeding in public, the emotional and practical benefits of continuing to breastfeed and of conflicting feelings when breastfeeding an older baby. They consider support from partners, families and groups to be important, acknowledging that long-term breastfeeding is more likely to take place in private with approval for breastfeeding progressively changing to surprise or disapproval the older the baby becomes.

Liminality
The concept of liminality and how it has been used to discuss health issues, including breastfeeding, is outlined in the following paragraphs. This is followed by an explanation of the use of the concept by Mahon-Daly and Andrews (2002), thus providing context for the discussion of the findings from this study in the latter part of this article.

Health 21(1)
Liminality usually refers to two related ideas -being 'betwixt and between' (Turner, [1969: 95, building on Van Gennep, [1960] 1909 and 'matter out of place', that is, ideas of pollution (Douglas, [1966(Douglas, [ ] 2002. Van Gennep identified three main stages to rites of passage -separation (pre-liminal), transition (liminal) and incorporation (postliminal) -each associated with particular rituals (Turner, [1969(Turner, [ ] 2009Van Gennep, [1960] 1909. During separation people move from their previous way of life towards the liminal state (Madge and O'Connor, 2005) where their existence is neither how they were before nor how they will be afterwards. This stage introduces 'the possibility of moving to a new structure or back into the old' (Jackson, 2005: 333). Incorporation returns the person to the 'secular' world (Teather, 1999) but usually in a different social state (Turner, 1979). In different types of social passage, one or other of these stages might be more important than others (Czarniawska and Mazza, 2003). The focus here is on the liminal phase, which is usually short but could be prolonged; sometimes people remain in a liminal state for the rest of their lives. Turner ([1969] 2009) argued that culturally prescribed and shared rites of passage enable people to move through separation, transition and incorporation and to develop new identities. Today, there are fewer shared rites of passage than before, and when movement is not straightforward … these persons elude or slip through the network of classifications that normally locate states and positions in cultural space. Liminal entities are neither here nor there; they are betwixt and between the positions assigned and arrayed by law, custom, convention, and ceremonial … (Turner, [1969(Turner, [ ] 2009 Liminal states suggest danger and threat because the previous identity is replaced by ambiguity, separation and a different order, that is, 'declassification without reclassification' (Navon andMorag, 2004: 2338): … Danger lies in transitional states, simply because transition is neither one state nor the next, it is undefinable … (Douglas, [1966(Douglas, [ ] 2002 Douglas ([1966] 2002) addresses the social consequences of crossing boundaries and of being in liminal states/places, both 'Purity and Danger'. She is known for her examination of 'matter out of place', that is, human body products that are considered dirty when separated from the body (e.g. menstrual blood), and how these 'me' and 'not me' products (Leach, 1964, quoted in Jackson, 2005 are 'betwixt and between' and considered taboo (Dowling et al., 2012).
Turner described the common space inhabited by those in the liminal phase as communitas; this is a way of living rather than a common place (Madge and O'Connor, 2005: 93), 'a shared sense of alterity' (Czarniawska and Mazza, 2003: 273) rather than identity. Communitas has three inter-related areas: 'in between (liminality), on the edges (marginality), and beneath (inferiority)' (La Shure, 2005). Liminality and communitas are also associated with the concepts of structure and anti-structure. 'Anti-structure' is used to describe the way social groupings outside the mainstream are still dependent on it for their position and impact; communitas exists because of its outsider relationship with other social structures (Turner, [1969] 2009).

Using liminality to understand health issues, including breastfeeding
Although Van Gennep is acknowledged as the originator of this concept, this article draws more on recent, and perhaps more fluid, interpretations of liminality. Van Gennep's focus on rituals and reincorporation is not always evident in later discussions although ideas of 'betwixt and between' and 'matter out of place' have been applied to a range of health issues. These include the sick role, refugees' status, living with chronic pain and fertility treatment (Allen, 2007;Jackson, 2005). Sometimes explicit links are made between liminality and being between two social identities (Madge and O'Connor, 2005) and with place and powerlessness; liminal people, such as those with mental health problems, can be 'banished' to 'marginal spaces and unloved places' (Wolch andPhilo, 2000: 144, quoted in Warner andGabe, 2004: 389), thus associating liminality with 'threat or unease'. Little et al. (1998) develop and use liminality to consider its unsettling nature. In their work, the stages are not always clearly defined -survivors of serious illness enter and remain in some form of this state for the rest of their lives, in an 'enduring and variable state' (Little et al., 1998(Little et al., : 1490. The powerlessness of people in liminal states is also recognised, when they cannot go back to what they were, nor see how they can evolve into a new identity. Navon and Morag (2004) relate this to men having hormonal therapy for advanced prostate cancer, who believe themselves to be 'not temporarily unclassified but permanently unclassifiable' (p. 2344).
Liminality has been used to explore women's health issues, such as cervical cancer screening (Forss et al., 2004), childbirth rituals (Hogan, 2008) and premature birth (Taylor, 2008). The transition to liminality can be unintentional, with difference not always visible (Forss et al., 2004) (e.g. having an abnormal cervical smear), and people find themselves 'at once no longer classified and not yet classified' (Turner, 1967, quoted in Forss et al., 2004. In the breastfeeding and early motherhood literature, liminality is often used as a contributing idea rather than a central concept, discussed in relation to time and childbirth, the purity and danger of foetal liminal status, the transition to parenthood (McCourt, 2006(McCourt, , 2009, and pregnancy and childbirth as rites of passage (Longhurst, 1999). Bartlett (2005Bartlett ( , 2010 recognises social constraints to women's ability to breastfeed as and where they like but also finds liminality an unhelpful concept: … breastfeeding cannot just be a liminal time or space which women occupy while waiting for 'normal' time to be returned. To understand breastfeeding as such is to devalue and constrain it to a nether-land outside of normal life. For women who breastfeed multiple children for months or years at a time, breastfeeding can be more 'normal' than not breastfeeding … (p. 126) We argue later that however 'normal' it becomes for individual women, it remains far from normal for those around them.

Liminality and breastfeeding
Mahon-Daly and Andrews (2002) use liminality to examine space and place in women's experiences of breastfeeding, and they apply the concept of 'rites of passage' to analysis of contemporary experiences. Rites of passage have been used to examine maternal and child health, but liminality has not been used specifically by others to theorise breastfeeding experiences. Although Mahon-Daly and Andrews' article is frequently cited, the liminality aspect is rarely used by others as a central concept. Groleau et al. (2006) identify the importance of rites of passage when 'abandoning' breastfeeding, and Dykes (2006) refers to breastfeeding in many communities as a 'marginal and liminal activity, rarely seen and barely spoken about ' (p. 206, also 2003, 2009). Others use Mahon-Daly and Andrews' findings more peripherally, for example, Sachs et al. (2006), Boyer (2010), (2011), (2012) and Sachs et al. (2006).
Mahon-Daly and Andrews (2002) acknowledge Van Gennep's work but draw on a reconceptualization of this by Wilson (1980) and on work by Davis-Floyd and Sargent (1997). They particularly focus on the second, transitional stage of rites of passage and using ethnographic data, discuss breastfeeding and liminal/transitional experiences in relation to three 'levels ' (2002: 65). The first refers to the post-natal period where women are not pregnant but neither have they returned to their 'normal' bodily state. This lasts until women stop lactating and reintegration occurs. For women who do not breastfeed or who stop breastfeeding early on, this is very brief and they soon return to their 'normal' bodily state. For women who breastfeed long-term, this liminal phase might last for some time; each woman's experience will be different as breastfeeding will end at different times.
Second are the ways in which breastfeeding changes women for life. They reach 'a new understanding of themselves and their bodies' (2002: 65) and how they communicate this to others. The third level of explanation considers breastfeeding itself, including the behavioural rituals associated with particular places through which women move as breastfeeding mothers. Breastfeeding is seen as liminal in terms of time and space and is a temporary state -breastfeeding is not viewed as a 'normal' activity because women want to return to a 'normal' life. It is in this third level of explanation of transition that Mahon-Daly and Andrews consider rituals -an important aspect of Van Gennep's conceptualisation -particularly relating these to spatial behaviour.
Breastfeeding as pollution ('matter out of place') is also addressed in relation to liminality. Breastmilk is described as 'a potentially impure body fluid' (Mahon-Daly and Andrews, 2002: 69). Participants discuss breastmilk leakage and stains as 'perceived dirtiness' and note different reactions to spilt breastmilk and formula milk. Viewing breastmilk as polluting reinforces breastfeeding as a liminal state. Mahon-Daly and Andrews' work does not address long-term breastfeeding, but provides a useful platform to explore women's experiences of this. Their work focuses more on the second phase of Van Gennep's model than on separation and reincorporation, and it is this aspect that is specifically drawn on here.

Methods
Our intention was to explore women's experience of successfully breastfeeding their babies for over 6 months and whether this could be used to help more women to breastfeed for longer. We wanted to learn more about the difficulties that women experienced, sources of support, what helped them continue when others stopped and how liminality could be used to think about their experiences.
A qualitative design using micro-ethnographic methods (Erickson, 1992;Lutz, 1981) -participant observation in breastfeeding support groups, face-to-face (FTF) interviews and online asynchronous interviews (OAI) via email (Dowling, 2011) -generated the project data (author 1 (S.D.) was breastfeeding long-term during the study). The complementary data collection methods generated extensive rich data about women's long-term breastfeeding experiences. Participant observation increased understanding about how women support each other during long-term breastfeeding. OAI allowed in-depth exploration of women's experiences, supported by researcher reflection. FTF interviews explored similar issues, helping participants to reflect on their breastfeeding in a different way from support groups. Triangulation was used to increase the richness and diversity of data (Denzin and Lincoln, 2008).
Data collection took place from January 2008 to March 2009. Participant observation first took place in LLL meetings to increase the likelihood of meeting long-term breastfeeding women. S.D. was a member and had previously attended meetings. Observation in two other (non-LLL) breastfeeding support groups was also negotiated to meet women from a wider range of socio-economic backgrounds, living in areas with lower rates of breastfeeding. Observation in these three groups enabled contact with over 80 women breastfeeding new-borns to 4 year olds. Some mothers were in stable heterosexual relationships; single mothers and lesbian mothers also attended the groups. Women's ages ranged from early 20s to late 40s. Some women worked outside the home full-time or part-time in paid or voluntary work; others were full-time mothers. Recording social and demographic information on all the women met through these three support groups was not possible as membership was fluid and contact variable.
Six women were interviewed FTF. Four of these women were accessed purposively via LLL and two via other support groups and snowballing. Their ages ranged from 27 to 42 years, between which they breastfed 11 children. Breastfeeding duration ranged from 18 months to 4.5 years, and they had experience of breastfeeding while pregnant and tandem feeding (the term used to describe the practice of breastfeeding more than one baby/child during the same time period). OAI was used with four women using the same recruitment methods as above. One of these women was an LLL attendee; the other three were not. These women breastfed four children for between 14 months and 6.5 years. They were slightly older than FTF interviewees (OAI range: 35-47) and had higher educational qualifications.
Most data collected were from observations in support groups. This is recognised as a limitation as it does not represent the views of women who were not accessing this type of formal support. In-depth data from OAI and FTF women offered other perspectives. Half (5) of interviewees (FTF and OAI) were regular attendees at support groups, but half (5) were not. LLL attendees are acknowledged to be perhaps an unrepresentative group of breastfeeders; observation in two other non-LLL groups allowed for different perspectives. As is usual in small-scale qualitative research, the intention was not to achieve representativeness or generalisablility although, as noted, attempts were made to explore the experiences of a broad range of women, thus increasing the transferability of the findings.
Research ethics approval was granted by the Research Ethics Committee of the University of the West of England, Bristol. All interviewees received written information about the project and had the opportunity to ask questions before consenting to participate. Project information was available at all breastfeeding support group meetings, and introductions were made at each meeting to newcomers.
Permission to audio record LLL meetings was sought at the beginning of each meeting and from any late arrivals. All these meetings were audio-recorded and transcribed.
Recording was not possible in other groups as the meeting structure encouraged simultaneous conversations, so field notes were made immediately after these observations. All FTF interviews were recorded and transcribed and all emails from OAI were imported into Microsoft Word. Pseudonyms are used to anonymise all participants. All documents were imported into NVivo (V.8, QSR International) for thematic analysis (Braun and Clarke, 2006). The findings are discussed here in relation both to liminality and to the four main themes identified through analysis -'Deciding', 'Commitment', 'Challenges' and 'Being supported'.

Deciding to breastfeed long-term
There appeared to be a number of elements contributing to the decision to breastfeed and then to carry on breastfeeding. In this study, participants talked about 'always knowing' they wanted to breastfeed:

… I just always had it in my mind that I would … I just assumed that I would … (Mandy, interview)
They also talked about their knowledge and understanding of health and other benefits and a perception of breastfeeding as a 'natural' activity: … it's not necessarily the health benefits … I just thought it's a natural thing … why would I want to bottle feed? I just thought the milk that a mother produces for her child is perfect for the child … (Josie, interview) The decision was also sometimes related to the influence of others (although some participants had known no other breastfeeders previously) and to pre-existing knowledge of the benefits of breastfeeding.
Participants talked about their own pre-natal ideas about when breastfeeding would end and how others clearly conveyed their expectations about the end of the breastfeeding period ('still breastfeeding?'). Some talked about feeling pressurised into stopping breastfeeding once their child reached 6 months and that the duration of breastfeeding implied by health promotion messages was relatively short (see Dowling and Brown (2013), for further discussion of this point).
A small number set out to breastfeed long-term, but for most it happened gradually; some associated this with the developing relationship with their child and with the child's will. Many participants talked about finding long-term breastfeeding 'shocking' or 'surprising' before they themselves became long-term breastfeeders: … I remember being quite shocked at the time … I remember then when I was pregnant thinking 'I wonder if that will be me?'… (Tina, interview) … it was surprising. It's … a funny visual … thing, the first time you see it … a big boy, with, you know, boots … long legs … I remember being, kind of, impressed, you know, thinking 'probably might not do that' … (Mandy, interview, talking about the first time she saw someone breastfeed an older child) Liminal states are not culturally acceptable when their duration cannot be controlled (Hogan, 2008), and there are societal expectations about rites of passage (Taylor, 2008). This becomes an issue for women who continue to breastfeed despite societal pressure to wean. The way breastfeeding is talked about suggests it is good for babies, unusual but tolerated up to 12 months and unnecessary, inappropriate and increasingly bizarre beyond this point. Comments about the appropriateness of breastfeeding are common once teeth erupt or children walk and talk. As research participants acknowledge having these opinions before breastfeeding, and feeling shocked and/or surprised at first exposure to longterm breastfeeding, this indicates the extent of these societal expectations.

Commitment (living with long-term breastfeeding)
Participants talked about living with breastfeeding in relation to 'physical and emotional commitment', 'commitment to a way of life' and 'commitment to a style of motherhood', often in ways that they had not anticipated. There is some overlap between this theme and the next with challenges identified in both.
As breastfeeding continues, for some women there were difficult consequences for personal well-being, relationships and paid employment. This is despite all saying that they wanted to continue. Women talked about this commitment setting them apart from other mothers, as those in a liminal phase are from the rest of society. They are marked out as different by decisions made early on about feeding babies and children -such as how children are helped to sleep, where they sleep (for some, long periods of co-sleeping), how mothers behave with them (e.g. extended periods of carrying) and by long-term decisions made by some (e.g. home education) -which set these women apart from their peers. Isolation was experienced emotionally and physically and was referred to as a 'cost' (Sarah, email).
Participants felt that they were 'in a phase' and in a different place to other mothers. They talked about being committed to a way of life that they cannot see their way out of -this is a phase with no end: … as this time continues without a clear end in sight … as the mothers of similarly aged children are returning to some state of normalcy … (Sarah, email) With no apparent exit, other life-choices were difficult, that is, having another child, returning to paid work or returning to their previous self.
Women talked about being different on many levels, recognising that they are in a different place when other mothers are 'returning to normal'. Being in a particular place by virtue of choosing to mother one child in this way, makes any other place, for example, having another child, appear difficult to move into. Women described how they were in a Health 21 (1) different place from where they were before they became mothers; for some this was related specifically to breastfeeding, for others more generally to parenting. Women did not know what or how they would be in the future: … When I began this I did not expect that I would end up staying at home taking care of a small child … some of my frustrations with the lack of separation extend from a sense of never having made a decision to quit and stay at home … (Sarah, email) This commitment places women in a liminal space by virtue of the decisions they make and because they perceive that society views their behaviour as transgressive and not worthy of support.
Breastfeeding women acknowledged that they did not have the personal space or bodily freedom experienced by their non-breastfeeding peers. In this 'betwixt and between' time and space, women often expressed mixed emotions about the benefits of the emotional and physical closeness they experienced through breastfeeding and the feeling of being 'tied' to their children and to a way of life: … the emotional clinginess … thinking 'I need, I just want some space' and because there's just no space at work, or studying or running the home … I felt quite claustrophobic … breastfeeding could feel like that sometimes … (Tina, interview) Expectations about set patterns or behaviour rules are internalised; long-term breastfeeders break these rules but are relatively powerless about how their behaviour is viewed. They may not be happy in their new liminal state, but cannot easily return to their previous state. This is illustrated by the ways women in this study talked about the difficulties despite making a positive decision to breastfeed long-term, here talking about lack of personal space: … they're mine, I've got to say 'they're my boobies' and he says 'no, my boobies. And I say "they're my boobies and I've got to have them to myself for a minute' … you just need some time don't you …I'm not just a pair of boobs … (Woman breastfeeding a 2-year old, LLL observation, our emphasis)

Challenges in breastfeeding long-term
Two main sub-themes in relation to challenges are addressed in these data: breastfeeding in public and responding to/knowing about cultural unease with the practice. Women who breastfeed long-term do so knowing how it is perceived by wider society, both by what is said to them and how they behave with others. Participants talked about both actual and anticipated reactions to breastfeeding in public. Most said they were happy to breastfeed older babies and children in public -and few had received negative comments -but nearly all developed strategies to do this comfortably, including avoiding eye contact and thinking carefully about how they positioned themselves: … I never was worried when Mark was with me. I just looked at him and Janie and didn't look around … on my own I felt uncomfortable and worried but I did the same thing -I just never looked up, hoping that if I didn't make eye contact no one would talk to me or say anything … (Sarah, email) … I just kind of ignore people around me, when I'm doing it … sometimes I do try and go in a bit of a quieter place … you don't really want to be the centre of attention when you're doing it anyway … (Sam, interview) At times participants also interpreted spaces in their homes as public, echoing Pain et al. (2001) and Scott and Mostyn (2003). Conversely, they made public spaces feel private by using strategies, as described above, to make long-term breastfeeding in public more acceptable and talked about this in terms of their comfort and managing others' discomfort.
Participants talked about long-term breastfeeding in ways that demonstrate 'matter out of place'. They talked about times when they expected negative reactions (with surprise when they were not forthcoming), showing their understanding of taboos about breastfeeding and where it might be considered socially inappropriate: … Once I nursed her on a walk and she was having what I call melt down right by the Catholic Cathedral … So I nursed her there on the steps until she fell asleep. A very friendly priest came up and had a chat with me, not at all fazed that a 2+-year-old was sleeping and nursing. That was the only time someone actually approached me when I was nursing in public (I was prepared for a rather different response) … (Sarah, email) There were clearly expectations that others would find breastfeeding older babies and children challenging, and unpleasant or difficult comments were anticipated. Many women think of 'feeding publicly as breaching a cultural taboo' (Stewart-Knox et al., 2003: 267). Participants talked about feeling more uncomfortable breastfeeding in public as their children grew older and had an increasing awareness of others' potential reactions. Other examples of times when their lactating breasts and/or their breastmilk were considered 'matter out of place' are noted where breastfeeding felt comfortable, and who it was comfortable to breastfeed in front of. Other women talked about feeling 'on display' or not feeling comfortable breastfeeding with certain people present as their child grew older, even in their home: … you can't just leave it all out … [referring to her breasts] … (Sam, interview) Uncontained breastmilk may be interpreted as women's lack of control over their bodies (Battersby, 2007), and women who breastfeed long-term may also be seen to lack control by not conforming. In this study, Sam talked about being told that she had … gone beyond the call of duty… (Interview) A participant at a LLL meeting commented that she had been told: … you've done it for so long now you don't have to put yourself through it any more', and 'you've punished yourself enough by breastfeeding her' … (LLL observation) These comments suggest the women's behaviour was transgressive, and possibly damaging, but also noble, self-sacrificing and within their control to change.
For many, the longer breastfeeding continues, the more likely it is to become secret or hidden. In part, this is because long-term breastfeeding often takes place at times which are more private (first thing in the morning or before sleep), but it is also in reaction to actual or perceived responses from others. The secrecy related to long-term breastfeeding for many women is partial -they may tell close friends or family but not be open generally about it in public, or they may tell close friends but not their wider family. Most women recognised that they were choosing to behave in a way that set them apart from others: … you're sort of out of synch with the common view aren't you? Cos I'm really proud that I'm still feeding her, I think it's brilliant and it feels like totally the right thing to be doing. But it is sad, it's sad to be in a situation where you don't feel you can be proud of it because people don't understand … … Now he's eighteen months they just assume that I've stopped, nobody's mentioned it for a while … (LLL observation) To a certain extent, by 'allowing' others to think that they have stopped breastfeeding (maintaining the secrecy discussed above) these women can also be seen to be contributing to the invisibility of long-term breastfeeding, although they are also challenging cultural expectations through their behaviour.

Being supported in long-term breastfeeding
Participants talked about four main categories of support: families (including partners), friendships, support groups and health professionals. Support was sometimes expected but not forthcoming. Support through books and from the Internet was also identified by women as significant.
Some women breastfed long-term with support from friends and family, but for others it was isolating and set them apart from their peers and usual sources of support. Supportive friends and family were those who facilitated breastfeeding, did not criticise decisions about breastfeeding and/or overtly agreed or made joint decisions about continuing to breastfeed and other related parenting decisions. Lack of support from families was expressed through negative comments and attitudes addressed to the breastfeeding women, usually from mothers or mothers-in-law. One woman said that she no longer visited her parents with her 2-year-old son because her mother said that continued breastfeeding was 'disgusting', and not something that she wanted in her home. Being in a liminal state can be unsettling, for those in it and friends and family around the liminal person (Hogan, 2008). Women also talked about the difficulty close relatives had with their long-term breastfeeding because it was perceived as inappropriate and stopped others (particularly fathers and grandparents) from relating to children.
Many sought alternative sources of support and become involved with networks in a sub-culture of like-minded people; often this was specifically related to breastfeeding but also connected to other choices such as home-educating or attachment parenting: … in the beginning, all I had was my NCT friends … I found it really uncomfortable because they were so different to me … a lot of them had stopped breastfeeding within four to eight weeks … So after that I gravitated towards people that were more like minded because it made me feel better and therefore I was a more calm person to be around [son] … (Josie, interview, authors' emphasis) … there is an active home ed. group that we go to twice a month … it is like an LLL meeting because there are so many people breastfeeding both babies and toddlers! There is a lot of breastfeeding in the home ed. Community … (Christine, email) Breastfeeding groups were referred to as comfortable places for receiving support for breastfeeding long-term in ways that other places were not, particularly if also in relation to tandem nursing or co-sleeping. Associating in this way helped many women remain in a secure liminal space for some time. The idea of communitas -people in the liminal phase inhabiting a common space (Deflem, 1991) -may be relevant to some long-term breastfeeders. Breastfeeding and associated decisions become part of how some women identified themselves and how they identified with others, leading to a sense of belonging to a community and a 'shared sense of alterity' (Czarniawska and Mazza, 2003: 273). This seems to be the case for LLL attendees and women attracted to ideas informing attachment parenting, the 'like-minded people' referred to by Josie (above).
This shared sense of community can be experienced without a shared physical presence. 'Sarah' talked about how it may be acquired through reading and the Internet -as well as through 'actual contact' (email). Madge and O'Connor (2005) also note how useful the liminal space of cyberspace is to new mothers. The proliferation of Internetbased support for long-term breastfeeding mothers suggests that cyberspace is becoming an important source of support and contact for many women, for example, Facebook groups, Twitter threads, websites such as those of netmums.com and the Association of Breastfeeding Mothers.

Liminality and long-term breastfeeding
Our main focus in this article is on space and time and the idea of being 'betwixt and between'. In Mahon-Daly and Andrews' schema the first liminal phase is post-natal. Those who choose to breastfeed (for however long) are not immediately reintegrated into society: … total reintegration may not occur so long as lactation is active and positive. Breastfeeding bodies are indeed physically different, are relatively uncommon, and are experienced by a minority … (Mahon- Daly and Andrews, 2002: 65) Most women experience reintegration fairly rapidly; however, breastfeeding women in this study remain in a liminal state for a period of time. As others have shown (e.g. Little et al., 1998;Navon and Morag, 2003), reincorporation is not always straightforward.
These women are 'betwixt and between' as their experience continues to be different from those who stopped breastfeeding and those who never breastfed. For some, the liminal phase did not lead to reincorporation into society as before, but incorporation into a new society as a changed being -especially attachment parents and/or home educators, where decisions about how they lived and behaved separated them from most other parents. For some, the liminal phase is clearer as they intend to return to their former life; most, however, recognise that life would never be the same again. Some used this to build a new identity as a long-term breastfeeder and breastfeeding counsellor, or birth educator or other related activity.
Mahon-Daly and Andrews' (2002) second level of explanation of liminality relates to the way breastfeeding changes women -a transition to a new understanding of themselves and their bodies, and the way they let others know about their 'new world' (p. 65). Our participants talked about communicating to others their decision to carry on breastfeeding, their changed perceptions and the qualities that enabled this. They talked about their determination, not allowing others to criticise their behaviour, and their pleasure and delight in what they were doing: … I love this! I don't want it to stop! … (Woman at LLL observation) They referred to their confidence in communicating with others and deciding to breastfeed long-term, as well as having a self-confident personality.
In their final level of explanation, Mahon-Daly and Andrews highlight how breastfeeding women use places to ensure a comfortable experience, demonstrating liminality in time and space. In this research, women talked about where they breastfed, who was present and about moving in and out of different spaces for breastfeeding. Societal expectations about the duration of breastfeeding affected women's practices when breastfeeding in public.
These findings support the application of Mahon-Daly and Andrews' insights to women who breastfeed long-term. There are two sorts of liminality: the ritualised transitions between states ('rites of passage' meaning) and less clearly defined states (Jackson, 2005). Long-term breastfeeding can be seen as both/either 'not-quite-either' or 'some of both' (Jackson, 2005: 345).
Liminal people are 'at once no longer classified and not yet classified' (Turner, 1969, Mahdi et al., 1987. All the women interviewed were asked what name they gave to their continued breastfeeding. Most women commented on the range of terms available, but some had no words for what they did, usually because it was not discussed with others. Women who breastfeed long-term appear to be between social identities, with a temporary identity for which our culture has no name, and it may be that they are 'permanently unclassifiable' (Navon andMorag, 2004: 2344). Coming to an understanding of their social identity is hard for women whose behaviour is discussed negatively in the media or who are unable to talk about it to those around them. The social anxiety associated with liminal people (Warner and Gabe, 2004) may be seen with long-term breastfeeding, for example, media attention following the Time Magazine (2012) cover showing a woman breastfeeding a 4-year old. Many people commented negatively about this on discussion boards, newspaper websites and television programmes, demonstrating the social unease around long-term breastfeeding. Women who openly breastfeed long-term can be interpreted as culturally transgressive and a threat to order (Smale, 2001).

Conclusion
This study's findings concur with Gribble (2007Gribble ( , 2008, Stearns (2011) and Faircloth (2009, 2010a, 2010b as well as with older findings, discussed previously. The overall picture is of a group of strong-willed and determined women with a clear sense of purpose and of 'doing the right thing'. Most of the women whose experiences are discussed here were involved in a range of support networks. Increasing understanding about the experiences and feelings of these women is important although, as not all women are able or want to access such support, other (potentially different) experiences may warrant further research.
Liminality has been used to think about this experience, drawing on and extending the use of the concept by Mahon-Daly and Andrews (2002). Increasing understanding of the experience of these women might help in encouraging more women to breastfeed for longer (in line with current policy guidance) by both normalising and de-stigmatising the practice.