I just saved you – does it matter what I look like? Reading and discussing feminist fairy tales with a group of 12-year-old girls

Using a post-structural lens which situates gender as discursively produced, this study investigates how four 12-year-old girls read feminist fairy tales and what feminist issues and concerns they discern and relate to in these texts. The study used a dialogic approach in Action Research informed reading groups to stimulate their thinking and explore their experience of gender constraints. The article explores how the girls drew on discourses and discursive practices to position themselves in relation to the texts, often breaking the frame of the text and creating ‘counter-fictionals’ and using performativity to construct their versions of femininity. The author argues that the book group acted as a liminal space that allowed the girls a safe space for exploring and subverting gender constructs and constraints.


Introduction
This study investigates how four young girls read feminist fairy tales and what feminist issues and concerns they discern and relate to in these texts.My questions arose from my research interest in children's literature and from my role as a mother to an adolescent girl.In the first instance, my interest was sparked by reading Davies' 1991 account of how young children were not able to question gender roles and that when they were shown feminist fairy tales which reversed stereotypes, they resisted these strongly.Thirty years on, I wanted to investigate these feminist issues and concerns for myself with a group of adolescents, an age population 'whose views on gender relations and sexuality are often underrepresented in the literature' (Moffatt and Norton 2008, 107).
A post-structuralist feminist framework which situates gender as discursively produced emphasises the importance of critically examining the construction of and resistance to gender within children's texts.By reading short texts with a fairy tale motif and strong feminist ideologies with a small group of 12-year-old girls, including my daughter, I sought to analyse and evaluate their understanding and experience of gender issues through their interactions with the text and with each other.Assuming a role as Parent-as-Researcher, I encouraged dialogical discussions to amplify opportunities for exploring the complexities and contradictions within the selected fairy tales.I was particularly interested in witnessing how the setting allowed the girls to reflect on and question gendered norms and draw parallels to their own afflictive experiences.As my study progressed, I was fascinated to watch not only that the girls read resistantly (Lehr 2001) but to see exactly how this resistance unfolded through their engagement with the texts.As this article explores, the girls drew on discourses and discursive practices to position themselves in relation to the texts, often breaking the frame of the text and creating 'counterfictionals' (Skidmore's [2000] term for conjectures about alternative storyworlds) and using performativity to construct their versions of femininity.Since our sessions took place outside of a school context and amongst a group of girls who were not part of each other's tight social circle, I argue that a liminal space (not-school, not-social) was created, which allowed the girls to share and combat the emotional impact these expectations had on them.

Feminism, gender, and dialogic practices
This research is positioned in the intersection between gender studies (specifically as it relates to girlhood), feminism and children's literature.Children's literature is one of the most useful sets of cultural products for investigating values imposed on children (Baker-Sperry and Grauerholz 2003).Consequently, I use literary texts to explore how gender structures are depicted and how, in turn, these are perceived by the readers themselves.Gender inequality is still a problem in children's books, as shown by Ullah and Naz's critical review of empirical studies of gender representation (2014).Recent studies focus on gendered meanings and stereotypes in children's books (McCleary andWiddersheim 2014, Ullah andNaz 2014) and in fairy tales (e.g., Marshall 2004, Parsons 2004, Kuykendal and Sturm 2007) and how these are perceived by child readers (e.g., Trousdale and McMillan 2003, Moffatt and Norton 2008, Cossens and Jackson 2020).As such, my research is part of the continuum of research that seeks to understand current relationships between literature, identity and power.However, recent gender studies involving only female children focussing on issues of feminist concern in children's literature appear scarce.Further, the group I was interested in, tweens, are poorly represented, a fact also remarked on by Moffatt and Norton (2008) and Cossens and Jackson (2020).This research thus specifically adds to our understanding of how a group of adolescent girls perceive gender constructs in children's literature.
The feminist lens I mobilised in this research is post-structural and discourse based.Post-structural feminism theorises the ways in which cultural discourses shape knowledge, meaning, subjectivity and identity (McCann and Kim 2003).A post-structural approach does not attempt to map out universal laws concerning women's or men's experiences.Instead, it aims to describe how gender is socially constructed and performed and how gender inequality is maintained and reproduced but also challenged.From this perspective, discourse, in this case in children's literature, is seen as highly significant in the reproduction of, and resistance to, gender inequality.As such, this invites an analysis of how gender is discursively produced (Marshall 2004) and by what means this is variously acknowledged and resisted by the reader.The way children relate to and interpret the text and how they insert themselves into the narrative are thus critical issues for a feminist reading of children's literature.As a reader reads, she is simultaneously engaged in the processes of reading, interpretation and criticism (Ketter and Lewis 2001).The text itself does not contain the meaning but holds possible meanings which are brought to life by particular readers in particular contexts and times.Meaning-making 'lies in the intersection between reader, text and context' (Maine 2014, 151), which makes reading a situational event (Maine 2014).Alexander's work on dialogic teaching has as its central tenet that we need to 'harness the power of talk to engage children, stimulate and extend their thinking, and advance their learning and understanding' (Alexander 2018, 1).As my research was conducted in a book group setting, meaning and learning was dialogically co-constructed in the intersections between the text, the individual girls and the group.

Action based research
The study was an action research project.It was carried out in one-hour-long book group sessions, once a week for four weeks.It involved four girls, one of whom is my daughter, in a small group reading context which, Maine and Hofmann say, 'offer the ideal dialogic "situation" for reading' (Maine and Hofmann 2015, 46).As a methodology, Action Research (AR) bridges the divide between research and practice (Somekh 2006).It is a cyclical process of observation/reflection/plan/action with critical reflection and evaluation taking place throughout the project and where the findings are continually fed back directly into practice which is changed accordingly (Somekh 2006).As I do not have a teaching background, I felt AR would be a helpful framework to facilitate and support my engagement with the participants as it 'affords a rigour and relatively straight-forward methodology for improving one's own' practice and reflecting on it' (Gilbert and Macleroy 2021, 11).As such, I used the AR approach to enhance and deepen my research practice, continually reflecting on what had worked well and what less so and experimenting with different approaches for each of the four book group sessions.My research process was thus emergent and exploratory rather than entirely pre-planned.

Ethics and the role of parent-as-researcher
Feminist research principles of reflexivity, collaboration, respecting women's voices and 'do no harm' (Ramazanoglu 2018, Jenkins, Narayanaswamy, andSweetman 2019) are central to my research values, and they supported and enhanced all aspects of this research project.Given that my interest originated from my parental role as a mother of a girl, it felt natural to include her in my research.As a result, I took on the role of 'parentas-researcher' (PAR), which allowed me to focus on topics that are meaningful to my life while at the same time using my 'research skills to pursue the understanding and analysis of personal interest' (Adler and Adler 1996, 48).The notion of reflexivity as a feminist research principle calls into question the role of the researcher and the method itself, especially with respect to issues of power and representation which were particularly pertinent given the age of my research participants and my role as a parent to one of them.The PAR role does bring to the fore several potential ethical and moral responsibilities, most critically the question of direct authority over one's own child involved in the research.Research involving a PAR may feel like an intrusion into the child's life, the role bifurcation may be confusing for all the research subjects or 'it may enhance the researchers' relationships with their children and others by deepening their intimacy, involvement and understanding (Adler and Adler 1996, 46).I found that including my daughter in my research allowed me to draw on Fine's (1994) notion of 'working the hyphens', which she argues gets better data and aligns this research with feminist and queer methodologies, which seek to resist hierarchical separation between researcher and participant (Tillmann-Healy, 2003).Of course, there are latent disadvantages to 'working the hyphen' as a PAR, including the possible impact our pre-existing mother/ daughter relationship might have had on my daughter's willingness to share her views and the potential for conflict between us.However, I found the Action Research methodology helped me reflect both on my role as a PAR and on my own feminist views and positionality and how these changed as a result of the research process.Further, the fact that the research timeframe was so short and circumscribed meant I could take advantage of the ease of entrée and reflected identity afforded by my daughter's presence without (m)any of the disadvantages associated with the PAR role.
Several factors shaped the setting of the research and the selection of participants.Firstly, there were practical considerations: I felt that the 'dialogic situation' of the small group reading context required the intimacy facilitated by face-to-face meetings and as this research took place in the UK in the spring of 2021, I needed a setting I could use subject to lockdown restrictions.One of the advantages of the PAR role is the ease of entrée, which the 'complete membership role' offer the researcher (Adler and Adler 1996).As such, my daughter acted as a 'wedge' (Levey 2009) to allow me to recruit local girls who were in the right age group and who would be able to adhere to lockdown restrictions by meeting at my house.While the participants were friends of my daughter's, and their mothers were my friends, the girls did not ordinarily socialise as a group.The participants were not representative of society at large; they were middle-class, highability, white girls, three of whom attended single-sex schools.They were also habitual and committed readers.Table 1 'Participants' sets out the four participants, their ages and their school situation at the time of the research study.It also sets out their response to questions designed to understand how femininity and expectations of school and society were perceived by them.
My feminist views are widely known by the girls and by their parents, who share them to a large degree.I asked the girls to fill out a questionnaire which probed their interest in and understanding of feminist issues, particularly gender inequality.From this, it was clear they all actively identified as feminists.Over and above the fact that we were reading feminist fairy tales, the shared feminist view naturally influenced our conversations in the book group somewhat.However, I went to considerable lengths to mitigate unduly Strongly agree Disagree influencing their responses.For instance, I aimed to play the role of participant-observer rather than the adult-facilitator-teacher by deliberately using dialogic practices aimed at encouraging 'questioning and exploration rather than a singular, correct, or closed response' (Maine 2014, 151).Further, I was also aware that the participants might be more reticent in talking about sensitive topics in front of an adult whom they already knew or who was their parent.To mitigate this, I removed myself from the discourse, either directly, providing an 'adult-free' forum by including a semi-structured peer interview session in the second book group session, or indirectly, by using a 'Philosophy for Children' (Vansieleghem and Kennedy, 2011) approach to determine what we were going to be discussing in session four.Few researchers use children as collaborators, particularly in the data collection phases and the few who do rely most often on adolescents rather than children under 12 (Levey 2009).For me, it was immensely important to centre and respect the girls' voices within the gender discourse, both as a feminist imperative to 'respect female voices' but also because of my personal relationship and involvement with the girls, which by its very nature endured beyond the research timeline.In essence, this research was collaborative through its book group setting and the research principles and tools I deployed, but I also made sure to brief the participants to obtain their verbal consent, and the girls were shown this article to ensure they approved of my portrayal of their conversations.Pseudonyms were assigned to participants to ensure confidentiality.In my research, BERA (2011) ethical guidelines were followed, and ethical approval was granted by Goldsmiths University.

What we read and how data was collected
I chose four short fairy tale/mythological texts with strong feminist themes: 'The Pea and the Princess' (Grey 2003), 'The Paper Bag Princess' (Munsch 1980), 'Hansel and Greta' (Winterson 2020) and the poem/painting 'Not my best side' (Fanthorpe 1978).These four texts are 'contaminated' tales, part of a larger intertextual corpus of feminist fairy tales where 'retelling (. ..) not only implies rereading but most importantly, rethinking' (Joosen 2004, 36).In contrast to more complex and multi-layered novels, these works' brevity, multimodality, and thematic clarity made them very suitable for one-hour book group sessions.Table 2 sets out the texts we read and the approach I took in each of the four book group sessions.P4C approach using the poem as stimulation for identification of themes for discussion I collected data by a range of methods: group discussions, group worksheets and peer interviews as well baseline and exit questionnaires, adding or jettisoning data tools as the action research went along, and I learnt both from the interactions and from the data.Data transcription and analysis took place as soon as possible after each of the four book group sessions to inform and prepare for the next session.Reflecting on the analysis of the first two transcripts made me realise there were recurrent key themes we had not explored sufficiently in the group.Consequently, I designed a set of semi-structured questions around expectations and role models, which the girls used in peer interviews in the third session.The girls loved interviewing each other, and the interview transcripts provided key insights for me to reflect and build on.I developed sociological feminist codes through a thematic-based sifting of the data based on Ritchie et al'.s(2014) five-step qualitative research practice model.Through a repeated process of refinement, I extracted three overarching and overlapping themes: expectations of feminine appearance, expectations of feminine behaviour, and everyday misogyny in an educational setting.In this paper, I draw from the first two themes to explore how the girls broke the narrative frame of feminist fairy tales to discuss issues of perfection, anger, gender, and sexuality in the liminal space of the book group setting.

Issues of perfection
One of the strongest themes of our book group -the pressure to be perfect -was raised from the start in our conversation about 'The Pea and the Princess'.This is not surprising given the pervasiveness of the beauty ideal in fairy tales (Baker-Sperry and Grauerholz 2003).While I had found the text amusing, most, except Violet, failed to appreciate the intertextual humour which stems from the clash between the romantic and the feminist frame and which is underpinned by the illustrations of a number of famous fairy tale princesses in the following section: [The prince] met princesses of all shapes and sizes with a wide range of hobbies and interests, but none of them seemed really right somehow.One was too loud, too quiet or too funny, too grumpy, too sleepy, too energetic, too scruffy, too tidy, too pink, too scary.(The Pea and the Princess, np) Ignoring the intertextual play between the narrative frames, the participants instead articulated their resistance to the discourse around the heterosexual quest for female perfection by taking exception on behalf of younger girls and worrying about what it might do to their mental health: Megan: I do not like the fact that the women in the story have to be perfect.Sensitive to even the slightest thing to be good enough [. . .]Emma: it is just setting requirements for girls today who read this book [. ..]Even if they don't consciously know it [. ..], they will think, I need to be perfect [. . .]Mathilda: It's lowering kids' self-esteem, really [. ..] (Audio-recording, session 1) Aside from deconstructing the impact the dominant societal narrative of female beauty has on girls, the participants also problematised the effect the notion of perfection has on boys' expectations of girls: Emma: Yeah, you also make boys think that they need the 'perfect girl' [. ..] (Audio-recording, session 1) By moving the conversation to the general 'boys' and not the text-specific 'prince' Emma has drawn a parallel between the text and her social reality; by doing this, she not only critiques flaws in the original plot but also recognises the complex workings of discourse as anything but theoretically constrained by the text.
It is interesting to note that other than talking about how physical perfection might raise boys' expectations, at no point did they discuss if and how beauty standards might be aimed at attracting 'a man' (or a partner of any persuasion) for themselves.Moffatt and Norton's (2008) research with adolescents found that even when participants produced feminist constructs of gender relations, they invariably constructed masculine and feminine sexuality as uniformly, de facto, heterosexual.My participants, in contradistinction, further played on the paradigmatic changes in 'The Pea and The Princess' and were quick to queer the prince by questioning why he was on a quest for a princess, why not a prince or even a pea?While they were open-minded to different permutations of sexual desire and gender, it did not appear to flow into their identity quest yet.
The theme of perfection recurred in a number of the discussions, and consequently, I asked them to explore it in more depth in the peer interview sessions in book group 3.As Emma and Megan were exploring questions about the expectations in terms of looks, they made clear that this was not just a rhetorical concern on behalf of younger children but real and felt by themselves: Megan: [. ..] it is inbuilt in society.I think it's important that they teach people that you can be different [. ..] that you don't have to be perfect -'cause nobody's perfect, everybody is different and beautiful in their own way [. . .]Emma: Like flowers Megan: Yes, like flowers.(Audio-recording, session 3) Seeing difference and imperfection as beautiful is encapsulated in the image of the flowers and was a way for them to wrest some power back from a discourse they felt was damaging.Megan, in particular, who had stated that she never changed her behaviour to fit in with what was expected of her as a girl, determinedly maintained that 'nobody's perfect'.This insistence that 'everybody has their flaws' (audio, session 1) is a plea for building understanding and agreement in the group that the discourse of perfection as a means to attain successful femininity is impossible.By our third session, the trust had built between the participants to the extent that Emma, who had said that she did change her behaviour to fit in with what was expected of her, was able and willing to share the impact of these expectations on her: Sometimes I can -especially on a bad day when it's dark, and I'm really tired [. ..] -I will just be lying there in my bed, in the dark and alone, and thinking that I'm not good enough and PEDAGOGY, CULTURE & SOCIETY then I'll ask myself 'why am I not good enough' and the answer usually comes up as 'because I don't look nice enough' or sometimes it's just simple 'because I am a girl' and I feel like a lot of other girls have to go through this.And no-one [. ..] boys or girls or any other gender, should have to go through that.(Audio-recording, session 3) This confession indicates the damage that expectations of looks do to girls and connects with a recent educational report from the Department of Education which found lower levels of happiness and higher levels of anxiety in girls compared to boys (Department of Education 2020).
In my own initial reading of 'The Pea and the Princess', I had perceived it as an irreverent eco-feminist tale that employed some of the traditional subversive fairy tale tropes such as inanimate objects made animate and the underdog winning through (Zipes 1982).I had particularly liked the book's uncoupling of Hans Christian Andersen's original association of a 'real' princess with hyper-sensitivity showing the 'princess' instead as practical and down to earth.However, the girls had a different, more critical perspective.For them, it was not enough that the paradigmatic structure of the fairy tale had changed -they articulated clearly that the text still contained too many damaging gender models and expectations which impacted the girls' sense of self-worth.
It would be possible to situate these conversations in a post-feminist discourse, but that would only be the latest in a long line of rhetorics that all focus on the same constructions of 'perfection' encapsulated in the figure of the princess.The princess is an archetypal figure in Western European fairy tales.She embodies the feminine beauty ideal -the socially constructed notion that physical attractiveness is one of women's most important assets and something all women should strive to achieve and maintain -which is still a dominant construct of fairy tales and children's books (Baker-Sperry and Grauerholz 2003), even in the feminist retellings we discussed in the group.The limitations that the notion of 'a real princess' imposes on both males and females are encapsulated in Violet's perceptive comment: [T]he fact that it says 'the real princess' kind of means that he's just looking at princesses instead of everybody, so he might just be looking in the wrong place?(Audio-recording, session 1) Violet highlights that the traditional configuration of the princess, even if subverted by contemporary references and values, is a limitation of these books.As such, she points to the fact that, paradoxically, both 'The Pea and the Princess' and 'The Paperbag Princess' are based on the same stereotypical constructs they explicitly seek to challenge.They are, in fact, what Kuykendal and Sturm (2007) term 'fractured fairy tales' and, therefore, not entirely successful as feminist fairy tales.
Feminist retellings essentially seek to transfigure and disrupt the established genre of fairy tales by either making paradigmatic or syntagmatic changes (Zipes 1982).While the syntagmatic changes inherent in 'The Paperbag Princess' were apparent to the members of the book group when pointing out stock tropes 'well, stereotypically dragons, eat princes and princesses' (Megan, Audio, session 2), they did not passively accept these either -they did their own retellings.By making syntagmatic changes to the stories, inventing backstories and alternative endings -what Skidmore (2000) terms counterfictionals -they extended the narrative beyond the text in an act of emphasising or resisting themes of the text in front of them: Emma: [. ..] if someone were to [. ..] rewrite that book in a modern setting, she would go to law school and become a lawyer to rescue the prince from a legal situation.But then he's like, 'you're not girly enough', and she's like, 'Okay, I have money; I can live happily.Goodbye!' Megan: [. ..] at the end [. ..]I think she's skipping off [. ..]I hope . . . .I imagined [. ..] she comes home and she changes things, and she becomes Queen without a husband [. ..] to show young girls that they don't need a man.[. ..]Mathilda: I would rewrite it and say that the girl stays single forever and [. ..] maybe she got a job, but maybe also she just stayed at home and didn't have a job or just stayed at home and maybe like became a painter or something [. ..] then she gets to go off and live her life however she wants it to.She stays single . . . .[. . .]Megan: [. ..]Then I would write that she went home [. ..] . ... she argued with her parents and, like, just try to change laws so that she could rule without having a prince, and then she did that, and then she became Queen, and she never had a man again because she didn't need them.She was strong on her own.[. ..] (Audio-recording, session 2) As the girls are constructing emancipatory scenarios for Elizabeth in which she variously goes to law school or negotiates changes to the line of succession to allow her to become Queen in her own right, they emphasise themes of independence -whether this is expressed as financial independence, artistic endeavour or simply the freedom to live a life of her own choosing.While these are scenarios of empowerment ('I hoped, I imagined'), it is potentially problematical that in order for them to envision a girl being free, she has to be alone -'single' and 'strong on her own'.The societal discourses of success that these girls are exposed to are also evident; by linking becoming a lawyer and monetary success with freedom, Emma's point is closely associated with the postfeminist ideals of economic capacity described by McRobbie (2009), while Mathilda's visions of alternative ways in which to live your life are possibly more reflective of a post-Covid social GenZ ideal.We did not explore either of these points further in the session, nor did it come up again.In future, this would be an interesting area to explore, particularly since, as the participants repeatedly acknowledged by referencing 'younger girls', children's literature plays a direct and important role in modelling gender constructs, whether these are generally socially approved or not.
In the same conversation, the participants discussed what happened to Ronald: Ronald's imagined end became more violent as the participants crystalised their dislike of his response to Elizabeth; at first, he is merely extradited from a sticky legal situation, then he is left behind, then he is 'punished' by the dragon, and finally, he is definitely eaten by it.Sipe and Ghiso (2004) counsel us to pay attention to transgressive interplay or readings.
Aside from the creative joy the girls clearly derived from this, I interpret these syntax and paradigm disruptions as part of their resistance to the text.By questioning and changing the paradigm of the stories, they used the dialogic space not only to create meaning through story-making but also to challenge and rupture the frame of the text.As such, my participants' counter-fictionals constituted a playful yet profound resistance to the hegemonic narrative structures and conventions of fairy tales -even of feminist retellings.

Anger as an emotion of resistance
Psychologists attribute a double role to anger: it is both an inwardly directed signal to overcome a negative situation and also 'an outwardly directed communicative signal establishing differentiation' (Williams 2017, 1).This dual role of anger was evident in the conversations my participants had, notably in their discussions of 'The Paper Bag Princess'.Davies (1991) used 'The Paper Bag Princess' to conduct seminal research with a group of 4-year-old children.She found that 'most are unable to see Elizabeth as a genuine hero' (60), concluding that the children saw it as a 'fairy-tale-gone-wrong'; they wanted 'a different, a "proper" ending' (68).Essentially, they wanted Elizabeth to scrub up and look like a princess so that Ronald would want to marry her.However, it was very clear from the book group discussion that these views were not shared by my participants.I found that, at age 12, my participants had reached an adequate level of maturity and had sufficient discursive histories to critique stereotypes and envision alternative positioning.Kuykendal and Sturm (2007) argued that 'The Paper Bag Princess' was unsuccessful as a feminist fairy tale because it was 'fractured' and 'that children [. ..] do not identify with their one-dimensional protagonists' (41).However, while my participants were not uncritical of 'The Paper Bag Princess', they had no problems identifying with Elizabeth.These girls essentially thought Ronald was ungrateful and undeserving of being saved and admired Elizabeth for her resourcefulness (Emma: 'I mean, the paper bag dress -she should be a fashion designer, to be honest, to make that out of a big paper bag is very impressive', audio-recording, session 2), bravery and cleverness in tricking the dragon: The girls approved strongly of Elizabeth's anger as an impetus for change in the storyand they amplified it by actively voicing out various responses that they inferred from the text and the images.This voicing out is another example of performativity (Sipe and Ghiso 2004) vis a vis the text, which aided the girls in expressing their resistance to Ronald's (and, by extension, society's) expectation that girls' looks are more important than their character and actions.Thus, the emotion of anger was also used by the girls to establish differentiation.The girls approved of Elizabeth's anger not only because they felt it was appropriate to the situation but also because the emotion of anger itself runs counter to the stereotypical portrayal of princesses: Emma: I like the fact that she's angry because, at the time, a lot of princesses of fairytales would sort of have been breaking down sobbing and praying for another prince to come along and help her [. ..]I like the fact that she's looking angry.She's going to do something about it -not 'I'm going to break down, and somebody comes to fix this for me because I'm a weak little thing' . . . .(Audio-recording, session 2) Further, the fact that anger was portrayed at all was picked up in the interviews between Violet and Mathilda.In the following transcript, I had asked them to interview each other to discuss which kind of princess they would like to write or to be themselves: Mathilda: [I would not be] one of those like pretty ones that sit around all day.
Violet: Perfect, and stuff.Mathilda: yeah, definitely not perfect (. ..) if I had to write a fairy tale, I would just make myself angry just to prove that princesses don't have to [. ..]I would just make myself have a temper because I feel like girls often are like silenced because they think they're not allowed to have like a temper (Audio-recording, session 3) For Mathilda, anger and perfection are antithetical; she defines her ideal protagonist as angry instead of 'pretty' and 'perfect'.This is a sophisticated form of resistance to a hegemonic definition of successful femininity as docile and the sexism that requires girls to be nice to be accepted or well-regarded (Reay 2001).As such, she is making an explicit point about how these expectations silence girls.
This conversation about anger led to a more general discussion about the behaviour expected of girls and how these gender expectations were stopping these two from 'being themselves'.They, characteristically, both advocated for girls to be accepted as quirky or shy if that is what they were: Mathilda: Some girls are just naturally shy girls [. ..]Certain people would just say, 'Oh no, you don't have to act like that anymore, you can speak your mind, you can be brave' [. ..]Um, and I think some girls can just be shy [. ..]It's not always the men who are making her uncomfortable [. . .] for example it could be the girls in the environment, or it could also just be [. ..] a new environment [. ..] it makes me wish that girls could just be how they are and people wouldn't jump to these conclusions like, Oh, she thinks that she can't speak like this we need to change that, we need to change her'.(Audio-recording, session 3) This discussion could be situated in the fact that Violet and Mathilda are both at all-girls schools where the discourse is firmly aimed at empowering the girls to speak up for themselves.Indeed, in contrast, Emma's 'anecdotes' from school life were centred around being silenced and intimidated by boys ('They were just shouting at me the entire time and telling me to shut up and do what I was told.The entire time!' Audio-recording, session 4).However, as Mathilda articulates clearly, these expectations -even if aimed at empowerment -can also be limiting for naturally reserved girls.Perhaps, as facilitating adults, whether parents or teachers, we should aim to remove barriers for girls to be outspoken rather than requiring everyone to be so.

Methodological learnings and book group as a liminal space
The fact that the research took place outside of a school context caused me some concern during the planning and execution stages.While there were many advantages to avoiding the structure of an institutionalised framework (see also below), I missed the support it gave as far as safeguarding was concerned.Though no such issues were raised, I had resolved I would utilise the relevant school's safeguarding process.However, a more likely scenario for a PAR pertained to issues of 'guilty knowledge' (Adler and Adler 1996) -what to do if issues were raised in our sessions which I felt the parents should know about.This issue did come up when one of the participants related a real-life experience of genderspecific inequality and misogyny which had taken place in the classroom.In this instance, I stepped out of the researcher role and into the parent role, firstly to affirm and validate the girls' feelings and experiences in the instance of telling and subsequently to confirm with the relevant parent that they were aware of this incident.
I had also been concerned that the girls would feel my role as a parent overshadowed that as a researcher and consequently feel reticent about expressing their views.Adler and Adler (1996) argue that PARs have a diminished role pretence and a greater relational identification with their children-as-subjects compared to other research roles.This gives them a reflected identity such that the child subjects may reveal things they would not ordinarily reveal to an adult.In my research, certainly, this also seems to have been the case.In the exit questionnaire, Violet commented: 'I think [you being . . . .'s mother] helped me to talk more about my thoughts because I already know you, and therefore I felt more comfortable'.The combination of the mitigating factors that I put in place and the fact that trust pre-existed in the relationship thus appears to have deepened the dialogue rather than the reverse.
As action research practices informed the study, I tried out different approaches to engaging with the text in each session, moving from a tightly focussed textual page-bypage walkthrough in the first session to merely using the text as a starting point in a 'Philosophy for Children' inspired conversation in the last session.I found the participants had the most interesting and spirited conversations in sessions which were structured, but loosely so, and where the adult lens was de-centred in such a way that the girls themselves could direct the dialogue either through discussion of key themes they had identified in their reading (as we did when discussing 'The Paper Bag Princess' in session two) or in the peer interviews (session three) at which I was not present.In other words, the more opportunities the girls were given to collaborate or co-direct the research, the greater the opportunities for profound insights.In her review of the role of children in ethnography, Levey (2009) states that few researchers use children as collaborators in conducting interviews, for instance, and that there is still a potential bias against taking their knowledge and opinions as seriously as adults.This is probably less the case in the field of educational research, but here research is often subject to institutional gatekeepers (Adler and Adler 1996) who may impose restrictions on the types of conversations, the type of activities or the learning outcomes expected.Further, schools are constrained by curriculum requirements, and there might be little space for subversive conversations.For example, Emma mentioned that she had been taught 'The Paperbag Princess' four years previously when she was nine, and at no point had they discussed any gender issues.To her, it was undoubtedly crucial that the book group was not in a school setting -she resisted completing formal preparation for the sessions because, she said, 'Book group is fun; I don't want it to be like school' (captured in my notebook, 10/6/21).
All the girls participated enthusiastically in the sessions; indeed, the conversations were interspersed with frequent utterances of 'yeah!' or 'I know!' and 'Yes, right?' as the girls echoed and supported each other's views and experiences.In this way, the social 'context of the reading activity' (Maine 2014, 151) enhanced the participants' motivation.Jones points out that power and identity are central to whether readers feel 'any sense of entitlement' (Jones 2013, 220) to analyse and critique dominant discourses.In this sense, empowerment is a product which could result from my research.Since the girls all identified as feminists at the outset of this project, I had wondered whether the book group discussions would make a difference to how they saw the world and each other from a feminist point of view.However, the exit questionnaire I asked the girls to complete clearly showed that they had both enjoyed the sessions and learnt from them.The sessions thus worked on multiple levels.On the textual, reader-response level, the sessions enabled the girls to voice a critique of the current supposedly feminist children's stories, and the girls came away with an increased awareness of the textual limitations: 'I think I need to be more alert for these kinds of small anti-feminist jabs' (Mathilda, Exit Questionnaire).
The sessions also worked on the social and political level and fostered a sense of solidarity and increased understanding of problems faced by their peers: I learnt that we all had quite similar views about the treatment of women and how it should be changed [. ..]I found out about the problems that others around me face in and around school, whether it is homophobic friends or misogynistic classmates (Violet, Exit Questionnaire) These comments show that the book groups challenged and enhanced the participants' literary and general critical skills.
From the briefing process, the girls were aware that we would be discussing gender equality and feminist fairy tales, and this inevitably framed the conversations.Combined with the fact that the girls knew that I am a feminist, this probably allowed them to express their views more confidently and clearly.I deliberately set out to try to create a safe space without the usual constructs of adult-child authority and in a setting which was different from the usual spaces for social interaction (not-school, not-social).The space provided by the group, physically and mentally, and my presence there (as a parent researcher, a feminist woman, but not a teacher) arguably helped the girls articulate their views in ways that might not be possible in a classroom or social setting.The shift in power and agency that happened as the girls actively collaborated in the research led to the growing sense of 'entitlement' to critique dominant gender norms referred to by Jones (2013).As such, I would argue that the book group functioned as a kind of liminal space -a space away from the everyday life of school and family, stripped of the expectations inherent in these established social contexts.It provided a space where the four girls were free to explore how they and some of their peers were affected by gendered issues and, in some small way, helped them prepare for the transition from 'tweens' to teenagers.

Conclusion
In this article, I have focussed on two emergent issues raised by the girls: the pressure on women to be perfect and anger as resistance to expected feminine appearance and behaviour.These were interconnected; the expectation of 'perfection' encapsulated in the 'perfect princess' as quested for in 'The Pea and the Princess' or Prince Ronald's request for Elizabeth to 'come back when you are dressed like a real princess' ('The Paper Bag Princess', np) angered them.The girls used this anger to resist and circumvent the conformist gendered discourse of these expectations.They were able to position themselves in the literature, resisting, rupturing and challenging stereotypes through empathy, performativity, and relating the themes of the fairy tales to their day-to-day lives.Resistance was brokered by the girls offering 'counter-fictionals', extending the narrative beyond the plot in the text to explore their ideas further.This included imagining different endings to stories, adapting plots to reflect their values, while also critiquing and analysing flaws in the original plots.
In this research, I occupied the role of Parent-As-Researcher. I worried about this in advance of the project, having found 'home-schooling' my children during the Lockdowns of 2020 less than joyful for all concerned.However, the power differential between the researcher and research subjects was partly addressed by the collaborative nature of the research and partly by the fact that I was genuinely interested in their views without holding any kind of teaching agenda.I found conducting collaborative research with the girls such an overwhelmingly positive experience that we have extended our collaboration to another project.I found that the book group offered potential as a liminal space -not an education setting but also not their usual social setting -acting as a site for different kinds of reflection and discussion, exploring and subverting existing gender constructs and constraints.Further, within the setting of the book group, the benefit offered by the explicitly feminist orientation of the conversations created space and supported the girls in their exploration and expression of social and political critique.Through the discussion of feminist fairy tales, the girls within the study were able to draw parallels to their own navigation of complex gendered narratives and expectations of girlhood.
Overall, my findings underline the challenging conditions under which these girls forge their understanding of the relationship between literature, identity and power and how this reflects in their lives.Of course, children are not passive recipients of their gender roles: 'children resist, rework, create and influence adults as well as each other' (Thorne 1993).However, the girls' discussion of the gendered expectations they experience supports Reay's findings that conformist discourses continue to exert more power than transgressive or transformative ones (2001).In contrast to her students, though, my participants were not silenced by it -at least not when given the dialogic space -instead rebelling against the discourses of gender, power, and culture through their engagement with the text.Nevertheless, this is not to imply that a resistant reading entirely negated any issues or feelings of inadequacy.The participants increasingly drew parallels to their own experiences as the sessions progressed, culminating in some very frank and heartbreaking comments on the impact of misogyny and gender expectations on these young girls.However, given that much of the research with girls and young women suggests that inequities borne of sexism do not surface in research discussions (Pomerantz, Raby and Stefanik 2013), the fact that they were discussed at all is encouraging and, in some ways, liberating.

Table 2 .
Overview of the texts.
is waving like come help me, help me, but then when [she] comes, he's like, Oh no, you're terrible, go back and come back when [. ..] you look like a real princess, and it's like if you want her to help be nice to her! [. ..]Mathilda: I think she looks really angry because she's just come all this way to save the prince that she's like in love with, and he couldn't care less about her, really, and then I think she's starting to realise that they're not a [. ..] a good match.
Megan: [I]t also looks like the princess defeated the dragon [. ..] the prince came all the way here, tricked a dragon to save you, and you're being angry with me', and it's like[...]at the end her voice might raise a bit like she's angry [. ..]She doesn't look sad.[. ..]She looks like she's fine.[Acting out with a different voice]: I look fine!I just saved you.Does it matter what I look like?Mathilda: She looks fine for having her house burnt down and having come all that way!