‘Ordinary’ and ‘diverse’ families. A case study of family discourses by Finnish early childhood education and care administrators

ABSTRACT The increased family diversity is a major global trend. Although family configurations are also diverse in contemporary Finland, it has been argued that Finnish family policies and institutional understanding of family life continues to focus on the heteronormative two-parent family with a native Finnish background. To address this issue, we analysed Finnish family discourses through qualitative interviews with early childhood education and care administrators (n = 47), applying a discourse analytic framework. Our results suggest that families are discussed through two divergent but interwoven discourses, i.e. the discourse of ordinary families and that of diverse families. The former focuses on heteronormative two-parent native-born Finnish families, perceived as ‘ordinary’ and familiar, and the latter on, especially, LGBTIQ and immigrant families, perceived as ‘new’, confusing and strange. We demonstrate how diversity is produced discursively by othering families that diverge from the ‘ordinary’. These discourses, reflecting wider cultural understandings of family, may have implications for both families and institutional practices. We conclude by arguing that the conventional and heteronormative understanding of family remains entrenched in Finnish family discourses.


Introduction
For decades, social scientists have argued that the meanings, possibilities and boundaries for family and family life are profoundly defined, negotiated and constructed in institutions such as schools, workplaces, health care and religious communities (Giddens, 1991;Gubrium & Holstein, 1990;May, 2013;Smith, 1993). For example, three decades ago, sociologists Jaber Gubrium and James Holstein (1990) in their pioneering book What is Family? illustrated how institutional and organizational settings contain builtin conceptions, ideals and practices regarding what a family is or what it should be. To date, however, the encounters between institutions and families have predominantly been studied through the accounts and experiences of individuals and families, as institutional understandings on families and family diversity have remained underresearched. In this article, we contribute to filling the research gap by presenting a case study on family discourses produced in the institutional context of Finnish early childhood education and care (henceforth ECEC).
While recent research on families and personal relationships has highlighted the increased diversity of family lives and the transformation of cultural ideals and concepts of family, it has also been argued that some families continue to be perceived as more culturally valued, accepted and legitimate than others (Heaphy, 2011;Lahad et al., 2018;Ribbens McCarthy et al., 2013;Walsh, 2018). That is to say, the socalled nuclear family, comprising a male and female parent and their children, continues to flourish as the cultural ideal of a 'proper' family (Chambers, 2001). We can also expect this cultural ideal to be present in institutional understandings of families and family diversity. For example, in Finlandthe context of our study and an example of a Nordic society -, despite its long history of promoting the equality and wellbeing of families, there is evidence that normative ideas of family continue to exist in institutional settings (Homanen, 2013;Onnismaa, 2010). It has been argued that Finnish family and childcare policies remain targeted at native-Finnish heteronormative two-parent families with the mother in the role of primary caregiver (see, e.g. Moring, 2013;Yesilova, 2009).
In this article, we study family discourses in the institutional context of Finnish ECEC through qualitative interviews with local ECEC administrators (n = 47). In Finnish society, ECECa first step in Finland's educational systemis an institution which almost all children attend at some point before starting formal education at the age of seven (Karila, 2012). Municipalities in Finland have considerable autonomy over how they organize their ECEC services. Local administrators guide parents when it comes to making decisions on ECEC, such as whether to continue caring for their children at home with the help of a home-care allowance or to opt for ECEC in either a public or private ECEC centre. Local administrators are also in charge of drawing up the local ECEC curriculum, based on the National Core Curriculum for ECEC, and drafting suggestions concerning where and how ECEC services will be located and organized in their municipality. Thus, they have a powerful position in defining and constructing family in both local ECEC policies and praxis.
Our research question was: What are discourses of family produced by Finnish ECEC administrators when they describe municipal ECEC services?
Qualitative interviews with administrators, gathered in ten municipalities across Finland in 2016, were discourse-analyzed. This case study on family-related discourses produced by Finnish ECEC administrators contributes to the fields of family studies and the sociology of family life by (1) extending knowledge on institutional understandings of families in the Finnish and Nordic context and (2) demonstrating how cultural ideals and understandings inform institutional family discourses.
We begin by discussing recent global changes in family life and the cultural ideals and research paradigms related to families in the institutional context. Next, we familiarize the reader with the Finnish context. We then present the data and method, followed by the results. Finally, results are discussed and conclusions drawn.

Background
Changing families: diversification, ideals and the research paradigm It has been argued that global changes in both family relations and the cultural ideals of families have taken place in recent decades (e.g. Carrington, 2002). In parallel, a shift in the research paradigm has also been noted (e.g. Chambers, 2012). To set the context of this study, these developments are discussed below.

Family diversification
A wide array of social science research has pointed to marked changes and trends in families and personal relationships over recent decades (see e.g. Lahad et al., 2018). First, the diversification of family lives and family forms is generally regarded as a major global trend (Beck & Beck-Gernsheim, 1995;Chambers, 2012;Forsberg & Nätkin, 2016;Giddens, 1992;OECD, 2011). It has been shown, for example, that low fertility, the entry of women into the labour force, voluntary childlessness and postponed parenthood have reformulated families globally since the 1970s (Forsberg & Nätkin, 2016;OECD, 2011). Second, and more recently, attention has been paid to divorce, remarriage, post-divorce families, blended families, single parenthood and joint custody, the increasing number of non-heteronormative families and growing global migration as key diversifiers of family relations in the 'global north' (Chambers, 2012;Gahan, 2018;Lahad et al., 2018). In sociological terms, Chambers (2001) basing her work on theoretical claims about the individualization and democratization of intimate and personal relationships (Beck & Beck-Gernsheim, 1995;Giddens, 1991Giddens, , 1992 has described this phenomenon as 'rising postmodern family diversity'.

Family ideals
Cultural narratives, ideals and discourses create, mediate and restrict the opportunities that individuals have to live, organize and narrate their family and personal relationships. Carrington (2002) has described how, in the mid-twentieth century in western societies, the nuclear family became the culturally dominant 'proper' and 'normal' family model (Carrington, 2002, p. 73). After that, family ideals and discourses have expanded and diversified. 1 According to Chambers (2001, p. 20), cultural discourses reflect the democratization and individualization of intimate relationships, and a wider variety of relationships have begun to be accepted as family. However, several researchers (e.g. Chambers, 2001;Heaphy, 2011;Lahad et al., 2018;Valiquette-Tessier et al., 2016;Walsh, 2018) have argued that the heteronormative nuclear family continues to flourish as an ideal, or as Chambers (2001, p. 1) puts it, 'as a symbol, discourse and powerful myth within the collective imagination'. According to Heaphy (2011;see also Chambers, 2001;Ribbens McCarthy et al., 2013), some families are perceived to be more 'normal' and legitimate than others while certain unconventional types of family life are pathologized, stigmatized, or perceived as troubled, in a way or another. For example, a meta-analysis of studies published after the turn of the millennium on family stereotypes shows that nuclear families represent an ideal against which other family types are compared (Valiquette-Tessier et al., 2016). As Gahan's (2018) study on Australian separated same-sex parents, Lahad et al. (2018) work on blended families in Israel, and Walsh's (2018) research on migrant families in the UK demonstrate, ideals of family put lot of pressure on people living in unconventional or non-idealized family configurations, as they need to display and perform their families as 'proper' and 'decent' to play down their position as 'other'.

The research paradigm
The latter half of the twentieth century saw a major shift of research paradigm in family studies away from the uniform, nuclear-like family (Parsons, 1955) towards more fluid and diverse conceptualizations of peoples' close and intimate relationships (Beck & Beck-Gernsheim, 1995;Giddens, 1992). This change of paradigm originated in the late 1960s, when feminist scholars criticized the persistence of the ideology of the (white) nuclear family as an assumption in family studies (Allen et al., 2009). Initially, the notion of 'natural' gender roles in family life was critiqued. Later, in the 1970s, scholars drew attention to the diversity of family constellations and the impact of gender, generation, race, ethnicity, and sexual orientation on family relationships (Allen et al., 2009). In the 1980s and 1990s, many of the ideas and concepts introduced by feminist researchers were incorporated into the mainstream of family sociology (Chambers, 2001). Since then, alongside the concept of family, social scientists have studied families, family lives and personal relationships with conceptualizations and sociological tools such as belonging (May, 2013), relatedness (Carsten, 2000;Smart, 2007), personal life (May, 2011;Smart, 2007), practices of intimacy (Jamiesson, 2011) and linked lives (Bengtson et al., 2002). The concepts of 'displaying' and 'doing' family (Dermott & Seymour, 2011;Finch, 2007) both based on David Morgan's (1996) concept of family practiceshave also been drawn on to better understand the performative, discursive and constructive but also multifaceted, fluid and intersectional, character of family life. Hence, the focus in recent sociological family studies has primarily been on the personal and intimate nature of family life. Encounters between institutions and families have also, with rare exceptions, mainly been investigated through the accounts, narratives and experiences of individuals, families and family members (see e.g. Walsh & Mason, 2018).

The Finnish context
In this section, we first provide a brief overview of families in Finland, along with a glance at the discursive 'familistic turn' in Finnish society in the early 2000s. We then discuss Finnish ECEC and recent research results on Finnish families in institutional settings.

Families in Finland
Statistics Finland defines a family as a unit of a married or cohabiting couple or persons in a registered partnership 2 and their children living together; or either of the parents and his or her children living together; or a married or cohabiting couple and persons in a registered partnership without children. (Statistics Finland, 2020a) On this definition, the total number of families in Finland was 1,476,000 at the end of 2016 3 (Statistics Finland, 2017). While the total number of families has steadily increased, the number of families with children has decreased, mainly due to decreasing fertility. In 2016, the total number of families with children was 570,000. Approximately four-fifths of these families were married couples (59%) or cohabiting couples (20%) with children. Approximately 10% of the families were categorized as blended families, that is, family units with at least one child from one or both parents' previous relationships. Singleparent families accounted for approximately one-fifth, of which almost nine out of ten were mother-child(ren) families. The vast majority of the two-parent families were with other-sex parents. Of the married couples with children, only 733 were same-sex couples (approx. 0.2%). While the proportion of family arrangements diverging from the 'nuclear model' has increased during the 2000s (Statistics Finland, 2020c) and statistics alone do not give a complete picture of contemporary family diversity, the numbers suggest that the heteronormative other-sex parent family with children has remained the commonest family form.
Ethnic and cultural minorities have always existed in Finland. The 'old minorities' include the indigenous Sami, Romani and Swedish-speaking populations (Raento & Husso, 2001). Whereas the Swedish-Finns have been a 'strong minority' with clear institutional status and considerable power in politics, culture and the economy, the Sami and the Romani minorities have experienced social, political and cultural marginalization (Raento & Husso, 2001). However, unlike many other European countries, Finland was for long ethnically and culturally rather homogeneous with low levels of inmigration until the late 1980s (Saukkonen & Pyykkönen, 2008). Between 1990 and 2017, the Finnish population with an immigrant background increased tenfold (Statistics Finland, 2020b). This has meant a growing number of families with immigrant parents. Currently, the population with an immigrant background accounts for approximately seven per cent of the general population. Thus, demographic change has been swift, and multiculturalism is now part of present-day Finnish society, at least in the major cities and urban areas (Statistics Finland, 2020b; see also Saukkonen & Pyykkönen, 2008). The largest foreign populations in Finland at present are speakers of Russian, Estonian, Arabic and Somali (Statistics Finland, 2020b).
Finland, one of the Nordic countries categorized by social policy researchers as dualearner welfare states, has a long history of supporting equality between children, genders, families and different social groups (Leira, 2002;Forsberg, 2005). For example, according to Forsberg (2005), the 'nuclear family model' -(i.e. father as sole breadwinner and mother exclusively caring for the children)never became dominant in Finland (Forsberg, 2005). Some scholars have, however, argued that contemporary Finnish family policies are dualistic, i.e. they promote gender equality but also cement the gendered division of care (e.g. Rantalaiho, 2010). The latter has been seen as an outcome of the child homecare allowance (Rantalaiho, 2010), which supports mothers' informal home-care arrangements (Kröger et al., 2003). In addition, the criticism has been raised that Finnish family policies are targeted mainly at native-Finnish heteronormative twoparent families (see, e.g. Moring, 2013;Yesilova, 2009; for more information on Finnish family policies, see Ministry of Social Affairs and Health, 2013).

Public interest in family life
In the national public and political arenas, families, family lives and parenting have been hot topics throughout the 2000s (Jallinoja, 2006;Sihvonen, 2020). Based on her analysis of family discourses in the Finnish media, Jallinoja (2006) has argued that an ideological shift known as the familistic turn took place in Finland in the early 2000s. In her account, the familistic turn was manifested in the Finnish media in an increase in discourses on the family as the key social unit and in growing societal concerns about parental skills and children's wellbeing (Jallinoja, 2006). Jallinoja shows how this new wave of familism was constructed from both modern and traditional discourses and ideas about family life. That is, although 'modern' features of family lifesuch as diversity, the individualization of personal relationships and fathers' increased involvement in caregivingwere emphasized, it also upheld traditional values, such as maternal primacy and the nuclear family as the norm.
Although the familistic turn took place almost two decades ago, since then families, children and parenting have remained widely discussed both in public and political arenas in Finland. During the 2010s, attention has been paid especially on men's role in families (Author's own), parenting support and declining birth rate (Sihvonen, 2020). When the interviews for the study reported in this article were gathered, public attention was paid especially on immigrant families and families of same-sex parents, mainly due to the European refugee crises in 2015 and the law allowing same-sex marriages taking effect in early 2017 (for more information on the European refugee crisis in Finland, see Seppälä et al., 2020; on the gender neutral Finnish Marriage Act, see Lahti, 2019, p. 17).

Finnish early childhood education and care
Early childhood education and care (ECEC) refers to integrated services for 0-6-year-old children. In Finland, it is attended by 98% of all children at some point before the start of formal education at age seven (OECD, 2018). Public ECEC is the most common institutional ECEC arrangement in Finland, as municipalities are obliged to organize the service for children whose parents request it, regardless of parental employment status (Author's own). Municipalities can either provide the services themselves or they can subsidize private sector ECEC by offering vouchers or by topping up the private daycare allowance to which families are entitled if they wish to use private ECEC services.

Families in institutional settings
Some relatively recent research on Finnish families in institutional contexts has found that the heteronormative two-parent family with children is commonly understood as the family norm and ideal. For example, Homanen (2013) found that while awareness of the diversity of family forms has increased, the two-parent heteronormative family continues to be perceived as the ideal in prenatal clinics and that deviation from the heteronormative model can be considered a risk for the family and unborn child. Parallel findings in ECEC settings were reported by Onnismaa (2010), who concluded that heteronormative two-parent families are considered 'normal families' and that families not fulfilling these criteria were often seen as deficient or problematic in one way or another. These studies also highlight how, in addition to family form, features such as the parents' mutual commitment and both parents' participation in working life were attached to the idea of a 'decent' family. As Forsberg (1998) notes in the contexts of child welfare and social work, these beliefs, norms and ideals about family are invariably present in the work of experts working with their clients. Thus, it is important to ask how the cultural ideals of family are reflected in the discourses produced by ECEC administrators.

Interviews with ECEC administrators
In this article, we study the discourses on families produced in interview talk by Finnish ECEC administrators. Our empirical data comprise 47 qualitative interviews conducted in 2016 with municipal ECEC administrators. The interviewees were mainly mid-level administrators, responsible for the municipal ECEC. Their job titles included Director of Preschool, Service Advisor and Pedagogical Coordinator, and most were in contact with families on regular basis. A few high-level administrators, such as heads of municipal education, were also interviewed. The semi-structured qualitative interviews focused on municipal ECEC policies and service provision. The interviews were gathered as a part of the multidisciplinary research consortium Finnish Childcare Policies: In/equality in Focus (2015-2021) focusing on potential sources of inequality in Finnish childcare policies and early education. Recently, for example the interpretative frames and discourses pertaining to the organizing and provision of local ECEC have been examined based on the interview data used in this article (Eerola et al., 2020). The consortium was funded by the Strategic Research Council of the Academy of Finland.
The interviewees were drawn from ten municipalities selected for the research project owing to certain important differences in their populations, demographics and geographical location. As this was a qualitative study, the interviewees were not recruited as a representative sample of Finnish ECEC administrators. Instead, their invitation to participate in the study was purposive, the aim being to ensure wide contextual variation.
With the exception of five male interviewees, all the interviewees were women, largely because ECEC at the municipal level in Finland is predominantly administered by women. All the interviewees had at least a BA-level education. Participation in the interview was voluntary and all the interviewees gave their informed consent to participate in the study. Since most of the interviewees were female, they are all referred to as she to protect their anonymity, particularly in the case of the male informants.
The interviews were conducted by a team of nine researchers, including all the present authors. Mean interview duration was 90 min. To ensure consistency in the implementation of the interviews, two joint training sessions for the interviewers, all of whom had previous experience in qualitative interviewing, were held. The first sessions took place before commencing the interviews, and the second after the first few interviews had been conducted.

Analysing the administrator interviews
Our data analysis method draws on a diversity of discourse analytic approaches (Antaki, 2008;Van Dijk, 2011) and thus can be characterized as 'generic discourse analysis', which Antaki (2008) described as a working procedure that aims to make sense of a specific domain or topic through, for example, interview transcripts. By taking this stance we are committed to an idea that the use of language is political and it (re)produces culturally constructed views of the world (Gee, 2017).
In the analysis, our aim was to identify the discourses of family produced by Finnish ECEC administrators. Here, discourse refers to the composition of sentences that constitutes shared meanings and a collective way of organizing our knowledge and experiences of the world. The study of discourses is examining language use and its active role in meaning-making, and maintaining normativity (Gee, 2017). We understand discourse here as 'the ways in which meaning is mobilized for the maintenance of relations of domination' (Thompson, 1984, p. 5). This aligns with a poststructuralist Foucaultianinfluenced approach, which places emphasis upon power and it's material-discursive effects. The use of language inevitably privileges some people, while silencing others. It thus creates a lens through which families are discussed. We analyse interviews of ECEC administrators in an effort to identify the characteristics that are constructed as necessary for a group of people to achieve the status of a 'family'. In addition, we examine what kind of family types administrators construct, what kind of families are given privileged position and what kind of families are positioned in the margins. Thus, the analysis can be characterized as feminist: it aims to understand how cultural constructionsin this case, constructions concerning familyare connected to inequality (see Gill, 2009).
We commenced our analysis by reading through the interviews and using Atlas.ti to select all the fragments containing references to families and family life. These excerpts were then analysed by focusing on how the interviewees conceptualized family and discussed family diversity, including the kinds of statements about families that were taken for granted, the kinds of categorizations of families that were constructed, and how the interviewee positioned herself in relation to different types of families, for example, by using the pronouns 'we' and 'them' (see Gee, 2017). We continued analysing, what kinds of positions in the institutional context of ECEC were given to the different types of families. The analysis yielded two key discoursesordinary family discourse and diverse families discoursethrough which families were discussed. In accordance with the social constructionist approach, we are aware that the identified discourses are our interpretations and were produced together in the interview talk of the interviewees and interviewers.
We conducted the analysis with the original Finnish interview transcripts and translated the excerpts chosen for the manuscript into English. Translating interview transcripts from one language to another is a delicate process that requires the researcher to make fine-grained choices. As Nikander (2008) notes, this process is often hidden and rarely visible to the reader. It is also a lonely process in which researchers need to come up with solutions of their own, as translating interview transcripts has remained a rather neglected topic in qualitative research literature (Nikander, 2008). We translated the chosen samples with the aim of conveying both the general sense of the talk and the most important discursive nuances such as pauses and moments of hesitation.

Family discourses by the ECEC administrators
Next, we discuss the results of our analysis by exploring the two main discourses of family produced by the interviewed administrators. These discoursesthe ordinary family discourse and the diverse families discourseare interwoven and dependent on each other and produced in parallel by the interviewees. This resonates with Billig et al. (1988, p. 24) notion of the dilemmas that can emerge when contradictory themes arise as 'a dialogue within the self about opposing values'. In the ordinary family discourse, the archetype of an ordinary family consists of two other-sex parents with children. The diverse families discourse, in turn, while it implies a broader understanding of families and family forms, is, however, produced by othering families that diverge from the ordinary family. The two discourses emphasize how, especially through factors such as heteronormativity and ethnicity, understandings of family life and its familiarity and strangeness are constructed. We argue that together, these discourses construct family in a way that reflects wider cultural ideals and norms of family life and personal relations in institutional settings in present-day Finland.

The ordinary family
Despite the increased understanding of family diversity, it is apparent from the research literature that two other-sex parents with children remains the dominant cultural ideal throughout western societies (see, e.g. Chambers, 2001). According to our analysis, this ideal is reflected in the discourse of the ordinary family, which portrays heteronormative two-parent families with children and a native Finnish background as the 'ordinary' family archetype. Our analysis shows how the 'ordinary family' as the family norm is constructed by perceiving family diversity as something that in one way or anothermainly in terms of heteronormativity and ethnicitydeviates from the 'ordinary'. That is, whereas the ordinary family is spoken about as something familiar that need not be explained, reflected on or questioned, other types of family life are understood according to their deviation from the 'ordinary'. In this discourse, the ordinary family is also a family that is recognized by and seen as suitable for the local ECEC services. It represents parents as responsible actors who, for example, contribute to society by working (fathers) and caring (mothers) and who actively make decisions on childcare, and child and family wellbeing.
The heteronormative understanding of family has a strong foothold in the ordinary family discourse. That is, though family diversity is a fact in Finnish society, it is nevertheless assumed that families have two different-gender parentsa mother and a father. For example, in our data this assumption was hardly ever questioned when the administrators discussed parents' childcare solutions. This can be seen in the excerpt below, in which an ECEC administrator from a mid-size Finnish city talks of infants' homecare. The excerpt shows that the assumption of two different-gender parents is taken for granted when parents' childcare solutions are discussed: I think it's mainly because women are paid less, so that's why they usually stay at home. There's some kind of economic reasoning behind it … Sure, there might be other reasons too, but … […] Maybe fathers would stay at home too, but women's pay is just not enough to feed the family.
In this excerpt, the interviewee contemplates why most of the parents who care for their child at home tend to be female. She suggests that in general females are paid less than males and therefore the reason might be economic. Her use of the term 'fathers' (in contrast to 'women', which she uses when referring to female parents) when speculating that males might also want to stay at home to care for their children hints at the family orientation of the male parent. The two pauses could indicate hesitation or defence (Gee & Handford, 2013). What remains self-evident and unquestioned in this construction is that a family has two different-gendered parents. Andrews (2002) has drawn attention to the importance of mothers in contemporary cultural narratives of family life. This understanding of mothers as primary caregivers and fathers as their assistants and breadwinners is reflected in the 'ordinary' discourse in the emphasis on gendered care and parental practices as part of heteronormative family dynamics. While the discourse encompasses variation in mothers' rolesgood motherhood is not defined, for example, by the mother's status as a working mother or stay-at-home mother -, maternal primacy in care work and emotional labour is broadly accentuated. The father, in turn, is described as the mother's assistant in care and the primary breadwinner. The gendered parenting roles foregrounded in the discourse are in evidence in the following two excerpts, where the interviewees are discussing the division of parental care work and upbringing responsibilities. In the first, the interviewee uses the normative term 'normal' in referring to the situation where the mother cares for her child at home. Thus, she constructs the situation not only as typical but also as 'right' and 'standard'and hence the opposite of abnormal. In the second excerpt, the interviewee emphasizes the mother's primacy in upbringing by narrating how exactly mothers (and for example, not 'parents' or 'mothers and fathers') are 'offloading their own upbringing responsibilities' onto institutional early education: Let's say if the father is unemployed, then usually the mother goes to work. But, if it's the normal situation, it's the mother who stays [at home].
In some families, the child's best interest are not always taken into account, mothers are just offloading their own upbringing responsibilities onto the day-care centre.
In addition to heteronormativity, the discourse posits Finnishnesswhich goes undefinedas another key characteristic of the 'ordinary family'. That is, the discourse distinguishes between families interpreted as Finnish from families of other ethnic origin. Our analysis revealed two discursive ways in which this was done. First, families with a native Finnish background were often simply referred to as 'families'. In turn, families perceived as 'non-Finnish' were often referred to as, for example, 'immigrant families' or 'multicultural families'. Second, our analysis revealed that the distinction between the two types of families is often manifested in talk about 'our' native Finnish families but 'those' or 'these' immigrant families (see Gee & Handford, 2013). These distinctions made through word choices commonly occurred when native Finnish families and immigrant families were being compared. This issue is exemplified in the following excerpt, in which an ECEC administrator in a major Finnish city refers to 'the Finnish families' as 'our main group' in municipal early education: As I see it, there's also lots of variation in these multicultural families too, just as much as in … as … this sounds a bit funny but, you know … in our main group … like, the Finnish families … The excerpt also indicates that the administrator is aware of the discursive tensions present in her choice of words. First, the interviewee's speech manifests difficulty of finding the correct words to explain differences between families based on their ethnicity. Second, by saying 'this sounds a bit funny', she implies her awareness of family diversification and the inappropriateness of labelling families. This indicates a tension between here-and-now interaction and established discursive practices and pinpoints the moment when dominant frames are called into question (Gee & Handford, 2013).

Diverse families
Whereas the discourse of the ordinary family draws on the familiar and traditional, the discourse of diverse families focuses on families that are perceived as 'new', confusing or strange in one way or another. Our analysis showed that diversity is a product of othering families and relationships that diverge from the 'ordinary', as the administrators recognized diversity by applying the yardstick of the 'ordinary'. The discourse also, however, supports the aim of encountering all families as equal. The discourse focuses especially on LGBTIQ and immigrant families, as they challenge the ideas of heteronormativity and Finnishness that inform the ordinary discourse. While the administrators discussed LGBTIQ families as unproblematic with respect to ECEC services, working with immigrant families was narrated as more problematic and challenging.
The discourse of changing family lives and 'new' family forms has appeared broadly in research and in public and political discussion throughout the 2000s. It is also reflected in the diverse families discourse, which is informed by a shift away from 'traditional' families towards 'new and diverse' families. That is, family diversity is perceived as a contemporary phenomenon linked to broader trends in society. The interviewees narrated the diversification of families primarily as a positive trend and the equality of families as a key aim of ECEC services. For instance, in the next excerpt a pedagogical coordinator discusses the importance of understanding family life 'outside the heteronormative box' and encountering all families as equal. The excerpt also indicates that family diversity is a new phenomenon that is not well known by ECEC staff: The world around us is changing, so what's the family nowadays? The concept of family is really broad, and I see it as really crucial to give education about this to our staff non-stop. We need to educate our personnel, to learn how to show a positive attitude, and learn how to approach diversity open-mindedly, free from prejudice that the family is always the traditional mother-father combination […] And our objective should be to strengthen equality between all families, regardless of who the family members are. It doesn't matter if there's two mothers or two fathersthey are the child's parents.
Although the diversification of family lives was mainly represented as a positive trend, some interviews revealed a yearning for the 'good old times'. For example, in the following excerpt a day-care centre director sees nuclear families as 'normal' and 'traditional' and blended families as confusing and unfamiliar, as manifested by the interviewee's difficulties in finding the right words. The excerpt also shows that the boundaries between discourses are blurred and mobile (see also Gee & Handford, 2013). That is, although blended families were generally included in the ordinary family discourse, they were also linked with the trend towards family diversification: In the old days we had these so-called normal, traditional families, but now … to be honest, there's quite a lot of these, for examplewhat's this new fancy term … these blended families.
The accounts also contained some inconsistencies, such as in the next excerpt in which the interviewee points out the importance of understanding family diversity in society but, at the same time, in her choice of words ('those are something that … '), separates same-sex parent families from 'us', thereby othering families that diverge from the heteronormative form (see Gee & Handford, 2013). Moreover, the interviewee's account of not participating in an education sessions implies a dismissive attitude towards family diversity, especially LGBTIQ families: Well, the families with same-sex parents, for example … Those are something that we need to really think about … And we actually just had a really good education session on that issueor so I heard, as I wasn't there, but anyway … the message was, we have to learn to think in kind of new ways, that there's lots of different kind of families … The coexistence of the aim of equality for all families and marginalizing those that diverge from the Finnish and heteronormative family ideal lies at the core of the diverse families discourse. Our interpretation is that both broader cultural ideals in society and concepts and understandings of the family in Finnish ECEC lack a vocabulary for discussing family diversity and diverse families in an inclusive way. Our analysis indicates how othering is often produced by discussing 'these' or 'those' families when describing families that challenge the ideal represented by the ordinary family discourse. This is evident in the following excerpts, in which immigrant families and families of same-sex parents are discussed. In the first excerpt, an interviewee discusses 'these families' and clearly differentiates them from 'ordinary' Finnish families. The second excerpt portrays families with same-sex parents as an example of 'some of these things', othering them from the norm: Among these immigrant [the ethnic group mentioned by the interviewee has been anonymized] families, there's been things, like, when we've had some events or happenings, they've just decided not to participate.
Some of these thingsyou know, parents of the same sex and other such stuffyou rarely see them [in the municipality].
Our analysis found some variation between the administrators from urban and those from rural areas. First, those from rural areas quite often represented diversity as existing somewhere else, where it does not affect them. This is the case in the previous excerpt, in which the administrator mentioned the rarity of same-sex parent families in the municipality. This can also be seen in the next excerpt, in which the interviewee first acknowledges the diversity of families but then continues by describing families diverging from the native-Finnish ideal as exceptions: Of course, all the families are different, but … In our village, for example, multicultural families are exceptions, you really don't see them here.
Second, the interviewees from the urban areaswhere diversity can be expected to be more commonsometimes spoke of 'diverse' families as more problematic than those in the ordinary family discourse. For example, an administrator from a major Finnish city described the 'bad attitude' of a particular group of immigrants: We have tried to develop our services together with these familiesmothers with this backgroundbut … […] They have problems in their attitudes, they just want to get their child into full-time day care, and they want it right away … And then they want to have all the benefits too, but … That's not the way it works.
In this excerpt, the administrator is replying to a question on whether immigrant families have been involved in service development. The conjunction 'but' indicates a contrast between two sentencesthe first sentence is about ''we', 'our' and 'us' and the later one about 'them' (see Umbach, 2004). She contrasts the attempt by her (and the group she belongs topresumably other administrators of Finnish origin) to involve immigrant families in service development with the unreasonable wishes of those families, referred to as 'they', i.e. other.

Conclusions and discussion
In this case study, we analysed discourses on family produced by Finnish ECEC administrators. We identified two interweaving and overlapping discoursesthe ordinary family discourse and the diverse families discourse. In the discourse of the ordinary family, heteronormative two-parent native-Finnish families are represented as 'ordinary', while the diversity discourse focuses, in particular, on LGBTIQ and immigrant families. Based on these results, according to which 'diverse' families are perceived as other, we argue that the conventional and heteronormative understanding of family retains a strong hold in discourses of family in Finland. The two discourses, informed by both significant changes and continuities in contemporary family life, also echo wider cultural understandings of family. Our results reflect recent findings in the UK context, according to which, despite the increased understanding of family diversity, same-sex couples and immigrant families are still often referred to as unconventional families (Walsh, 2018) and white families with a native background continue to be understood as the norm (Heaphy, 2011).
Our results revealed the presence of tensions in the administrators' talk about family diversity. First, the analysis showed that the notion that all families are equal and that family diversity is understood against the yardstick of the norm were produced in parallel. Second, in discussing family diversity in the interviews, the administrators searched for appropriate expressions, choosing their words carefully to avoid violating the ideal of equality. These observations show that tensions in talk about equality between family configurations are also present in societies like Finland where the idea of equality is strongly embedded in societal values. Hence, these findings showing that discourses on family reproduce established hierarchies between family types merit consideration, as it is crucial that the ideals of equality are to be more fully realized.
The ordinary family and diverse families discourses used in the institutional settings may have implications for families as well as institutional praxis. If diverse families and the children living in them are perceived as 'other', their needs may go unrecognized. As our examples showed, this could lead to ECEC services being developed mainly to serve the needs of families perceived as 'ordinary' while the needs of other families might get neglected. As families have diverse needs and wishes, it is likely that some of them contradict with current institutional praxis (see Billig et al., 1988). The fact that it is culturally accepted to use discourses of 'ordinary' and 'diverse' families allows administrators to deal with this ideological dilemma by placing families with needs that conflict with current institutional praxis in the margins instead of changing the institutional practices.
Increased public interest in family lifethe familistic turnwas seen in Finland in the early 2000s. More recently, immigrant families and families with same-sex parents have, in particular, been regularly discussed, for example, in the public and political arenas. During the years 2016 and 2017, when the interviews for this study were conducted, the recently passed law allowing same-sex marriage and a sudden inrush of asylum seekers were 'hot topics' in Finland. We expect that the public debate has also informed ECEC administrators' perceptions of families and family diversity. National debate aside, the discourses produced by our interviewees reflect wider cultural ideas and narratives of family. Previous research has looked in detail at the impact of increased individualism on family diversity in late modernity in western societies (Chambers, 2001). Family diversity has even been described as one of the key characteristics of present-day family life (Chambers, 2001). However, as our results show, hierarchical discourses of family persist.
Previous research has demonstrated that understandings and ideals of family are in constant flux and that discursive shifts are slow. Although using a cross-sectional case study design to make sense of societal trends is complicated, we argue that our results hint at more varied and diverse ways of understanding families. That is, we interpret the current diversity discourse as indicative of a slow but steady shift towards the recognition of LQBTIQ and immigrant families as part of contemporary Finnish family diversity. Similar cultural and discursive shifts were observed during the latter half of the twentieth century, when cohabitation gradually became established as an accepted and recognized form of family life in both Finland and other western societies (Jallinoja, 1997). In parallel, however, our results also reflect othering, discrimination and even racism that immigrant and LQBTIQ families face their daily mundane life in the contemporary Finnish society. We suggest that early childhood teacher education and interservice training keeps providing tools for educators and administrators so that they are able to critically reflect and challenge hegemonic discourses that might marginalize immigrant and LQBTIQ families.
Conclusions should not be drawn on the prevalence and generalizability outside Finnish ECEC of the discourses identified in our qualitative analysis. However, with data collected from ten diverse municipalities from across Finland, the results comprehensively illuminate how families are discussed and understood by local Finnish ECEC administrators. As we also demonstrated, the discourses reflect wider as well as narrower cultural understandings and ideals of family. Hence, to better understand the connections between institutional thinking and family life calls for both theoretical and empirical studies in different institutional and societal contexts. Notes 1. Although broadly recognized, cultural narratives on family are also have a local dimension.
That is, meanings attached to the concept of family may differ between, e.g. Nordic and southern European societies or between urban and rural municipalities. 2. Registered partnership was replaced by same-sex marriage in 2017. 3. The total population of Finland in 2016 was approx. 5.5 million.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was supported by the Strategic Research Council of the Academy of Finland (grant number SA314317).