It … he [the Moon] didn’t look tasty! Transnational children’s use of metaphors as a transcultural tool to gain symbolic power in family discourse

ABSTRACT The aim of this study is to set out to explore how children living in Polish-English transnational families in the UK develop symbolic competence and symbolic power in home settings. By focusing on the children’s use of metaphor in family discourse when talking about the Moon, from a cognitive discursive perspective, this study explores the children’s creative use of language in achieving their social goals. It moves beyond seeing language, cognition and culture as bounded systems but rather powerful meaning-making resources that are part of the children’s complex discourse-worlds. The findings show how transnational families provide children rich opportunities to develop intercultural or ‘transcultural’ competence.


Introduction
This study focuses on the spoken discourse of Polish-English transnational children from two families living in Glasgow, UK. The findings suggest that transnational family environments provide rich opportunities to develop intercultural competence, in particular symbolic competence (Kramsch, 2006). 1 This notion refers to an ability that goes beyond a simple decoding of the language to an understanding of the dynamic, ongoing meaning-making processes of the interactions taking place during any communicative act (Kramsch, 2006;Kramsch, 2020). In a country where monolingualism is the prevailing ideology (Phipps & Fassetta, 2015), and where multilinguals are often framed in the deficit (Dakin, 2017;Safford & Drury, 2013), multilinguals' cultural and linguistic resources are often left untapped to the detriment of both the individual and to our multicultural society (Bailey & Marsden, 2017). This study's exploration of transnational children's often creative use of language shows them as having competences that are becoming ever more necessary in our dynamic multilingual world.
Transnational children are used to navigating their way in and between the heritage and majority language and culture and they frequently know aspects of the host's cultural norms and values more than their parents. Additionally, they have been shown to take a strong active role in the socialisation process, which can lead to traditional parent-child roles in the family being disrupted (Luykx, 2005;Renzaho et al., 2017). Whilst recent transnational studies have examined these changes through the lens of children's agency from a sociolinguistic perspective (Said & Hua, 2017), I am not aware of any multilingual or transnational studies that have explored what language reveals about the cognitive architecture of the interlocutors during family interactions. Thus, this paper aims to explore the language children use to express their thoughts in family discourse and the effect it has on other family members. By focusing on this aspect of language use, it becomes possible to gain a deeper understanding not only of the social but also the conceptual motivations transnational children have to bring about such changes to the family dynamics.
The present article focuses on the spoken discourse of four children (aged between 7 and 10) of two Polish-English speaking families, living in Glasgow, who participated in the AHRC funded Creative Multilingualism project: The Moon in narrative, metaphor and reason: a multilinguistic perspective. As the Moon is an object of significant cultural interest, with each cultural community attaching its own symbolic meaning to it, it is a particularly useful lens to explore transcultural variation between different communities. Not only did the project reveal what cultures share, and how the meaning of the Moon varies across cultural groups (Zacharias, 2019), but it also explored how the variation was profoundly influenced by the participants' social position within the family, as well as their worldviews and what texts the participants had engaged with, in other words, their knowledge, memories and ideologies they had.
Culture can be thought of as a set of shared ways to frame concepts that characterise groups of people. Many of these cultural understandings are reflected in the metaphors used by the members of a cultural group (Kövecses, 2005(Kövecses, , p. 2, 2015Sharifian, 2011), whereby a metaphor can be understood here as a device 'through which we perceive or experience one entity in terms of another' (Littlemore, 2019, p. 1). Due to the cultural significance of metaphors, the focus of the analysis in this study is on the metaphors the participants used to talk about the Moon. Focusing on metaphors allowed me not only an insight into the more established cultural resources that the participants were drawing from but also how and why they created novel language, in this case metaphor, to respond to the situation they were in, a key aspect of the concept of symbolic competence (Hult, 2014;Kramsch, 2006).
The innovative aspect of this paper lies in the approach of analysis, which fuses elements of sociolinguistics (Bourdieu, 1991;Kramsch, 2006Kramsch, , 2020Reid & Ng, 1999) and metaphor theory (Kövecses, 2005(Kövecses, , 2015Littlemore, 2019), which belongs to the larger field of cognitive linguistics, an approach to understanding the relationship between conceptual thinking and language. Recent metaphorical research has called for a more ecological approach to understanding metaphor use, in which its use is inherently situated in and emerges from particular cultural, social, individual and environmental circumstances (Gibbs, 2020, p. 18). This approach that integrates insights from both cognitive linguistics and sociolinguistics moves beyond seeing language, cognition and culture as bounded systems. In other words, rather than shying away from the messiness and complexity of understanding metaphor in discourse, this paper, by acknowledging the situated, dynamic, contextual and, interactional factors at play, provides a more holistic view of the meaning-making processes involved. Thus, a fuller understanding of the symbolic competencies that emerge in the transnational children becomes possible.
By focusing on metaphor as a discourse phenomenon, then, it becomes possible to trace how the transnational children negotiate their position of power in family interviews to be able to affect changes through their own (re) interpretation of the cultures they belong to.

Research context
This study was part of the Open World Research Initiative project: The Moon in narrative, metaphor and reason: A multilinguistic perspective, that was funded by the four-year (2016-2020) AHRC 'Creative Multilingualism' research programme, whose primary aim was to investigate the rich connections between linguistic diversity and creativity in contemporary Britain. Central to the research programme was the notion that there exists a fluid interplay between the different languages spoken in Britain that becomes a creative force that shapes how we live our lives. The programme advocates that being multilingual is part of the human condition , whether it is the ability to speak more than one official language such as Polish or English, or being able to switch between registers: from telling a joke in the local dialect to a family member, to talking to a pet dog, or to writing an essay for school. The programme advocates that successfully adapting to our ever-changing environment requires all of us to creatively draw on, and shift between, the different cultural, cognitive and linguistic resources that we have at our disposal to collectively meet our current intellectual and social challenges . This paper focuses on this phenomenon in transnational families, in particular the children, who by having to skilfully negotiate their way between cultures develop a translingual metalinguistic awareness in that they start to recognise the power of the languages they are in contact with. They become aware of the various effects of the languages they use and how they bring about changes in the others' thoughts, emotions and behaviour, in other words, as Bourdieu (1991) described it, the 'symbolic power' behind the language used.

Transnational families
Due to the recent intensification of globalisation, the number of transnational families worldwide has increased significantly. A transnational family maintains a sense of familyhood, despite the fact that the members are unable to be physically close to each other (Bryceson & Vuorela, 2002). Poles are the largest white migrant group in Scotland. Since the enlargement of the European Union in 2004, Scotland has seen a sharp rise in the number of Poles settling in the country, with the number of Polish-speaking children in schools at approximately 17,000 (Scottish Government, 2020).
With the increase in the numbers of transnational families, there has also been an increase in the amount of research carried out on transnational family life. Various topics have been studied, for example, identity (Chamberlain & Leydesdorff, 2004); emotions and belonging (Skrbiš, 2008); and intergenerational relations (Parreñas, 2005). More recently, attention has been given to the socialisation process both from the perspective of the caregivers and the children, with a focus on the participants' agency (Fogle & King, 2013;Said & Hua, 2017). Echoing Said and Hua's (2017) research on language socialisation, this paper also focuses on the children's ability to bring about change in thought and action, however, instead it uses the concepts of symbolic competence and symbolic power (Bourdieu, 1991;Kramsch, 2006Kramsch, , 2020 to make sense of this process.

Symbolic competence and symbolic power
The transnational context of the project reflects the dynamics of increased contact between people from different cultural backgrounds. These situations have called for new notions of 'competence' that go beyond the functional notions of accuracy, efficiency and appropriateness of Hymes's longstanding communicative competence (Kramsch, 2006). Even in interactions in which the same linguistic form is used, alternative interpretations will arise due to the different personal, social and historical trajectories each communicator is part of (Kramsch, 2006;Sharifian, 2013). As Kramsch (2006, p. 251) reminds us: Language learners are not just communicators and problem solvers, but whole persons with hearts, bodies, and minds, with memories, fantasies, loyalties, identities. Symbolic forms are not just items of vocabulary or communication strategies, but embodied experiences, emotional resonances, and moral imaginings.
Kramsch refers to the ability of multilinguals to identify these resonances and imaginings as symbolic competence (2006). Within this notion of competence lies not only the ability to negotiate and understand the conventions of the social context but also what the language reveals about the minds; the interlocutors' motivations and identities, for example, and how they not only are able to respond to the situation but actively shape the context they are in (Hult, 2014). Symbolic power is defined as the ability to actively affect others' thoughts, emotions and actions (Bourdieu, 1991;Kramsch, 2020). In the analysis which follows, symbolic power is seen to be a dynamic discoursal phenomenon that emerges from the family member's ongoing pursuit of their social goals by collaborating with others through the interactions taking place. It is reliant not only on the language itself but also on whether the person who says the words is seen to be in a legitimate position to use them (Bourdieu, 1991;Kramsch, 2020). In transnational families whose children speak the official language of the host country more proficiently than their parents, the covert power dynamics may take on new directions. This is what this paper sets out to explore.
Much linguistic research in this area has focused on language teaching contexts. In Li Wei's (2014) study, for example, he analyses multilinguals' language practices in a Chinese complementary school in the UK, observing discrepancies between the teachers' and the learners' symbolic competencies, with teachers wishing to 'transmit certain cultural values' that are met with resistance from the pupils leading them to use their languages in creative ways. Application of the concept of symbolic competence to out-of-classroom settings is rarer. One notable exception is Hult's study (2014), which employed ethnographic and sociolinguistic approaches to explore the insider/outsider identities of covert bilinguals in Swedish speech settings. Both these studies illustrate a strong sociolinguistic tendency to explore symbolic competence. Yet, whilst Kramsch (2011, p. 357) notes in her paper that symbolic symbols 'are to be seen as conceptual categories, idealized cognitive models of reality that correspond to prototypes and stereotypes through which we apprehend ourselves and others (e.g. Fauconnier & Turner, 2002;Lakoff, 1987)', there is a paucity of studies which explore symbolic competence and symbolic power from a cognitive-discursive perspective. This study hopes to make a contribution to fulfilling this gap.

Transcultural environments and prototypical understandings of the Moon
Cultures and languages, we are part of, offer us symbolic symbols (Kramsch, 2011), typical, 'prototypical' concepts at our disposal that act as common reference points for members of a community to use when communicating. Part of becoming a member of this culture is learning what these reference points are. To understand each other successfully in a transcultural environment is the ability to recognise that the variation in how we conceptualise the world depends, to a large extent, on the language varieties we speak and the cultures we are part of (Sharifian, 2013).
As the languages we use have evolved through the interactions with both our material and social environments our conceptualisations (e.g. schemas, categories and conceptual metaphors) are largely cultural in nature (Kövecses, 2015;Sharifian, 2011). In turn, how we, as members of particular cultural communities use language to express our thoughts provides cultural meaning and variation (Sharifian, 2013). Each language community builds its own shared prototypical representation of its cultural symbols. The Moon is a particularly suitable lens for comparing and understanding different cultures as it features so strongly in all cultures. It acts as 'a cultural mirror' as every culture has stories to tell about the Moon (Jerram, 2017). In Polish culture, for example, the Moon frequently symbolises the mind, frequently an unstable mind and sometimes humour too, as reflected in the expression 'spadł z księżyca' (fall from the Moon) that is used to describe someone's strange behaviour.
Many of these cultural symbols are conventional expressions that can be understood metaphorically. For example, in the expression 'the Moon is made of green cheese', that has found its way into children's story books across the globe, the Moon is taken to be a round cheese wheel. Importantly, for this study, due to the cultural prototypes' ubiquity in the language, they belong to people's shared, everyday knowledge within each cultural group (Moon, 1998). Such stable associations, that specific cultural groups have about the Moon and reflect 'the community's conventional patterns of thought or world views' (Boers, 2003, p. 235). This shared knowledge was reflected in the group interviews and was used in the analysis as one means to infer the meaning of the participants' utterances.

Methodology and analytical approach
To explore the use of metaphor in the spoken discourse of the two families I designed, set up and administered two group interviews in each of the families' homes. A purposive sampling approach was adopted and participants were recruited through personal contacts. Each family group interview lasted between 45 minutes to an hour. To create a relaxed atmosphere in which the family members, in particular the children felt at ease, the interviews were conducted at the home of the participants. The interview questions were designed to elicit from the participants their memories and experiences of seeing the Moon, any associations they made between the Moon and any cultural artefacts, such as books they had read or films they had seen. As different cultures ascribe a different gender to the Moon, there was also the opportunity to comment on whether they understood the Moon to be masculine or feminine. The participants were reassured that they were not going to be tested on their knowledge about the Moon; there were no 'right' or 'wrong' answers. The interview was semi-structured in design and included questions such as: . Can you describe the Moon? . Can you describe the last time you saw the Moon? . Would you say the Moon is masculine or feminine? . What does the Moon remind you of? Whilst I, the interviewer, spoke only English, the interviews were conducted in both Polish and in English. The participants were encouraged to speak both languages as they normally would to each other at home. The first family 'A' had both a Polish mother and father who had lived in the UK for 14 years. Their two daughters were both born in the UK, and were Ania (pseudonym), 10 years, and Hanna (pseudonym), 8 years, respectively. The second family 'B', had a Polish mother and a Scottish father. Only the mother was present at the interview. The older sibling, Janek (pseudonym), 9 years, had spent the first four years of his life in Poland, whilst his younger sister Marta (pseudonym), 7 years, had spent just 11 months. Both children attended the local Polish community school regularly on Saturday mornings. The two families actively maintained Polish and spoke both languages at home.

Positionality of the researcher
The flexible nature of the interview allowed me to respond and follow up on the prepared set of questions and my responses reflected my own desires, needs and emotions as I interacted with the family members. The social roles and relationships which were enacted throughout the interviews can be conceived of as the 'positionality' of the researcher (Coffey, 1999;Copland, 2015). At the start of the interview process, a couple of the adult participants expressed they were anxious about being interviewed on a topic they felt they had little knowledge about, and that they might not be able to provide answers that I, the researcher, was hoping for. Another source of anxiety amongst the adult participants was their concern about how to interact with their children, in Polish and English, to a guest in their home who was there to carry out research. Being a parent from a bilingual household myself, I empathised with the parents, yet was also aware of the unique opportunity I had to collect data for my project. This tension was most acute at the start of the interviews but eased a little once an interview pattern had established, and the children appeared to take it all in their stride.

Approach to analysis
All the interviews were audio recorded and later translated and transcribed independently for the first stage of a thematic analysis (Braun & Clarke, 2006). I was mainly interested in sections of the transcript that linguistically represented a change brought about by the children and therefore examined passages in the transcript that showed changes in behaviour, attitude or emotion initiated by the children's speech. Within these episodes, I investigated the children's symbolic competence and symbolic power by exploring how the children utilised their apparent knowledge of the two languages and cultures to pursue their social goals. In addition, as I anticipated the participants might use metaphor to describe the Moon, for reasons outlined above, I drew on metaphor theory (Cameron & Maslen, 2010;Kövecses, 2005;Littlemore, 2019) and cognitive linguistics, e.g. conceptual blending theory (Fauconnier & Turner, 2002), more broadly, to explore these metaphors further and the impact they had on the social and cognitive dynamics.
As the analysis took place after the interviews, the sections of the data that were selected and closely analysed and later presented were lifted from their 'original' context. This resulted in me construing the social realities of the participants in different ontological domains: one at the interview as I remembered it and the other as I read the transcripts. Although these two versions of the event were occasionally realised in my mind as two separate realities, they often blended in my mind to form one reality. The constant toggling between these different layers of realities, meant a cautious approach to the analysis was necessary. My aim, however, was to make my inferences as transparent as possible to allow for verification to enhance the trustworthiness of my claims.
One further aspect of the process is worth mentioning. As I read the transcript, I developed a further sense of who the participants were through a process of characterisation. By reading the transcript and soft-assembling other snippets of knowledge I gained of the participants, for example their age, position in the family, and profession, I developed a mini-mental model (Stockwell, 2020) of the participants and I used this schematic knowledge to interpret the cognitive and social dynamics of the interaction. To help overcome any potential biases that might have arisen from this characterisation process, I was able to check my interpretations with family members and a Polish speaker.
One complication was that I do not speak Polish and therefore I was heavily reliant on the English translation of the events that took place. To overcome this obstacle, my interpretations were also read and checked by a Polish speaker. Our combined linguistic and cultural knowledges, including mental-models of the participants, constituted our discourse-worlds (Gavins, 2007;Kramsch, 2006;Werth, 1999), which allowed for a more rigorous interrogation of the data. In sum, there was myself, an English speaker, who had carried out the interviews and who approached the analysis of the English written texts with a vivid recollection of the interviews, and there was a Polish/English speaker, who, although not present at the interviews, drew on her knowledge of both languages and cultural prototypes to help me interpret the texts.
The analysis of the data involved a number of stages. As the focus of the analysis was on spoken discourse, the analysis followed the steps of the discourse dynamics approach outlined by Cameron and Maslen (2010). Here, after first reading through and familiarising myself with the entire transcript and recalling the events of the interview, I identified any potential metaphors. Rarely, did the entire metaphor reveal itself explicitly in the text. However, from knowing the topic of conversation and the context, the topic of the metaphor was, in most cases, the Moon itself. The entity that the Moon (or Moon's light, shape, etc.) was being compared to, surfaced in the language as a 'lexical unit'; either an individual word or phrase (Cameron & Maslen, 2010, p. 103). Sometimes, they were expressed as a direct metaphor, or a simile, where the speaker made it clear that one entity is being compared to another: 'A is like B'. The lexical unit was labelled as metaphoric if its meaning in the context contrasted to a more basic meaning. During the analysis, I noticed that the metaphors were either conventional or more creative. Seeing the degree of creativity of the metaphors used in the spoken discourse to be both an individual and social phenomenon, I took creativity to be 'the interplay between ability and process by which an individual or group produces an outcome or product that is both novel and useful as defined within some social context' (Plucker et al., 2004, p. 90), and used this definition to guide my analysis. The findings were then discussed and verified with the Polish/English speaker.

Analysis one: mothers, daughters, moons and croissants
The following episode is from the group interview with family 'A'. Here, the family were asked to describe the Moon and also, what it looked like the last time they saw it. The mother and father started responding to my questions in both English and Polish, with Ania, the eldest daughter, listening first in the wings before offering her own answer. Hanna, the youngest daughter, fidgeting a little, had just been asked to sit next to her father. This episode is near the start of the interview, when the participants were still settling down. Possibly for this reason, the mother takes quite an assertive role in steering the conversation by asking her own questions to both myself and her husband. The conversation moves at a fast pace. Have you seen a lunar eclipse?
The interaction between the family members comes to its conclusion with an exchange between the mother, 'it looked a bit like a small croissant' and her 10-year-old daughter, Ania, 'yes, but it … he didn't look tasty'. The exchange caused the participants to laugh. The comic effect produced by 'he … it didn't look tasty' an extension of the mother's croissant metaphor, can be analysed in terms of conceptual blending (Fauconnier & Turner, 2002), a cognitive process whereby two distinct conceptual or mental spaces, the croissant input space and the Moon input space, form a newly emerged blended space (represented by the circle at the bottom of Figure 1). The Polish word rogalik (a Polish bakery item with a crescent shape) may trigger in the daughter's mind the English mental space of 'croissant'. Rogalik is often used in Polish to refer to the Moon's shape. It is a conventional term and not particularly novel. However, here on hearing rogalik, Ania, who has lived all her life in Britain, draws on her experiential knowledge of eating croissants. This is possibly the most accessible knowledge structure she has for the concept rogalik. The resulting comic effect relies on the incongruity of the fact that we do not normally eat Moons. What is a conventional metaphor in Polish has become a novel metaphor in English. Although this cognitive analysis explains why this metaphor is funny, it ignores the symbolic power behind Ania's use of Polish and English, in other words, Ania's desire and potential ability to affect the participants' emotions through her creative use of metaphor to achieve her social goals. Based on my schematic knowledge of the participants, and observation of the co-text, we have a mother-daughter interaction, in which the mother is quite assertive, and comes across as a 'cultural carrier' of Polish. Ania responds to her mother by humouring her through her use of metaphor, reinforced through negation 'but, it … he didn't look tasty', that cancels out her mother's claim that the Moon and rogalik look similar. The use of 'he' is also intriguing as it could be interpreted as an additional means of humouring her mother: The switch from the grammatically correct 'it' to the grammatically incorrect 'he' in English as it assigns gender, is noteworthy as it indicates a possible deliberate intention to borrow the Polish grammatical gender marking whilst speaking English to tease her mother (see analysis three for further discussion of gender markings).
This episode explores how Ania uses humour to release some of the tension and to possibly assert some of her own authority, through her developing symbolic competence. She does this by using her knowledge of the two languages and cultural domains creatively to form a type of 'hybrid' mixing, in her ongoing pursuit of her social goals in collaboration with others.

Analysis two: remembering the Moon with her brother
The next episode is the response given by the mother from family 'B' and her two children, Janek and Marta, after I asked them to describe the Moon. The episode begins with the mother interacting with Janek. She then directs her attention to Marta (7), who, according to the mother, is quite shy and does not like speaking Polish. In this episode, the mother takes the lead by first eliciting from Janek details about the Moon.  Janek: In the latter half of the episode, the mother is trying to elicit a response from Marta. The mother draws her daughter's attention or foregrounds the shape of the Moon with 'what other shape can the Moon have? It can be completely round, right?' This observation is not grounded in any particular scene and has a high level of abstraction. To provide a clearer coordinated mental reference she then provides access to a particular scene that her daughter can more easily conceptualise. She does this by evoking a world-switch (Gavins, 2007), a mental flash back, to a specific past event: 'Remember we were?' We can see here how the mother skilfully manages the conversation, by employing questions to direct her children's attention and to elicit information.
Janek, Marta's older brother, is the more talkative of the two children. He re-enters the conversation with 'taki bananowy' [sort of banana shaped] despite the fact the question was directed at the daughter. It's not clear if the boy was present at the scene the mother is evoking but he provides a direct metaphor, flagged with 'taki' -'sort of', that concurs with his mother's question. This is a useful metaphor to use, as banana and the Polish banan are cognates and it is therefore easy for the daughter to understand. Also, Janek might think Marta is familiar with the cultural prototype of SHAPE OF THE MOON IS FOOD. Evidence for this is found three minutes later in the conversation when the mother and Janek refer to this cultural prototype through the linguistic realisation: 'the Moon is like cheese': Mother: Takie dziury jakby prawda? Takie dziury w serze czasami się tak robi, że księżyc się tak rysuje, że wygląda jak ser. [Like holes, right? Like cheese holes, sometimes people draw the moon like cheese with holes.] Janek: Dlatego ludzie mówią, że księżyc to ser. ' The moon just made of cheese'. [That's why people say the moon is cheese.] <laughing> Mother: Aaa, bo ma takie dziury, prawda? [Aaa, because of the holes, right?] Given that one function of metaphors is to enable people to talk about hard-to-describe situations or objects (Kövecses, 2010;Low, 2020, p. 42), such as the physical features of the Moon, they are often used in pedagogical settings. One possible purpose of the metaphor 'sort of banana shaped', then, is pedagogical in that it is used by Janek and the mother to scaffold the interaction between herself and her daughter Marta, in a similar way to how it might be used in a pedagogical context. Finding Janek's response credible, the mother re-uses 'the Moon is a banana' metaphor in the next turn. Taking creativity to be the production of both a novel and useful entity as defined within some social context (Plucker et al., 2004, p. 90), this metaphor stands out to be creative in that it is novel in the sense that it is new to the discourse (it was not found earlier in the transcript) and it is perceived by the mother to be useful: she re-uses the metaphor in the following turn. It is also useful for Janek, as it may serve his personal desire to assert his own agency to gain recognition, but also to realise certain social goals, such as to please his mother, to support his sister by scaffolding the interaction, or even just to fill a gap in the conversation. With his creative use of metaphor, Janek's demonstrates his developing symbolic competence, through his awareness of the effect of the language he uses, and thus earns and probably re-affirms his position in the family.
Analysis three: A man? No, it's like a poshy lady wearing a silver dress! The last analysis shows how the father and two children of family 'A', draw on their cultural knowledge to categorise the Moon according to gender. In the following analysis, the father's perspective, which represents, here, the conventional cultural norms that he is part of, will be presented first, and then compared to the eldest daughter's perspective.
The Moon was frequently referred to as an animate object throughout the interviews especially in passages in which the participants were talking about the Moon and the Sun together. By way of an example, the father in family 'A' said 'I do tend to compare it with the Sun and I think I see it as a slightly passive entity'. Here, he uses the adjective 'passive' whose basic sense is: 'accepting what happens without trying to control or change events or to react to things' (Rundell & Fox, 2007). Personification results in this example because to be 'passive' is, according to this definition, the attribute given to an agent: only humans can accept and try. The personification is reinforced when the participant goes on to say 'yeah, yeah it's kind of a minor key if you like … a minor player'. The father then referred to the Moon and the Sun in terms of psychological traits: In this example, he brought extra cultural knowledge of humans such as stereotypical male and female characteristics of rationality and irrationality to his description. This extra knowledge is known as metaphorical entailments (Kövecses, 2010, p. 122) and is a particularly useful way of understanding how the father extended and embellished through the linguistic expressions 'more like the subconsciousness' and 'more rational and more connected with what is visible' the Sun-Moon metaphor. The participant drew on his cultural knowledge, understanding of psychology and human behaviour to elaborate the metaphor. The fact that he personified the Moon as passive and feminine reflects the findings of MacKay and Konishi's (1980) corpus study of a large collection of children's literature in English, in which the Sun was represented as an active male and the Moon as a passive female. Once again, the way he personified the Moon appears to be influenced by what he had read, seen and heard about the Moon previously. Interestingly, when asked later in the interview, if the Moon was either masculine or feminine the father and the mother both replied, with certainty, that the Moon was masculine, whereas both of the children said that the Moon could be both masculine and feminine. One possible explanation why the father chose to classify it in this instance as masculine is that the question triggered his declarative knowledge that the Moon is assigned a masculine grammatical gender in Polish. In contrast, Ania and her sister Hanna have not been exposed to much formal schooling in Polish, in which they might be taught this. Furthermore, research suggests as English does not have any grammatical gender markings, the acquisition of gender markings in Polish by heritage language speakers may be negatively affected (Cornips & Hulk, 2008). At a first glance, then, knowledge about grammatical gender markings appears not as foregrounded in Ania's mind when asked directly by me, the interviewer, whether she thinks the Moon is either masculine or feminine. Evidence for this is debateable, however, as she referred to the Moon as 'he' in English in analysis one, an indication that she has at least a good awareness of the significance of grammatical gender markings.
Drawing on my own knowledge about the family dynamics and considering Ania's age (10), hearing her parents making reference to this aspect of language knowledge might have triggered some resistance in Ania towards them. In response to her mother's statement: 'Because in Polish "księżyc" [the Moon] is male gender', Ania shifts the conversation by not referring to a grammatical category but by referring to a book she has read. As Reid and Ng (1999) note, a change in the topic in intergroup casual conversations indicates an attempt by the speaker to control the conversation. Her desire to take control may explain why she chooses to shift the topic towards her knowledge of books, to ascribe gender through the use of personification, instead of accepting the information provided by her parents.
Ania, then, builds on her younger sister's position that the Moon can be 'both [masculine and feminine] always'. Without acknowledging her parents' claims that the Moon is masculine, Ania makes explicit reference to the [Polish] book and the symbolism in the book to explain her reasoning. She recognises a metaphorical connection between the Moon and a lady in a silver dress and uses the simile, 'it's like a poshy lady' to justify that it can be feminine. Her response demonstrates an ability to reflect on the non-arbitrary nature of the metaphor by reflecting on her social and sensorimotor experiences. This ability to reflect on the meaning-making process is further evidence of her developing symbolic competence (Kramsch, 2006). Although her parents do not explicitly acknowledge her response, I, the interviewer, respond to Ania two lines later with 'Oh that's interesting!'. With this comment, I see Ania's response as legitimate and as a credible source of information. It is not just the words which hold symbolic power, but the credibility the speaker holds due to their social position (Bourdieu, 1991, p. 170;Kramsch, 2020, p. 6). I viewed Ania during the interview as an expert in Polish, a position she may not frequently hold in the eyes of her parents.

Discussion and conclusion
Recent studies on intercultural understanding of transnational children in home and school contexts have shown that these children play an active role in their own and the other interlocuters' socialisation process often through creative means (Said & Hua, 2017;Wei, 2014). By negotiating their way in and across cultural and linguistic landmarks, transnational children develop strategies and competencies that go beyond that of achieving efficiency and clarity in communication to ones which involve an understanding of the meaning-making process itself; its symbolic competence (Kramsch, 2006(Kramsch, , 2020. Whilst all humans have developed this competence through navigating their way around different registers, and are in this sense 'multilingual' , transnational children's position as re-interpreters of two (or more) languages appears, as is shown in this paper, to amplify this competence. Although hitherto transnational studies have highlighted and convincingly demonstrated transnational children's ability to affect change, their focus on solely the social dynamics of the interactions limits their scope in understanding the conceptual work taking place to bring about these changes. Despite the small data set and the difficulties, at times, in capturing and representing the context, this paper set out to contribute to these discussions by providing a cognitive-discursive account of these processes.
The analysis in this paper aimed to demonstrate how Polish-English transnational children use language, in particular metaphors, to bring about change in family interviews when talking about the meaning of the Moon, an elusive object of deep significance to all cultures. Seeing these changes as a collaborative effort of pursuing one's own as well as others' social goals, I have adopted the notion of symbolic power (Bourdieu, 1991;Kramsch, 2006Kramsch, , 2020 to frame the changes taking place during family interactions. Although there exist moments of collaboration and at times tension in all families, my findings have set out to show how this power game plays out and is reflected in the interactions held in multiple languages spoken by transnational families. Transnational family members may be perceived within their family to know aspects of the host and heritage cultures and languages in different ways than other families, which can lead to new role reconfigurations and potential tensions. These points of tension surfaced in my data, for example, with Ania in collaboration with others using her knowledge of Polish and English to create language (it … he didn't look tasty; like a poshy lady) to tease and humour her parents. Noteworthy, were the several moments of co-operation, for example, in episode two, where Janek assists his mother with her elicitation.
Although I anticipated the participants to use metaphor in the interviews, I had not anticipated the extent to which the children would use metaphor so creatively to achieve certain social goals. Despite the small data set, and a limited number of participants with this analysis I explored how and why the children created this new metaphorical language to respond to the situation they were in, a key aspect of the concept of symbolic competence (Hult, 2014;Kramsch, 2006). This paper also reflected an important shift in our understanding of metaphor use in multilingual contexts. The creative metaphors analysed in this paper could have been interpreted as errors and might have been framed therefore as 'problems in need of remedy' (MacArthur, 2016, p. 133), if interpreted against native speaker norms. However, by taking a cognitive-discursive approach, language use can be viewed more positively, as being examples of 'hybrid' metaphors (MacArthur, 2016, p. 133). This finding supports what others have noted, that adopting strict native speaker norms can be unhelpful when analysing speakers of multiple languages (Seidlhofer, 2001). Framing the metaphors as hybrids enabled me to see the metaphors as transcultural meaning-making tools to bring about change to the children's social position.
I finished my analysis by drawing attention to how Ania's credibility as an expert in Polish was seen through the interviewer's eyes. To me, a non-Polish speaker, she had earned the status of an expert, not only through her understanding of Polish but also her developing symbolic competence. In a country in which bilinguals are often portrayed as a problem, this short analysis provides some further evidence, that transnational households can provide rich linguistic tapestries that afford symbolic competencies and linguistic creativity that should be valued in schools and in the wider society. The wider implications of this study's findings are that transnational children can be seen as assets. Yet, the question remains whether these competences are transferable to other contexts, if the children are unaware that they have them. And, if the children do not have the opportunity to critically reflect on them, would these transcultural competences lay dormant and eventually cease to exist. Note 1. In metaphor studies, it is convention to represent the conceptual metaphor with small capitals (see e.g. Kövecses, 2010).