Cultural features of performance in Norwegian children’s narratives

: This article presents part of a research project whose aim is to investigate Norwegian 6–7-year-old children’s language use in story-telling. The data consist of the children’s oral texts based on a wordless picture book. By studying creative and stylistic features of the children’s narratives, I attempt to shed light on cultural interaction and socio-cultural factors as fundamental contributors to those narratives. The aim of this article is to contribute to the exploration of a culturally anchored perspective on narratives in light of the reading culture prevailing at Norwegian kindergartens. evaluative (a) of mind, references to characters’ and and behaviours; (b) character speech, which to attribute an intentional state to a story character; distancing


Introduction
Narratives involve the use of language for communicative purposes. For this reason, many investigators consider narrative ability to be the "gold standard" in research into children's language development (Miller, Heilmann, & Nockerts, 2006). Since the typical content, structure and purpose of narratives vary across cultures (Heath, 1982;Slobin & Berman, 1994), many researchers attempting to avoid cultural and linguistic biases have chosen to study narratives based on wordless picture books (Botting, 2002). One such picture book which has been and remains a frequent tool for collecting narratives for research purposes is Frog, where are you? (Mayer, 1969).
Research on children's narratives is characterised by tension between two perspectives: a cross-linguistic one and a cross-cultural one (Uccelli, 2008). Cross-linguistic studies explore typological differences between languages and suggest ways in which language-specific grammatical features may influence children's narrative performance and development, while cross-cultural studies focus on culture as the source of variation, documenting the impact of culturally valued patterns of communication on children's narrative performance and development. Most of the research based on narrations of Frog, where are you? (Slobin & Berman, 1994;Verhoeven & Strömqvist, 2004) takes a cross-linguistic perspective and tends to be based on the assumption that an episodic and chronological structure is the norm to be aimed for. This research operates within the well-established cognitivist paradigm, using traditional quantitative methods to map the course of narrative performance across languages and age ranges. However, I find one study based on narrations of Frog, where are you? that are cross-cultural in nature, and it's findings underscore the necessity of looking more carefully at how children actually perform the activity of narrating (Reilly, 1992).
This article presents part of a research project, rooted in the cross-cultural tradition, whose aim is to investigate Norwegian 6-7-year-old children's language use in story-telling. The data consist of the children's oral texts based on the above-mentioned picture book Frog, where are you? The inspiration for the article comes from the fact that, in my analysis of the children's narratives, I have encountered numerous cases where a narrator clearly manifests familiarity with important aspects of narrative while at the same time more or less completely ignoring structural norms for narratives (e.g., episodicality and chronological order). This apparent disparity in the Norwegian children's narratives calls into question the cultural neutrality of Frog, where are you? as an instrument for assessing children's language and narrative development, and it seems interesting to find out what a cultural perspective can bring to the investigation of linguistic and narrative skills. By studying creative and stylistic features of the children's narratives, I attempt to shed light on cultural interaction and sociocultural factors as fundamental contributors to those narratives. The aim of this article is to contribute to the exploration of a culturally anchored perspective on narratives in light of the reading culture prevailing at Norwegian kindergartens, and to consider what supplementary action, if any, is needed to ensure that Frog, where are you? can be seen as a relevant basis for assessing Norwegian kindergarten children's language development.

Background
Research based on narratives, both within cross-linguistic studies and in the context of assessment, has traditionally focused on norm-referenced measures of structure. Many such traditional studies are inspired by the analytical framework for oral presentations of personal experiences proposed by Labov andWaletzky (2006 [1967]), which distinguishes between the referential and the evaluative function of narrative. 1 The referential function is a technique that involves constructing narrative units that match the temporal sequence of an experience. The evaluative function, by contrast, mediates between what happened and the narrative, displaying a personal interest determined by stimuli in the social context. According to Labov and Waletzky, any meaningful narrative will include both functions. The referential function, which has gained most attention in narrative research, represents a traditional structural approach to narrative and narrative analysis. It is also the basis for most "Frog Story" research (Justice, Bowles, Pence, & Gosse, 2010;Miller et al., 2006;Petersen, Gillam, & Gillam, 2008;Slobin & Berman, 1994;Verhoeven & Strömqvist, 2004). 2 The evaluative function, by contrast, is of greater importance in cross-cultural studies.

Cross-cultural research based on Frog, where are you?
Frog, where are you? represents a traditional narrative consisting of 24 pictures and following a prototypical story sequence. First there is an introduction where we are introduced to the key characters: a boy, a dog and a frog kept in a jar. In the middle part the frog runs away, and the boy and the dog search for it. Towards the end they find the frog. However, earlier research relating to the overall structural level of the same narratives as studied here (Hoel, accepted) reveals that, of the 70 texts recorded in the full research project, not a single one has a complete and coherent plot structure.
The children seem to be rambling in and out of the main structure, with explicit reference being made to the overall plot structure ("on-plotline narrative") in some episodes but not in others ("off-plotline narrative") (see Appendix). But this dynamic alternation between on-plotline and off-plotline narrative is not due to a lack of experience in the narrators. In fact, the main findings from this research are that there are certain pictures in the book that hardly any of the children narrate in an on-plotline manner and that there are elements in the graphic material that may explain why the children go off-plotline. These findings also indicate that the reading culture in which the children participate may be part of the reason why they deviate from an overall plotline orientationmeaning that a wordless picture book may not constitute a culturally independent basis for narrative macrostructure, as is often assumed by researchers choosing to use such books as elicitation tools.
There are earlier studies focusing on culture as the source of variation in children's narratives. Rollins, McCabe and Bliss (2000) account for the influence of cultural differences, including their implications and suggestions for culturally sensitive assessment, in the description of a three-step process for assessing pre-school children's narrative discourse, which is based on an exploration of Labov and Waletzky's framework for narrative analysis. The method used by Rollins et al. involves detecting cultural differences by means of "[e]valuation statements or words [that] indicate the speaker's opinion or thoughts about what is being described" (p. 227).
The development and use of evaluative expressions is also explored by Chen and Yan (2010) in their study of English narratives elicited from 80 Chinese-English bilinguals and 80 American monolingual peers at four ages (five, eight, ten and young adult) using Frog, where are you? The results revealed age-related growth in the development and use of evaluative expressions in both categories of narrators, but bilingual children differed from monolingual children in the quantity and quality of the evaluative clauses used. In their analysis, Chen and Yan distinguish five types of evaluative devices: (a) frames of mind, which includes references to story characters' mental and emotional states and behaviours; (b) character speech, which tends to attribute an intentional state to a story character; (c) hedges, used as distancing devices; (d) references to negative states and actions; and (e) causal connectors.
Although both Rollins et al. (2000) and Chen and Yan (2010) see culture as a source of variation, both studies represent the cross-linguistic perspective in their focus on evaluative expressions; for example, neither of them devotes any attention to the use of paralinguistic devices as a product of cultural influences.
There are also studies underscoring the necessity of looking more thoroughly at how children perform the activity of narrating. In a study of Spanish-speaking children's evaluation and temporality in the construction of personal narratives, Uccelli (2008) concludes that individual differences may overrule the traditional chronological/episodical structure, pointing out that "[m]eaning resides not only in what we say but also in how we say it" (p. 176).
The influence of children's (unconscious) strategies to make Frog, where are you? narratives engaging has been investigated by Reilly (1992) in two studies comparing the development of storytelling (performance) and story construction (coherence). In the first study, 3-4-year-olds and 6-8year-olds narrated Frog, where are you?; and in the second one, 7-8-year-olds and 10-11-year-olds retold the story to a 3-year-old. 3 In her first study, Reilly defined "performance" as the ability to engage the listener and to convey the narrator's perspective and attitude through linguistic and paralinguistic channels (characterisation, quoted speech, evaluative comments, facial expressions, gestures, prosodic features (pitch, volume, voice quality and length) and lexical/phonological stress).
Her analyses show that the older group's narratives are both longer and more complex than those of the younger children, and that the younger children use significantly more affective elements than the older ones. Reilly claims that her findings are in line with Labov and Waletzky's suggestion that narrative serves both a referential and an evaluative function: "[f]rom this perspective, affective expression may be viewed as one aspect of the narrator's evaluation" (p. 356).
Two wordless picture books have been used in a cross-cultural study of creative and stylistic features (Gorman, Fiestas, Pena and Clark, 2011), examining narratives by 60 African-American, Latino and Caucasian first-and second-graders with a view to analysing the effects of culture and genre (referred to as "creative and stylistic devices") on the content and structural organisation of narratives. 4 Gorman et al. performed analyses of organisational style and creative features (dialogue, reference to character relationships, embellishment and paralinguistic devices). They found many similarities and differences between the ethnic groups. The results were not statistically significant as regards organisational style or use of paralinguistic devices, but they were in relation to other creative features, reflecting home cultures and socio-cultural expectations. The authors conclude that culture influences children's production of narratives: "[a] 'good' narrative may be one that not only contains a well-organized representation of events, but also captures the listener's attention and interest" (p. 16). Similarly, Bloome and Champion (2003) describes how structure and performance may not always go together: for example, a coherent and traditionally structured narration may not be presented in an engaging way, or an engaging narrative may not necessary involve a clearly structured text. 5 By contrast, Gorman et al. (2011) found that a picture book constitutes a culturally independent basis for narrative macrostructure, something they believe may be due to the influence of school (the average age of the participants in the study was 7.65 years), genre and data-collecting procedures. They conclude that the narrating of picture books "goes beyond the idea of narratives as exact reproductions of what the pictures portray" (p. 18), also claiming that the children's texts reveal something about the children's "interpretations of the how and why of storytelling" (p. 18).

Culturally valued narrative styles
Various approaches to narrative analysis seek to capture the complexity of different narrative styles. Engel (1995) maintains that children's narrative styles reflect "individual, familial, and community values and ways of telling" (p. 16). In a comparative study focusing on bedtime story reading, Heath (1982) describes different patterns of book-related language use in three literate American communities. These patterns of language use depend on the social and cognitive aims prevalent in each socio-cultural group, and Heath claims that her study shows that there is no one-and-only way of acquiring communicative competence, nor a universally applicable model of development.
In an article concerning the shift from a dominant cognitive perspective to a more pragmatic orientation in the field of child-language research, Bamberg (2009) highlights language gamessuch as question-and-answer routinesand differently framed book-reading activities as sources of motivation for children's narratives. Mjør (2009) describes the processes of meaning-making that take place as young Norwegian children (12-24 months) and their parents read picture books together, as well as the impact that the adults' degree of motivation exerts on the children's ability to experience picture-book texts as coherent, meaningful and interesting.
Children tell stories that are valued by their particular community, because they narrate in the context of their social interactionswith their parents, as described in several studies, and in Norway also within the kindergarten culture. Most Norwegian children, from the age of 10 months until they are six years old, spend eight hours a day, five days a week at kindergarten, with the exception of five weeks of holiday each year. 6 The Norwegian National Curriculum (Kunnskapsdepartementet, 2011) lays down that all kindergarten programmes must be built on a holistic philosophy of teaching, with care, play, learning and "formation" (danning) at the core of activities (p. 5). 7 Play -particularly "unorganised play" (frilek)has a prominent place at the Norwegian kindergarten as an important aspect of children's lives, not least as a basis for development and learning. Unorganised play is internally motivated and is not controlled by external forces. Children have no purpose in playplaying is an end in itself. Characteristic of play is that it is a voluntary activity that the child chooses to participate in. Play is "pretend" -it is beyond what we perceive as the "real" world. Play is of importance for the well-being of the children and constitutes a fundamental aspect of life and learning (Kunnskapsdepartementet, 2011).
Another key word in Norwegian kindergarten culture is "participation" (medvirkning). This means that kindergarten teachers have to listen to the children's views and take them into account in their planning. Participation is explicitly mentioned in the National Curriculum (p. 15) and there is also a separate booklet on child participation (Bae, Eide, Winger, & Kristoffersen, 2006). Children's participation is also prominent in Norwegian kindergarten research (Borg, Kristiansen, & Backe-Hansen, 2008). A project studying the characteristics of conversation practices at Norwegian kindergartens, concluded in 2010 (Gjems & Løkken, 2011), found that adult conversation practices are very important for the development of children's language and their learning in general, particularly the ways in which adults invite children to participate in linguistic activity.
Reading activities are also referred to in the official kindergarten curriculum, which lays down that the staff must "create an environment in which children and adults every day experience excitement and joy through reading aloud, telling stories, singing and conversation" (p. 34). Picturebook reading is a traditional kindergarten activity, and the picture-book genre is one of the most important literary styles (alongside the fairy tale) for Norwegian children (Birkeland & Mjør, 2012). However, despite being well anchored in the National Curriculum, the reading activities are often unsystematic and initiated by the children. When organised systematically, reading activities are generally carried out as "shared reading" in groups where reading can easily be disturbed by children who are anxious or distracted by each other or other things around them. It is therefore important for the kindergarten teacher to hold the children's attention and engage them in the reading activity, for example by talking to them about the things they see in the pictures and asking them questions about the content of the book. Hence, these reading activities are mainly dialogic in nature, with both the adult teacher and the children contributing actively. It should also be noted that while the National Curriculum refers to reading activities necessarily taking place "every day", there are no official guidelines for how long the activity should last. Some Norwegian kindergartens also use more structured methods for reading activities, such as "dialogic reading" (Hoel, Wagner, & Oxborough, 2011) and "literary conversation" (Solstad, 2008(Solstad, , 2011, where the goal is to stimulate language learning and literate knowledge, respectively. Under both methods, the kindergarten teacher actively initiates and maintains conversations in which the children contribute their thoughts and opinions about the content of a bookboth pictures and written text. Among other things, the adult encourages the children to see connections between what happens in the book and the children's own experiences. Both methods assume that the children will participate actively and that all children in the group will be included in the dialogue. The involvement of the kindergarten teacher is important because the teacher conveys the text to the children, facilitates the dialogue (by asking questions about the content of the book) and supports and extends the children's language (Whitehurst et al., 1988).
The interactive way of reading is a cultural method existing in Western culture: it is something that care-takers such as parents and kindergarten teachers do in the company of children (Rhedin, 2004, p. 53). 8 This cultural method is characterised by the use of dialogue and by the adult's encouragement of the child to be active and to participate (Mjør, 2009). The children become familiar with the idea of being actively involved in reading activities through the reading traditions they encounter at kindergarten and through reading activities organised by their parents (Mjør, 2009). Adults who are reading with children will express their sympathies and antipathies when conveying the text and communicating with the childrenlaugh, smile or chuckle when the text is funny, use a special voice to express whom they feel sorry for or whether they think a character is tough or sweet, etc. Both the reading situation and the adult reader introduce the children to the ways in which the kindergarten culture relates to and uses texts by modelling practices. The interactive reading culture is also commented upon in the work of Bus (2001), where it is claimed that young children hardly ever encounter texts through recitation; the author's words are always embedded inand inextricably linked tothe social interaction between adult and child. To sum up, such a dialogic nature is what characterises the reading culture prevalent at Norwegian kindergartens.

Method
The aim of present project is to investigate 6-7-year-old children's language use in narrations of Frog, where are you? elicited by the researcher. The data-collection procedures described in the "Frog Story" literature were used, meaning that the children were first told to browse through the full book and then asked to tell the story in their own words and at their own pace, while simultaneously turning the pages (Verhoeven & Strömqvist, 2004, pp. 4-5). The children narrated individually in familiar surroundings at each child's own kindergarten or school. All the children participating had met the researcher in advance, reading books together and the children telling the researcher about their reading habits and favourite stories. To make the children comfortable with the setting and to familiarise them with the task, each child first looked at a different picture book and the researcher initiated a conversation involving the child's favourite storyfrom a book, a film or a game. During the task, the researcher sat across from the child to avoid being seen as a fellow reader, and to promote child language, minimise pointing and encourage the use of explicit labels for characters, objects and actions (Miller et al., 2006). The session ended when the story had been told, and the children were given small gifts in return for their participation.
Participants for the study were selected by teachers at six Norwegian kindergartens. The criteria were: 40 children (20 girls and 20 boys) (1) who had attended the same kindergarten full-time for the past three years, (2) whose language development was within the normal range 9 and (3) whose parents had Norwegian as their mother tongue. 10 In addition, the kindergarten teachers were interviewed in order to ascertain whether the group selected needed to be adjusted and to find out about the usual procedures for reading activities at each kindergarten.
A total of 35 children (five turned out not to meet the criteria) participated in the study (18 girls and 17 boys), and each child produced a narrative looking at the pictures in Frog, where are you? on two occasions: first during the last month of kindergarten (at the age of six) and then again during the last month of the first year of primary school (at the age of seven). Both kindergarten and primaryschool texts will be explored in this article, but no comparisons between them will be made here.
The children's oral texts were recorded in audio and video and have all been transcribed using the CHAT standardised transcription system (http://repository.cmu.edu/psychology/181/) (MacWhinney, 2000). For the purposes of this article, parts of them have also been translated from Norwegian into English.
In order to explore the different ways in which the children in this material perform their narratives, I have examined creative and stylistic features in the light of cultural interaction and sociocultural factors. Each utterance in the children's narratives has been analysed for established features of performance -i.e. features documented in previous studies of children's personal and fictional narratives, especially Reilly (1992) and Gorman et al. (2011) as well as additional features of performance reflecting the specific narrative culture in which the child participates. 11 The established features include (a) speech, both direct and indirect; (b) reference to characters' relationships, including descriptions of the nature of a relationship, the invention of names and reference to conduct not shown in the pictures; (c) embellishment, such as expressions of fantasy, suspense and conflict; and (d) paralinguistic devices, including expressive sounds and exclamatory utterances. In addition to this, two features not yet described in the cross-cultural literature will be explored: a paralinguistic device referred to in this article as reading voice, and the children's use of genre-specific wordings or phrases, referred to in this article as fixed wording.

Results
In the following, in order to document the impact of culturally valued patterns of communication on the children's narratives, I will present various features that have been described by the researchers referred to above as well as the two above-mentioned additional ones and the category of "Outside comments", illustrating them with examples from the Norwegian children's narratives. Summary information about the number of instances of each feature is presented in Table 1 (kindergarten) and Table 2 (primary school). 9 To identify children within the normal range of language development, the teachers used an observational material, "TRAS" (Espenakk et al., 2011), developed for use at Norwegian kindergartens. The language areas covered by TRAS are: interaction, communication, attention, comprehension, language awareness, pronunciation, word production and sentence production. 10 It was also intended to obtain information about the children's socio-economic background in terms of the extent of their mothers' education, but this turned out to be difficult. Kindergartens do not have access to such information because of the strict Norwegian rules on the provision of sensitive personal information. The kindergarten teachers were requested to ask the mothers about their education, but the teachers considered this question to be too personal. However, compliance with the other three criteriaconcerning length of kindergarten enrolment, assessed language development and Norwegian-speaking parentsare strong indications that the participating children do not belong to any at-risk group as regards their language development. For the overall research project, however, the absence of this information means that I am unable to draw conclusions based on socio-economic factors. 11 A single utterance may include several instances of the same performance feature and/or instances of several different features.

Speech
The use of direct speech means that the narrator takes the character's perspective (Gorman et al., 2011;Reilly, 1992). Example: "He says, 'Oh, an owl'" [Han sier:"Oi, en ugle"] (Boy 4, six years old). Direct speech (including onomatopoeic) can also be used of animals: "Croak, croak, can I have my friend?" [Bak, bak, kan jeg få vennen min?] (Boy 4, six years old). In the case of indirect speech, the child uses verbs such as talk, say, tell and ask to indicate character speech. Example: "then he told the dog to be quiet" [så sa han til hunden at han skulle vaere stille] (Boy 9, seven years old). The effect of using character speech is to make the narrative more dramatic. By contrast, in several narratives the boy and/or the dog just "call" for the frog: "He calls for the frog" [Han roper etter frosken], which comes across as more prosaic. The verbal search (as opposed to actions such as looking at places or lifting things) characterises many of the narratives.

Reference to characters' relationships
When the narrator refers to the nature of a relationshiphow characters are related or acquaintedthis produces an effect of empathy (Gorman et al., 2011). For example, the use of words such as mother and child indicates that someone needs to be taken care of: "and then there was a frog mummy and a frog daddy and then there were many frog babies" [og så der var det en froskemamma og en froskepappa og så var det mange froskeunger] (Girl 6, six years old). The present material includes examples of several ways of specifying the nature of the frog's relationships. Boy 3 (six years old) refers to the family connection: "Then he found the whole family" [Så fant han hele familien]; Boy 3 (seven years old) emphasises the marriage: "who was married with lots of small babies" [som var gift med mange småunger]; and Girl 3 (seven years old) highlights the love connection: "since he had jumped to his girlfriend" [siden han hadde hoppet til kjaeresten sin]. The relationship between the boy and the frog is referred to using terms such as "his pet" [kjaeledyret sitt] (Boy 1, seven years old), while Girl 1 (seven years old) simply says that the boy "had" [hadde] the frog. An utterance such as "then they took him home and up to their bedroom" [så tok de han med hjem og opp på rommet sitt] (Boy 2, six years old) places the boy within a relational framework of the kind the narrator knows from home: his own room in his family home. Words that are only implicitly relational, such as "cute" [søt] (Boy 3, six years old) or "the little frog" [den lille frosken] (Girl 1, six years old), also create an effect of empathy. The same is true of references to the boy's reactions to the frog's actions, such as "the boy was sad" [ble gutten lei seg] (Girl 1, seven years old) or "is completely dejected" [blir helt overgitt] (Boy 10, seven years old). A feature known from other studies (Gorman et al., 2011) is the invention of names for characters. Even though there are no examples of children inventing names for the characters in this material, there is an interesting point to make in relation to cultural differences in how the narrators refer to some of the characters, for example depending on where in Norway they live. Children from Central Norway talk of "frog" [frosk] and "elk" [elg] while children from the west coast more often talk of "toad" [padde] and "deer" [hjort]. Boy 13 (six years old) calls the dog "the sheep" [sauen] in his narrative. This is clearly intentional, because when he forgets and says "the dog" [hunden] he immediately corrects himself: "while the dog had got, I mean the sheep had got its head inside the jar" [mens hunden hadde fått, jeg mener sauen hadde fått hodet sitt inni glasset]. This choice of word adds humour to the narrative, which is probably also the intended effect. There are in fact several examples of humorous word choicesoften associated with bodily referencesapparently intended to surprise, entertain or attract attention: "his bottom" [rompen] (Boy 2, seven years old); "right on his head" [på hue] and "fell on his face" [et tryn] (Boy 7, six years old) and "straight in his mug" [rett i trynet hans] (Boy 4, seven years old). 12 Another example of intentional use of humour is when Girl 13 (seven years old) says, "they put on clothes, and the dog the jar" [de tar på seg klaerne, og hunden syltetøyglasset], whereupon she chuckles.
The performance feature of conduct, finally, occurs in cases where the narrator refers to ways of behaving or treating others appropriately in relationships, such as helping or apologising (Gorman et al., 2011). Examples: "The parents were so nice" [så snille de foreldrene var] (Boy 5, seven years old) and "then he looked after him" [så passet han på han] (Girl 3, seven years old).

Embellishment
The performance feature of embellishment comprises fantasy, suspense and conflict. Fantasy occurs when the narrator transcends what is evident in a picture, using original ideas (Gorman et al., 2011). This feature contributes to making a narrative interesting, and it reflects the narrative culture in which the child participates.
In the present material there are examples of fantasy in both on-plotline narratives and offplotline narratives, even though Gorman et al. (2011, p. 17) and others claim that on-plotline narrative emphasises factual and sequential descriptions of events. One example: in Picture 7 of Frog, where are you?, the traditional standard plotline reading is: the jar is broken, the boy picks up the dog and the dog licks the boy's face (Slobin & Berman, 1994).
Figur 1 Picture nr 7. Reproduced from Mayer (1969) with the permission of the publisher.
Girl 3 (six years old) says, "there the dog had lost the jar that they had the frog in. Then he got it on his head. He believes the dog ate it." [der hadde hunden mistet krukke som de hadde frosken oppi. Så fikk han den oppå hodet. Han tror at hunden har spist den opp.] The plotline point of the broken frog jar, which hardly any of the narrators in this material mentions (Hoel, accepted), is present. But in addition to this, the narrator presents an interpretation of the picture that goes beyond what is actually shown in it by claiming that the boy suspects the dog of having eaten the frog.
Off-plotline narrative elements sometimes include interesting pieces of narrative, and some of them add something unique to the narrative, as with Girl 5 (seven years old) narrating Picture 1. ]. This is description rather than narrative, and it is clearly off-plotline. The traditional standard plotline reading of this picture is: a boy and a dog look at a frog in a jar (Slobin & Berman, 1994). Girl 5's narrative has much in common with the picture-book genre, which as mentionedis one of the most important literary styles for Norwegian children (Birkeland & Mjør, 2012). In that genre, the verbal text does not normally describe the illustration. Instead, picture books combine two ways of tellingone verbal and one visual. The actual content of the book is realised through the interaction between these two ways of telling. Girl 5's narrator uses the first person, which is not unusual in children's literature (Birkeland & Mjør, 2012), and also gives the narrative a poetic touch. Off-plotline fantasy features of this kind in the material tend to build on the visual text, but do not always do so.
One example of off-plotline narrative not based on the picture can be seen in Girl 8 (six years old) narrating Picture 8. She says, "and then he sang a song so the bees came" [og så sang han en sang sånn at biene kom]. A year later, when she is seven years old, she is still on the same theme: "and then he sings so that bumblebees come out of their house" [og så synger han sånn at det kommer humler ut av huset sitt]. According to the Frog Story tradition, the boy is calling for the frog (Slobin & Berman, 1994). The visual text contains no obvious explanation for the fantasy (even though it should be noted that the boy's mouth is open). The bees are neither foregrounded nor presented in any dramatic way (some children even see them as raindrops or flies). And still there is somethingan experience, a film, another bookthat gives rise to two off-plotline narrations, one year apart, about the boy singing and attracting the bees/bumblebees. What is also interesting here is that, in my opinion, these off-plotline narrations are not of lower narrative quality than the narrations that stick to the plotline to a greater extentrather, they are of a different narrative quality. In these two specific examples, the narrator uses a method that is culturally valued in her everyday setting: she adds an original idea. If anything, the singing comes across as an imaginative and adventurousmaybe even poeticfeature in the narrations, certainly capable of awakening a listener's interest.
Suspense is the term used for situations where the narrator intentionally leaves out information, asks a question or hints at something that the listener does not know (Gorman et al., 2011). The aim of suspense is to create ambiguity. Examples: Girl 13 (seven years old) asks, "and can you guess what he found there?" [og aner du hva han fant der?]. Boy 14 (six years old) predicts a conflicted reaction when he says, "But the frog goes out" [Frosken går jo ut]. Both Boy 4 (six years old) and Girl 3 (six years old) urge their listener to "Look!" [se då!] to prove that their narratives are adequate. Girl 3 (six years old) asks the listener, "Do you think so?" [Tror du det?] about the ongoing actions. Boy 11 (six years old) gives clues about how the story should be understood, saying twice "that looked scary" [det så skummelt ut], and then "he just thought it was a tree" [han trodde bare at det var et tre] and "I think he's going to fall down, but he won't" [jeg tror at han kommer til å dette ned, men det kommer han ikke til å gjøre]. Boy 16 (six years old) anticipates the results of the action: "I bet he got wet. I think maybe he got tired too…" [Jeg vedder på at han ble bløt. Jeg tror kanskje han ble trøtt også…]. Some children (Girl 14, six years old; and Boy 17, seven years old) begin their utterances with "I think" [jeg tror], but this might reflect insecurity about the task rather than an attempt to create suspense.
When a narrator indicates that characters are creating problems, for example by being unkind, this represents the performance feature of conflict. According to Gorman et al. (2011), the aim of using this feature is to capture the listener's interest. Inappropriate conduct, unlike appropriate conduct, is recorded as an invitation to conflict. Frog, where are you? includes a number of potential conflicts. According to the "actantian model" (Greimas, 1983) 13 , the frog's escape from the boy is what gets the story started -it is the initiating conflict. This is reflected in wordings such as "escape" [rømme] (Girl 1, six years old) and "ran away" [stakk av] (Girl 1, seven years old). Indeed, this conflict is what the Frog Story tradition depends upon, and it is very close to the traditional plotline understanding. Other conflicts are also narrated or hinted at, such as "is bored" [kjeder seg] (Boy 4, six years old) and "then the boy got angry" [så ble gutten sint] (Boy 2, six years old), which are interpretations based on the boy's facial expression. Other characters involved in conflicts are the owl and the deer: "the boy tries to hide behind the rock. The owl takes the boy" [gutten prøver å gjemme seg bak steinen. Uglen tar gutten] (Boy 1, seven years old) and "and then the elk got angry" [og så ble elgen sint] (Boy 2, seven years old).

Paralinguistic devices
The performance feature of expressive sounds occurs when the narrator produces non-word sounds or sound effects (Gorman et al., 2011;Reilly, 1992) such as "he falls 'bang, smash, crunch'" [han detter "bang, knus, knas"] and "then he flies off, 'splash'" [så fyker han av, "skaplaesj"] (both Boy 8, six years old). Expressive sounds are a way to create involvement. This is a well-known feature of both cartoon films and children's play.
Another performance feature that can be used to create involvement is exclamatory utterances, which is when the narrator produces an emphatic sentence or interjection at high volume (Gorman et al., 2011;Reilly, 1992). These are often accompanied by other expressions of involvement such as smiles, laughter, a sad voice or a happy or high-pitched voice, such as when Boy 4 (six years old) indicates surprise by saying "Oh, an owl" [Oi, en ugle] and "Oh, a big rock" [Oi, en svaer stein]. When the narrators express involvement through humour or empathy, I consider it their intention to establish the experience of the text as a joint realm for narrator and listener. One example of this is how Girl 17 (seven years old) uses qualities of her voice to express empathy when saying "The dog sees a frog" [Hunden ser en frosk]. These voice qualities give expression to a friendly, maybe even devoted, way in which the dog is looking at the frog. 14 Figur 4 Picture 22. Reproduced from Mayer (1969) with the permission of the publisher.
When the same narrator says, "with many babies" [med mange unger], she also uses a voice expressing empathy. 15 Signals of involvement are often sent out in situations where the listener may express recognition of the narrator's intention, for instance by chuckling, smiling or making a facial expression similar to the narrator's own at about the same time as the narrator. Girl 16 (six years old) is aiming for involvement when she chuckles after saying, "and then he falls down from the elk" [og så detter han ned fra elgen]. The listener responds by laughing, which encourages the narrator to continue. On such occasions, the listener's response will verify that the narrator has succeeded in establishing a joint realm for the narrative.
14 That the dog does not see the frog as pray or as something unknownwhich it could easily have done in real lifealso shows the narrator's fictional competence. The story need not be reasonable or realistic. The narrator establishes a platform for the story using the tools in her possession, and the result is a new realm: the realm of the narrative. 15 This is something that many children do when talking about babies (in this case, frog babies), perhaps because, in the language culture which these children participate in, it is common practice to speak to young children (and animals) in a high-pitched voice; this is in fact one characteristic of "child-directed speech" (Snow, 1995).

Established features of performance: a summary
In their narratives, the Norwegian children use performance features that reflect, in various ways, the narrative culture in which the children participate. To capture their listener's interest, the Norwegian children typically use the features of conflict and fantasy. Conflict creates excitementboth in the narrator and in the listenerand is a well-known feature both in children's literature and in everyday speech. The conflicts that the narrators refer to relate both to the overall plotline and to individual pictures or episodes in the book. However, it appears to be the picturesparticularly the main character's facial expressions (bored, angry)that trigger references to conflicts. All narratives except two contain at least one instance of conflict, and there are a total of 281 instances of conflict.
Fantasy occurs in both on-plotline and off-plotline episodes in 22 narratives; there are a total of 38 instances. Fantasy is used to express both subjective relevance and inter-textual references. For example, the episode where the boy is singing to the bees (see above) may represent subjective relevancethe narrator may have found something in the text that will enhance its relevance in a subjective setting, such as an experience, a theme, a problem or something else that the narrator can use to connect to the text (Smidt, 1989). Sources of subjective relevance will, naturally, vary between narrators, and they can be found both in the text itself and in the general situation. This underscores that the meaning of a text does not reside solely within the text but also comes into existence as an experience within the narrator, meaning that the dynamic relationship between the text and the narrator is essential. Hence it cannot be assumed that a given text corresponds to a fixed or "one and only true" narrative or narrative structure. Depending on the basis for their narration, the narrators in the present material ramble in and out of a main structure, including things they notice, are inspired by, come to think of, have experienced, have learned, etc.and still they manage to finish their narrations. These perspectives open for interesting and new analytical possibilities in relation to offplotline narrative. In addition, fantasy may also be a result of inter-textual comprehensionof the narrator bringing traces from one text into another. In any case, the examples help to emphasise that there is no one accurate way to understand a text, and that off-plotline narrative is thus not a flaw (Hoel, accepted). In other words, from a performance perspective, what the narrator is aiming for may be an interesting narrative rather than a one that is strictly "by the book".
Involvement is highly valued during reading activities at Norwegian kindergartens, and the children are familiar with adults modelling involvementvisibly or audiblyby adopting expressions of sympathy, antipathy and humour. Expressive sounds are used in only nine of the seventy narratives on a total of 12 occasions, which may indicate that the children's ideas about book-based narrative diverge from those about cartoon films, etc. On the other hand, however, exclamatory utterances are frequent: 100 instances in 38 narratives. Empathetic statements are often associated with the feature of relationship. For example, someone may be presented as being so small that he or she needs to be taken care of. The feature of nature of characters' relationships occurs in 52 of the children's narrations of Frog, where are you?, with a total of 144 instances being recorded. The numerous different ways in which the narrators indicate the nature of relationshipsfamily, friendship, love connection, ownershipsuggest that they are accustomed to focusing on the ways in which the main characters of a book are related to one another, and that these relationships influence the ways in which books are read and discussed. When the narrators refer to ways of behaving or treating others appropriatelyusing the feature of conduct (which occurs in 11 narratives)this is also connected to their focus on relationships.
Almost all the narrators dramatise their narratives using speech. There is indirect speech in 60 out of 70 narratives (147 instances) and direct speech in 32 narratives (60 instances). In fact, speech is frequently used in Norwegian children's literature, and it is also characteristic of how literature is presented to the children: as a dialogue between children and an adult reader. The dialogic nature of the reading culture at Norwegian kindergartens is also reflected in the children's use of the feature of suspensei.e. when the narrators seek to involve the listener by asking questions related to the pictures or to the listener's expectations, just as they are used to doing in their own reading practiceswhich occurs in 23 narratives. The Norwegian children's narratives also display an additional performance feature belonging to the category of paralinguistic devices: reading voice. This finding contrasts with those of Gorman et al. (2011), who found minimal use of paralinguistic devices, ascribing this to the effect of genre by claiming that storybook narrative tasks tap children's decontextualised narrative skills.

Reading voice
In my analysis of the children's narratives I identified an additional performance feature that I have called reading voice. Reading voice is characterised by a pronunciation that differs clearly from the children's everyday speech, such as a standardised "literary" pronunciation or the dialect of the Norwegian capital (which is predominant on national television and in similar contexts). Some narrators change the quality of their voice (making it squeaky or silly), some add a marked stress on single words and dramatise by varying their volume, and others seek to remove the intonation of their everyday speech, striving for a more "flat" and factual style, similar to the "non-involved" voice of news anchors. Boy 2 (six years old) constitutes a good example of the use of reading voice: he uses flapped (dental) r instead of guttural r; his reflexive pronoun is /sei/ not /seg/ as is usual in his dialect; he changes suffixes: /sengen/ instead of /sengå/ 'the bed', /luften/ instead of /loftå/ 'the air', /frosker/ instead of /froskar/ 'frogs'; he uses the determinative pronoun /noe/ instead of /noge/ or /någe/ 'some'; he uses the indefinite article /en/ not /ein/; he uses hypercorrections such as /letet/ instead of /leita/ or /lette/ 'searched'; he pronounces normally mute consonants: /var/ instead of /va/ 'was'; he uses stress to emphasise single words: "then he saw a ELK" [så så han en ELG]; and he generally uses a dramatic narrative voice.
Reading voice has much in common with what Heilman et al. calls a literate style of speaking (2010), and Reilly (1992) also highlights prosodic features such as voice quality and lexical stress in her analysis. In addition, the children's use of reading voice also strongly resembles children's use of their voices in role play, which is an unorganised voluntary activity where the child both participates in and directs play (Kunnskapsdepartementet, 2011) and where a child may alternate between "play voice", which indicates that the utterance belongs to the realm of play, and normal voice, which indicates that the utterance belongs to the real worldfor instance, planning the further development of play. The Norwegian children's narratives contain similar switches between reading voice and normal voice. One example is Girl 1 (six years old). During her narrative, there is a word she cannot remember. She stops telling her story and addresses the listener, thus no longer positioning herself as a narrator, and therefore no longer using a special voice: "I don't remember what it's called" [eg huske ikkje ka de hette igjen], she says in her normal voice. Vedeler (1999) says that Norwegian children's language during role play becomes "more literary than in other contexts" [mer litteraert enn i andre sammenhenger] (p. 69my translation). According to Vedeler, this is because play places demands on participants to convey meaning within the theme of the play, which is taking place in the children's imaginations and can only be communicated linguistically. Analogous demands are placed on the children in the present material: they are asked not only to narrate but also to be the narrator, and the narrator position must be communicated linguistically.

Fixed wording
Another additional performance feature found in the Norwegian children's narratives is genre-specific wording or phrases, which in this article is referred to as fixed wording. I have also included some literary-language syntax in this category. The effect of fixed wording is a literate style of speaking (Heilmann et al., 2010), and this is a way to contribute to communal meanings and shared understanding (Engel, 1995).
Fixed wording is characteristic of a narrative communicative situation. To some extent it is culture-specific and hence not directly transferable between the Norwegian and English languages, but in some cases there are direct equivalents. Examples include the formulas used to begin a fairy tale: "Once upon a time" [Det var en gang] (Boy 2, six years old); or end it: "and they lived happily ever after" [snipp, snapp, snute, så var eventyret ute] (Boy 16, seven years old) (note, however, that the ending formulas are equivalent in function only, not in verbatim content). Repetition patterns, such as "they search and they search" [de leter og leter] (Girl 7, six years old) are familiar from narratives that children encounter, as are wordings like "and then he goes back home" [så drar han hjem igjen] (Boy 17, six years old) because the home-out-home structure is common in children's literature (Birkeland & Mjør, 2012). An expression such as "suddenly" [plutselig] (Boy 2, six years old) represents a common way for children to create excitement in a narrative. Statements such as "then they were happy" [så ble de glade] (Boy 11, six years old) reflect the happy-ending culture of children's literature and have therefore been recorded as fixed wording. So have formulations such as "and that's the end of the entire book" [så var hele boka slutt] (Boy 13, seven years old), "finished" [ferdig] (Boy 2, seven years old) and "that's the end of the entire book" [der var hele boka ferdig] (Boy 17, six years old), because these are things that adults often say when ending reading activities with children.

Outside comments
Outside comments are also included in the analysis because they often indicate that the child is leaving the position of narrator. In several cases, outside comments arise from uncertainty about what to call the animals in the pictures, depending on the local fauna. The narrators seem afraid of naming the animals incorrectly, even though this would not affect the narrative. For example, Girl 5 (seven years old) says, "What animal is that?" [Hva er det for et dyr?]; Boy 6 (six years old) says, "What's that? An eagle?" [Hva er det? Ei ørn?]; and Girl 1 (six years old) says, "I don't remember what it's called" [Jeg husker ikke hva det heter igjen]. This uncertainty, no doubt made stronger by the fact that it is sometimes unclear what species the animals in the illustrations belong to, is a disruptive factor in the Norwegian children's narrative performance.

Additional features of performance: a summary
The feature of reading voice is used in 22 narratives, where a total of 121 instances are recorded. Some of the children use reading voice consistently throughout their narratives while others do so only in parts of their narratives. The nature of the reading voice variesthere is not a single reading voice used by all narrators. However, the specific features described above are used in several narratives, even though individual narrators combine them in different ways. The use of reading voice reflects a "pretend-as-if" situation which is highly valued in the Norwegian kindergarten culture; the child is interacting with the listener in a "fantasy realm" common to them both.
Fixed wording seems to have much in common with the "discourse markers", "cue phrases" or "pointers for narratives" of writing research, which are described as genre-signalling words or phrases that ensure that a text is understood as the narrator intends it to be (Senje & Skjong, 2005). Fixed wording is often used to begin and end stories and to mark changes or turning points (i.e. "suddenly" [plutselig])that is, as stylistic markers of macrostructure. Fixed wording is used in 54 of the 70 narratives, and a total of 129 instances are recorded in the material.
Both of the additional featuresreading voice and fixed wordingserve as significant indicators of literacy influence. The children use their experience with both adult narrators and texts when solving the task that they are given. The adult narrators serve as models for the role play of "being a narrator" while familiar wordings from previously experienced texts ensure that the narratives will be perceived as intended.

Recognising children's narrative: between scientific study and face-to-face action
Through their narratives, the Norwegian children come across as narrators who are familiar with important aspects of narrative but also more or less ignore the canonical structural norms of the Frog Story tradition. What could it be that motivates children to narrate in one way instead of another? From a cross-linguistic perspective, the answer is to be found within the narrators and in their language systems, but this is obviously not a sufficient answer considering that the narrators clearly draw on their experiences from kindergarten. In fact, the Norwegian children construct their narratives in the light of criteria that are given value in the culture of which they are part (Bloome et al., 2003).
What types of cultural interaction and what socio-cultural factors at Norwegian kindergartens may be considered as fundamental contributors to the children's narratives? The Norwegian kindergarten culture is rooted in a holistic philosophy of teaching that entails a minimal focus on standardised or norm-based methods for carrying out reading activities along with a greater emphasis on creating excitement and joy. The reading culture at Norwegian kindergartens facilitates dialogue, participation, visible involvement, embellishment and the expression of empathy. In picture-book reading, the pictures are of great importance to Norwegian children's meaning-making in relation to the narratives, and both adults and children focus on the pictures. The adults invite the children to study the pictures in detail, to make descriptions, and to use their imagination to elaborate on what they seeoften in the form of short narratives within the narrativeand the children are encouraged to link what they see in the pictures to their own thoughts and experiences. These activities are carried out in order to make the children participate actively in the reading activity, to make them engage in and contribute to linguistic activities. Elaborations of the kind that the children in the present material often makein several cases off-plotline but still narrative in natureare clearly appreciated and valued by the adults.
The Norwegian children's literacy experience is clearly visible in their interest in reading and elaborating on pictures and in their use of fixed wording and reading voice. The kindergarten is part of a literacy culture that must be taken into consideration in any discussion or assessment of Norwegian children's language development. The dialogic reading culture at Norwegian kindergartens values participation, engagement and reflection -features that are also emphasised within a "text as performance" perspective. The tension between narrative-as-text and narrative-as-performance may be the result of different views on what language is and how language should be usedand these are culturally based features. Bloom et al. (2003) claims that Western culture is dominated by a dualistic language culture that separates text from context, where meaning is assigned to the symbol rather than to the contextualised use of the symbol. 16 At Norwegian kindergartens the norm based assessment represents this perspective in particular.
The dualistic approach to language and narrative has been criticised by researchers and practitioners alike, both generally for its hegemonic status in assessment contexts and specificallyin relation to narratives -for how the understanding of young children's narratives is presented as restricted to the assessment of the children's use or non-use of a finite set of opportunities. One result of a dualistic language ideologywhere narrative is seen as a text and not considered in the light of its context and usemay be that narrative performance and the use of narrative in face-to-face situations are considered unimportant and not valued. In this way, the dualistic language perspectivewith its focus on narrative structure as a decontextualised standard for children's narratives rather than on language as used in the light of its social and cultural dynamicruns counter to the dialogic reading culture at Norwegian kindergartens.
Despite this dialogic reading culture, however, the narrative as a well-structured text is generally what is most highly valued in norm-based assessments of children's narratives. The narrative theory underpinning Frog Story research is a set of strict rules for how to analyse children's narratives, and these rules exert a powerful influence as an "exemplary model". The present research suggests that there is an excessive focus on assessing children's narrative achievements by studying the ways in which they either follow or ignore certain set rules in their narratives and that there is therefore a need for different ways to study and describe children's linguistic and narrative skills.
Norwegian children's focus on dialogue and participation during reading activities, especially as regards the pictures in the books, does affect the way they narrate picture books. Children learn their narrative forms from the narratives they experience, are asked about and are encouraged to make. Narrative structure is shaped by cultural forces, as are content and function. Narrative structure constitutes a fundamental principle within the cross-linguistic Frog Story tradition, but there is no systematic communication of plotline in picture-book reading at Norwegian kindergartens. This is one of the reasons why the Norwegian children appear to be inexperienced narrators according to that tradition. Considering this, if Frog, where are you? is to be used to assess Norwegian kindergarten children's language development, we must take account of aspects other than those traditionally focused on, such as references to characters' relationships, embellishments and paralinguistic devices. of a given language practice, but also of the power relations that follow from the imposition of a standard. This is why it is important to draw attention to tension between cultures rather than assuming that the situation is static.