A Developmental Study of the Narrative Components and Patterns of Chinese Children Aged 3–6 Years

Abstract

To investigate the narrative development of Chinese-speaking children aged 3–6 years, 80 children were prompted to tell personal stories. High point analysis was used and both narrative components and overall narrative patterns were analyzed. In terms of narrative components, Chinese children were more skillful in using complicating action, orientation and evaluation than they were in resolution, direct and reported speech. With age, their narratives were longer and richer in content. No gender differences were found in any of the seven narrative components. For narrative patterns, the most frequent was the two-event pattern at age 3 and the chronological pattern from age 4 to age 6. However, the chronological pattern at age 5 and 6 differed from the chronological pattern at age 4. The proportion of leap-frog, two-event and one-event patterns decreased with age, but most Chinese children at age 6 have yet to develop the ability to tell a classic narrative that includes resolution. These findings provided evidence for a developmental pattern of the narrative development in preschool Chinese children and suggested more attention should be paid to improve children’s ability to resolve a story, telling listeners how everything worked out in the end.

Introduction

Over the past 50 years, narrative has attracted much attention from scholars in different fields. Narrative is oral discourse that recapitulates a series of real or imaginary events, usually in temporal order. Narration is part of everyday life. The ability to narrate is critical to forming relationships with others, for acquiring literacy, and for testifying in court, to name only a few important uses (Chang 2009; Jordan et al. 2000; Snow et al. 2007).

Narrative Components

Labov (1972) described the structure of African American teenagers’ oral personal narratives: all such narratives were organized around a certain heavily evaluated high point (the climax of the experience and the point of telling the narrative). He further identified five basic narrative components: orientation (i.e., descriptive information about the setting of an event), evaluation (i.e., the narrator’s feelings about who, what, when, where, and why something happened), complicating action (i.e., discrete, sequenced past tense events that together constitute some larger happening), resolution (the result of the story; the way things turned out), and codas (expressions that close the narrative). Labov’s approach, though put forward almost 50 years ago, endures as a pioneering and well-established framework for analyzing personal narrative. Many newer research methods are based on it, and it has been found appropriate for analyzing the personal narratives of children from 3 to 9 years of age (McCabe and Bliss 2003). Peterson and McCabe (1983) analyzed over a thousand personal narratives of 96 children aged from 4 to 9 years and made refinements to Labov’s classification of narrative components. They substituted the broader category of “appendages” (the abstract at the beginning or attention-getters, and codas at the end of narratives) instead of simply “coda.” Chang (2009) further developed Peterson and McCabe’s framework of narrative components and added direct speech and reported speech based on Taiwanese children’s narrative data, though these categories are not exclusive to Taiwanese narration and have been found in European North American narration (Ely and McCabe 1993). Ely and McCabe (1993) found girls cited their own, their mothers’ and other children’s past speech more than did boys, and did so more directly. These investigators corroborated Labov’s finding that young children’s narratives are composed of orientation, complicating action, evaluation, appendages and speech. Children became better narrators with age, employing more types of narrative components and offering more detailed information in regard to each component (Chang 2000; Minami 1996; Peterson and McCabe 1983; Rollins et al. 2000; Umiker-Sebeok 1979; Xu 2007). McCabe (2017) reviews research documenting differences in high point structure that are due to cultural variation; for example, the high point component of evaluation is much more prominent in narratives from African American and European North American children than in Taiwanese children. As another example, Japanese children tell narratives that are very succinct, consist of narration that can usefully be depicted as consisting of primarily 3-line stanzas, and that combine telling about multiple similar experiences in one coherent narrative—a very different pattern than American English-speaking children (Minami and McCabe 1991). This cross-cultural variation provides a compelling reason for examining 3- to 6-year-olds’ narration in other cultural contexts.

With regard to children in Mainland China, Wu et al. (1984) studied their picture-book narrative abilities, yet only analyzed syntactic structure. Shi (1986) asked children aged 3–7 to observe monkeys in the zoo and then asked them to narrate what they had observed in the zoo the next day, but the research did not analyze any specific narrative dimensions. Li et al. (2006) studied children’s picture-book narrative ability and found that age 4–5 was a crucial period for the development of children’s picture-book narrative ability. The features and hidden meaning of pictures both had an effect on children’s narratives. Li (2013) conducted a longitudinal study of the narrative developmental trajectory of micro- and macro-structure in preschool children aged 4–6 using personal narratives. The analysis of the four macro-structural dimensions showed that complicating action was a dimension that children aged 4 would have to spend 6 months to develop substantially, while evaluation and dialogue were dimensions that children aged 4 would need to spend 12 months to progress significantly. Children could barely use dialogue at age 4 but started to use it at age 5 and could actively use dialogue at age 6. Among all macro-structural dimensions, children’s skill in using orientation showed no significant progress over the whole research period. Wu (2014) investigated the narrative structure of 10 children aged 9–11 telling a story of their own choosing (a story learned from a storybook; a story told by an adult, a sibling or classmate; a story made up by themselves), and the results showed that Chinese children aged 9–11 only included three of five narrative components from Labov’s fully-formed narrative structure model: orientation, complicating action and resolution.

Narrative Patterns

Peterson and McCabe (1983) found Labov’s conception of a narrative’s overall high-point structure to be a useful starting point for characterizing younger children’s productions. They discerned seven narrative patterns (classic pattern, ending-at-the-high-point pattern, chronological pattern, leap-frog pattern, miscellaneous pattern, two-event pattern and one-event pattern) in European North American children. McCabe and Rollins (1994) later adapted the research on narrative patterns into a clinical research tool for use by speech-language pathologists in appraising children’s narrative skills, which proved to be applicable to children from some cultures other than African American or European North American.

Lai et al. (2010) indicated the most frequent narrative pattern of Taiwanese middle-class children at both age 4 and 5 was a leap-frog pattern, meaning that they did not tell events in sequential order and left out critical events. However, Xu’s (2007) research on the personal narratives of 48 Taiwanese children aged from 3 to 5 showed that the most frequent pattern at age 3 was a two-event pattern, a chronological pattern at age 4 and a classic pattern at age 5.

In view of the contradictory findings regarding the developmental progression of narrative structure in young Chinese-speaking children, the present study was intended to explore the matter further, if not altogether to resolve the discrepancies.

The Present Study

As introduced above, we see that more attention has been paid to picture-book narratives (Li et al. 2006; Wu et al. 1984) than to personal ones (Li 2013). Among different types of narratives, personal narratives are most frequently used by children and are integral to all cultures and ages (McCabe and Bliss 2003). Therefore, personal narratives were chosen to be collected and analyzed in this study. Some previous studies are case studies (Wu 2014), which cannot fully represent the whole picture of Chinese children. To date, research has not examined children over the whole preschool period from age 3 to 6 years, and no research has been done on narrative patterns of Chinese children in Mainland China. In addition, Chinese children and European-American children are nurtured in very different cultures. The narrative development of children in Chinese language culture is likely to present a picture that differs from their English-speaking peers. Therefore, the aim of the present study is to explore the developmental trend of narrative component usage and overall patterns of personal narratives of Chinese-speaking children aged 3–6 and to probe the gender differences in these two dimensions. High Point analysis (HP) will be employed. HP analysis was chosen because it looks at both narrative components and patterns (Peterson and McCabe 1983). It has also been widely employed by researchers to assess children’s narrative abilities across different languages and cultures (Champion et al. 1995; Crago et al. 1997; Minami and McCabe 1991; Silva and McCabe 1996).

Method

Participants

The participants were 80 Mandarin-speaking preschool children ranging in age from 3 to 6 years old (20 children for each age group with a balance of gender). They were the only children from middle-class families. Most mothers had graduated from a university, while a few reported that they completed their education in a vocational school. All children were recruited from a kindergarten affiliated with a prestigious university in Shanghai, China. They were assessed as typically developing children with no learning disabilities and no speech or hearing or visual problems when they were accepted. Table 1 presents a summary of participant demographic data.

Table 1 Demographic data for children from age 3 to age 6

Procedures

Before interviewing the children, the experimenter developed an interactive relationship with each one with the help of toys. The same experimenter interviewed children in their kindergarten to elicit narratives following the procedure developed by Peterson and McCabe (1983). First, the experimenter provided the child with a number of prompts, which consisted of brief personal narratives with specific topics like doctor visits, injuries, spills, fighting experiences, and family outings, etc. Then the child was asked to tell a similar story based on his/her own experience, during which the experimenter prodded the child by using neutral and non-leading phrases like “And then?” “Yeah,” “tell me more” or merely repeating what the child had just said. No more specific follow-up questions were asked because children’s stand-alone skill in narrating was to be examined. Stickers were given to children as rewards at the end of the conversation.

All interactions were audio- and video-taped. Four or five narratives were generated in response to prompts. Each child’s two best examples of narrative were used for analysis by referring to the structure, evaluation and temporality of the story. Previous studies (Chang 2000, 2004; Peterson and McCabe 1983) on children with typical development usually select the longest narrative for analysis. Re-examination of the data in this study indicates that the choice of best narratives among typically-developing children by selecting their longest narrative does not differ from a comprehensive assessment of the structure, evaluation and temporality of all narratives produced. Narratives were divided into propositions, defined as each clause with a verb and its arguments.

The language samples were transcribed from videotapes in Chinese characters following the guidelines for Codes for Human Analysis of Transcripts (CHAT), part of the Child Language Data Exchange System (CHILDES; MacWhinney 2000), parsed at the clause level (Berman and Slobin 1994). To objectively examine the children’s grammatical and independent narrative abilities, utterances such as fillers (ums, ahs, etc.), imitations, self-repetitions and utterances that were supported by the experimenter were excluded from analysis.

The transcripts were completed by four university students trained in the transcription process but blind to the hypotheses of the study and then checked a second time by the first author, who had conducted all the interactions with the subjects and was familiar with the children’s speech patterns.

Basic Language Productivity of Children Aged 3–6

Narrative discourse requires general oral language competencies. After transcription and check, Computerized Language Analysis (CLAN) was used to count the number of four basic language measures: number of utterances, mean length of utterance (MLU), number of words, and number of different words.

Narrative Components

The two best narratives produced by each participant were analyzed using the first step in HP analysis, which assigns propositions to one of the following narrative components depending on its primary function (Peterson and McCabe 1983; Chang 2009). Altogether seven types of narrative components were categorized. The definitions and examples are as follows:

• 1. Complicating action

Complicating action refers to the series of specific events which build up to the high point of the narrative. Note that complicating actions in past tense are also called events; events/complicating actions comprise a total experience or happening that is narrated. This is the backbone, the information component of the narrative. E.g.

• 一 只 蟑螂 一下子 跑到 了 电脑 上面 [^c].

• yīzhī zhāngláng yīxiàzi pǎo dào le diànnǎo shàngmiàn [^c].

• A cockroach suddenly ran to the top of computer.

• 2. Resolution

The resolution refers to specific events that occur after the high point and resolve the high point action or crisis, clearing the stage or capping off the experience. For example, when a child tells stories of his visiting a doctor, he makes a high point of getting a shot. After the shot, he resolves the high point by saying the following sentence.

• 然后 我们 就 回 家 去 了 [^c].

• ránhòu wǒmen jiù huí jiā qù le [^c].

• Then we went back home.

• 3. Appendages

This is a composite category which includes narrative comments that are attached to either the beginning or the end of the main body of the narrative. They are superfluous niceties of narration, and are of four types.

Abstracts are defined as summaries or encapsulation of the whole narrative that appear at its outset. E.g.

• 昨天 我 把 腿 摔 断 了 [^c].

• zuótiān wǒ bǎ tuǐ shuāi duàn le [^c].

• Yesterday I broke my leg.

Attention-getting devices are located at the beginning of a narrative, in order to attract the attention of listeners. E.g.

• 你 猜 我 昨天 干了 什么 了 [^c]?

• nǐ cāi wǒ zuótiān gànle shénme le [^c]?

• Guess what I did yesterday?

Prologues are remarks that bring a narrative’s importance or consequences to the present time, but occur at the beginning of a narrative. E.g.

• 我的 手 还 疼 呢 [^c].

• wǒde shǒu hái téng ne [^c].

• My hand still hurts.

Codas are endings of a narrative, signaling nothing else happened relevant to the story line. E.g.

• 就 没 了 [^c].

• jiù méi le [^c].

• That’s all.

• 4. Orientation

Orientation is descriptive information about who, what, where, and when something happened. For instance, the first two phrases of the following sentence are both orientations.

• 春 游 的 時候 [^c] < 我 > [/] < 我 > [/] 我 五岁 时 [^c] 去 < 一个 [/] 一个 海洋馆 [^c].

• Chūn yóu de shíhòu [^c] < wǒ > [/]< wǒ >[/] wǒ wǔsuì shí [^c] qù < yīgè > [/] yīgè hǎiyángguǎn [^c].

• When spring outing, I [/] I [/] I at the age of five, …went to an acquarium.

• 5. Evaluation

Evaluation refers to statements that tell the listener what to think about a person, place, thing, event or the entire experience. E.g.

• 我 就是 不敢 把 屁股 露出來 [^c] < 因为 > [/] 因为 打 下 去 太 重 了 [^c]!

• wǒ jiùshì búgǎn bǎ pìgǔ lùchūlái [^c] < yīnwéi > [/] yīnwéi dǎ xià qù tài zhòng le [^c]!

• I dare not expose my ass [^c] Because [/] because the shot was too rough.

• 6. Direct speech

Direct speech refers to clauses that include speech directly uttered by persons involved. E.g.

• 然后 老师 说 [^c] < 杨海龙 打 人 > [“] [^c] < 然后 你 也 打 他 > [“] [^c].

• ránhòu lǎoshī shuō [^c]< yánghǎilóng dǎ rén >[“] [^c]< ránhòu nǐ yě dǎ tā >[“] [^c].

• Then the teacher said [^c]“Yang Hailong hit you.” [^c] “Then you should hit him too.” [^c].

• 7. Reported speech

Reported speech refers to clauses which include indirect speech of persons involved. E.g.

• 医生 < 就 > [/] < 就 > [/] 就 叫 我 不要 哭 [^c].

• yīshēng < jiù > [/] < jiù > [/] jiù jiào wǒ búyào kū [^c].

• The doctor then asked me not to cry.

Narrative Patterns

The second step of HP is to determine the overall narrative pattern—the macrostructure of the narrative as a whole. In order to do this, it is necessary to identify the narrative pattern by answering each one of the questions presented in Fig. 1. Answering yes allows the person to proceed to the next question. Answering no results in the determination of the type of narrative pattern displayed in the right column. A more detailed description of the seven narrative patterns is shown in “Appendix A”.

Fig. 1

Adapted from McCabe and Rollins (1994)

High point analysis.

All narratives were scored by the third author. Codes of narrative components were entered on the computer directly below each clause line and the narrative pattern was evaluated at the end of each story. 50% of the transcripts were checked on a line-by-line basis for coding accuracy by the first author who was very familiar with the coding system. The inter-rater agreement was estimated at 90% for narrative components and 88% for narrative patterns.

Results

Basic Language Productivity of Children Aged 3–6

Descriptive statistics of four basic measures—the number of utterances, mean length of utterances (MLU), number of words, and number of different words are displayed in Table 2 for four age groups.

Table 2 Means, standard deviations of children’s basic language productivity at each age

Table 2 shows that the number of utterances, MLU, number of words, and different words are all increasing with age. Four separate one-way ANOVAs were conducted. The number of utterances (F (3, 76) = 30.60, p = .00, η² = .55), MLU (F (3, 76) = 7.49, p = .00, η² = .23), number of different words (F (3, 76) = 24.30, p = .00, η² = .49), and number of words (F (3, 76) = 25.32, p = .00, η² = .50) all developed significantly from age 3–6. Post Hoc analysis (LSD) showed there remained significant differences between age 3 and 4 in number of utterances (p = .001), MLU (p = .001), number of words (p = .002) and number of different words (p =.00). Significant differences were found between age 5 and 6 in number of utterances (p = .00), number of words (p = .00) and number of different words (p = .00). Thus, children’s grammatical and semantic ability both developed significantly from age 3 to age 6. Especially, they developed significantly from age 3 to 4 and from age 5 to 6.

Narrative Components

Table 3 indicates children at age 3 could tell on average 4.75 events, provide 1.75 orientation clauses, and 1.25 evaluation clauses. Only one child offered a coda and 3-year-olds did not offer any resolution or include direct or reported speech when telling their past experiences. At age 4, the mean number of events children could tell was 9.25. Orientation and evaluation clauses rose to 5.3 and 3.2 respectively. Five children ended their narrative with a coda. Three children included direct speech and one child used reported speech in their narratives. One out of twenty children provided resolutions. Five-year-old children could tell on average 14.85 events, providing 8.65 orientation and 3.9 evaluation clauses and four out of twenty children provided resolutions. At age 6, the mean number of events children could tell was 14.90, which was similar to 5-year-olds. The orientation and evaluation clauses increased to 8.9 and 5. Six children offered codas. Eight children used direct speech and seven children included reported speech in their narratives. 50% of the participants (ten children) resolved the high point action.

Table 3 Means and standard deviations for the number of clauses of each type of narrative component

Table 3 also shows an overall upward trend in frequency of all seven narrative components from age 3 to age 6. Seven separate one-way ANOVAs were conducted. The complicating action (F (3, 76) = 10.29, p = .00, η² = .29), orientation (F (3, 76) = 10.84, p = .00, η² = .30), evaluation (F (3, 76) = 5.92, p = .001, η²= .19) and resolution (F (3, 76) = 6.82, p = .00, η² = .21) all developed very significantly from age 3 to 6. The direct speech (F (3, 76) = 3.51, p = .02, η²= .12) and reported speech (F (3, 76) = 3.25, p = .03, η² = .11) developed significantly from age 3 to 6. Post Hoc analysis (LSD) showed there remained significant differences between age 3 and 4 in complicating action (p = .04), evaluation (p = .04) and orientation (p = .02). Significant differences were found between age 4 and 5 in complicating action (p = .01) and orientation (p = .02). Significant differences were also found between age 5 and 6 in resolution (p = .03). No significant differences (p > .05) were found in appendages, direct or reported speech between adjacent ages. With age, preschool Chinese children are getting more adept in narrating their own experiences and their narratives are richer in content.

Hoff-Ginsberg (1992) made a persuasive argument that both frequency and proportion of language measures should be assessed. To explore the core types of narrative components, the proportion of different types of narrative components for every child at each age group was calculated with the number of each narrative component divided by the total number of seven narrative components. The mean proportions of different types of narrative components for each age group are presented in Fig. 2.

Fig. 2

Mean proportion of different types of narrative components for age groups 3–6

Figure 2 shows the mean proportion of complicating action, orientation, evaluation and appendages slightly fluctuated from age 3 to age 6. However, the mean proportion of resolution, direct speech and reported speech showed a trend of increasing. Seven separate one-way ANOVAs were conducted. The proportion of complicating action (F (3, 75) = 1.97, p = .13, η² = .07), orientation (F (3, 75) = 1.33, p = .27, η² = .05), evaluation (F (3, 75) = .61, p = .61, η² = .02) and appendages (F (3, 75) = .24, p = .87, η² = .009) did not develop significantly from age 3 to 6. The resolution (F (3, 75) = 4.09, p = .01, η² = .14), direct speech (F (3, 75) = 2.87, p = .04, η² = .10) and reported speech (F (3, 75) = 3.43, p = .02, η² = .12) developed significantly from age 3 to 6. Post Hoc analysis (LSD) showed there remained significant differences between age 3 and 6 in the mean proportion of complicating action (p = .02), resolution (p = .005), direct speech (p = .007) and reported speech (p = .008). Significant differences were also found between age 4 and 6 in resolution (p = .01) and reported speech (p = .02).

Narrative Patterns

To work out the core patterns used by preschool Chinese children, the proportion of different types of narrative patterns for every child at each age group was calculated with the number of each narrative pattern divided by the total number of narrative patterns. The mean proportions of different types of narrative patterns for each age group are presented in Table 4.

Table 4 Mean proportion of different types of narrative patterns (%) for age groups 3–6

Table 4 shows the mean proportion of end-at-high-point, leap-frog and chronological pattern fluctuated from age 3 to age 6. However, the mean proportion of classic was moderately increasing and the miscellaneous and two-event pattern were decreasing. Seven separate one-way ANOVAs were conducted. The classic pattern (F (3, 76) = 4.48, p = .006, η² = .15) developed significantly from age 3 to 6. The end-at-high-point pattern (F (3, 76) = 2.15, p = .10, η² = .08), leap-frog pattern (F (3, 76) = 1.70, p = .17, η² = .06) and chronological pattern (F (3, 76) = 2.33, p = .08, η² = .08) did not develop significantly. The mean proportion of miscellaneous pattern (F (3, 76) = 6.42, p = .001, η² = .20) and two-event pattern (F (3, 76) = 3.43, p = .02, η² = .12) decreased significantly from age 3 to 6. Post Hoc analysis (LSD) showed there remained significant differences between age 3 and 4 in leap-frog pattern (p = .03) and miscellaneous pattern (p = .00). Significant differences were also found between age 4 and 6 in classic pattern (p = .005).

As is shown in Table 4, the one-event pattern only occurs in 3-year-old’s narratives. The more frequent occurrences of the two-event (40%), miscellaneous (25%), and one-event patterns at age 3 compared to older age groups indicates that 3-year-old children have more difficulties with coherent language production. At age 4, the most frequent narrative pattern was the chronological pattern (42.5%), which consisted of temporally sequential event clauses with no particular climax evaluated. Leap-frog narratives, which violate sequential order in telling and often also leave important events out, are more frequent at this age than any other ages. The chronological pattern was also the most frequent pattern at age 5 and amounted to 62.5%, which indicated that 5-year-olds tend to merely recount the events chronologically without dwelling evaluatively on a particular high point event. At age 6, the three most frequent narrative patterns were the chronological pattern (40%), classic pattern (20%) and end-at-high-point pattern (17.5%). This indicates that children can narrate longer narratives that comprise more events by this age. The increasing proportion of end-at-high-point narratives indicated that children at age 6 are progressing towards well-organized classic narrative patterns.

Gender Differences

A one-way ANOVA on the frequency of each type of narrative components and patterns with gender as factor was conducted. The results are presented in Table 5.

Table 5 Comparisons of gender differences among children aged 3–6 on the number of clauses of each type of narrative component

Given the low frequency of some types of narrative components and their number of comparisons shown in Table 5, the Bonferroni correction, with 7 comparisons, a significance level of .007 (.05 divided by 7) has been used. No gender differences were found in complicating action (F (1, 78) = .64, p = .43, η² = .008), orientation (F (1, 78) = .17, p = .68, η² = .002), evaluation (F (1, 78) = .78, p = .38, η² = .01), resolution (F (1, 78) = .16, p = .69, η² = .002), appendages (F (1, 78) = .44, p = .51, η² = .006), direct speech (F (1, 78) = .22, p = .64, η² = .003), and reported speech (F (1, 78) = 5.91, p = .02, η² = .07).

In terms of narrative patterns, only one 3-year-old boy produced a one-event pattern. The mean number of leap-frog pattern for boys and girls is the same. No gender differences were found in other types of narrative pattern with classic pattern (F (1, 78) = .69, p = .41, η² = .009), end-at-high-point pattern (F (1, 78) = 1.24, p = .27, η² = .016), chronological pattern (F (1, 78) = .07, p = .79, η² = .001), miscellaneous pattern (F (1, 78) = 0.51, p = .48, η² = .006) and two-event pattern (F (1, 78) = .10, p = .75, η²= .001).

Discussion

Overall Development

With age, the narrative components and patterns of Chinese children are both developing. In terms of narrative components, our cross-sectional exploration of the basic high-point structural components confirms that the preschool years are a period of remarkable increase in the frequency of narrative components produced in personal narratives of Mainland Mandarin-speaking Chinese children; while 3-year-old children did not include resolutions or direct or reported speech in their narratives, 4-year-olds included all seven narrative components. All participants at each age group offered complicating action in their narratives. The frequency of complicating actions in narratives increased significantly from age 3 to 5. Children substantially develop an awareness of the importance of embedding orientation in their narratives to aid their listeners’ comprehension of events. Similarly, Peterson and McCabe (1983) found that older English-speaking children used significantly more types of orientations in their narratives than younger children. More than 60% of the subjects at each age group in the present study offered evaluation in their narratives. Evaluation developed significantly from age 3 to 4. The proportions of direct and reported speech developed significantly from age 3 to 6 although they constitute less than 5% of the components. For narrative patterns, the proportion of classically patterned stories—an epitome of narration in some cultures—increased significantly from age 3 to age 6 while the occurrence of two-event and miscellaneous pattern decreased significantly.

This orderly development with age may be closely related to children’s basic language level and overall cognitive development, for example, consistent with lexical development. As noted, the most common pattern in Mandarin-speaking children’s narration at age 3 was the two-event pattern. This result is in line with that of Xu’s (2007) study of Taiwanese children. This relatively modest narrative accomplishment may be attributable to the limited general language productivity of 3-year-old children. The number of words and number of different words and the number of utterances and MLU of 3-year-old children in this study were significantly lower compared with older children. Li (2013) found that Mandarin-speaking children at age 3 could use about 1000 words and could mainly use simple or incomplete sentences to express their meanings. In view of the relatively limited language production of children at age 3, it is not surprising that 3-year-olds mainly produce one-event, two-event or miscellaneous patterns. This result may be consistent with the level of cognitive development as well. At age 4, the most frequent production of leap-frog narration among Chinese children compared to other ages is similar to the result of Peterson and McCabe’s (1983) and Lai et al.’s (2010) studies. Recent research has consistently reported significant relationships between verbal ability and executive functioning measures among preschoolers (Blair, 2003; Hongwanishkul et al., 2005). Executive control skills have repeatedly been found to improve gradually during childhood, with a pronounced developmental progression between 5 and 9 years of age (Best and Miller 2010; Welsh et al. 1991). Four-year-old children’s executive ability of inhibition, updating and task-shifting is developing, but rather limited because their prefrontal cortex is not mature yet. The study of the cranial nerve mechanism shows that the prefrontal cortex is an important physiological basis for participating in the execution process (Roberts and Pennington 1996). Children aged 4 cannot manipulate and transform information in a logical way, failing to understand the true relationship between cause and effect. Therefore, they are most likely to jump around from event to event, adding irrelevant information, omitting an occasional key event and in general placing a great burden on listeners.

Of note, the most frequent narrative pattern among preschool Chinese children was the chronological pattern from age 4 to age 6. However, the chronological pattern at age 5 and 6 differed from the chronological pattern at age 4 (Examples are shown in “Appendix B”). In some cases of chronological patterns, according to Peterson and McCabe (1983), children merely tell a series of events that occur at roughly the same time instead of narrating an integrated experience. The chronological pattern of Chinese children at age 4 resembles this one. Yet children at age 5 tend to include more narrative components like evaluations and orientations in narratives, although they still lack clear high points or resolutions. Hence, we can see that the chronological pattern at age 5 was richer, more evolved, and more complete compared with that at age 4. The chronological pattern of Chinese children at age 6 is similar to that at age 5.

Gender Differences

In the current study, there were no gender differences in the frequency or proportion of any of the seven narrative components, a result that was also found in the narration of English-speaking North American children (Peterson and McCabe 1983). Previous studies on other English-speaking children mainly targeted gender differences in children’s emotion talk, finding that girls produced more emotion talk than boys (Dunn et al. 1987; Adams et al. 1995; Buckner and Fivush 1998; Peterson and Biggs 2001), while no gender differences were found on evaluation in this study. Chinese girls did not produce more evaluation than boys. Such a discrepancy is likely to reflect cultural differences between the parents and their sons and daughters in the way of communication. The study done by Zhou (2002) on pragmatic development of Mandarin-speaking children indicated Chinese mothers and children during their interaction seldom used Marking, Discussion of Hearer’s feelings and Discussion of Speaker’s feelings. As a result, when Chinese children tell their own stories, both boys and girls may only occasionally express feelings of themselves or of others. This study for the most part documented gender similarity, not gender difference in narrative components and patterns. Possible factors are the differences in methodology between studies: parent–child conversations (Adams et al. 1995; Buckner and Fivush 1998) versus semi-structured narrative talk with prompts. The content of narrative topics also varied: talk about ongoing or present situation (Dunn et al. 1987) versus talk about past events.

Cultural Characteristics

The results of this study show that the development of Chinese children’s narration presents both cultural universality and culture uniqueness.

Regarding cultural universality, it’s found that appendages (prologues, abstracts and attention-getters) were comparatively rare at all ages, similar to that of Hudson and Shapiro’s (1991) study of English-speaking children. Perhaps this is due to the fact that in interviews that were specifically designed to elicit personal narratives from them, children found no need to use prologues and attention-getters to arouse listeners’ attention. Codas, however, were more commonly used. Although only one out of twenty children at age 3 used codas, around eight children on average at age 4 to age 6 added codas to end their stories. As children grow older, they have more awareness of actively ending their stories and have better pragmatic abilities by taking listeners into consideration, clearly notifying them that their narrative is over. Cultural universality is also manifested in the most frequent components. Based on the results of the proportion of narrative components, the three most frequently used components of preschool Chinese children’s personal narratives were consistent from age 3 to 6: complicating action, orientation and evaluation, which amounted to more than 80% of all the narrative clauses. It is not surprising that these three components occur most frequently in narratives, since complicating action is the backbone informational component of narratives, which basically tells listeners a sequence of events. According to Labov and Waletzky (1967), orientation serves a referential function of narration, while evaluation fulfils the affective function—staving off a hypothetical listener’s query about why the narrator thought those events were important enough to tell. That complicating action is the most frequently used narrative component is in line with many previous studies of English-, Japanese-, and Chinese-speaking children (Chang and Chang 2000; Chang 2004; Minami 1996; Peterson and McCabe 1983; Xu 2007; Zhang 2012).

In terms of cultural uniqueness, Chinese children for the most part do not use resolution until they are 4 years old. Even at age 6, 70% of the participants still did not resolve their narratives. Thus, Chinese children may not be said to have mastered the usage of resolution during their preschool years, although the proportion of resolutions did increase significantly from age 3 (0) to 6 (2.05%). It is much lower than that of English-speaking American peers in Peterson and McCabe’s (1983) study (4.2–7.7%) or that of Mandarin-speaking Taiwanese peers in Xu’s (2007) study (0.89–8.56%). Possible reasons are two-fold. One is the kindergarten education in China. According to an interview the first author had with kindergarten teachers of the children, personal narration is seldom practiced in daily teaching. Teachers are more willing to engage in picture-book reading with children because it is easier to conduct. Also, teachers seldom model a good personal narrative. The other reason is the style of parenting at home. Previous studies show young children’s independent narrative skills originate in dialogic contexts between children and competent adults, in many cases their mothers (Fivush 1991; Fivush and Fromhoff 1988; McCabe and Peterson 1991; Reese and Fivush 1993). Yue et al. (1997) found out that Chinese parents were more devoted to teaching children knowledge and useful skills in order to prepare children for their future academic development than in encouraging children to talk about their personal experiences. Miller et al. (1997) observed that Chinese mothers had a tendency to convey social rules and moral standards when narrating with their children, thus contributing to an “overly self-critical interpretive framework.” It seems that Chinese mothers’ greater emphasis on information, knowledge, social rules and moral standards rather than on causal relationships (e.g. why-questions) or requests about what children were feeling contribute to fewer resolutions in Chinese children’s narratives.

Educational Implications

Chinese parents and teachers could be encouraged to pay more attention to improving their children’s ability to resolve a story, telling listeners how everything worked out in the end. Parents might prompt or scaffold children to offer resolutions and dialogue by asking relevant questions. Kindergarten teachers could also give preschool children specially adapted training in language-related activities and model more classic patterns so that children may learn to produce more developed patterns by themselves.

Limitations of the Study

Given the fact that all participating children of the study were recruited from a kindergarten affiliated with a prestigious university in Shanghai, a developed city in China and most mothers had graduated from a university, the findings of this study perhaps cannot be generalized to all Chinese children.

Hudson and Shapiro (1991) found that topic differences affect narrative structure. This study showed complicating action did not develop significantly from age 5 to 6, which differed from Li’s (2013) study. Li found preschool Chinese children achieved significant progress in producing complicating actions every 6 months from age 4 to age 6. This difference might be due to different data (cross-sectional versus longitudinal) and/or different topics of narration. Follow-up studies may want to have broader narrative topic design.

Cultural variation in the structure of narrative has been documented for many cultures (McCabe 2017). This study of Mandarin-speaking Mainland Chinese children found a developmental increase in details of their experiences (events, background information and emotions) with age, but many fewer resolutions, direct and reported speech in their narration compared to English-speaking children. Chronological narratives that tell a sequence of actions without a marked high point or resolution are most common in the narratives of Chinese children aged 4, 5, and 6 years, which may be attributed to parents’ style of talking. Therefore, future work could go further in studying the dialogic contexts between children and their main care-givers and help children establish sound narrative components and narrative patterns through appropriate conversation.

Standing back from this work, cultural differences in form and exact age of acquisition of narrative components warrant inquiry before generalizing findings from past studies to new cultural settings. Nonetheless this paper adds to the body of work that enables us with caution and careful attention to draw two overarching conclusions. First of all, High-point analysis may be usefully applied to narration in many cultures. Second, narration undergoes substantial development beginning at the age of 4 years.

Appendices

Appendix A: Narrative Patterns in High Point Analysis (Based on Peterson and McCabe 1983)

Classic Pattern: The narrator may begin with an abstract (“Did I ever tell you about when I cut my hand?”) or attention-getter (“You know what?”). The narrator then provides orientation information about who, what, when, and where some experiences occurred, followed by a series of complicating actions that build to a high point (You can’t believe the worst part.”), and then give additional series of events that resolve the problem (“I had to go to the hospital and had it bandaged.”).

Ending-at-the-High-Point Pattern: The narrator gives a chronological sequence of events, as in the classic pattern above, but simply builds to a high point and ends the narrative. There is no resolution in this pattern.

Chronological Pattern: The narrative is a simple description of successive events (e.g. “I ate yogurt. I put it on the table and it spilled. I cleaned it with napkin.”).

Leap-frog Pattern: The narrator gives more than two complicating actions that are components of a single experience, but does not give them in chronological order and/or omits key events. That is, the narrator jumps from one event to another and leaves events out.

Miscellaneous Pattern: The narrator gives more than two past events but without a logical or causal sequence to these events in the real world.

Two-event Pattern: The narrator only tells us two events that constituted an experience (e.g. A child shared his experience of seeing a doctor: “I put on trousers and then left.”).

One-event Pattern: The narrator only tells us a single event that constituted an experience (e.g. “One time I fell from the bike.”).

Appendix B: Chronological Pattern at Age 4 and Age 5

Chronological Pattern at Age 4

@Bg::

travel

*EXP::

那 你们 出去 玩 的 時候 [^c] 发生 了 什麼 事 [^c]?

nà nǐ men chūqù wán de shíhòu [^c] fāshēng le shí me shì [^c]?

What happened when you travel?

*CHI::

我们 出去 玩 也 到 动物园 [^c]

wǒmen chūqù wán yě dào dòngwù yuán [^c].

We also went to the zoo.

*CHI::

但是 发现 一 个 厉害 的 # 一 个 大 豹 [^c].

dàn shì fāxiàn yīgè lìhài de # yī gè dà bào [^c].

We found a big leopard.

*CHI::

卷起 肉 来 [^c]

juǎnqǐ ròu lái [^c].

They rolled meat.

*CHI::

给 他 吃 [^c].

gěi tā chī [^c].

And fed the leopard.

*EXP::

猎豹 啊 [^c]?

liè bào ā [^c]?

Cheetah?

*CHI::

金钱 豹 [^c].

jīn qián bào [^c].

Panthera pardus.

*EXP::

谁 给 他 肉 吃 的 呀 [^c]?

shuí gěi tā ròu chī de ya [^c]?

Who fed the leopard?

*CHI::

爸爸 [^c]. [+ R]

bà bà. [^c].

Dad.

*CHI::

找 一 根 棍子 [^c] 把 肉 卷 起來 [^c].

zhǎo yī gēn gùnzǐ.[^c] bǎ ròu juǎn qǐlái.[^c].

Dad found a stick and rolled the meat.

*CHI::

放到 豹子 跟前 [^c].

fàng dào bàozǐ gēnqián [^c].

Put it in front of the leopard.

*CHI::

豹子 吃 的 一 口 耶 [^c].

bàozǐ chī de yī kǒu yē [^c].

The leopard ate the meat.

*EXP::

嗯 然后 呢 [^c]?

èn ránhòu ne [^c]?

Then?

*CHI::

然后 爷爷 就 給 他 喂 吃 的 [^c].

rán hòu yéyé jiù gěi tā wèi chī de [^c].

Then grandpa fed the leopard.

*CHI::

结果 豹子 看 了 好 长 时间 [^c].

jiéguǒ bàozǐ kàn le hǎo cháng shíjiān [^c].

We spent a lot of time watching the leopard.

*CHI::

看 了 牛 [^c].

kàn le niú [^c].

We saw a cow.

*CHI::

看 了 马 [^c].

kàn le mǎ [^c].

We saw horses.

*CHI::

然后 就 坐 游览 车 [^c].

rán hòu jiù zuò yóulǎn chē [^c].

Then we took the cable car.

*CHI::

回 家 了 [^c].

huí jiā le [^c].

And went home.

*EXP::

然后 坐 游览 车 回 家 啦 [^c]?

ránhòu zuò yóulǎnchē huíjiā la [^c]?

Then you took the sightseeing bus and went back home?.

@Eg::

travel

Chronological Pattern at Age 5

@Bg::

birthday

*EXP::

你 告诉 我 [^c] 上个 月 你 过 生日 的 事情 吧 [^c].

nà nǐ gàosù wǒ [^c] shànggè yuè nǐ guòshēngrì de shìqíng ba [^c].

Tell me about your birthday.

*CHI::

嗯 # 上个 月 过 生日 [^c]. [+ RE]

èn # shàng gè yuè guò shēngrì

We celebrated my birthday last month.

*CHI::

就 是 我 生日 蛋糕 啊 [^c]. [+ N3]

jiùshì wǒ shēngrì dàngāo a [^c].

I had a birthday cake.

*EXP::

<对 啊 > [/] 对 啊 嗯 [^c].

<duì ā > [/] duì ā èn [^c].

Yes.

*CHI::

有 一 天 < 我 > [/] 我 [/] 我 晚上 睡觉 起来 [^c] 然后 奶奶 说 [^c].

yǒuyītiān < wǒ > [/] wǒ wǎnshàng shuìjiào qǐlái [^c]. ránhòu nǎinǎi shuō [^c].

One day when I woke up. Grandma said to me.

*CHI::

今天 我 过 生日 [^c].

jīn tiān wǒ guò shēngrì [^c].

Today was my birthday.

*CHI::

然后 就 到 大华 家 去 [^c].

ránhòu jiù dào dàhuá jiā qù [^c].

Then we went to Dahua.

*CHI::

然后 < 大华 家 准备 > [//] 大华 家 买 了 一个 蛋糕 [^c].

ránhòu < dàhuá jiā zhǔnbèi > [//] dàhuá jiā mǎi le yīgè dàngāo [^c].

And bought a birthday cake.

*CHI::

到 晚上 [^c].

dào wǎn shàng [^c].

Then evening came.

*CHI::

回到 爷爷 奶奶 家 [^c].

huídào yéyé nǎinǎi jiā [^c].

I went back to my grandparents’ home.

*CHI::

<奶奶 > [/] 奶奶 说 [^c]. < 快 打开 > [“] [^c].

<nǎinǎi > [/] nǎinǎi shuō [^c]. < kuài dǎkāi > [“] [^c].

Granma said. “Open it.”

*CHI::

<看看 究竟 是 什么样 的 蛋糕 > [“] [^c]?

<kànkàn jiūjìng shì shímeyàng de dàngāo > [“] [^c]?

“And look at the cake.”

*CHI::

然后 打开 [^c] 给 她 看 [^c].

rán hòu dǎkāi [^c]gěi tā kàn [^c].

Then I opened the box and showed it to grandma.

*CHI::

她 看 上面 还有 两 只 糖 [^c].

tā kàn shàngmiàn háiyǒu liǎng zhītáng [^c].

There were two candies on the cake.

*CHI::

像 两 只 公主 一样 的 糖 [^c].

xiàng liǎngzhī gōngzhǔ yīyàng de táng [^c].

They were in the shape of a princess.

*EXP::

哦 # 有 像 两 只 公主 一样 的 糖 啊 嗯 [^c].

ò # yǒu xiàng liǎng zhī gōngzhǔ yīyàng de táng ā èn [^c].

Two princess-like candies.

*CHI::

还有 一个 我 在 大华 家 吃掉 了 [^c].

háiyǒu yīgè wǒ zài dàhuá jiā chīdiào le [^c].

I ate one of them at Dahua.

*EXP::

<嗯 > [/] 嗯 [^c] 那 然后 呢 [^c]?

<èn > [/] èn [^c] nà ránhòu ne [^c]?

Yes. Then?

*CHI::

+, 然后 < 吃 好 蛋糕 点 > [//] 吃 好 饭 [^c].

ránhòu < chī hǎo dàngāo diǎn > [//] chī hǎo fàn [^c].

After we finished dinner.

*CHI::

然后 就 吃 蛋糕 了 [^c].

ránhòu jiù chī dàngāo le [^c].

We ate the cake.

*CHI::

给 灯 关上 [^c].

gěi dēng guānshàng [^c].

We turned off the light.

*CHI::

拿 点 蜡烛 [^c].

ná diǎn làzhú [^c].

We lit the candles.

*CHI::

然后 我 就 去 < 到 > [/] 到 厨房 间 去 [^c].

rán hòu wǒjiù qù < dào > [/] dào chúfáng jiān qù [^c].

Then I went to the kitchen.

*CHI::

叫 [^c] 爷爷 开 一点 火 [^c].

jiào [^c] yé yé kāi yīdiǎn huǒ [^c].

I asked grandpa to light candles.

*CHI::

然后 吱 这样 [^c].

rán hòu zhī zhè yàng [^c].

Then it like “zhi…”

*CHI::

然后 爷爷 关掉 火 [^c].

ránhòu yéyé guāndiào huǒ [^c].

Then grandpa turned off the stove.

*CHI::

我 就 插 在 蛋糕 上 [^c].

wǒ jiù chā zài dàngāo shàng [^c].

I put the candles on the cake.

*EXP::

嗯 # 把 蜡烛 插 在 蛋糕 上 [^c] 对 吧 嗯 [^c].

èn # bǎ làzhú chā zài dàngāo shàng [^c] duìba èn [^c].

You put candles on the cake, right?

*CHI::

然后 我们 切 了 [^c].

rán hòu wǒmen qiē le [^c].

Then we cut the cake.

*CHI::

然后 还有 说 [^c] < 生日 快乐 > [“] [^c].

ránhòu háiyǒu shuō [^c] < shēngrìkuàilè > [“] [^c].

Then they said “happy birthday”.

*CHI::

<然后 > [/] 然后 吃 好 了 [^c].

<rán hòu > [/] rán hòu chīhǎo le [^c].

Then it was over.

*CHI::

奶奶 说 [^c] < 不要 吃 了>” [^c].

nǎinǎi shuō [^c] < búyào chīle>” [^c].

Then grandma said “Don’t eat any more.”

*CHI::

<这个 糖 等 明天 再 吃 > [“] [^c].

<zhègè táng děng míngtiān zài chī > [“] [^c].

Leave the candy till tomorrow.

*CHI::

我 到 明天 吃掉 了 [^c].

wǒ dào míngtiān chīdiào le [^c].

I ate the candy the next day.

*EXP::

哦 # 然后 到 了 明天 [^c] 把 蛋糕 吃 了 [^c] 是 不是 [^c]?

ò # ránhòu dào le míngtiān [^c] bǎ dàngāo chī le [^c] shì búshì [^c]?

Oh you ate the cake the next day, right?

*CHI::

给 那个 糖 吃掉 了 [^c]. [+ RE]

gěi nàgè táng chīdiào le [^c].

I ate that candy.

*CHI::

吹 蜡烛 这 样子 吹 [^c]. [+ R]

chuī làzhú zhè yàng zǐ chuī[^c].

We blew the candles like this.

*CHI::

呼 [^c].

hū

“huhu”

*EXP::

嗯 # 好 [^c] 嗯 那 你 有 没有 许 一个 愿 啊 [^c]?

èn # hǎo [^c] èn nà nǐ yǒu méiyǒu xǔ yīgè yuàn ā [^c]?

Oh, and did you make a wish?

*CHI::

<爸爸 > [/] 爸爸 说 许 一个 愿 [^c]. [+ R]

<bàbà > [/] bàbà shuō xǔ yīgè yuàn [^c].

*CHI::

我 怕 难为情 [^c].

wǒ pà nánwéiqíng [^c].

I was too shy.

*CHI::

就 在 桌子 下面 了 [^c].

jiù zài zhuōzǐ xiàmiàn le [^c].

I was under the table.

*EXP::

跑 到 桌子 下面 [^c] 去 许愿 啦 [^c].

pǎo dào zhuōzǐ xiàmiàn [^c] qù xǔyuàn lā [^c].

You made a wish under the table?

*CHI::

不 对 [^c]. [+ N3]

bú duì

No.

*CHI::

我 怕 难为情 咧 [^c]. [+ RE]

wǒ pà nánwéiqíng liě [^c].

Because I was too shy.

*EXP::

哦 # 你 还 怕 难为情 啊 [^c] 哦 # 嗯 # 好 [^c] 嗯 # 好 哦 [^c].

ò # nǐ hái pà nánwéiqíng ā [^c] ò # èn # hǎo [^c] èn # hǎo ò [^c].

Oh you were too shy.

@Eg::

birthday