行政院國家科學委員會專題研究計畫 成果報告
漢語學齡前兒童敘事中評論語言的發展：跨敘事類型的研究
計畫類別： 個別型計畫 計畫編號： NSC94-2411-H-004-033- 執行期間： 94 年 08 月 01 日至 95 年 07 月 31 日 執行單位： 國立政治大學英國語文學系 計畫主持人： 薩文蕙 報告類型： 精簡報告 處理方式： 本計畫可公開查詢中 華 民 國 95 年 10 月 23 日
摘
要
評論是敘事中不可或缺的一環，評論語言透露敘事者對該事件所抱持的態 度、想法，並凸顯出事件主旨及其可資報導的理由。是以，兒童在發展其敘事能 力時，必須學會針對不同類型的敘事活動，適切地使用評論語言，以傳達其看法 及事件的重點。儘管近年來文獻上對兒童敘事能力的發展多有著墨，我們對兒童 敘事中評論語言的使用與發展仍所知有限。實則上，研究觀察指出兒童敘事中評 論語言的使用頗為頻繁。鑑於此，本研究旨在探討漢語學齡前兒童敘事中評論語 言的發展，以期對兒童敘事能力之進展有深一層的認識。 本研究觀察的對象為十二位學齡前兒童，其中六位男童、六位女童。研究者 在孩童平均年齡 5;5, 5;8 以及 5;11 時進行觀察採樣。我們以長期觀察的語料為 據，將評論語言加以分類，剖析不同類別的評論語言其各自的發展軌跡；此外， 隨孩童年齡的逐漸成長，我們試圖找出評論語言在整體敘事語言中所佔比例的消 長，以縱觀其整體發展趨勢。我們以年齡、性別與敘事活動類型為獨立變項，除 檢驗各獨立變項對評論語言發展的影響，並觀察各變項間的互動，以剖析評論語 言發展過程的各種樣貌。 研究結果顯示，隨兒童年齡成長，其評論語言使用的總量與多樣性均呈增加 的趨勢。在各類評論語彙中，僅兩種評論機制(CAS, SOM)呈現年齡主效應；僅 一項評論機制(PAR)顯現性別主效應。此外，兒童使用評論語彙時，會參酌敘事 活動之類型而做調整，此亦反映出他們已認知到不同類型的敘事活動間之差異。 一如我們所臆測的，年齡、性別與敘事活動類型等變因，會影響兒童對某些評論 語彙之倚重程度。 本研究之結果，讓我們對漢語學齡前兒童其評論語言的發展有進一步認識， 同時，亦為台灣的孩童在 Frog story 的敘事研究上，留下珍貴的漢語語料。ABSTRACT
Evaluative language plays an important part in the production of a narrative, which consists of the non-factual, perspective-building linguistic expression referring to emotion, attitude, belief and affect. It not only expresses the narrator’s attitude, provides clues to help the listener to interpret the story, but makes the story more engaging and vivid. Thus, while developing narrative competence, children need to acquire the ability to use evaluations appropriately within various narrative activities.
Despite widespread interest in emerging narrative competence, however, we lack information on how young children develop evaluative language skills. As a matter of fact, children’s narratives bristle with evaluation. Therefore, the present work tries to explore how Mandarin-speaking children develop their evaluative language skills in narration.
Twelve Mandarin-speaking preschoolers, six boys and six girls, participate in the present study. The narrative data are collected when the subjects are of mean ages 5;5, 5;8, and 5;11. We divide evaluative language into several sub-types and analyze the developmental trajectory of each sub-type. Three independent factors are considered in this work: age, gender, and narrative task. Their main effects and interactional effects on the development of evaluative language are assessed.
The findings suggest that preschoolers employ increasing amount and variety of evaluative devices over time. Among various evaluative devices, merely two evaluative devices, CAS and SOM, display Age main effect; only one device shows Gender main effect. Regarding Task effect, we note that preschoolers use evaluative devices differently in different narrative tasks, which implies that they are aware of task difference in their early life. As hypothesized, the preference for each evaluative device changes over time, and varies between genders and between narrative tasks.
The outcome of our work not only advances our understanding of the development of evaluative language in preschoolers, but contributes to the sample pool of the frog story some valuable narrative data from Mandarin-speaking children in Taiwan.
1. Introduction
Earlier findings suggest that language development does not finish with the development of sentence structure (Tager-Flusberg and Sullivan, 1995). Thus, the study of children’s language development needs to take into account their emerging ability in narration. According to Labov, narratives encompass two important functions: reference and evaluation. The former helps to orient the hearer to whom the narrative is about and where and when the action takes place. The latter reveals the narrator’s attitude and lets the hearer know why the narrative is told and what its point is. Despite widespread interest in emerging narrative competence, however, we lack information on how young children use evaluative language skills in narration. As a matter of fact, children’s narratives bristle with evaluation. Therefore, the present work tries to explore how Mandarin-speaking children develop their evaluative language skills in narration.
2. Literature Review
2.1 Evaluative Language and Narrative Development
Labov (1972) values the role that evaluation plays in a narrative. His work has a profound impact on the studies of evaluation in narrative development. To begin with, Kernan (1977) and Umiker-Sebeok (1979) observe that children provide more and more contextual and extra-narrative elaboration and evaluation with increasing age. In particular, Kernan notices that, with increasing age, children realize that the understanding and interpretation of narratives is assured through the use of contextual and extra-narrative elaboration rather than the narrative events.
Peterson and McCabe (1983) investigate personal narrative from ninety-six children, aged 3;6 to 9;6. They find that nearly half of children’s narrative comments express some evaluation via various linguistic devices and that there are significant differences in the popularity of different evaluative devices. They also note that the variety of evaluation increases over age, though the overall incidence of evaluation fails to exhibit age effect. The most striking finding in their assessment is the lack of significant gender differences.
In another research, Miller and Sperry (1988) investigate how young children’s early talk evolves into personal experience narrative. According to them, the 2-year-olds’ narratives carry exceedingly more referential expression than evaluative language; however, the children increase their use of evaluation over time. One crucial finding of this study is the extent to which young children, at the youngest ages, can convey their attitude toward the past event.
In line with previous studies, Bamberg and Damrad-Frye (1991) explore children’s ability to provide evaluative comments. They detect that 5-year-old
children begin to provide evaluative comments in their narratives and that references to ‘frames of mind’ are preferred by children of older age.1 They also note that the variety of evaluative devices used by children do change with age.
In a more recent study, Ukrainetz and her colleagues (2005) analyze the development of expressive elaboration in children’s narratives. They find that evaluation is the most frequent type of expressive elaboration and that it shows significant age changes.
2.2 Narrative Task and Evaluative Language
Narrative is not just one type of discourse, but a family of narrative genres, ranging from ongoing descriptions of everyday activities to lengthy fictional accounts. Researchers suggest that different narrative genres draw upon different content knowledge, different linguistic structural knowledge and carry various functions (Hudson and Shapiro, 1991; Hudson and Nelson, 1986; Wolf, Moreton and Camp, 1994), which underlines the importance of genres in the study of narrative development.
Wolf, Moreton and Camp (1994) relate narrative genre to the use of evaluative language in narrative discourse. To uncover possible genre influence on evaluative language, Shiro (2003) examine Venezuelan children’s developing abilities to use evaluative language in different narrative tasks based on different genres. She (2003) finds that age and socio-economic status have a great influence on the use of evaluation in fictional stories than in personal narratives.
Difference between personal and fictional narrative genres regarding the use of evaluative language is further documented by Bamberg (1997). He suggests that the preference for evaluative language tends to be genre-specific since different genres have fundamental differences in perspective building, particularly regarding self representation in the narrated world. When narrating a fictional story, for instance, children need to adopt not only a narrator’s stance but the character’s perspective in the story world. In personal narratives, however, a narrator simply adopts a narrator’s stance. Accordingly, telling fictional narratives tends to be more cognitively demanding for young children than telling personal stories.
The above-mentioned findings lead us to speculate that children’s narrative competence cannot be assessed in a single narrative task, given the importance that different narrative tasks may affect children’s narrative performance differently. In light of these findings, the present project aims to explore task effects on preschoolers’ use of evaluative devices by using two narrative tasks based on two
1
Bamberg and Damrad-Frye (1991) defines ‘frame of mind’ as the description referring to emotional states, mental states or mental activities.
narrative genres: personal experience narrative and fictional narrative.
Researchers on Mandarin-speaking children’s language development have examined various aspects of narrative development (Cai, 1996; Chang, 2000; Hsu, 1992; Huang, 2003; Wang, 1998). Yet, few studies have explored the development of evaluative language in Mandarin-speaking children’s narratives. Among them, Chang (2000) employ longitudinal data to examine children’s use of evaluative language. She finds that her subjects include a larger number and variety of evaluation in their personal experience narratives with increasing age.
Unlike Chang’s (2000) longitudinal approach, Chang (2001) draws on cross-sectional data from personal narratives to explore Mandarin-speaking children’s development of evaluative language. In agreement with Chang’s (2000) findings, Chang’s (2001) results indicate that children use increasing amount and variety of evaluative devices over time.
Another research using cross-sectional data from Mandarin-speaking children is by Huang (2002). In her fictional narrative data, she notes that older children provide more background information and explain more about story characters’ internal motivations. Hence, they use a wider variety of evaluative devices than their younger counterparts.
In sum, these researchers find that Mandarin-speaking children use a larger variety of evaluative devices with increasing age, which confirms the findings from children of other language background (Chang, 2000; Chang, 2001; Huang, 2002). Although these researchers have given attention to the development of evaluation in preschoolers’ narratives, they collect data from narrative tasks based on one narrative genre: personal anecdote or fictional story. As Schneider (1996) points out, task-related factors may affect children’s narrative production. Thus, the present work tries to explore the interaction between narrative-task demands and children’s use of evaluative language skills. We aim to observe how preschoolers employ evaluative devices to convey their internal experience and world view in different narrative tasks based on different narrative genres.
3. Method
3.1 Research Question
While developing narrative competence, children need to acquire the ability to use evaluative language appropriately within various narrative activities. By answering the following research questions, the present study attempts to explore the development of Mandarin- speaking preschoolers’ evaluative language.
1. Do preschoolers use more evaluative devices with increasing age? 2. Do preschoolers employ a larger variety of evaluative devices over time?
3. Do different narrative tasks interact with preschoolers’ use of evaluative devices?
4. How do Age and Gender effects manifest themselves in the development of evaluative language in preschoolers’ narratives?
3.2 Hypothesis
Based on the afore-reviewed literature and the preliminary findings of my pilot study, the general hypotheses are put forward.
1. Preschoolers include more evaluative devices and apply a greater variety of such devices in their narratives with increasing age.
2. Preschoolers use evaluative devices differently in different narrative tasks. 3. There are differences in preschoolers’ use of each evaluative device over
time and between genders.
3.3 Subject
Twelve Mandarin-speaking preschoolers, six boys and six girls, chosen from nursery schools, participate in the present study. All subjects are normally developing children from similar middle-class socio-economic backgrounds. The mean age of the children is 5;5 months at the first session of data collection, and 5;11 months at the last session.
3.4 Material
This project includes two narrative tasks. One of them involves fictional narratives. In order to control the content of the fictional narratives, we use a twenty-four-page story book entitled Frog, where are you (Mayer, 1969) as the material to elicit fictional narratives from subjects. In the other task, the experimenter elicits personal experience narratives related to physical harms or pains.
3.5 Data Collection
The narrative data are collected at three time points when the subjects are of mean ages 5;5 (Time 1), 5;8 (Time 2), and 5;11 (Time 3). Order of the narrative tasks children asked to perform is randomized. The entire interviews are audio-taped and subsequently transcribed.
3.6 Coding System
Adapted from previous studies of evaluation (Bamberg and Damrad-Frye, 1991; Chang, 2000; Chang, 2001; Lin, 1993; Miller and Sperry, 1988; Peterson and McCabe,
1983; Wang, 1998) and developed from the collected data, seventeen types of evaluative devices are classified in the present project: character speech (CAS),
comparison (COM), conditional connective (CON), compulsion words (CPW), causal explanation (CSE), evaluative word per se (EVA), evaluative connective (EVC), exaggeration (EXA), expression of intentions (EXI), explicit negation (EXN), frames of mind (FOM), hedges (HEG), intensifier (INT), repetition (REP), sound modification (SOM), and verbal qualifier (VRQ).
3.7 Data Analysis
The transcriptions follow the guidelines of Codes for the Human Analysis of Transcripts (CHAT) of the Child Language Data Exchange System (CHILDES). Cohen’s kappa statistic is used to assess inter-rater reliability. Inter-rater agreement result is 91%.
A number of Child Language Analysis (CLAN) programs are used to analyze the data. Then, the three-way ANOVA (an analysis of variance) and LSD post hoc comparisons are employed to assess differences between variables.
4. Results and Discussion 4.1 Quantity and Variety
The first two research questions concern about the variety and overall incidence of evaluative devices. Results from ANOVA for number of evaluative devices exhibit Age main effect (F = 14.520, p < 0.01), but no main effect for either Gender or Task. There is only one interaction (Age x Gender) reaching significant level (F = 4.545, p < 0.05). The significant Age effect reflects that preschoolers use more and more evaluation in their narratives over time, which is consistent with the findings for Mandarin-speaking children from Chang’s (2000) and Huang’s (2002) studies. Post hoc comparisons further reveal that male preschoolers steadily increase their use of evaluative devices throughout the time span, whereas female preschoolers display the growth in the first phase (Time 1 to Time 2), but exhibit a leveling-off of growth in the second phase (Time 2 to Time 3).
With respect to the variety of evaluative devices, ANOVA yields Age main effect (F = 14.757, p < 0.01) and Task main effect (F = 17.905, p < 0.01), but no main effect for Gender. The only significant interaction is between Gender and Task (F = 5.526, p < 0.05). Post hoc comparisons indicate that preschoolers use a larger variety of evaluative devices in the fictional narrative task than in the personal one, which suggests that the same child may use evaluative devices differently in different narrative tasks.
The Age effect indicates that there is an age-related increase in variety of evaluative devices, which is in consonance with the findings of previous studies (Chang, 2000; Chang, 2001; Peterson and McCabe, 1983). Regarding the interaction between Gender and Task, we note that, in the fictional narrative task, female preschoolers use a greater variety of said devices than their male counterparts.
To sum up, preschoolers employ increasing amount and variety of evaluative devices over time, which implies that children have better linguistic abilities with increasing age. The Task effect and Gender x Task interaction reflect that preschoolers seem to know how to respond to different narrative task demands. Consequently, when the variety of evaluation is discussed, possible task influence should be taken into account.
4.2 The Development of Each Evaluative Device
The third and fourth research issues concern about the impact from Age, Gender, and Task on the use of each evaluative device. The results from three-way ANOVA show that seven out of the seventeen devices display Task main effect (i.e., CAS, COM, FOM, INT, REP, SOM, and VRQ). Among them, CAS, REP, SOM, and VRQ are more pronounced in the fictional narrative task, whereas the other three are more pronounced in the personal narrative task. The outcomes suggest that different narrative tasks may have different impacts on children’s use of evaluative devices, which also implies that the same child may use different evaluative skills in different narrative tasks.
The major difference between the two narrative tasks in our project is that they base on different narrative genres. Though it is not proper for us to claim that the Task effects are exclusively resulted from genre difference, we may indirectly infer that genre differences may possibly influence the way preschoolers use evaluative devices. Given the fact that differences exist between fictional and personal narratives with respect to their underlying assumptions, communicative purposes, perspectives and topics (Shiro, 2003; Allen et al., 1994), we suggest that different linguistic knowledge is probably drawn upon in narrative tasks of different genres, and that there are differences in the ways children develop task-specific narrative skills. The task effects also imply that preschoolers are aware of different task demands. One implication obtained here, thus, is that children’s narrative competence cannot be assessed in a single narrative task.
Regarding Age main effect, merely two devices, CAS and SOM, exhibit such effects. What’s more, five other devices display an increasing tendency, though their effects fail to reach significance (i.e., COM, EVA, EXI, FOM, INT). For the rest of the devices, their occurrence rates are still fluctuating. Taken together, our results
seem to imply that preschoolers from age 5;6 to age 6 are still exploring and doing experiment about the evaluative language skills they acquired.
Among the 17 evaluative devices, PAR is the only device which demonstrates Gender main effect. We note that female preschoolers use exceedingly more PAR than their male counterparts. This finding is congruent with that from Peterson and McCabe (1983), which shows that girls are more likely to interject exclamations and verbal stress. The overwhelming lack of significant Gender main effect in our work confirms the findings from Peterson and McCabe (1983) and Chang (2001).
5. Conclusion and Suggestions for Future Research
The present study aims to assess the developmental trajectory of the evaluative language skills in preschoolers’ narratives. The main findings are summarized as follows.
The first two research questions address the developmental pattern for quantity and variety of evaluative devices. Our results show that, in general, preschoolers employ increasing amount and variety of evaluative devices over time. Male preschoolers steadily increase their use of evaluative devices throughout the time span, whereas female preschoolers display growth in the first phase but exhibit a leveling-off of growth in the second phase. The interaction between Task and Gender shows that female preschoolers use a greater variety of evaluative devices than their male counterparts in the fictional narrative task.
The third and fourth research questions focus on the influence from Age, Gender, and Task on preschoolers’ use of each evaluative device. Merely two evaluative devices, CAS and SOM, display Age main effect. The task-specific preferences suggest that evaluative language skills are employed differently in the personal and fictional tasks and that preschoolers are aware of task differences. The impact from Gender is less pronounced than that from Task or from Age on the use of evaluative language skills, for only one device PAR exhibits Gender main effect.
Great care has been taken to minimize the potential flaws in the present work; there remain, nevertheless, several limitations. To begin with, our sample size is far too small, and hence we yield only limited amount of information regarding the research topics. The second limitation is that our preschoolers are selected from a middle-class community. Actually, children from different socioeconomic conditions may experience different set of social interactions and related narrative genre practice (Michaels, 1981). Third, the time span for observation in the present work is not long enough. Though we note growth in certain aspects at certain time points, it could be merely the peak of a minor tremor, or the seemingly growth might turn out to be just a minor fluctuation in the extended developmental progression.
To better understand children’s evaluative language skills and to present significant growth in the development of evaluative language, future research should include a larger number of subjects of varied socioeconomic backgrounds and follow these developmental paths over a longer period of time.
6. Bibliography
Allen, M., M. Kertoy, J. Sherblom, and J. Pettit. 1994. Children’s narrative productions: a comparison of personal event and fictional stories. Applied
Psycholinguistics 15: 149-76.
Bamberg, M. 1987. The Acquisition of Narrative: Learning to Use Language. Berlin: Mouton de Gruyter.
Bamberg, M., and R. Damrad-Frye. 1991. On the ability to provide evaluative comments: further explorations of children’s narrative competences. Journal of
Child Language 18: 689-710.
Cai, Q. F. 1996. Story-telling: A Study of Children’s Use of Referring Terms. M.A. thesis, Graduate School of Western Languages and Literature, Providence University.
Chafe, W. 1994. Discourse, Consciousness and Time. Chicago: University of Chicago Press.
Chang, C. J. 2000. Narrative Performance across Contexts and over Time:
Preschool Chinese Children and Mothers. Unpublished doctoral dissertation,
Graduate School of Education, Harvard University.
Chang, J. W. 2001. A Developmental Study of Narrative Structure and Evaluative
Devices. M.A. thesis, Graduate School of English Language, Literature and
Linguistics, Providence University.
Hsu, J. 1992. Story-telling: A study of the development of communicative competence. Jing Yi Ren Wen Xue Bao 4: 1-17.
Huang, J. R. 2002. A Developmental Study on Children’s Evaluative Comments in
Narration. M. A. thesis, Graduate School of Linguistics, Fu Jen Catholic
University.
Huang, C. C. 2003. Talking about past events in conversation: An analysis of Mandarin mother-child and adult-adult discourse. Taiwan Journal of
Linguistics 1(1): 123-156.
Hudson, J., and K. Nelson. 1986. Repeated encounters of a similar kind: Effects of familiarity on children’s autobiographical memory. Cognitive Development 1: 232-271.
Hudson, J., and L. Shapiro. 1991. From knowing to telling: the development of children’s scripts, stories and personal narrative. In A. McCabe and C. Peterson
(eds.), Developing Narrative Structure. Hillsdale, NJ: Erlbaum.
Kernan, K. T. 1977. Semantic and expressive elaboration in children’s narratives. In S. Ervin-Tripp and C. Mitchell-Kernan (eds.), Child Discourse, 91-102. New York: Academic Press.
Klecan-Aker, J. S., G. K. McIngvale, and P. R. Swank. 1987. Stimulus
considerations in narrative analysis of normal third-grade children. Language
and Speech 30: 13-24.
Labov, W. 1972. The transformation of experience in narrative syntax.
Language in the Inner City, 354-396. Philadelphia: University of Pennsylvania
Press.
Labov, W., and J. Waletzky. 1967. Narrative analysis: Oral versions of personal experience. In J. Helm (ed.), Essays on the Verbal and Visual Arts, 12-44. Seattle: University of Washington Press.
Lin, A. 1993. Intensity Devices in Spoken and Written Chinese Narrative. M.A. thesis, National Taiwan Normal University.
Maccoby, E., and C. Jacklin. 1974. The Psychology of Sex Differences. Stanford Univ. Press.
Mayer, M. 1969. Frog, where are you? New York: Dial Press.
Michaels, S. 1981. “Sharing rime”: Children’s narrative styles and differential access to literacy. Language in Society 10: 423-442.
Miller, P, and L. Sperry. 1988. Early talk about the past: the origins of
conversational stories of personal experience. Journal of Child Language 15: 293-315.
Peterson, C., and A. McCabe. 1983. Developmental Psycholinguistics: Three Ways
of Looking at a Child’s Narrative. New York: Plenum.
Schneider, P. 1996. Effects of pictures versus orally presented stories on story retelling by children with language impairments. American Journal of
Speech-Language Pathology 5: 86-95.
Shiro, M. 2003. Genre and evaluation in narrative development. Journal of
Child Language 30: 165-195.
Tager-Flusberg, H., and K. Sullivan. 1995. Attributing mental states to story characters: a comparison of narratives produced by autistic and mentally retarded individuals. Applied Psycholinguistics 16: 241-256.
Ukrainetz, T., L. Justice, J. Kaderavek, S. Eisenberg, R. Gillam, and H. Harm. 2005. The development of expressive elaboration in fictional narratives. Journal of
Speech, Language, and Hearing Research 48: 1363-1377.
Umiker-Sebeok, D. 1979. Preschool children’s intraconversational narratives.
Wang, X. P. 1998. A Developmental Study of Acquisition of Narrative Skills by
Children in Taiwan. M.A. thesis, Graduate Institute of Linguistics, Fu Jen
Catholic University.
Wolf, D., J. Moreton, and L. Camp. 1994. Children’s acquisition of different kinds of narrative discourse: genre & lines of talk. In J. Sokolov and C. Snow (eds.),
Handbook of Research in Language Development Using CHILDES, 286-319.
Hillsdale, NJ: Erlbaum.
Wooldridge, P., L. Nall, L. Hughes, T. Rauch, G. Stewart, and C. L. Richman. 1982. Prose recall in first-grade children using imagery, pictures, and questions.
Bulletin of the Psychonomic Society 20: 249-252.
7. Self-evaluation of the Project (計畫成果自評)
The present work examines the development of Mandarin-speaking preschoolers’ evaluative language skills in narratives. As hypothesized, preschoolers use a larger number and variety of evaluative devices over time. In addition to the influence from Age, the impacts from Gender and Task on the use of evaluative language skills are also shown.
We try to assess Mandarin-speaking preschoolers’ use of evaluative language skills via two different narrative tasks on different narrative genres. The results suggest that the same child may use different evaluative language skills in response to demands from different narrative tasks, which implies that these preschoolers are aware of task difference and try to employ task-specific narrative skills. Thus, one implication from this study is that, among the various factors that may have impacts on narrative performance, different narrative-task demands should also be taken into account, when narrative development is considered,
In addition, the knowledge obtained here may enhance educators’ and parents’ understanding of children’s narrative competence. Given the possibility that different narrative tasks do not have equivalent task demands, it is necessary to compare the results from different tasks, and to analyze whether they provide different kinds of language information. Therefore, care should be taken to provide narrative-elicitation situations that are likely to generate narrative of distinct nature when teaching and assessing children’s narrative abilities.
This study has another significance of providing narrative data based on the frog story, the worldwide research tool. In the pool of samples from nearly 50 languages, our work contributes valuable data from Mandarin-speaking children in Taiwan. Hence, findings based on this study will be written out and submitted as journal paper.
Although our tasks draw upon different narrative genres, it is not feasible for us to assume that the obtained task effects originate from genre difference solely.
Actually, there exist other possible confounding variables between the personal narrative and the fictional narrative tasks, which render interpretation of the task effect complicated.
To begin with, we observe that fictional stories are far more interesting than personal experiences to these preschoolers, which probably results in the richer elaboration in fictional narratives. According to Schneider (1996), different narrative tasks may generate different estimates of children’s narrative abilities. In the personal narrative task, our preschoolers are asked to describe personal experiences. They have only the structured but non-elaborated anecdotes as story models. In the fictional narrative task, however, children are provided with visual aids from the picture book, which renders the task less abstract than the personal task. Moreover, the pictures from the story book may possibly enhance evaluative expressions produced, which in turn may lead to the fictional task advantage. Earlier studies suggest that stimuli used to elicit narratives may have impacts on children’s narrative productions (Klecan-Aker at al., 1987; Wooldridge et al., 1982). Thus, to provide sound explanations for the task effect, we need more investigations to tell whether pictures actually make narrative performance easier or harder.
One more issue on stimuli regards the prompts that are used. To make the comparisons between research findings feasible, we follow Chang’s (2001) data collection procedure, in which the open prompting questions are first used to elicit personal narratives. If the open prompts fail to elicit data, the structured anecdotes are then used. Actually, one possible function of structured prompts is to provide story models for young children. As a result, Chang’s procedure might lead to imbalance in linguistic stimulation for subjects, for some subjects could be inspired by the structured prompts, while others have no access to such story models. To avoid this inequality in prompts, future research designs should consider differential effects from prompts with or without narrative models.
Another factor relevant to the task effect is the knowledge shared by the narrator and listener. The extent of shared mutual knowledge assumed by the narrator may have differential effects on the information he or she provides (Schneider, 1996). In other words, when the listener is assumed to share mutual knowledge, the narrator will probably provide less information. On the other hand, when the listener is not supposed to have mutual knowledge, the narrator is obliged to provide a more complete story. Given this, we need to further clarify the difference in mutual knowledge status of the listener embedded in our two narrative tasks.
Given the importance that task-related factors seem to have on narrative production, before generalizing our findings in task effects, we need to verify these findings by experiments of better design, from which more credence will be gained.
