Handbook of Language and Literacy Development - a Roadmap from 0 to 60 Months

children image


Narrative Development (37-48 Months)click to print Print
Research Review / Parent

Written by: Carole Peterson, Memorial University of Newfoundland

Introduction to Narrative Development

During the year that children are 3 years of age, they already have some of the important foundation skills that are needed for narrative development, such as a cognitive sense of self that is becoming more sophisticated than it was when they were first developing it as 1 year olds. As 3 year olds, children's sense of self is beginning to become more psychological as they try to understand their own and other people's motivations and knowledge states. They are learning more about their world and the sorts of predictable relationships there are between events. They have acquired the ability to readily talk about the then-and-there. That is, their language is freed from the constraints of the current environment and they now talk about what happened yesterday, last month, and long ago as well as what will happen in the future. Their language is more complex and they are continuing to learn how to linguistically encode objects, actions, events, and relationships among them. But they still have far to go in terms of learning how to structure comprehensible narratives. Parental interactions as well as knowledge of cultural values are important influences in helping children acquire narrative skill.

Child Developmental Abilities That Support Narrative Development

Children's narrative development depends upon continuing development of their memory, theory of mind, thought structures, and language.


As stated before ('Narrative Development in 24-35 Month Olds'), autobiographical memory is memory for specific episodes in one's past that have personal significance or meaning (Nelson & Fivush, 2004). Although there is debate over whether 2 year-olds have autobiographical memory because these memories may not yet be well enough embedded in personal meaningfulness (see Nelson & Fivush, 2004), some memories of 3 year- olds can clearly be called autobiographical. Furthermore, the events in their lives are readily verbally accessible and can often be remembered for many years. For example, researchers who have investigated children's memory for highly salient events (such as injuries, hurricanes, trips to Disney World, etc.) that occurred when children were 3 years of age find that the children have extensive recall of those events many years later (see reviews in Bauer, 2006, and Peterson, 2002). Some of these memories can even be recalled when the child has become an adult. Researchers who study the earliest memories of adults (a phenomenon known as childhood amnesia) find that adults' first memories typically date from the time they had been 3 years of age (Bauer, 2006). Clearly, being able to remember events for long periods of time as well as being able to readily represent those memories linguistically are important tools for narrative development.

Theory of Mind

Theory of mind is the understanding of people as mental beings who have thoughts, feelings, desires, and intentions. By the end of their first year of life, infants understand that people are agents who have intentions, and 2 year-olds begin to talk about the feelings that are associated with various events. However, children who are 3 years of age still have a naïve theory of mind (see reviews in Astington and Dack, 2007, and Flavell, 2004). If, for example, they look into a box of crayons and see that there are candies in there, they will assert that a new person entering the room (who has had no opportunity to see what is in the crayon box) will think that there are candies, not crayons, inside. That is, they do not understand that someone can believe something that is different from what they know or believe. Three-year-olds still have a primitive grasp of their own mental processes, and are poor at understanding other people's thoughts. In short, they are poor at 'mind reading', i.e., detecting their own and others' perceptions, feelings, desires, and beliefs.

Nevertheless, children do make some strides in their understanding of minds during the year that they are 3 years of age. They are beginning to understand that people have thought processes and by this time they begin to use a variety of mental verbs, including "think," "remember," and "pretend" (Wellman, 1992). An understanding of people as mental beings is an important foundation for narratives because the episodes of our lives that we talk about are commonly created or influenced by these inner thought processes, and children begin to acknowledge this in their narrative language as 3 year olds.

Thought structures

As stated previously (see 'Narrative Development in 24-35 Month Olds'), two important cognitive developments that foster narrative acquisition are (1) improvements in reasoning about causal and temporal relations, and (2) developing schemas of how narratives are organized.

Reasoning About Causal and Temporal Relations

Understanding the world requires us to recognize the relations between events. So important is this understanding that Hume (1748/1955, cited in van den Broek, 1997. p. 324) calls it "the cement of the universe." As reviewed earlier (see 'Narrative Development in 12- 23 and 24-35 Month Olds'), even infants are able to observe simple temporal and causal relations between events (see review in Bauer, 2004). However, children's understanding of these relations continues to improve with age. In terms of temporality, 3 year-olds must learn about both absolute and relative time (last Monday vs. yesterday), and they must come to understand a host of markers that indicate the temporal order of events (e.g., before, after, since, when, then).

In terms of causality, children have an easier time understanding and talking about psychological causality (the relationship between intention or emotion and behaviour, e.g., "I wanted one too so I grabbed one," "I cried because I was scared") than physical causality (the causal relationship between objects in the world or objective conditions, e.g., "the bumper got bent because the other car hit it," "it snowed because it was cold") (McCabe & Peterson, 1997; Hood & Bloom, 1979). But there are many types of relations between events, including conditional ones, enabling ones, contingency relationships, and so on (van den Broek, 1997). Events in the world are seldom isolated but rather are inter-connected in a host of ways, and it takes many years for children to understand, much less linguistically encode, all of these complex relationships.

Schemas and Scripts

Through experience with repeating or predictable events in the world as well as through listening to storybooks, children develop schemas (also called scripts) about how events are organized (Nelson, 1986). That is, they develop mental frameworks that can be used to structure narratives (Naremore, 1997). As stated previously (see 'Narrative Development in 24-35 Month Olds'), schemas consist of an organized series of events, and through schemas, children come to understand that there is a macro-organization to an event series: there are beginning events, there are middle events, and there are ending events. As children begin telling personal experience narratives, they must come to understand that these too are composed of a series of events with an overall macro-organization or schema, with beginnings (e.g., orientation to the context of the events such as where and when they took place), middles (such as sequences of events describing what happened, problem complications, actions to achieve goals, etc.), and ends (resolutions, goal achievement or failure, etc.) (Labov & Waletzky, 1967/1997; Nelson & Fivush, 2004; Peterson & McCabe, 1983).

Narratives that do not provide an orienting context to where and when events took place are difficult to understand (Snow & Imbens-Bailey, 1997), and 3 year-olds are continuing to learn the necessity of providing such orientation. In a longitudinal study of middle-class Caucasian 2 - 3½ year olds, Peterson and McCabe (1994, 1996) found that children 2 - 2½ years old spontaneously included an orientation to when and where events took place in 7% and 29% of their narratives, respectively. At 2½ - 3 years of age, this rose to 19% and 40% for when and where. And at 3 - 3½ years old, they spontaneously provided when and where 32% and 42% of the time. However, in those narratives about events that took place away from the child's home (the children's location when talking), 3 year-olds provided an orientation to where 54% of the time (Peterson, 1990). Thus, children are steadily improving the comprehensibility of their narratives by providing an orientation to the context of the events being talked about.

However, 3 year-olds have much more difficulty figuring out and incorporating other aspects of narrative schema. Comprehensible narratives describe the series of events that make up the middle section of a narrative in correct temporal order (Labov & Waletzky, 1967/1997; Peterson & McCabe, 1983), but 3 year-olds often have difficulty doing this. In fact, the independent narratives of 3 year- olds (i.e., those in which an adult is not prompting) typically conform to a leapfrog structure (Peterson & McCabe, 1983). That is, children hop around from event to event in relatively random order when narrating about past occurrences. Nor do they typically provide an ending to their narratives, but rather leave the story events unresolved (Peterson & McCabe, 1983). Both of these features make it more difficult for adults to understand children's narratives, and such structural short-comings will also be problematic for teachers in school settings. Thus, children still have a long way to go in figuring out and successfully utilizing appropriate narrative schema.


Vocabulary and grammar. Children's vocabulary continues to increase substantially, adding many words per week (see section on Vocal Skills). This of course allows them to encode a lot more objects and relations with language. The grammatical complexity of their sentences is also increasing substantially (Tomasello, 2006), allowing them to communicate more complex representations. Having the appropriate words as well as the grammar (such as complex tense construction, relative clauses, etc.) in which to verbally encode events and relations between events help children not only talk about those events but also to remember them (Haden, Ornstein, Eckerman, & Didow, 2001; Nelson & Fivush, 2004).

Linguistically encoding causal and temporal relationships. We have seen that children are acquiring a better understanding of causal and temporal relations between events but learning to linguistically encode them is a new challenge, one that 3 year-olds work hard at. Two-and three-year-olds are learning to use 'because' and 'so' to mark causal relations and these linguistic markers are mostly used accurately at age 3 (Hood & Bloom, 1979), although mostly indicating psychological rather than physical causality. Children are also learning to encode both relative and specific temporal locations (e.g., yesterday, last Sunday), as well as learning how to use markers that indicate temporal order or relative time between events (e.g., then, before, after, while). They also begin to talk about past and future events relative to a reference time other than now ("I played with playdoh when I was at grandma's") (Tomasello, 2006).

Coherence of noun identification. An important key to understanding someone's narratives is to know who and what the speaker is referring to. For example, consider the following by a 3 year-old: "I told him to don't put it. I told her what he was doing. And when I was doing it I turned and pushed him." This narrative is incomprehensible because the pronouns are unspecified: him, he, her, and it are all unidentified. Worse, it is not clear whether each it refers to the same thing. Three-year-olds understand and correctly use pronouns in their here-and-now talk, but they often have difficulty using them appropriately in extended narratives (Menig-Peterson, 1975; Peterson, 1993; Pratt & MacKenzie-Keating, 1985). For example, in a longitudinal study of middle-class children, fully a fifth of new referents introduced into their narratives by 3 year-olds were ambiguous pronouns (Peterson & Dodsworth, 1991). Such confusion in identifying referents seriously compromises the coherence of children's narratives (Bennett-Kastor, 1983; Shapiro & Hudson, 1997).

Narrative Production

Narratives not only report what happened in the past, but they also place the events within space and time. They also place the events "in a coherent framework that explains how and why events happened as they did, and what these events mean for the self" (Fivush, Haden, & Reese, 2006, p. 1570). The narratives of 3 year- olds, like those of 2 year-olds, are seldom independent tales of personal experience. Instead, they are mostly co-constructions that take place within an interactive context with parents and others (Eisenberg, 1985; Miller & Sperry, 1988; Nelson & Fivush, 2004). However, an increasing proportion of these narratives are initiated by children themselves, and during the co-narrations that ensue, children are contributing increasing amounts of new information. They are also responding more appropriately to the questions that their conversational partners ask and they are increasingly elaborating on their own and their partners' contributions. These narrative skills are strongly influenced by parents, culture, and even gender.

Parental Influences On Narrative Development

Amount of Talk About Past Events

Parents typically direct a lot of talk toward their preschool-aged children. For example, Cameron-Faulkner, Lieven, and Tomasello (2003), in a longitudinal study of 2 and 3 year-olds, found that children heard an estimated 5000 to 7000 utterances per day. However, there is considerable variation between children in how much talk is directed toward them (Hart & Risley, 1995 - see 'Narrative Development in 12-23 Month Olds'). Moreover, talk that is specifically about the past is particularly likely to vary between families (Hart & Risley, 1995; McCabe & Peterson, 1991; Ratner, 1984). It is such talk about the past that is especially important for the development of both memory and narrative skill (Fivush et al., 2006).

Style of Parental Talk

Considerable research has shown that parents differ in their style of talking about the past. When researchers began to study parental styles of talking with their children about the past, there was unexpected convergence from several laboratories (Fivush & Fromhoff, 1988; Hudson, 1990; McCabe & Peterson, 1991). All of these investigators independently identified differences in how extensively or elaboratively parents fostered their children's talk about past events. These style differences have now been widely replicated (see review in Fivush et al, 2006). Elaborative parents encourage extensive narration about the past events being talked about. They ask lots of questions, especially open-ended wh- questions ("What did we do at the zoo?"). They also, through both their questions and their statements, embellish, supplement, and thus add a lot of new information to the narrative. For example, if their child said that they saw a bear at the zoo, the parent might embellish by saying that the bear was big and fuzzy, too, and then ask for more information about the bear. Thus, they provide elaboration about the events being talked about, and encourage and support their children's elaborations. Parents who are low in elaboration tend to ask specific questions for which they want specific answers, and repeat the same questions if not answered 'correctly.' They are also more likely to have short discussions of each past event as well as readily change to a new topic (i.e., 'topic-switching') if their children do not respond the way they want them to (McCabe & Peterson, 1991). These different styles of parental talk have been shown to have far-reaching effects on children's narrative development. In addition, they have been shown to influence a number of other aspects of both cognitive and socio-emotional development (see review in Fivush et al, 2006). For example, a more elaborative reminiscing style has been linked to autobiographical memory skills, literacy acquisition, attachment relationships, understanding of the self, and theory of mind development.

Helping Children Learn Narrative Structure

The socio-cultural theory of Vygotsky (1978) is the one typically used to account for the role of parents in shaping children's narrative development. In this theory, complex cognitive tasks are learned within an interactive social context. Furthermore, the inter-psychological processes that take place within that interaction are subsequently internalized by the child (or in Vygotsky's view, become intra-psychological processes). In other words, children learn how to structure various cognitive tasks in accordance with how parents (or other knowledgeable people) have prompted them to do these tasks. As stated previously (in 'Narrative Development in 24-35 Month Olds'), adults who are helping children learn new tasks provide extensive guidance and feedback, thereby providing the structure and much of the content of the task. Thus, parents provide the basic scaffold for the child's narrative by the questions they ask and the comments they make. Good scaffolding by an adult is sensitive to the child's level of accomplishment, and as a child gets more competent, the scaffold changes accordingly (Haden, Haine, & Fivush, 1997; Reese, Haden, & Fivush, 1993). Essentially, feedback from children about increasing skill competence leads to more elaborate concepts being scaffolded, so that both partners engage in increasingly complex narration.

When parents are elaborative when their children are 3 year-olds, their children produce more sophisticated narratives later. This has been replicated in numerous laboratories - see review in Fivush et al, 2006. As one example, in a longitudinal study documenting children's developing narrative skill between 2 and 6 years of age in European Canadian children, Peterson and McCabe (see Peterson & McCabe, 2004 for a review of this body of work) found that children whose parents were elaborative when they were 2 and 3 year olds were able to tell longer narratives when they were asked to produce narratives on their own at later ages (Asking children to produce the entire narrative on their own, with no adult scaffold, assesses children's internalization of narrative skill.) They were also more likely to produce narratives that had complex structure when they got older. Specifically, their narratives were more likely to conform to schemas that represent good narrative structure, with beginnings, middles, and ends. Moreover, children of parents who used an elaborative reminiscing style were more likely to orient their listeners to the context of the narrated events, to explicitly connect events causally and temporally, and to embed their narratives in an evaluative context. In short, parents help their children learn narrative schemas through the interactive conversations they have about past events (Fivush et al, 2006; Peterson & McCabe, 2004).

Conversations that foster the development of narrative skills include the following:

Gender Influences

Although parents mostly talk to their sons and daughters similarly, some differences have been found. When such differences are documented (and most of this research has been done with North American Caucasian samples), parents tend to be more elaborative when talking with their daughters than their sons (Fivush et al, 2006; Nelson & Fivush, 2004), and they are more likely to talk about emotions and social relationships with daughters. But whether and how much gender differentiation there may be is probably mediated by culture as well as the child's temperament and interests.

Cultural Differences

Although narratives and stories are universal, there are differences in the value assigned to them as well as their function within different cultures (Wang, 2004). This in turn leads to differences in how parents engage in narrative-talk with children. This recognition of cultural differences has fostered a growing body of research on parent-child narrative conversations, and much of it focuses on 3 year-olds and their mothers. For example, the Maori (aboriginal people of New Zealand) highly value the transmission of stories within their culture, and thus in comparison to European New Zealanders, Maori mothers engage in telling more elaborative stories about significant events of their cultural group and of their children's lives (Reese, Hayne, & MacDonald, 2008). In contrast, Japanese value subtlety in communication and the ability to infer the intentions of others, as well as discourage excessive talk about the self (Minami & McCabe, 1991, 1995). Thus, Japanese mothers engage in fewer conversations about past experiences with their young children, and when they do, they foster conciseness and conformity to social roles (Minami & McCabe, 1991, 1995). Korean mothers, like Japanese, engage in substantially fewer narrative conversations with their 3 year-olds than do European American mothers (Mullen & Yi, 1995), nor do they encourage children to contribute information or introduce their own topics. Rather, they are more likely to prompt children to simply confirm information provided by mothers (Choi, 1992), information that is more likely to emphasize the child's social networks and roles. In contrast to the above cultures, Latino households are characterized as gregarious, with frequent multiparty conversations (Melzi, 2000), and Central American mothers place a greater emphasis on their children reporting on prior conversations.

Considerable research has focused on differences between cultures that foster individuality versus those that emphasize the person as embedded within a web of relationships, i.e., an emphasis on independence versus interdependence of the self (Wang, 2004). To foster an autonomous self-identity, European American and Canadian parents initiate lengthy memory conversations about events that have significance to the child, and depict the child as a central character in the co-narrated story (Wang, 2001, 2004). However, to emphasize the child as an interdependent, social being, Chinese parents instead emphasize group activities and relationships in narrative conversations. When talking about emotional events with their 3 year-olds, American mothers engage in an "emotion-explaining style" that emphasizes the reasons for children's emotional reactions (i.e., focusing on the self), whereas Chinese mothers engage in an "emotion-criticizing" style that emphasizes child proper behaviour and maintenance of relationships (Wang, 2001, Wang & Fivush, in press). Cultures also differ in terms of the directive function of narration (Wang, 2004), i.e., how much a culture uses narratives to teach lessons from the past. For example, Taiwanese mothers use talk about misbehaviour to teach children moral lessons about future behaviour (Miller, Wiley, Fung, & Liang, 1997), whereas American mothers seldom use narratives in directive ways to point out and correct children's transgressions. Rather, such narratives are more likely to be used for entertainment.

Thus, narratives serve different functions in different cultures. And these are reflected in how parents socialize their children through narrative. There are considerable differences between cultures in how much narrative talk parents foster as well as in how they engage in such talk. Nevertheless, across cultures, parents who engage in more frequent and more elaborative conversations about past events with their children have children who produce longer and more elaborated narratives (Fivush et al, 2006). Furthermore, an elaborative parental reminiscing style has been linked to children's language and literacy skills as well as memory.


Children have many of the cognitive foundations they need for narrative skill acquisition, such as memory, language, self-concept, and so on. They are learning to understand and encode many relationships between events, but they still have far to go in terms of learning structural aspects of competent narration. Four-year old children still require a great deal of help from parents or others to tell their narratives, and they still have difficulty with many aspects of learning and using appropriate narrative schema. Parents (both mothers and fathers, as well as other salient adults in the child's life) play an important role in helping children acquire these skills through how frequently they reminisce, how elaboratively they do so, and the sort of narrative scaffolding they do. How parents structure their conversations with their children when talking about past events is also mediated in important ways by their culture, both in terms of the values assigned to narration and the construal of the self within their culture.

Peterson, C. (2008). Research Review: Narrative Development 37-48 Months. In L.M. Phillips (Ed.), Handbook of language and literacy development: A Roadmap from 0 - 60 Months. [online], pp. 1 - 8. London, ON: Canadian Language and Literacy Research Network. Available at: Handbook of language and literacy development