The Nature of Young Children’s Authoring

Deborah Rowe. Handbook of Early Childhood Literacy. Editor: Nigel Hall, Joanne Larson, Jackie Marsh. Sage Publication. 2003.

What is authoring? How does one recognize it? At what age do children begin to exhibit activities that we might call authoring? Is authoring confined to written language or does it include children’s ways of weaving together talk, print, play, art, and music to express their meanings?

Imagine for a moment that you are new to the field of early literacy research and in search of answers to these questions. You pick two monographs from a list of recommended readings. You first tackle Ferreiro and Teberosky’s Literacy before Schooling (1982). You easily conjure up the research scenario. Researchers dictate words and ask young children to write them. Children wrinkle their brows as they consider how to map oral language to print. Ferreiro and Teberosky’s analyses convince you that authoring occurs in the mind of the child. To understand it, researchers must track children’s hypotheses about the nature of the writing system. Next you pick up Dyson’s Writing Superheroes (1997) and begin to read. Here the tale unfolds in a different direction. Dyson allows you to peep over the shoulders of Tina and Sammy as they negotiate writing activities in their classroom. Here young authors are making choices about what to write, who to write with, and how their texts will eventually be received by their teacher and classmates. Dyson’s children are almost certainly making decisions about spelling and other encoding issues—you can see this in the unconventional texts reproduced in the monograph’s illustrations—but this is not what seems most important. Instead, Dyson’s analyses highlight how these children negotiate access to meanings and social roles that are available (or not) to persons of their gender, age, ethnicity, and class. Authoring seems to occur not so much in the mind of the child but between children. You come away convinced that borrowing from popular culture texts like X Men positions young authors as certain kinds of writers in the official world of the classroom and certain kinds of kids in peer culture.

The divide between Ferriero and Teberosky and Dyson seems a large one, not because their work is irreconcilable, but because they look at children’s authoring from such different vantage points. Even cursory reading leaves the strong impression that there is more than one ‘story’ of young children’s authoring. The theoretical perspectives adopted by these researchers serve to differentially frame the authoring act—as to who or what is observed, under what conditions, and what aspects of the authoring event are subjected to intensive analysis, interpretation, and theorizing.

Scope of the Review

My goal in this chapter is to explore the theoretical vantage points researchers use to draw boundaries around authoring and child authors. This review began with a search of international research literature published between 1990 and 2002 and focused on authoring of children from birth to age seven. Authoring was broadly defined to include children’s construction and expression of meanings in a variety of communications systems. However, in order to limit the scope of the review, I included only studies that in some way explored children’s writing (i.e. their attempts at recording linguistic messages on paper), though some studies also included authoring in other symbol systems such as talk, drawing, gesture, and dramatic play.

While there are many possible frames one might apply, in this review, studies were categorized according to major two properties of authoring: location and semiotic system. With regard to location, investigations of childhood authoring have taken two broad approaches. Researchers working from cognitive, psycholinguistic, sociocognitive, and sociopsycholinguistic perspectives have located authoring in the mind of the child and designed studies focusing on children’s hypotheses about print and their formation in interaction with others. Researchers working from sociocultural, situated cognition, and New Literacy Studies perspectives have located authoring as occurring collectively between children and others in their communities. This research has focused on describing socially situated authoring practices, and on analysing the ideological underpinnings of authoring. While scholarship on early literacy learning has given most attention to cognitive and cultural dimensions of authoring, researchers’ decisions about a second property of authoring—its semiotic boundaries—have also been influential. Therefore, this review concludes with a discussion of studies adopting multimodal views of young children’s authoring.

Locating Authoring: Cognitive and Sociocognitive Perspectives

Authoring as a Cognitive Process

In the last three decades, much of the research on young children’s authoring has focused on describing the cognitive processes involved in text production. In general, this work has taken a constructivist view of young children’s authoring. That is, children are seen as building their own understandings about the nature of the writing system (Ferreiro, 1990). They do so by constructing and testing hypotheses about print (Ferreiro and Teberosky, 1982; Clay, 1991; Harste et al., 1984; Rowe, 1994; Teale and Sulzby, 1986) rather than learning through direct imitation of print in the environment. It is because young children’s hypotheses are ‘true constructions’ (Ferreiro, 1990; Ferreiro et al., 1996; Besse, 1996; Goodman, 1990; Kamii et al., 2001; Schickedanz, 1990) and do not entirely mirror adult views of writing that their products often look so unconventional to adult eyes. These writing behaviours that precede and develop into conventional literacy have been termed ‘emergent literacy’ (Sulzby, 1989; Teale and Sulzby, 1986).

Researchers working from an emergent literacy perspective, then, have defined young children’s writing in terms of children’s mental processes rather than textual products (Harste et al., 1984). Intentionality—the intention to create a text for a particular context (Hall, 1989; Harste et al., 1984)—rather than conventionality of the resulting product is viewed as the defining characteristic of authoring. Children become authors when they indicate that they have created a message with the intention to communicate, or at least that they understand that their marks might be meaningful to others if their audience knows how to interpret them. Thus, children who ask an adult ‘What did I write?’ are authors (Clay, 1975; Harste et al., 1984).

Further, analysis of child authors’ unconventional texts and talk reveal pragmatic, semantic, syntactic, and graphophonemic organization (Goodman, 1980; 1986; Hall, 1989; Harste et al., 1984) suggesting that young children’s authoring processes are not qualitatively different from those used by older writers (Harste et al., 1984; Teale and Sulzby, 1986). Young children have been shown to metacognitively monitor the adequacy of their written texts for particular audiences (Rowe, 1994; Sipe, 1998; Wollman-Bonilla, 2001a) and to use strategies to revise texts at the word and meaning levels (Calkins, 1994; Graves, 1994; Sipe, 1998). From this perspective, almost as soon as children pick up a pencil they become authors. Even in the years between birth and age five, many young children are already developing understandings of the functions and forms of writing, and accepting the social role of writer (Goodman, 1986; 1990: Teale and Sulzby, 1986).

Given the constructivist frame outlined above, it is not surprising that a considerable amount of attention has been devoted to describing children’s hypotheses about authoring including understandings about differences between drawing and writing, speech-print links, spelling, grammar, concept of word, written language functions, and genre features. (See Sulzby, 1991; Sulzby and Teale, 1991; Teale and Sulzby, 1986; Yaden et al., 2000 for previous reviews.) Research in the last decade has tended to confirm earlier findings and to extend them to different populations by designing studies with younger or slightly older children, speakers of other languages, and children with special needs.

Findings of a number of recent studies support claims of intentionality and organization in young children’s unconventional texts. Research has confirmed earlier findings that very young children have different physical action plans for drawing and writing (Brenneman et al., 1996), and distinguish writing from numerals (Landsmann, 1996) even though their products look unconventional to the adult eye. Though children’s texts often demonstrate unconventional spatial organization and serial order (i.e. random placement of words, right to left sequencing, and lack of spaces between words), like other aspects of writing, children’s hypotheses become more conventional over time (Clay, 1975; Sipe, 1998).

Considerable research attention has been devoted to confirming earlier descriptions of children’s hypotheses about orthography, both before and after children develop the alphabetic principle and begin to create ‘invented spellings’ (Bear and Templeton, 1998; Henderson and Beers, 1980; Gentry, 2000). A number of Piagetian researchers, working with different language groups, have provided support for Ferreiro and Teberosky’s (1982) description of a sequenced progression of hypotheses leading to the construction of the alphabetic hypothesis. These studies include work with children learning Italian and Spanish (Ferreiro and Teberosky, 1982), French (Besse, 1996), Portuguese (Grossi, 1990) and English (Kamii et al., 2001; Branscombe and Taylor, 1996). Other researchers (e.g. Sulzby, 1989) have described the progression of children’s emergent writing forms from scribbling to random and patterned letter strings (Bloodgood, 1999; Olson and Sulzby, 1991).

With regard to spelling, there is considerable consensus that spelling, too, develops in a fairly predictable way with children developing a series of qualitatively different hypotheses about how speech is represented in print (Fresch, 2001; Hughes and Searle, 1991; Korkeamaki and Dreher, 2000; Mayer and Moskos, 1998)—though there is some debate about the best scheme for dividing and labelling spelling stages (Bear and Templeton, 1998; Gentry, 2000; Read, 1971). (See Zutell and Scharer, in this volume, for a detailed review of this work.) Despite many differences in method (data collection techniques, instructional contexts, and different oral and signed languages), many studies support a developmental progression of learning to spell. Interestingly, Mayer and Moskos (1998) found that deaf children (ages five to nine) followed a similar trajectory.

While much of the last decade’s cognitively oriented research on young children’s writing has focused on children’s understanding of grapho-phonemic aspects of writing, studies of children’s understanding of written genre have given more equal attention to semantic and syntactic organization. Kress (1994) has argued that learning to write is not a generic process, but is instead a process of learning the demands and potentials of different genres. By age four or five many children are able to code-switch between oral and literate registers when they dictate a text (Cox et al., 1997). There is a growing body of work substantiating children’s understanding of genre features including spatial organization and syntactic and semantic structures (Donovan, 2001; Kamberelis and Bovino, 1999; Wollman-Bonilla, 2000; Zecker, 1999). Even six-and seven-year-olds have been shown to have considerable knowledge of micro-level features of story and information genre (Kamberelis and Bovino, 1999) including cohesion, tense, vocabulary, and word order (Donovan, 2001). Overall, children’s ability to produce genre-appropriate meanings outstripped their ability to record them in writing (Donovan, 2001; Zecker, 1999). In general, children’s stories were more conventional than information pieces such as science reports (Kamberelis and Bovino, 1999). Overall, their command of genre features appeared to be closely tied to experience and instruction (Wollman-Bonilla, 2000). Genre knowledge, like other aspects of children’s authoring, appears to develop early and become more complex with age (Chapman, 1995; Donovan, 2001; Kamberelis and Bovino, 1999). To summarize, while children’s hypotheses, forms, and meanings are not always conventional, they are clearly sensitive to the demands and potentials of different genres.


There is considerable research tracking the trajectory of children’s authoring hypotheses and textual forms, and there is little debate that, in general, children’s writing becomes more conventional over time. Nevertheless, there is disagreement among researchers about how stage-like and invariant such a progression may be. Research using Piagetian interviews has most often found strict adherence to stage progressions. For example, using clinical interview techniques, Kamii and her colleagues (2001) found an invariant progression in the sequence of children’s literacy hypotheses. Gentry (2000), whose spelling research is, in part, based in this tradition, agrees that spelling strategies will become more sophisticated with age, but contends that a range of spelling abilities may be displayed at any one time especially as children transition between stages—a finding supported by Korkeamaki and Dreher (2000).

Sulzby (1996) agrees with the notion that children’s trajectory toward conventional forms involves adding more sophisticated forms and hypotheses, but disagrees that there is a strict progression through stages. She and her colleagues (Olson and Sulzby, 1991) found that children retain older, less sophisticated hypotheses and forms in their repertoire, thereby producing texts that show evidence of more and less sophisticated forms. This pattern of creating texts with ‘mixed’ forms was particularly prominent when children were writing in the new environment of the computer. Other researchers (Clay, 1975; Harste et al., 1984) also challenge the notion of a fixed sequence of hypotheses or stages through which all children must pass. They suggest that differences in children’s hypotheses and the sequence in which they are formed may be related to differences in their experiences and because they attend to different aspects of their environment.

Thus, views of an early literacy learning trajectory range from those that hold a strictly sequenced stage theory, to those who emphasize individual variation related to differences in children’s experiences and personal inclinations. One likely reason for these differences is the effect of different research tasks. Only those researchers using standardized clinical interviews find invariant sequences of development. Once researchers begin to draw their data from a broader array of tasks, including observations of children’s writing in natural contexts, more variation appears. Secondly, researchers are not always discussing the same aspects of literacy development, making it possible that certain kinds of literacy knowledge may develop in more predictable patterns while others are more dependent on social experiences.

Individual Authoring in a Social World

Social interaction lurks just off-stage in most of the studies just reviewed. Following Piaget (1976), several emergent literacy researchers (Ferreiro, 1990; Goodman, 1986; 1990; Teale and Sulzby, 1986) argue that children construct literacy knowledge through a process of accommodation and assimilation as they interact with the environment. Sociopsycholinguistic perspectives (Harste et al., 1984) build on these views to highlight the inherently social nature of children’s cognitive processes. Both authoring processes and children’s texts bear the stamp of the social and cultural contexts in which they are formed. From this perspective, young children’s authoring is still seen as occurring in the mind of the child, but this learning is now seen as socially mediated and context specific (Vygotsky, 1978). Therefore, it is crucial to describe children’s individual authoring processes as they are formed and shaped in interaction with a social world (Rowe, 1994).

The importance of social interaction is demonstrated by research describing global links between social contexts and children’s authoring outcomes. Working in settings where children had entered the first years of formal schooling in the United States (Diffily, 1995; Freppon et al., 1995) and Italy (Formisano, 1996), researchers have documented variation in children’s writing in different instructional contexts (e.g. traditional skills-based instruction, emergent literacy instruction, whole language instruction). Differences include children’s definitions of writing (Diffily, 1995), amount and complexity of text generated (Freppon et al., 1995), genre of writing (Freppon et al., 1995), understanding of the functions of writing (Formisano, 1996) and development of hypotheses about speech-print relationships (Formisano, 1996).

Other research has taken a fine-grained look at specific features of young authors’ interactions with others, including parents, siblings, classmates, and teachers. With regard to adult interaction styles, research shows that when adults exert less control during a writing event, children express more interest in writing (Fang, 1999) and initiate more verbal interaction (Burns and Casbergue, 1992; Zucchermaglio and Scheuer, 1996) but also produce less conventional texts than when adults use a controlling style. However, it may be that once some children enter school they are already focused on conventions, and differences in adult style have less impact on conventionality of children’s texts (DeBaryshe et al., 1996). Adults appear to scaffold children’s writing by tracking the child’s progress and meanings and matching their contributions to the child’s current needs and independent writing level (DeBaryshe et al., 1996; Lancaster, 2001; Lysaker, 2000). Following Vygotsky (1978) researchers describe such interactions as developing shared consciousness with an adult and suggest that social interaction of this type supports children in crossing the zone of proximal development. Adult scaffolding allows children to do things collaboratively that they could not yet accomplish on their own (DeBaryshe et al., 1996; Lancaster, 2001; Lysaker, 2000).

When children interact with parents and other familiar adults during writing, there clearly is a relational component to their authoring processes. Adult-child interactions include physical closeness, shared rituals and celebrations of writing progress (Lysaker, 2000). Children engage in writing events to initiate and maintain friendships with peers and to communicate with present and absent audiences (Dyson, 1989; Rowe, 1994; Wollman-Bonilla, 2001a; 2001b). Dyson (1989) has argued that there is a dialectical relationship between cognitive and social aspects of authoring. Children’s literacy strategies are developed to accomplish social purposes and those social purposes in turn shape the strategies children develop and explore (Heath, 1991).

Several studies conducted in classroom settings with hearing (Condon and Clyde, 1996; Dyson, 1989; Labbo, 1996; Rowe, 1994; Rowe et al., 2001) and deaf children (Troyer, 1991; Williams, 1999) have described young children’s interactions with peers in writing events. When children were free to talk and write with others they took a variety of interactive roles including observing other authors; assisting another author by scribing, providing spellings, or sharing ideas; mirroring other authors’ texts and processes; sharing different parts of a writing task to complete a single text; and working collaboratively to co-author texts. Children used talk to negotiate and define their roles (Troyer, 1991), to request and provide help and information (Rowe, 1994; Sipe, 1998; Williams, 1999) and to challenge and question peers’ authoring practices (Rowe, 1994; Williams, 1999). When parents responded to children’s school writing through family message journals, they played many of the same roles including providing feedback and modelling appropriate genres (Wollman-Bonilla, 2001b).

One way social interaction has been shown to impact children’s authoring processes is by providing demonstrations of culturally appropriate authoring forms, processes, and meanings. A number of studies have recorded the ways that young children use ‘live’ authoring demonstrations and the resulting texts as a means of learning about genre-appropriate content, authoring processes, and purposes for authoring (Chapman, 1996; Rowe, 1994; Harste et al., 1984; Wollman-Bonilla, 2001b). One of the most frequent findings of these studies is that children linked their texts to those of other authors with whom they interacted. In some cases children used other authors’ demonstrations conservatively—sticking close to the form and content of another author’s text (Rowe, 1994). Kress (1997) and others (Dyson, 1989; Newkirk, 1989; Rowe, 1994) have argued that even in cases where children’s authoring processes appear to be imitative, they involve constructive work. There is no such thing as mere copying (Kress, 1997). Instead, children analyse other texts through the frames provided by their current hypotheses and reconstruct the forms, meanings, and functions in their own way. Children also use demonstrations as starting points or for authoring as they combine and recast elements of other authors’ texts to accomplish new purposes (Dyson, 1998; Rowe, 1994). What young authors choose to appropriate from a demonstration is motivated by their current hypotheses (Harste et al., 1984), individual purposes (Dyson, 1989) and social interest (Kress, 1997).

Social interaction also appears to play an important role in children’s construction and testing of literacy hypotheses. Constructivist theories of learning suggest that literacy learning occurs through a cycle of hypothesis testing in which children attempt to use their existing hypotheses to account for their experiences (Piaget, 1976; Harste et al., 1984; Ferreiro and Teberosky, 1982; Goodman, 1990; Short et al., 1996). When children cannot assimilate the new information in their existing schemes, the anomalies motivate a re-examination of the situation, and the construction and testing of new hypotheses. Social interaction becomes an integral part of children’s authoring in several ways. Sometimes children use interaction as a means of confirming their existing literacy hypotheses (Rowe, 1994). However, peer questions and comments also frequently challenge their understandings, pushing young authors to clarify, expand, and refine their intended meanings and the forms used to represent them (Condon and Clyde, 1996; Rowe, 1994). In the process, children shift stances to consider the audience’s perspective and monitor the effectiveness and appropriateness of their texts (Rowe, 1994; Wollman-Bonilla, 2001a).

Gregory’s (2001) study of siblings’ writing interactions in home settings provides an interesting complement and extension to classroom-based work. She found a kind of learning ‘synergy’ that went beyond Vygotskian notions of scaffolding where learning opportunities are thought to occur only for the less advanced partner. As in the classroom-based studies, social interactions around writing provided opportunities for the younger child to observe demonstrations and to ask for and receive help. However, the younger child also served as a kind of ‘trigger’ for the older child’s learning by asking them to think, explain, and assist. Social interaction around authoring, then, was a learning opportunity for both siblings.

Overall, these studies make a strong case for the social nature of literacy processes and knowledge. While general cognitive processes such as assimilation and accommodation or hypotheses testing are proposed as key elements of authoring, this line of research argues for the need to study authoring in relation to the social context in which it is embedded.

Authoring as Social Practice: Sociocultural Perspectives on Writing

The research reviewed thus far has viewed authoring as an ‘in-head’ phenomenon, albeit one that is shaped by people and social situations. Recently researchers interested in social and cultural aspects of authoring have challenged the notion of authoring as an individual mental act, suggesting, instead, that authoring occurs between people as they negotiate authoring processes, meanings, and textual forms as part of their everyday activities. As Barton and Hamilton argue: ‘Literacy becomes a community resource, realised in social relationships rather than a property of individuals’ (2000: 13). They note that at a micro level, literacy events are often accomplished jointly by a number of participants with the resulting literacy practices moving beyond the individuals’ understandings and meanings. At a macro level, communities create social rules and hold taken-for-granted assumptions about who can use and produce particular literacies under what circumstances (Barton and Hamilton, 2000; Santa Barbara Discourse Group, 1992).

From this perspective, authoring is seen as social practice—the accepted and valued ways of ‘doing’ writing (or art or talk) in a particular community (Barton and Hamilton, 2000; Lave and Wenger, 1991; Lemke, 1995; Solsken, 1993). While literacy practices are, in part, defined by observable behaviours involving print, they also involve values, attitudes, feelings and social relationships (Barton and Hamilton, 2000; Street, 1995). Authoring practices include definitions of text and authoring, the ways community members talk about authoring, ideological views of literacy, and participants’ socially constructed identities. Embedded in literacy practices are power relations that determine the use and distribution of texts and who has access to various positions in authoring events. Social practices are ideological in that they support the power of one social group to dominate another (Lemke, 1995).

Two major strands of research framed by the notion of young children’s authoring as social practice are discussed below. The first concerns itself with the ways that children and others co-construct situated literacies, and the second analyses the ideological nature of literacy practices.

The Social Construction of Situated Literacies

When authoring is seen as a socially situated act, it is no longer possible to discuss ‘the’ authoring process as if it were a generic characteristic of mind. Instead, young children’s authoring must be investigated as it occurs within the social practices of particular communities (Gee, 2001). Researchers working from this theoretical perspective have focused on culturally based variation in beginning literacy learning and the ways that different community practices lead to different patterns of literacy knowledge and processing. Heath’s (1983) classic study of variation in the literate practices of three communities in the Piedmont Carolinas is a precursor to more recent studies investigating the nature of young children’s participation in classroom communities. Kantor et al. (1992), for example, tracked the way that preschoolers and their teachers co-constructed meanings about literacy through everyday interactions. They concluded that each classroom activity’s materials, purposes, and participant structures framed literacy in distinctive ways. A number of other studies (Manyak, 2001; Larson, 1995; 1999; 2000; Power, 1991) have carefully described children’s participation in communities of practice formed around authoring events in American first-grade classrooms. Power (1991), for example, tracked the development of ‘pop-ups’ (i.e. portions of illustrations affixed so they stood out from the page) as a text convention in a first-grade classroom. By documenting the development of this text feature across time and across the texts of different children, she demonstrated how the classroom community co-constructed its own local definition of text. Unlike cognitively oriented studies of children’s genre knowledge that typically begin with researcher assumptions about conventional genre features, studies framed by a social practice perspective focus on the ways local literacy knowledge is shaped in face-to-face interactions and reflect the group’s values and experiences.

In addition to documenting the situated nature of classroom literacies, researchers have explored the consequences of classroom literacy practices for children’s participation and literacy learning. Following Lave and Wenger (1991), some early literacy researchers (Larson, 1999; Manyak, 2001) argue that children learn culturally situated literacies through legitimate peripheral participation in communities of practice. That is, children participate in literacy events by adopting authoring roles that approximate those of more experienced authors while providing additional support that opens up the practices to newcomers (Wenger, 1998). Through participation in the social life of families, schools, churches and other groups, children adopt community definitions of authoring, including ideas about acceptable meanings, and ways of constructing text.

The notion that literacy learning occurs through joint participation in literacy events has spurred investigators to document how children participate in authoring events. It appears that some classroom literacy events are accomplished through participation structures that allow children and teachers to flexibly shift between roles as experts and novices (Manyak, 2001), and as teacher, author, co-author, overhearer, and so on (Larson, 1995; 1999; 2000). Findings suggest that these shifts in participation allow children to take on new responsibilities for writing and learning. Though these studies do not directly analyse linkages between student positions and learning outcomes, the authors suggest that the nature of participation in authoring practices has consequences for literacy learning.

Authoring as an Ideological Practice

In the studies just described, literacy researchers were interested in describing the situated and distributed nature of authoring practices. Others working from a social practice perspective have studied the ideological nature of literacy practices; that is, they have focused on the ways that a group’s common-sense views of literacy create and maintain power over others, and are framed by relations of status and dominance within a larger political context (Lemke, 1995; Sheridan et al., 2000; Solsken, 1993). Not only are children’s authoring roles embedded within hierarchies of social relationships, but the ‘stuff of authoring is also ideologically loaded. When children construct written texts, they are doing more than selecting words. They are manipulating ideological symbols of power (or weakness) (Dyson, 2000a) and displaying an (often implicit) interdependence between meanings and the social and political positions they occupy (Lemke, 1995). When young children write, they appropriate cultural materials as a means of cultural production (Dyson, 2001).

From this perspective, authoring is co-constructed in local interactions and as part of larger social and political relations (Dyson, 1998; Solsken, 1993). As Solsken argues: ‘There is no politically neutral ground for literacy learning, either in families or schools’ (1993: 219). When they pick up the pen, children, their teachers, classmates, and families reproduce, resist, and transform hierarchies of social relations and their positions within them. For researchers working within this frame, authoring is seen as a process of negotiating social relationships and ideological views of the world. Children’s participation in authoring practices must be read against the structuring processes of race, gender, and class in the larger society (Dyson, 2000a). Authoring is also seen as an act of self-definition. Learning to write involves much more than adding new skills to the child’s cognitive repertoire. It requires them to take on new cultural identities and affects their sense of self in profound ways (Dyson, 2001; Rowe et al., 2001; Sheridan et al., 2000; Solsken, 1993). Key questions for these researchers are: what texts and positions are available for use by children of different genders, ethnicities, and social classes? What forms of agency do children exercise? How do they accept, resist, and transform textual practices and social positions?

Dyson’s (1989; 1993; 1997) research in culturally diverse American classrooms has been particularly influential in opening young children’s authoring practices to ideological analysis. Her work demonstrates how young authors use texts to construct social affiliations with their peers, as well as to accept and resist the ways they are positioned by others (Dyson, 1995; 2000b). Overall, Dyson found that children not only flexibly reframed culturally available signs, but also recast existing social relationships and social practices. Dyson’s (2001) analyses demonstrate that authoring involved assuming a ‘social voice’ that positioned children in particular ways in relation to the ongoing dialogue in their classroom community and in relation to the texts and practices of the larger society. She identified social and symbolic flexibility and recontextualization processes as key authoring processes used by child authors as they moved across symbolic, social, and ideological boundaries (Dyson, 1998; 2000b; 2001).

Ideological analyses have highlighted the intercultural nature of classroom authoring events (Lemke, 1995). Because children are simultaneously positioned in the overlapping communities of official and peer culture, the same authoring activities often have very different meanings for the children’s positions as students in the official world and as friends in the peer world (Dyson, 1993; Rowe et al., 2001).

Several researchers have also conducted ideological analyses of the gendered nature of authoring practices and the ways that writing may reify or transform the social positions occupied and available to young authors. Both Gallas (1998) and Henkin (1998) report that gender and ethnicity strongly influenced the ways children were positioned in authoring events by their peers, with girls and ethnic outsiders sometimes having limited access to powerful roles. Taking a different methodological approach, MacGillivray and Martinez (1998) analysed gender roles displayed in six-to nine-year-olds’ texts. They found that while many of the children’s story characters displayed traditional gender roles, some of the female students created texts where gender roles were not so rigidly categorized. For this latter group of girls, authoring became an opportunity to create worlds, explore multiple positions, and explore alternative definitions of power. Solsken (1993) also explored the ways middle class five-and six-year-olds’ orientations toward literacy were framed by gender and class relations in their families and the larger society. She found that boys’ and girls’ literacy biographies were impacted by societal ideologies about gender, especially the division of labour in the middle class homes where mothers took major responsibility for supporting children’s literacy learning.

Overall, defining authoring as social practice challenges the autonomous model of literacy, as outlined by Street (1995), that is implicit in cognitive and sociocognitive research on young children’s writing. Authoring is neither generic nor politically neutral. Children learn situated ways of making meaning with print that vary according to the literacy practices of their communities. When children write, they take up, adapt, or resist positions in existing systems of power relations. Negotiating their places in these cultural systems is a key part of authoring.

Expanding the Boundaries of Authoring: Semiotic Perspectives

Though discussion of cognitive and cultural aspects has dominated early literacy research, many researchers have commented on children’s tendencies to combine writing with other semiotic systems such as talk, drawing, gesture, and dramatic play. As Kress states: ‘Multimodality is an absolute fact of children’s semiotic practices’ (1997: 137). Authoring for young children involves language, vocalization, gesture, gaze, bodily action, and graphic production (Lancaster, 2001). While adults and older children are more likely to have adopted dominant views of writing as separate from other forms of communication, very young children have less cultural experience and so are less constrained by boundaries between sign systems (Kress, 1997). For young children, written language is intimately connected to other sign systems (Dyson, 2001).

The question for researchers is whether young children’s multimodal authoring practices should be viewed as immature processes that will one day be replaced by writing-only practices or whether children are, instead, at the beginning of a lifelong process of learning how their communities weave together writing, talk, art, drama, and other sign systems. Both theoretical positions are evident in recent research. As discussed earlier in this chapter, emergent literacy research continues to explore children’s abilities to distinguish writing from drawing, and to describe the development of writing along a path where multimodal texts are considered less sophisticated than those that rely only on print. Like other ideological positions, the definition of ‘good’ authoring as confined to written language is often unexamined even by researchers who systematically analyse other facets of young children’s authoring.

Though many school literacy events privilege authoring practices that isolate print from other sign systems, there are at least two important forces pressing early literacy researchers to broaden the semiotic boundaries for authoring. First, despite school and societal pressures to learn print-only practices, young children’s texts and authoring continue to be multimodal (Dyson, 1986; 2001; Harste et al., 1984; Kress, 1997; Newkirk, 1989; Rowe, 1994; Williams, 1999). To limit the focus of research to children’s writing is to ignore a large part of young children’s meaning making. Secondly, literacy research is being influenced by new technologies that produce adult texts that are increasingly multimodal. Internet websites link print, graphics, video clips, music, animation and more. Even traditional print media such as newspapers and magazines devote an increasing amount of space to photos and graphics, with a resulting decrease in print. In such an environment, young children’s penchant for ‘symbol weaving’ (Dyson, 1986) is increasingly being normalized and valued as part of dominant literacy practices (Harste, 2000; Kress, 1997).

Both classic and more recent studies document children’s flexible interweaving of semiotic systems (e.g. Berghoff and Hamilton, 2000; Clyde, 1994; Dyson, 1986; Gallas, 1994; Harste et al., 1984; Hubbard, 1989; Lancaster, 2001; Kress, 1997; Rowe, 1994; Newkirk, 1989; Upitis, 1992). Most often described have been authoring practices that combine writing, art, and oral language (Hubbard, 1989; Newkirk, 1989; Olson, 1992), but researchers have also noted children’s connections between writing, music, dance, dramatic play, and drama (Dyson, 1989; Gallas, 1994; Rowe, 1994; Rowe et al., 2001; 2003; Upitis, 1992). They argue that multimodal authoring practices allow children to draw on meanings formed in a variety of sign systems and to gain access to authoring events using non-linguistic forms of communication (Clyde, 1994; Harste, 2000). This appears to be particularly important for beginning writers and those whose strength is not language (Harste, 2000; Rowe et al., 2001).

Several researchers have described transmediation (Siegel, 1995) or transduction (Kress, 1997) as an important part of multimodal authoring activities. Harste (2000) argues that this movement of meanings across sign systems is part of all authoring events. Because there is often no one-to-one match between sign systems, children are encouraged to reflect on both meanings and forms (Siegel, 1995; Dyson, 2000b; Kress, 1997; Rowe et al., 2003).

When researchers broaden the semiotic boundaries of authoring, another outcome is increased attention to the embodied and material nature of authoring practices. For both hearing (Lancaster, 2001) and deaf or hard of hearing (Williams, 1999) preschoolers, authoring involves bodily action. As Williams observes, young children perform their early writings with gesture, facial expression, and pantomime. Embodied practices such as gaze and body posture carry important meanings, and are closely monitored by adults who interact with young authors (Lancaster, 2001). Multimodal authoring practices are also strongly influenced by the physical materials that are available in the environment (Dyson, 2000b). As Kress (1997) points out, the materiality of the objects is important in that children are adopting and adapting culturally significant elements of complex signs when they combine paper, writing tools, and objects from their environment, with gesture, talk, and drama.

Overall, it is clear that though young children initially see the boundaries between writing and other sign systems as more permeable than do adults, they quickly begin to explore print-centred views about meaning making. Nevertheless, throughout the early years, children’s authoring continues to be multimodal—a characteristic that some researchers consider a strength rather than a weakness. They view authoring as a material and embodied process through which children adapt and transform cultural resources, and which necessarily involves weaving together and moving across a variety of sign systems.

Provocative Questions and Future Directions

Cognitive, cultural and semiotic perspectives provide different answers to questions about the nature of childhood authoring posed at the beginning of the chapter. While these perspectives represent a rough map of new directions surfacing in the field over the last three decades, it should not be assumed that the initial frames have been abandoned. Instead, active research agendas are being pursued in each of these areas.

In general, research on childhood authoring began with attention to young children’s unconventional hypotheses about writing. This was followed by interest in the ways children constructed literacy as they interacted with researchers, parents, teachers, and peers. More recently, some researchers have used sociocultural theories to reframe descriptions of authoring in everyday settings as a collective rather than an individual act. This work has emphasized the situated and ideological nature of children’s authoring as they negotiate local definitions of literacy. The semiotic boundaries of childhood authoring are also increasingly of interest, with researchers broadening the definition of authoring to explore children’s strategies for creating multimodal texts. Overall, researchers are more frequently adopting multiple foci to look for the mutual shaping of cognitive, social, ideological, and semiotic aspects of authoring (Dyson, 2001) or at least acknowledging the impact of all aspects, even if not all are studied directly.

New theories provoke new kinds of thinking. They also suggest interesting directions for future research. It is clear that both individual and cultural theories make important contributions to our understanding of childhood authoring. Future research would profit from analysing research data from multiple theoretical perspectives. At the same time, researchers should work toward developing multidimensional theoretical frameworks that more powerfully account for both cognitive and cultural aspects of authoring.

Secondly, researchers may want to reconsider intentionality as a necessary condition for defining childhood authoring. If learning to write is co-constructed with others through participation in the social life of the community (Lave and Wenger, 1991), then the beginnings of authoring are likely to precede the young child’s individual intention to communicate through print. It is quite possible to imagine a situation in which very young children participate in writing events even before they recognize their social significance. In such a scenario, intention, like other literacy knowledge, should result from participation rather than preceding it, suggesting that we need to study even younger children and their families in order to witness the beginning of authoring.

Thirdly, theoretical perspectives that advance the notion of socially situated literacies expose the tendency of early literacy research to privilege the textual practices of mainstream children and families as ‘normal’ and marginalize the practices of non-mainstream families by considering these children ‘at risk’ (Dyson, 2000b). Research describing the diversity of community authoring practices and the relation of those practices to dominant literacies has the potential for opening new dialogues about the richness of non-mainstream practices and the ways that children who adopt them may be differentially positioned in school literacy events.

Finally, if authoring is more broadly defined as making meaning using all available semiotic resources, young children’s ‘symbol weaving’ (Dyson, 1986) becomes a viable focus for investigation and theorizing. Despite the interest of some researchers in multimodal authoring practices, literacy research remains largely print-centred with connections to other sign systems relatively underdeveloped. Given the changing face of communication technologies, it seems important to describe and nurture children’s capacity for multimodal authoring. We may find that, in this area, young authors are leading the way toward twenty-first century literacies.