Both practical experience and research indicate that there is a relationship between oral language impairment and reading disabilities. Many of the children who experience reading problems in school are also on the case loads of speech-language pathologists (SLPs). A number of studies have documented that young children with language impairment are at a higher-than-average risk of demonstrating later reading disabilities (Aram, Ekelman, & Nation, 1984; Aram & Nation, 1980; Catts, 1991, 1993). This literature indicates that language impairment may be a basic deficit that affects language function in both its oral and written forms.
In this report, we explore the concurrent language functioning of preschoolers in specific oral and written language situations. Awareness of the relationship between reading disability and oral language impairment in the school-age population has coincided with newer understandings of the reciprocal nature of oral and written language forms during the preschool years (Cox, 1994; Sulzby, 1985a, 1985b; Teale & Sulzby, 1986). These understandings have been central to research in the field of emergent literacy.1 Emergent literacy theory claims that children show knowledge and ability in written language through oral delivery forms, such as using structures suited to written discourse in an orally delivered version of a storybook while looking at the book's pictures rather than tracking and using cues from print (Sulzby, 1985a, 1994, 1996a). Conversely, children may import features typically suited to oral language situations into their storybook reenactments or their rereadings of scribbled or drawn compositions (Sulzby, 1986).
Integrating the theoretical orientation and concepts central to emergent literacy theory into understandings of oral language development can expand the scope of language-remediation services provided to young preschool children. This report is designed to aid that process. In this report we review the major theoretical concepts and empirical findings central to emergent literacy. Next, we examine oral and written differences in a variety of contexts, including our previous and ongoing work. Finally, we present examples from case study data drawn from our current research and discuss the implications of these data for SLPs who want to incorporate emergent literacy perspectives and practices into their language remediation with preschool children.
In our collaborative research program and in this report, we bring together perspectives from speech-language pathology and emergent literacy. Research in the field of speech-language pathology typically has focused on spoken language development of young children, whereas research in the field of emergent literacy has investigated young children's written language development. Researchers in both fields increasingly acknowledge the importance of studying the relationships between oral and written language acquisition. Sulzby (1985a) argues that children develop both oral and written language in an interrelated fashion during the preschool years prior to conventional literacy.2 Such a position suggests that investigation of the reciprocal nature of oral/written language forms during the preschool years is warranted.
Vygotsky's (1978) model of internalization of speech characteristics from social contexts has typically been used in parent-child interaction theory to mean that the child internalizes the speech characteristics of the more knowledgeable and present adult, usually the parent. Our study acknowledges this part of the Vygotskian model, but further acknowledges one broader social context of the child-parent dyad--that of the published storybook, or a nonpresent adult writer as mediated through the present adult (the parent). Storybooks, as Teale and Sulzby (1987) have pointed out, have a privileged status within mainstream U.S. and European cultural settings. Bus and her colleagues (Bus & Sulzby, in press; Bus & van IJzendoorn 1992, 1995; Bus, Sulzby, & van IJzendoorn, 1996) have found, when interviewing Dutch parents, that even those who were found later to read infrequently to children nevertheless strongly asserted that "of course" they read to their preschool-aged children.
Sulzby (1994) claims that the child's readings also have characteristics of Piagetian-like constructions not totally contained in the parental readings. Sulzby and Teale's (1987) study of parent-child interactions among low and middle income Hispanic and Anglo parent-child dyads showed that some, but not all, patterns of emergent storybook readings were seen first in interaction and later in children's independent speech. Our focus in this report is upon those language structures from the parental readings that appear first in interaction and then in the child's independent speech functioning, or the Vygotskian aspect of the Sulzby model. We analyze how this internalization process appears to be similar to or different from that which occurs during toy play interactions between parent and child.
In summary, our theoretical orientation is first of all formulated on the basic concept of a reciprocal relationship between oral and written language development during the preschool years. Second, we acknowledge the significance and impact of the parent-child interaction both in toy play and storybook interaction. Finally, because of its elevated status within our culture, we suggest that the language surrounding storybooks may be especially important to examine and particularly relevant to children with communication disorders.
In this section, we review key findings from emergent literacy research along with implications for practice that some researchers have recommended for typically developing young language learners. We also review findings from a small number of studies which have explored emergent literacy in children with communication disorders. We then review our own research program in children's emergent readings3 of storybooks, parent-child storybook reading, and elicitation of children's oral and written narratives.
Since children learn literacy content in a social context, the literacy environment of young children is critical to their successful literacy/language development (Morrow & Paratore, 1993; Sulzby, 1986). Children become "socialized" to the functions of literacy in their daily life (Heath, 1983; van Kleeck & Schuele, 1987). Preschool children who show a "high interest" in literacy activities tend to come from homes and schools where parents and teachers provided a supportive literacy environment (Hiebert, 1994; Morrow, 1983). These environments are not monolithic but vary across individual families and cultural groups (e.g., Anderson & Stokes, 1984; Moll, 1994; Taylor & Dorsey-Gaines, 1988).
Children learn many aspects of language during storybook reading, ranging from vocabulary to discourse structures. Studies have demonstrated that young children learn vocabulary (Moerk, 1985; Ninio, 1983; Snow & Goldfield, 1983; Whitehurst et al., 1988), syntactical constructions (Snow & Goldfield, 1983), and use of decontextualized language (Snow, 1983; Sulzby, 1985a). The development of such skills can lead to increased language performance (Chomsky, 1972) and increased achievement in school (Snow, 1983). One metanalysis indicates that, across studies, parent-preschooler reading is related to outcome measures such as language ability (most specifically acquisition of the written language register), emergent literacy competency, and reading achievement (Bus, van IJzendoorn, & Pellegrini, 1995).
Repeated reading of a storybook facilitates the child's internalization of written language structures (Sulzby, 1985a; Teale & Sulzby, 1987). Goodsitt, Raitan, and Perlmutter (1988) found that the interchanges between parents and children focused more on story content and less on labeling behaviors as familiarity with the book increased. Yaden (1988) reported that, with repeated reading, children began to ask fewer questions centered around the book's pictures and instead began asking more questions about the meaning of words and the story. Martinez and Roser (1985) suggested that the kinds of questions children asked after repeated readings demonstrated a greater depth of understanding. It has been suggested that reading a book with an adult many times permits a child to assume aspects of the adult language role (Snow & Ninio, 1986). Children have been found to begin emergent readings spontaneously when parents read books repeatedly (Sulzby & Teale, 1987; Bus, Sulzby, and van IJzendoorn, 1996).
Book reading routines appear to create a context that is highly routinized (Ninio & Bruner, 1978). There is some evidence that a mother's interactive speech during routines is more complex than when she is experiencing novel objects and activities (Snow, Arlman-Rupp, Hassing, Jobse, Joosten, & Vorster, 1976). Thus, book reading appears to create an enriched language environment.
Parents can vary greatly in the extent to which they use questioning behaviors and labeling behaviors during reading interactions (Heath, 1983, 1986; Ninio, 1980). These differences in parents' language performances result in differences in child language behavior. For example, Anderson-Yockel and Haynes (1994) reported that white mothers asked more questions of their preschoolers, resulting in greater question-answering communications by white preschoolers. African American mothers, on the other hand, asked fewer questions, which resulted in a greater incidence of spontaneous verbalizations by the African American preschoolers . SLPs need to be sensitive to these interrelationships and avoid prescriptive approaches that suggest to parents there is only one "right way" to read to their children (Kaderavek & Sulzby, 1995; Sulzby & Edwards, 1993).
A preschool program which provides for a the broadest usage of oral and literacy behaviors will be beneficial for the preschool child with language impairment (Schule & van Kleeck, 1987; Watson, Layton, Pierce, & Abraham, 1994). Early literacy interactions should aim to develop all modes of communication and provide normalizing experiences for children with specific language impairment (Watson, Layton, Pierce, & Abraham, 1994). Opportunities for long oral narrative performances allow young African American children, including those with language delays and impairments, to display oral and written language knowledge that might otherwise be overlooked (Hyon & Sulzby, 1994; Sulzby, Branz & Buhle, 1993).
Parents vary their teaching strategies in response to children's communication abilities. Pellegrini and colleagues found that parents of children showing typically developing communication skills appeared to be more demanding (high versus low cognitive demands) and used fewer support strategies (e.g., fewer task turns, less nonverbal management) than did parents of communicatively impaired children (Pellegrini, Brody, & Sigel, 1985; Pellegrini, McGillicuddy-DeLisi, Sigel, & Brody, 1986).
Even though children with developmental disabilities or severe speech and physical impairment may have some reduced ability to explore storybooks, all available opportunities including advanced technological solutions should be explored to maintain an enriched literacy environment (Koppenhaver, Coleman, Kalman, & Yoder, 1991; Pierce & McWilliam, 1993). For example, Koppenhaver et al. (1991) have suggested that one way to increase participation by severely impaired children is to provide communication symbols so that these children can ask adults to read aloud, request a page to be turned, or ask questions.
Katims (1991) examined the effect of using emergent literacy activities in a classroom of preschoolers with a range of cognitive, physical, emotional, behavioral, learning, and developmental disabilities. The emergent literacy activities included daily storybook readings paired with emergent writing activities, as well as the availability of a well-stocked classroom library center. The experimental group was compared with a control group of similar children in a preschool classroom that lacked these emergent literacy activities. The children who were exposed to emergent literacy activities were subsequently found to interact with books in more sophisticated and varied ways than the children in the control group did.
Many SLPs and teachers of young children already engage in many or most of the recommended practices. Others can be encouraged to include positive literacy environments for the language impaired children on their caseload or in the classrooms they serve. Recommended activities include stimulating communication in all modalities, exposing children to repeated readings of storybooks, and providing many opportunities for emergent writing activities (see, for example, Sulzby & Barnhart, 1992; Sulzby, Teale, & Kamberelis, 1989).
One advantage of incorporating these emergent literacy activities into classrooms and clinical practices is that they provide an opportunity for SLPs to observe a language-impaired child's ability to negotiate the demands of varying oral and written contexts. One aspect that distinguishes oral from written language productions is the varying demands of decontextualization between these contexts. Scollon and Scollon (1981) suggest that written language demands the use of linguistic forms, allowing a nonpresent audience to comprehend the text without the benefit of the context surrounding and/or imagined by the writer at the time of composition. We turn now to our research in emergent storybook reading, parent-child storybook elicitation, and children's oral and written narrative productions, in which we explore this issue of decontextualization more fully.
As we have described, independent emergent book reading elicitations from nonconventional readers allow researchers, teachers, and SLPs to observe the child's shifts between oral- and written-like language forms. The protocol we describe here is an outgrowth of Sulzby's research initiated in 1979 in which she asked young children to "read" from their favorite storybook.
The primary data used to develop the Sulzby Classification Scheme were obtained from 24 kindergartners in a middle to upper-middle socioeconomic status (MSES) suburb of a large Midwestern city. The children, whose ages ranged from 4 years 11 months to 5 years 11 months, were taking part in a year long study (Sulzby, 1985b) focusing on emergent reading and writing knowledge. Part of the data for this study (Sulzby, 1985a) included a beginning- and an end-of-the-year interview which included a discussion of a well-known book. As the children were taken individually to the interview, each was asked to select a personal favorite book from the classroom library. At the end of the interview, each child was asked to "read me your book." Children's transcribed narrative attempts were content-analyzed and classified in a schema from the least mature productions (i.e., picture-governed attempts with story not formed) to independent or conventional readings from print (see Figure 1).
A second study (also reported in Sulzby, 1985a) tested the scheme with younger children seen more frequently over a year long investigation. A total of 32 children aged 2, 3, and 4, all enrolled in a privately operated day care center in a suburb outside a large Midwestern city, took part in the study. Different numbers of children were present for different parts of the study. Half of the children were from lower socioeconomic status (LSES) and half from MSES homes. The classrooms were observed weekly. In addition, some of the children were seen quarterly and a subset were seen monthly and videotaped as the familiar researcher asked the child to "read me your book." A variety of books were used, including books read repeatedly in the classroom and "favorite storybooks" sent from home by the children's parents. During the analysis and classification of these younger children's emergent readings, no new categories or subcategories of reading behaviors emerged. Instead, it was found that (a) the categories appeared to be quite stable, (b) there was reasonable stability across storybooks for individual children, and (c) the behaviors appeared to be developmental in that the patterns differed predictably across ages. Additionally, none of the children in either the day care or kindergarten study was receiving formal literacy instruction, although all were being read to frequently in their care settings. More recently, the Sulzby Classification Scheme has been used reliably with U.S. populations including LSES white children (Reuning & Sulzby, 1984), LSES white children of Appalachian background (Elijah & Sulzby, 1992), LSES African American children (Elijah & Sulzby, 1992; Sulzby, Branz, & Buhle, 1993), LSES Mexican and Mexican American children (Sulzby & Zecker, 1991), and LSES and MSES Anglo and Mexican American children (Teale & Sulzby, 1987; Teale and Sulzby, 1989).
Book selection and the adult's words and affect are very important in eliciting emergent readings from all children, but particularly from children with language impairments. First, it is important to make certain that the book that is being used for the storybook elicitation is sufficiently familiar to the child. (As a rule of thumb we consider a book to be familiar if it has been read to the child on three to four occasions.) The book should also be one that the child shows evidence of enjoying through behaviors such as attentive listening, commenting on pictures or story, chiming in, echoing, or completing parts of the reading.
Another book-selection issue concerns the type of book that is used for the elicitation. This scheme was developed and used with narrative-type storybooks. (Although it has been used occasionally with other genres, adaptations were necessary and reliability and validity data are not available.) A narrative story is one that contains characters and a plot. Examples include books such as Are You My Mother? (Eastman, 1960), Clifford the Big Red Dog (Bridwell, 1985), and Where the Wild Things Are (Sendak, 1963). Research has suggested that by being exposed to complex narratives children have the opportunity to internalize discourse patterns, whereas with brief pattern books, such as Brown Bear, Brown Bear (Martin, 1970), children tend to rely on rote memorization (Sulzby, 1991) and show little variation on the Sulzby Classification Scheme.
The instructions that are used to elicit the emergent reading are a second important aspect to be considered. When asking the child to produce an emergent reading it is important that the SLP use a direct imperative request ("Please read me your book" or "Read me your book"). The SLP must be certain to ask for a reading. If the SLP asks "Can you read me the book?", this comment may be interpreted as calling for a judgment about the child's ability to read. Similarly, if the SLP says, "Tell me the book" or "Tell me about this book," these comments may be interpreted as requests for production of storytelling or an oral narrative.
Many children will eagerly respond to the request to read, while some children will be hesitant. There are some strategies that may help with hesitant children. First, use plenty of "wait time." Wait time is very important with young children when they are asked to perform an unfamiliar task. Attitude is important as well: use a calm tone and convey confidence that the child will begin "reading" through tone and glance.
If encouragement is needed, you might say, "It doesn't have to be like grown-up reading--do it your own way," "Try," or "Give it a try." If the child still refuses, provide support by saying, "Let me read a bit to you," then read the first page or two to the child. After two or three pages say, "Now it's your turn. You read it to me." If the child continues to refuse, then say, "Let's read it together." At this point begin an interactive reading at whatever level you feel is best for the child. When you feel the child might cooperate, possibly between pages four and six, you might try again to turn it over to the child. If the child continues to refuse to attempt the task, the behaviors and language observed still will be helpful in gaining insight about the child's understandings of written language. This concept will be explored more fully during the discussion of categories of emergent reading.
Once the SLP or teacher has obtained the emergent reading, it is possible to classify the language produced by the child (Figure 1) using an ordinal scale of 1-11. The following is a description of each of the levels of emergent reading, and behaviors and language of children that characterize emergent reading at each level. The Sulzby scheme is applied to the reading as a whole, not to individual utterances or short episodes, although the analysis of smaller elements may help making a decision about the whole. For the SLP, linguistic analysis of smaller units is important as a separate, but related type of analysis.
If these levels are gained in a valid manner with good rapport between adult and child, we consider them to be relative indicators of the child's development in a specific and important linguistic context; Sulzby (1994) explicitly cautions against interpreting these as stages in the general developmental sense. The levels can be used to indicate literacy progress but, for the SLP, examination of linguistic usage by the child within the situation is ordinarily of more importance than measuring literacy growth.
Transcribe the child's reading, including interactions with the SLP, with notes about the kind of intonation patterns used. Read the transcription carefully and make the categorization based upon the structure of the "text" as a whole. The following rubrics concern the structure of the entire reading attempt, not subparts. In contrast with the work of Elster (1994) judgments on subparts should follow, rather than precede, judgments on the whole. As can be seen in Figure 1, the structures are divided by a number of overarching distinctions: whether the child is looking at the print or pictures as if they are the source of the reading; whether the child's speech sounds like oral or written language; and whether the child's speech shows the structure of a story. Further information on this scheme can be found in Sulzby (1985a, 1985b, 1996a, 1996b).
Here the child's speech does not sound as if it is a story, but is primarily focused on isolated pictures. Sulzby (1985a, 1994) uses a very rudimentary notion of "story" --speech having a beginning, middle, and ending event structure--rather than the full blown episodic structure of story grammars. The speech is that of face-to-face oral interaction. The child appears to utter speech for each page separately without integrating the separate pictures into a cohesive story. The most obvious characteristic of this level is that the observer must look at the pictures in the storybook to understand the context of the child's speech utterances; in other words, the speech is very contextualized to the pictures and face-to-face interaction. The child often looks away from the book to the adult when uttering speech, although he or she may rapidly look back at the book while pointing. It is as if the child is attempting to ensure that the adult is looking at what the child is referring to. When the child makes an aside to comment about the reading event, to change to a different linguistic context such as telling about something happening in the environment or in the child's past, or to ask a question, often the adult cannot be sure that the speech is an aside or is part of the storybook reading event. (In later levels, children show appropriate linguistic and paralinguistic markings of asides.) Sulzby divides these no story formed reading attempts into two subcategories: labeling and commenting, and following the action.
If a child is producing behaviors which can be classified as labeling and commenting, he or she points to items on a page of the storybook and gives the label or descriptor for that item (e.g., "kitty-cat") or comments about the picture by giving some information ("wash hes [sic] hand") or even a long sentence, "That's a fighting guy up there." When the child points, it is usually directly at the picture being labeled or discussed, often with an emphatic punching or tapping motion. It has been observed that the speech of most children during picture-governed attempts is markedly less mature than the child's general usage. (It should be remembered that these data thus far have come from nonlanguage-impaired children.)
The second kind of oral production observable at the "story not formed" classification level are those that focus on the action taking place in the pictures--a level called following the action. Children performing at this level typically point or gesture toward the pictures and talk about the action in the picture in the present tense as if the action is immediately occurring. Children at this level frequently use their finger to trace the action in the picture, or gesture or perform the activity occurring in the picture. Sometimes these reenactments are accompanied by sound effects. In both levels within this category, the intonation patterns are that of conversational turn taking in oral speech. (Pitch levels from very young children in emergent storybook readings can occasionally be confusing in determining whether the intonation is reading-like or conversational. Young children often use a wide range of pitch, including very high pitch for some key words or phrases; these pitch shifts should not be confused with reading intonation.)
Although the child operating at these levels does not yet sequence the events of the book into a story, and does not decontextualize the language used, the behaviors exhibited at these early levels demonstrate significant advances in emergent literacy ability. We think this is particularly the case for language impaired children. When language impaired children are able to label a storybook picture, they have accepted an important "contract of literacy" (Snow & Ninio, 1986). Snow and Ninio (1986) suggest that when a child understands that "pictures are not things but representatives of things" (p. 126), and that "pictures are for naming" (p. 131) he or she has learned fundamental concepts needed for literacy development.
Some language impaired children have been observed to use the labeling or following-the-action patterns later chronologically than a child who is typically developing. This suggests that SLPs and parents need to be particularly sensitive to acquiring books that appeal to a child's interests and level of social-emotional development while continuing to facilitate his or her use of labeling behaviors. Label-type books designed for very young children may not be suitable for a communicatively impaired 3- or 4-year-old.
Watkins and Rice (1991) report that the use of verb particles constitutes a particularly challenging linguistic task for language impaired preschool children. When a language delayed child is able to use a variety of verbs and verb particles, such as those used in following the action reading attempts, a truly important literacy and language milestone has been achieved.
Again it is particularly important to use books that have high saliency for a particular child to maintain the motivation to perform a difficult task. One language impaired preschooler we observed was especially interested in trains, planes, cars and trucks. His mother observed that he would attempt to produce many more verb forms whenever she used storybooks that contained these high-interest items.
There are two basic subcategories at this level of emergent reading, dialogic storytelling and monologic storytelling. The basic characteristic of this category is that the speech is intonationally and structurally appropriate to oral face-to-face recounting of events, either in conversationally structured dialogue or extended discourse.
In dialogic storytelling it is possible for the listener to infer a weak storyline from the child's emergent reading of the book, but clarity of the story may be disjointed at times. Usually at the dialogic level the child looks at the book for more prolonged intervals. This behavior contrasts with that seen at the previous level of "no story formation," in which the child characteristically looked at the listener when giving information. Other behaviors noted to occur at the dialogic storytelling level include the use of "voices" to indicate the various characters and the use of dialogic comments directed toward the listener (e.g., "Look at this").
In addition to the behaviors noted above, the child's sensitivity to the needs of the listener begin to become evident at this level. Although the reenactment is typically contextualized to the picture, the child will begin to use some basic forms that assist in developing the weak storyline. This includes the use of present progressive and future tense verbs and a rough story sequence. Two patterns have been detected that will assist the SLP in identifying this structure. Some children create "voices" for characters shown in pictures but rarely use dialogue carriers such as "he said," or "the bird said to him," to indicate speaker identification. Dialogue may, however, be expressed as indirect quotation, in a typically oral format, "She says she don't wanna." The second pattern is to make conversational overtures and exchanges with the adult in giving the recount.
At the level of monologic storytelling the child begins to tell a complete story, still using characteristics of oral language. Although the story continues to be contextualized to the pictures (the child assuming that both participants can see the pictures), the child tells a complete story without requiring the listener's responses to help "keep it going." The child uses the intonation of face-to-face oral language as contrasted with a reading intonation that will appear at later levels. The child who performs at this level shows ability to sequence the events of a story readily in speech. This level is particularly important for language impaired children, because a relationship has been demonstrated between a child's ability to tell a sequenced story and later reading success (de Hirsch, Janksy, & Langford, 1966).
Sulzby and Zecker (1991) report that oral monologues are rarely found with MSES normally developing children in the U.S., but are more typically detected with LSES and minority children. In a study of LSES Mexican American children, they found a higher frequency of highly developed oral monologues and analyzed internal linguistic structures in these readings, particularly focusing on Spanish reading attempts.
At some point in emergent literacy development, children begin to demonstrate an understandings of written-language conventions. Their reading reenactments at this level fall into three subcategories: reading and storytelling mixed, reading similar-to-original story, and reading verbatim-like story. During reading performances at each of these subcategories, children will demonstrate some written-language linguistic features.
The first subcategory at this level, reading and storytelling mixed, reflects a child's performance that is in transition from an oral production to a production reflecting written-language characteristics. These types of reading are characterized by fluctuations between speech that sounds appropriate in an oral context and speech appropriate for a written context. There should be more than one switch between oral/written forms in the reading attempt. The child's narrative may differ from the actual story, yet the story is recited with a sense of attention to the listener's perspective and some decontextualization. Children from MSES backgrounds often go from dialogic storytelling to reading and storytelling mixed, skipping monologic storytelling as a level. It can be considered a transition subcategory, leading to the next two levels in which the child sounds like a reader for most of the reading attempt.
If a child performs at the second subcategory, reading similar-to-original story, he or she sounds like a reader while still looking at the pictures as though reading from them. Children at this level demonstrate patterns that are very similar to the chosen book. At this level children will decontextualize most of the language that is used and will use reading intonation patterns.
The final subcategory at this level is verbatim-like reading. At this level the child shows a marked internalization for stretches of the written text. This level is unique in that the child will appear to self-monitor and self-correct in order to duplicate the remembered text. Children at this level are often described as having "just memorized" the text but transcripts (Sulzby, 1985a) indicate that the cognitive task is more complex than a verbatim recitation. Instead the child seems to be attempting to reproduce an internalized representation of the written language contained in the book. Children using reading similar-to-original story language forms may sound more reader-like and keep the story closer to the original than do children at the verbatim-like reading level. This is because the overgeneralization and self-correction patterns that help define the verbatim-like reading may lead to some disorganization in local structure and signs of searching for specific wording. The examples below help to illustrate this phenomenon.
Example One and Example Two come from emergent storybook readings of the book Are You My Mother? (Eastman, 1960) used in Sulzby (1983, 1996a, 1996b). These children were read this and other storybooks repeatedly in a yearlong study conducted in a church-run day care center. All examples here and in the following sections use pseudonyms for the children. The examples are drawn from pages 22-25 in which a young bird is in search of his mother, whom he has never seen. He asks two animals, a hen and a kitten, "Are you my mother?" In making judgments about emergent literacy, key features of the given book need to be analyzed. In this book, for instance, the key characters, the mother and the baby bird, are introduced specifically in first few pages, then the baby bird becomes the protagonist throughout the book. The bird is referred to as "he," but young children often overgeneralize the need to give character specification through use of specific nouns or names. In the section these two children are reenacting, the bird is referred to as "he" whereas the new characters are specified by nouns. Additionally, the quoted speech is set off by dialogue carriers after the speech ("he said to the kitten," p. 22).
In Example One, Megan (age 3 years 10 months) gives a similar-to-original reading and, in Example Two, Brian (age 4 years 8 months) gives a verbatim-like reading. Words directly verbatim to the book are shown in capital letters.
Brian: CAME TO the kitty.
Said, "ARE YOU MY MOTHER?"
SAID the kitty TO--TO THE, um,
(pause, finger in mouth as he appeared to think).
"ARE YOU MY MOTHER?"
said the bird.
THE cat JUST LOOKED at him.
"Are you my mother?"
said--um, said the bird TO THE--HEN.
"NO, I'm a hen."
Megan's similar-to-original emergent reading follows the order of the book and sounds more organized than Brian's verbatim-like example, even though his is classified as the higher of the two. Consistent with a straight line increase in written language development, he places the dialogue carriers after the speech in contrast with Megan's preposing them. Other characteristics illustrate how his usage looks less developed on the surface while actually indicating higher development. For example, Brian's effort after his own notion of the book's wording is evident in his overgeneralizations about written language structures and his self-corrections, which show evidence of monitoring his speech in relation to an internal model. In the section found in Example Two, Brian over-specifies the dialogue carriers ("said the bird to the--hen"). He attempts to self-correct his attributing the question to the kitty and he finally gives up and rephrases the question and dialogue carrier.
For a language impaired child to reach levels of storybook reenactment containing written-language characteristics such as those shown in the previous three subcategories, he or she not only must understand narrative storytelling structure, but must also understand the language differences between oral and written language "codes" within the linguistic community. Tannen (1983) has suggested that children learn to use different language codes or registers depending on the communication goal. Children performing at the written-language level in an emergent storybook elicitation apparently understand the communication goals and perceive the need for increased decontextualization and different linguistic structures when "reading" a book--a milestone in language and literacy development.
The final four subcategories of the classification scheme reflect children's reading reenactments once they demonstrate a fairly stable awareness that it is print which people read, rather than the pictures (as was assumed at the previous levels of performance). The levels of reading influenced by print include refusing to read based on print awareness, reading aspectually, reading with strategies imbalanced, and reading conventionally. These categories are very important in the study of the transition into conventional literacy and are discussed at length in Sulzby (1996a, 1996b).
One characteristic indicating a child has become aware of the influence of print is the occurrence of specific "high-level" refusals to read emergently. High-level refusals are considered to be a result of the child's awareness of the skills needed for conventional reading. In a high-level refusal, the child might say "I don't know the words," "I don't know those sounds," or "I can't really read--I was just pretending" (Sulzby, 1985a). In general, the high-level refusals occur in the mature emergent reader as a sudden transition following productions of very complete written-like storybook reenactments (Sulzby, 1985a).
High level refusals can be contrasted with low level refusals. Low-level refusals can be a result of the child's unfamiliarity with the book, lack of confidence, or difficulty formulating the language to complete the task (a very real possibility for language impaired children). Low-level refusals are shown by the child's repeated silence or head shaking when asked to read and when encouraged. They are also shown by the child's subsequent low level of responsiveness in adult-child interactive reading (Otto, 1984) and continued refusal to take over the reading when offered, "Now it's your turn--you read."
Reading aspectually, the next subcategory, is important from a theoretical standpoint (Sulzby, 1996a). Here the child appears to be bringing together knowledge from other interactions with print, including emergent writing, word exploration, and phonologically oriented speech play and instruction. Three types of reading behavior appear to be functionally equivalent, but in all three types the child focuses on print as the source of the reading. The child attempts to read from print but focuses on only one or two aspects of the print to the exclusion of other features. Some children give a reading focusing on one aspect and others switch between aspects. The three aspects are a comprehension focus, a letter-sound focus, or a focus on knowledge of words. Children who focus on comprehension read much like the child at verbatim-like reading level but they point to the print or look closely at the print without tracking print accurately. One child (Sulzby, 1985a) pointed to print left to right, but bottom to top frequently in her reading; when she neglected pointing for stretches, she often added more speech and began to point even after she had covered the semantic content of the story text that had been read to her. Other children sound out text to nonsense words, seeming to ignore the semantic content. Others focus on known words by appearing to search through the text and just list words that sound like a random list ("a, and, the, a, a, the, the, the, too many the's, grandma!") Sulzby (1994) suggested this is an important level of development during which the child is practicing particular aspects of reading, performances which will be integrated in the next levels of reading development.
The final two categories describe print governed performances during which the child is becoming able to read "holistically" from print (Sulzby, 1985a). In both of these levels, the child is tracking the print by pointing or showing evidence of accurate tracking in reproducing a spoken word for most or all printed word units. Most children at this level point, but a close observation of the eyes and record of miscues and self-corrections help verify that it is the text that is being read. The first of these holistic levels is called reading with strategies imbalanced. This type of reading is different from the aspectual level because during reading with strategies imbalanced the child is able to use a variety of processes to track and decode text. However, at this level the strategies are not completely integrated or flexible. The child may sporadically use "known words" to substitute for unknown words, leave out words, or resort periodically to remembered text versus written text. The final level, reading conventionally, describes reading performances with an increased level of self-regulation and flexibility. When children at this level make a mistake, they more adequately self-correct the errors. Technically, this level is defined by evidence of flexible and coordinated use of the three aspects of reading shown in the two previous levels. Additionally children at this final level more frequently make comments that indicate awareness of the meaning of the text.
Children who begin to read conventionally may first piece together the aspects of reading with long and complex familiar texts; many of these children soon revert to reading very short and simple books that they had lost interest in previously but now again find interesting because these simple texts are easy to process conventionally. Many children spend much time appearing to "practice" this new skill, particularly orally. Others seem to develop fluency quickly and remain with complex texts. Some do both. Children also begin to show interest and pride in reading new or unfamiliar texts and talking about texts that they read.
It is at this final conventional level of reading performance that most traditional measures of reading achievement begin. However, the SLP or teacher with an emergent literacy perspective acknowledges the complex patterns of oral and written language understanding that are forerunners of conventional reading. Thus, the communicatively impaired child's mastery of yet another language context is an important milestone in language development.
In our most current research (Kaderavek & Sulzby, 1994) we examine the issue of oral-written language contrasts as it occurs with both typically developing and language delayed 2-, 3-, and 4-year-old white children from MSES homes near a middle-sized Midwestern city. The data collected from this research permits examination of oral-written language from two different perspectives. First, in order to examine interactive language patterns, we contrast parent-child interactive language as it occurs in a more oral-language situation (toy play) and a more written-language-like situation (storybook reading). Second, in order to examine children's independent language productions, we contrast children's productions of storytelling narratives (a more oral-language task) with children's emergent readings of a favorite storybook (a more written-language task). We will first discuss our observations of parent-child dyadic interactions in storybook reading and toy play contexts.
As previously mentioned, even young children respond to variations in context and alter their language and behaviors accordingly. For example, Sulzby (1986) demonstrated that 5- and 6-year-old typically developing children's language performances were different in response to more oral-like tasks (responding to a request to tell a story) versus more written-like tasks (responding to a request to dictate a story for an adult to record). Others have demonstrated differences in the interaction and language performance of the participants when comparing father-child versus mother-child conversational dyads of SLI children (Conti-Ramsden, Hutcheson, & Grove, 1995).
Our research compares toy play and storybook reading interactions for a group of typically developing and SLI children. In our preliminary analyses we have been particularly interested in comparing SLI children who enjoyed storybook reading with other SLI children who were unwilling participants in book interactions with their parents. Such comparisons may be particularly helpful in understanding the difficulty some language delayed children demonstrate in later reading achievement. To illustrate some of the variation we have noted in this regard, we present data from two language impaired preschoolers, John and Andrew. Both boys were observed as part of a larger research study over a 3-week period in 12 mother-child in-home interactions. Each child was videotaped interacting repeatedly with a storybook and toy. Although both John and Andrew showed evidence of enjoying their interaction with the toy (a city scene containing small figures of people, cars, pretend mail, and so on), John was observed to enjoy the book Sam Vole and his Brothers (Waddell, 1992), whereas Andrew showed some resistance to it.
All data presented here were obtained during the third mother-child interaction with the city scene or Sam Vole storybook. A 4-minute sample was taken from each interaction, with 1 minute taken from the beginning of the interaction, 2 minutes from the middle of each interaction and another 1-minute sample from the end of the interaction. In general, storybook interactions lasted 8-10 minutes in length while toy play interactions lasted for the full 30 minutes allotted. To keep measures such as mean length of utterance from being confounded by the book's text, the language data used for the storybook analysis used only the mother's interactive speech.
John, age 3 years 1 month, had been diagnosed as language impaired before being recruited into this study. He had received speech-language therapy prior to his participation in the study, but was not receiving therapy at the time of his in-home observations. In our research assessment protocol, John received a receptive language standard score of 91 (27th percentile rank) which placed his receptive performance within normal limits and an expressive language standard score of 65 (1st percentile rank) which indicated a severe expressive disorder as measured by the Clinical Evaluation of Language Function-Preschool Test. Consistent with a diagnosis of SLI, John scored within normal limits in cognitive performance using a nonverbal intelligence test and had normal hearing. John's phonological development was within normal limits and he was approximately 90% intelligible.
John's spontaneous language was pragmatically and semantically appropriate. His expressive language delay was characterized by immature syntax use and reduced sentence length. These aspects of his language delay are demonstrated in this excerpt (Example Three) from John's third toy play interaction with the city scene:
The data for John and his mother during toy play are summarized in Table 1. As the data demonstrate, both mother and child participated fairly equally in the toy play context, with John producing about 47% of the total utterances and his mother producing 53%. Forty-two percent of John's mother's conversation consisted of questions. Many of these questions were used to clarify the ongoing play. Some examples included, "Is this a girl baby or a boy baby?", "Is the fireman her Dad?", and "Where does that go?"
The balance of interaction between John and his mother was similar in the storybook and the toy play interaction, with John producing about 44% of the total utterances and his mother producing 56%. John's mean length of utterance (MLU) during the storybook reading stayed approximately the same as his utterance length during toy play, as did his mother's. John's type/token ratio (TTR) (a ratio of novel words compared to the total number of words produced) stayed within the same range during both interactions. John's mother's TTR was reduced in book reading, perhaps as a means of scaffolding the book's text to John's level. John's mother produced approximately the same number of questions as she did during toy play, 49%. Whereas the majority of her questions during toy play were mostly direct (requiring an answer), now her questions were more evenly divided between direct and rhetorical (not requiring an answer) types. Some examples of rhetorical questions included, "You think?" and "Is there?". Direct questions included, "Do you think he's scared?", "What are they all doing?", and "Who did he see?". Example Four represents an excerpt from John's storybook interaction.
We now turn to our observation of a second language impaired preschooler, Andrew, age 3 years 6 months. Andrew was enrolled in a special preschool for language delayed preschoolers through his local school district at the time of his participation in our research. Andrew received a composite receptive language standard score of 84 (14th percentile rank), which indicated his receptive language abilities were mildly impaired, and a composite expressive language standard score of 50 (below the 1st percentile rank) which indicated a severe expressive disorder as measured by the Clinical Evaluation of Language Function-Preschool Test. As was required for eligibility in our study, Andrew scored within normal limits in cognitive performance using a nonverbal intelligence text and passed a hearing screening test. Andrew had a phonological deficit in addition to his language impairment. He was 40-50% intelligible to a novel listener, but the investigator was able to understand 80-90% of Andrew's speech after observing him throughout the 3-week period of in-home visits. Andrew was also noted to have a higher-than-typical number of disfluencies characterized by part-word repetitions. No struggle or avoidance behaviors were noted in conjunction with these disfluent behaviors.
As indicated by his MLU in the toy play and storybook interactions, Andrew usually communicated using 1-3 word utterances. He showed deficiency in his use of verb forms, using only the unmarked verbs "do", "get", "go" and "turn", during his 4-minute sample of toy play interaction and was noted to omit plurals and articles. However, during toy play Andrew took an active part in the conversational interaction, producing 54% of the utterances. In fact, he assumed a leadership role within the toy play context in that he would often direct his mother's activity and disapprove of certain play sequences she would attempt. These aspects of his interactive style can be seen in the Example Five:
The data for Andrew and his mother in the toy play context are summarized in Table 2. Andrew's mother produced about 31% questions during the toy play interaction, the majority of which were rhetorical in nature. Some examples of her rhetorical questions included, "Should we put it away?" and "Are you sure she can fit?". Some direct questions included, "Where did that baby go now?" and "What do you want me to do with this?"
Whereas Andrew was eager to interact with the toys during his in-home observations, he did not appear to enjoy storybook reading. Andrew's overt behavior at these times consisted of verbalizing loud sighs, slumping down on the couch (sometimes laying down with his head away from his mother, or even sliding off the couch), and once during the book reading interaction analyzed here, putting his stockinged feet on the book obscuring the page.
We should note that both John and Andrew were observed interacting with an additional toy and book as part of the larger research project. This second set of items was selected by each subject's mother to make sure that we presented high-interest books and toys. John's and Andrew's reactions were the same in this second set of interactions, demonstrating that these data were not idiosyncratic to a particular book or toy item.
Example Six is illustrative of Andrew and his mother during a book reading interaction. Rather than selecting an excerpt in which Andrew did not respond at all to his mother's questions (which occurred frequently), we chose a segment during which he was most responsive. Even so, Andrew's interaction was quite different than that which occurred during toy play.
The balance of interaction between Andrew and his mother in the storybook interaction was quite different from the toy play interaction. Andrew's mother's percentage of utterances rose to 82%, while Andrew's dropped to 18%. Another difference was Andrew's mother's fairly significant increase in question production to 55% as compared with 31% during toy play. She used many more direct questions than she had during toy play. Some examples of the direct questions used by Andrew's mother in the book interaction include, "What's another picture you like?", "What kind are those?", and "Why does he do that?"
Andrew's MLU during book reading stayed approximately the same as was produced in toy play, but his mother's MLU rose from 4.0 in the toy play interaction to 5.347 in the book reading interaction. Andrew's use of a greater variety of words increased as measured by his TTR in the storybook reading. His TTR rose to .515 from a TTR of .396 in toy play. However since he produced only half as many utterances the smaller sample size during the book reading may have inflated this ratio. Like John's mother, Andrew's mother's TTR was reduced during the book reading interaction, suggesting an attempt to scaffold the book's text to Andrew's language level.
The overarching observation is that John's and his mother's interactive style stayed approximately the same between contexts (as measured by number of utterances produced by John and his mother and number of mother's questions) whereas these variables changed in Andrew's case (observed loss of balanced interaction and increase in mother's use of questions). Further, the changes occurring for Andrew and his mother during the storybook interaction did not appear to facilitate a positive response. Although these results cannot yet be generalized to the larger language impaired population, our case-by-case examination of these data has made us attentive to the difficulties that emerge for language delayed children when strong contextual shifts occur in parental expectations and language demands. It appears that, for some SLI children, the use of parental practices in relation to the child's development makes the transition from a more contextualized interaction (toy play) to a less contextualized interaction (storybook reading) unrewarding and sometimes even unpleasant.
These data suggest that we must exercise caution when we make recommendations to parents intended to facilitate mutual engagement in storybook reading. We are concerned about the potential dangers of certain "prescriptive" approaches to educating parents about storybook reading practices. We are particularly concerned about approaches that would make one set of narrowly defined recommendations for all children.
For example, some researchers in the area of emergent literacy have promoted a method of interactive reading in which parents are trained to use specific questioning strategies with their preschool children (DeBaryshe, 1992; Whitehurst et al., 1994). As part of this training, parents are asked to initially focus on asking "what"questions (e.g., "What is this?"), "attribute" questions (e.g., "What color is this?) and completion prompts. These are seen as being at a level of "low demand." With older children, parents are asked to increase the level of demand by asking more "distancing" questions, (e.g., "how" or "why" questions). Parents are asked to use approximately two to three of these strategies on each page of a storybook.
DeBaryshe (1992) studied 73 low income children (ages 26 to 60 months) and their parents. Seventy-eight percent were African American and the remainder were white. The 25% attrition rate for children and parents who began the study was quite high. These 18 children were noted to be significantly different from children who remained in the intervention study. They had significantly lower scores on the Illinois Test of Psycholinguistic Abilities and were rated by their mothers as not enjoying being read to and not asking for books to be read to them.
This dropout rate could be due to reactions similar to those we found with SLI children who disliked being read to and who responded negatively to their mothers when they used the kinds of questions or the high frequency of questioning that are part of the DeBaryshe (1992) and Whitehurst et al. (1994) procedures. For example, recommending this strategy to Andrew's mother might have encouraged her to try even harder to engage Andrew through the use of questions--an avenue unlikely to bring about Andrew's increased participation. It is premature to speculate on specific treatment recommendations in cases such as these. However, we can begin the process of understanding how to help children like Andrew by using some basic principles of emergent literacy practice in clinical settings.
How do these data influence SLPs in regard to the clinical application of emergent literacy perspectives? Most importantly, these data suggest that there are important reasons to include observation of parent-child book reading interactions into our ongoing work with communicatively impaired children. Speech-language pathologists have always carefully considered the communication dynamics between the language impaired child and his or her primary caretaker when designing an intervention program. A toy play interaction in which the parent is asked to play with his or her child while the SLP observes the interaction is a frequently used and valuable observational situation. We would suggest that observation of a shared book reading can provide key and complementary information. However, as noted in Table 3, in order to get the most information from a book reading interaction we must watch repeated book readings over time. It has been reported that children and parents frequently use more sophisticated language and ask different questions with repeated readings (Goodsitt, Raitan, and Perlmutter, 1988; Martinez & Roser, 1985; Snow & Ninio, 1986; Yaden, 1988 ). As we mentioned previously, repeated readings are one way that parents bring the world of written language to children. Our observations of storybook interactions can enhance our understanding of a parent's ability to interpret the book's text to his or her child and the child's internalization of these language patterns within a routinized language context.
1. Pick a narrative-type storybook and avoid brief pattern-type books. A narrative book is one that has characters, a setting, and a plot. Research has demonstrated that with narratives children have the opportunity to internalize complex patterns and written language constructions, whereas with pattern books children tend to rely on rote memorization.
2. Observe the storybook interaction in the home, or provide a homelike setting in your clinic or school. Literacy learning comes about because of the shared pleasure surrounding the literacy event. It's hard for an adult to enjoy the interaction when they are forced to sit at tables made for young children! Provide a couch or big chair--or even pillows on the floor--to observe the kinds of nonverbal interactions that occur around the storybook reading.
4. Observe more than one reading of the book. Children's concepts of the book will change, and this development will be reflected in their changing comments and questions. Watching repeated readings also allows the SLP to observe the parent's sensitivity to the child's changing levels of understanding.
One of the most significant methods that a parent uses to "negotiate" a book's text to his or her child's level of understanding is through language scaffolding. Scaffolding has been forwarded as an explanation of how parents interact with their children during storybook reading (Ninio, 1983; Snow, 1983; Sulzby & Edwards, 1993; Teale, 1984). Several aspects of scaffolding appear to be particularly important for a child's emergent literacy development. One factor is the parent's ability to encourage the child's verbalizations and to adjust his or her own speech to the child's gradually increasing competence with the written language forms in the storybook. A second factor is the parent's willingness to let the child lead the storybook interaction as soon as he or she is ready to "take charge" of the reading interaction to some extent. In general, many of these scaffolding behaviors appear to change as the child's developmental level and familiarity with the storybook increase (Teale & Sulzby, 1987). Overall, there appears to be a generalized change from a parental labeling of pictures towards a more frequent occurrence of reading the written text without modification.
A word of caution is needed in regard to tracking parents' changing levels of language scaffolding in their book reading interactions. Although we have observed parents generally moving from picture labeling (e.g., with very young children) to straight reading of the text, these behaviors do not occur in a strict hierarchy. A parent who is effective in engaging his or her child moves freely between levels as the child's interest and attention fluctuate. The implication of this finding is that SLPs will not only want to observe storybook interaction over a period of time, but will observe parents' flexible use of a variety of strategies that are responsive to the type of book, the child's understanding of the ideas contained in the book, and the child's willingness to participate in the book reading on a particular day. A list of the scaffolding interactions frequently seen in parent-child storybook interactions (Kaderavek & Sulzby, 1995) are listed in Table 4.
1. Labeling and commenting items in the pictures is one of the behaviors parents' demonstrate with very young children during storybook reading. While using this technique parents often use exaggerated nonverbal communication (gestures, motions and facial reactions) while naming the pictures.
2. Pausing behaviors are used by the parents to provide opportunities for the child to interact during the book reading exchange. We have observed that pausing seems to function as a means of keeping the child involved in the book without having to question directly or to "teach" the child.
3. Parents often using questioning behaviors during book reading. Although questioning can be an effective strategy if not overused, our preliminary data indicate that some language-impaired children do not react favorably to direct questioning during parent-child book reading.
4. Parents often engage in oral dialogue during their storybook reading. Rather than actually reading the book, they will present the storyline in an oral form. Clues to this kind of scaffolding are the use of "voices" to indicate the various characters and the use of oral- versus a written-language intonation characteristics. At this level the parent may also make connections between the story and situations or vocabulary with which the child is familiar.
5. Sometimes the parents begin to actually read the book but simplify the vocabulary and the syntax of the written text. The intonation patterns of the adult reader may begin to sound like written language--but the parent will alter the text to maintain the child's understanding and interest.
6. Finally, parents begin to read the text verbatim, not in long stretches but in attempts staggered between other scaffolding behaviors. Eventually, the child may ask for the parent to read the entire book verbatim, but we have not observed that with our children yet. Parents who read stretches verbatim also often use other internal scaffolding behaviors such as pausing, stopping with rising inflection (to encourage the child to complete the sentence), or stopping to discuss the story when the child appears interested or asks questions.
7. Parents who maintain a high level of the child's interest and engagement appear to be very sensitive to their child's reactions; they show flexible use of these scaffolding strategies in relation to the child's verbal and nonverbal behaviors.
Finally, we find that all book reading need not be "highly interactive" to successfully engage children and expose them to written language. We have observed that, in some cases, children actually prefer to have their parent "just read the book," particularly if they are accustomed to being read to and/or as they grow older. When one observes a book reading that is not highly interactive, one can monitor the verbal ("Read another book, Mommy!") and the nonverbal indicators (e.g., attending, running to get more books) to assess the success of the book reading. Additionally, the SLP can (as we do in our research) intermingle parent-child storybook observation with requests for the child to read the book emergently . The emergent readings provide information which help monitor the child's degree of internalization of the book's language.
Our second area of inquiry into oral-written language differences is the examination of the difference between narrative development in a more oral storytelling context versus the emergent reading task developed by Sulzby (1985a). Narratives are particularly important for observation because it has been suggested that narrative discourse development is significantly related to children's emergent literacy (Dickinson & McCabe, 1991) and to successful adaptation to school literacy (Feagans, 1982). This relationship was demonstrated by Roth and Speckman (1986) who found that older learning-disabled children (who had normal intelligence but had difficulties in reading, written expression and/or math) had difficulty with the construction of oral narratives.
In our research program we elicit two kinds of narratives--an oral storytelling narrative and an emergent reading, which we consider to be a narrative elicited in a written language context. The protocol for eliciting the emergent reading has already been described in this report. To elicit the oral storytelling narrative, we use the protocol suggested by Peterson and McCabe (1983). These authors suggest that in order to get a narrative one needs to tell a narrative. Following this protocol, the adult tells the preschooler several stories (mostly centering on occasions in which he or she was hurt or frightened--topics which appear to elicit the best responses from the children) and encourages them to respond with a story of their own (see McCabe & Rollins, 1994, for a recent discussion of this protocol). At least three narratives are elicited from each child and the best narrative obtained is used for analysis.
To illustrate some of the differences in oral- versus written-language narrative attempts, we present excerpts from narratives produced by Adam, a typically developing child, age 3 years 5 months, who participated in our research. The following (Example Seven) was Adam's oral narrative (The investigator told a story about when she was hurt and got stitches and then asked Adam the following question).
One of the most striking characteristics about this oral narrative was the lack of decontextualization. The narrative was told with motions and incomplete sentences. In fact, this narrative really does not stand alone, but actually appeared to function as a discourse "turn." Overall, it appeared to be a response to the investigator's story about being injured.
According to the narrative assessment protocol (McCabe & Rollins, 1994), Adam's oral narrative above was classified as an end-at-high-point narrative. He did not include the evaluative comments that would reveal the meaning that the event had for him as narrator. Similarly, this narrative did not have any resolution comments that served to wind up the crisis. Although this was not a classic narrative (a classic narrative contains all the features needed for a complete narrative), this kind of end-of-high-point narrative was very good for a child of Adam's age.
Contrast this with the narrative produced by Adam in his emergent reading of the storybook, Sam Vole and His Brothers. While Adam read emergently, he stared intently at the pictures, used a quiet voice for "reading" and a louder voice for his comments directed to the investigator, and commented on each picture as he turned the pages. His emergent reading (Example Eight) follows.
Adam's emergent reading above demonstrated characteristics of both written and oral language. Accordingly, using the Sulzby's Classification Scheme for Emergent Reading of Favorite Storybooks, this emergent reading was classified at the level of "reading and storytelling mixed." Some of the components that demonstrated Adam's switching between the oral (storytelling) and written language (reading) modes in the above text included his varying use of verb tense and his vacillation between clear and unclear introduction of characters.
At times Adam used verbs in a manner consistent with written language. Examples of this include his past tense use of "He singed" and "He jumped." Additionally, he used a stative form, as in "He be quiet," used to describe the characteristics of the protagonist, Sam. In contrast, at other times, Adam used verb forms consistent with oral language use. Examples of this included his use of the present progressive verbs, "He be picking." and "The bugs are eating."
Adam's second vacillation between oral and written forms occurred around his use of varying levels of decontextualization in his character introduction. Adam sometimes used specific and at other times unspecific introductions of the characters in the story. For example, towards the end of the story he stated, "And he saw his brothers!", a decontextualized form that is characteristic of written language use. (A less decontextualized form would be "And he saw them!"). However, an example of a more oral contextualized language form occurred in the initial sentences of the written narrative when Adam stated, "One day we're picking some flowers. And one day he was break some eggs."
It is enlightening to compare Adam's oral (Example Seven) and written (Example Eight) narratives. Several differences are obvious. The first is the difference in length, with the written narrative being quite a bit longer. Second, one can notice the variety of verbs that were used in the written narrative in contrast to the verbs "fall, move, and got" which occur in the oral narrative. Less obvious, but significant, was Adam's differing decontextualization in the written versus the oral narrative. In fact, if we had not elicited both types of narratives we would have been unaware of his ability to decontextualize. His oral narrative, "And I fall down like..." did not reflect his ability to tell a story in a way that permitted a nonpresent listener to understand the gist of what was happening in the narrative.
Finally, it was interesting to note that Adam completed his emergent reading almost completely independently. Other than the few acknowledgements that the investigator made when addressed directly (e.g., "Remember that? That's a cricket."), Adam was able to complete this task alone. This is especially significant when we contrast this emergent reading to that of an older language delayed preschooler who elicited much assistance in his emergent reading (see case history of "Timmy" in this chapter).
Adam still has some areas that will develop as he moves towards higher levels such as reading-similar-to-original and a verbatim-like story. Most significantly, his ability to decontextualize the story (e.g., introducing the characters first by name and then using pronouns when referring to them) and his ability to recreate longer sentences (that are similar to the written text) will improve. His ability to produce the written narrative that he did at the age of 3 1/2, however, was quite good.
The differences in Adam's oral and written narratives has implications for remediation of children with limited ability to produce narratives. It may be that children are able to explore more of the complexities of narrative structure when retelling a story that has been read to them as compared to the demands of creating a novel story in an oral storytelling mode. Future research data and ongoing input from SLPs who use emergent reading activities in their clinical practice will shed further light on the usefulness of emergent reading as a tool to develop increasingly spontaneous narrative productions. However, we feel that even this preliminary data indicated that eliciting children's narratives in response to written as well as oral contexts offered a more complete look at the children's understanding of language and their ability to vary language in response to contextual demands.
This research--along with work being done by other emergent literacy researchers and results of the clinical applications reported by SLPs--will continue to highlight the particular language forms and discourse patterns occurring in relationship to varying oral and written contexts. Overall, our preliminary data suggest that, even at the early ages of language development, children and their parents have clearly differentiated the demands of written and oral language contexts. This kind of examination will provide SLPs the opportunity to explore and observe the developing language of young children in even broader arenas.
This report provides an overview of emergent literacy issues that relate to the language development of preschool children. Several themes underlie this discussion. First, we illustrated that language performances in young nonconventional readers are not tied to a particular delivery form--preschool children can demonstrate their understandings of written language concepts orally. We believe that when parents, teachers, or SLPs use storybooks as stimuli for language production (and much of the time even when storybooks are not present), children are being asked to illustrate their understandings of written language concepts. This position suggests that those of us who concentrate on oral language development also need to be aware of children's emergent literacy development. Consequently, we highlighted some tasks that enable observers of children's language to more clearly understand and describe children's comprehension and use of written language forms.
Second, we described a process that can be used to elicit emergent readings. We provided guidelines for classifying these productions, and highlighted some specific language and behaviors important in the process of internalizing written language and becoming conventionally literate. This information can be used to track the development of children through their exposure to repeated storybook readings.
Third, we described parent and child behaviors that occur in both storybook interactions and emergent readings. This information can be useful to teachers and SLPs in their remediation of children with language impairments and in their efforts to understand how storybook language and behavior might be different than that which occurs during toy play. In other words, we have attempted to answer the question, "What are we looking for when we observe children and parents interacting with storybooks?" To aid this process, we provided guidelines for setting up parent-child storybook observations and a description of parental scaffolding behaviors that may occur during storybook reading.
Finally, we examined children's narrative language productions during emergent storybook elicitations and oral storytelling. Obtaining narratives in a variety of oral and written contexts helps us more carefully monitor children's narrative-production development, their awareness of the need for decontextualization in written language, and their ability to use linguistic features suited to oral and written language contexts. Understanding and documenting these changes broadens our perspective of the language capabilities of the children we serve.
In conclusion, we feel that one of the most important reasons to include emergent literacy perspectives and methods into clinical practice is that they provide a means for the SLP to participate in one of the central features of school instruction: literacy. The SLP's involvement aligns his or her knowledge of the child's oral communication development with understandings of the child's overall language development. If SLPs have a better understanding of a child's oral and written development, we will set up situations which facilitate that child's ability to move between modalities. The resulting interactivity across language forms would be designed to help the child continue to negotiate relationships between oral and written language, which is important for later school success.
Timmy is a 4 year 1 month old who has been diagnosed with specific expressive language impairment (SLI). He is from a European American, middle-economic level home. Timmy, his parents, his two sisters (age 13 and 2) and his young brother (age 2 months) live in a two-story older home in a small town near a mid-sized Midwest city.
There is no known cause of Timmy's language delay. He was born following an uneventful pregnancy and delivery. Other than his expressive language delay and some articulation delay, Timmy has been typically developing. At 4 years 1 month of age, he performed within normal ranges on a nonverbal intelligence test, and his receptive language abilities tested within the average range. His hearing was also normal. Timmy's utterances ranged from 1 to 6 words in length. He obtained a standard score of 3 on the Recalling Sentences in Context Subtest of the Clinical Evaluation of Language Fundamentals-Preschool Test (CELF-P) and a standard score of 5 on the Word Structure Subtest of the CELF-P. Timmy was within normal limits on all other subtests of the CELF-P.
Timmy's mother first noticed his language delay between 12 and 18 months. She discussed her concerns with her pediatrician, but was told to "wait and see what happens." By age 2 1/2, she had him evaluated at the speech and hearing clinic at a local university and went for one semester of therapy. She stated that she was taught to model language for Timmy and decided she would read to him every day "no matter what."
Timmy's mother described how she began an accelerated program of language stimulation, in consultation with the SLP. In the last year and a half she has "talked to him one-on-one all the time," along with daily storybook reading. She said that she feels that if Timmy is interested in a topic she will try to expose him to whatever they are discussing because he "needs to go out and actively see it."
In describing why she feels storybooks have been particularly important for Timmy, Timmy's mother stated, ". . . reading brings so many new words into light. Things you normally wouldn't talk about." She indicated that usually she and her son will talk about their most current storybook on and off throughout the day.
Along with his mother's efforts at home, Timmy has benefited from enrollment in a county preschool program for children with developmental delays. His mother indicated that the stimulation of the other children has been particularly instrumental in facilitating Timmy's desire to communicate.
As part of the authors' ongoing research Timmy was observed over a period of a month during storybook reading and toy play interaction and was asked to produce emergent readings (which we consider to be an opportunity to produce a written narrative) and oral storytelling narratives. The following is an example of Timmy's oral narrative. This story was produced in response to the investigator telling him a story of a camping trip she had recently completed.
(The investigator was assisted in interpreting this story by Timmy's mother. Timmy told about the day that his father took Timmy and his sister (Mayway) to the park. They played. It got dark, and they went home.)
(Please note that the examiner's repetition of Timmy's utterances is not necessarily a recommended practice. The utterances were repeated because Timmy was fairly unintelligible and the investigator was attempting to assist the transcription process that would occur later.)
What is most noticeable about Timmy's emergent reading was that although he struggled to communicate his ideas (and asked for the investigator's help), he demonstrated a great deal of written language knowledge. First, he was insistent on saying the names of the characters, indicating knowledge of written language's need for decontextualization, even though he had a great deal of difficulty remembering the names. Midway through the story, when he said "All sleeping" he continued to demonstrate that he should be specific and say "Sam not sleep", instead of "But he not" or "Him not." Additionally, he may also have demonstrated some metalinguistic knowledge at the beginning of his emergent reading when he said "Story all himself," perhaps meaning, "In this story, Sam wanted to go by himself."
When one compares Timmy's oral narrative to his written narrative, one can see differences in length, with the written narrative being quite a bit longer. Although the written narrative is not syntactically more complex,it does appear that his syntax is somewhat more well-formed in the emergent reading than in the oral narrative.
We are in the process of studying multiple levels of impact of storybook reading interactions for SLI and normally-developing children. Initially from these transcripts, one can infer that Timmy's motivation to talk about the stories he enjoys is one avenue for expanding and enriching his language world. The interactions in these two situations provide at least part of what is often called a literacy and language rich environment, the kind of environment which SLPs increasingly recommend.
2. Known to lay audiences as "real reading" and "real writing," conventional literacy encompasses those reading and writing behaviors that the members of a culture have implicitly or explicitly agreed upon as denoting traditional reading and writing. Conventiojnal literacy is often used to indicate the first signs that a child has made a transition from emergent to traditional literacy. See Sulzby (1996a, pp. 25-28, or 1989) for a detailed exploration of operationalizations of the terms conventional reading and conventional writing.
Anderson, A. B., & Stokes, S. (1984). Social and institutional influences on the development and practice of literacy. In H. Goelman, A. Oberg, & F. Smith (Eds.), Awakening to literacy (pp. 158-184). Portsmouth, NH: Heinemann.
Bus, A. G., & Sulzby, E. (1996). Becoming literate in a multicultural society. To appear in J. Shimron (Ed.), Literacy and education: Essays in memory of Dina Feitelson. Norwood, NJ: Hampton Press, Inc.
Bus, A. G., van IJzendoorn, M. H., & Pellegrini, A. D. (1995). Storybook reading makes for success in learning to read: A meta-analysis on the intergenerational transmission of literacy. Review of Educational Research, 65, 1-21.
Conti-Ramsden, G., Hutcheson, G. D., & Grove, J. (1995). Contingency and breakdown: Children with SLI and their conversations with mothers and fathers. Journal of Speech and Hearing Research, 28, 1290-1302.
Cox, B. E. (1994). Young children's regulatory talk: Evidence of emerging metacognitive control over literary products and processes. In R. B. Ruddell, M. R. Ruddell, & H. Singer (Eds.), Theoretical models and processes of reading (4th ed., pp. 733-756). Newark, DE: International Reading Association.
DeBaryshe, B. D. (1992). Early language and literacy activities in the home. U.S. Department of Education Field Initiated Studies Program. Final report for the project. North Carolina University, Greensboro. (ERIC Document Reproduction Service No. ED 351 406)
Elijah, R., & Sulzby, E. (1992, April). From emergent reading to conventional literacy: LSES black children's storybook reading. Paper presented at the annual meeting of the American Educational Research Association, San Francisco, CA.
Feagans, L. (1982). The development and importance of narratives for school adaptation. In L. Feagans & D. Farran (Eds.), The language of children reared in poverty (pp. 95-116). New York: Academic Press.
Goodsitt, J., Raitan, J. G., & Perlmutter, M. (1988). Interaction between mothers and preschool children when reading a novel and familiar book. International Journal of Behavioral Development, 11, 489-505.
Heath, S. (1986). What no bedtime story means: Narrative skills at home and school. In B. Schieffelin & E. Ochs (Eds.), Language socialization across cultures (pp. 97-124). New York: Cambridge University Press.
Hiebert, E. H. (1994). Becoming literate through authentic tasks: Evidence and adaptations. In R. B. Ruddell, M. R. Ruddell, & H. Singer (Eds.), Theoretical models and processes of reading (pp. 391-413). Newark, DE: International Reading Association.
Koppenhaver, D. A., Coleman, P. P., Kalman, S. L., & Yoder, D. E. (1991). The implications of emergent literacy research for children with developmental disabilities. American Journal of Speech-Language Pathology, 1, 38-44.
Moll, L. C. (1994). Literacy research in community and classrooms: A sociocultural approach. In R. B. Ruddell, M. R. Ruddell, & H. Singer. (Eds.), Theoretical models and processes of reading (pp. 179-207). Newark, DE: International Reading Association.
Pellegrini, A. D., McGillicuddy-DeLisi, A. V., Sigel, I. E., & Brody, G. H. (1986). The effects of children's communicative status and task on parent's teaching strategies. Contemporary Educational Psychology, 11, 240-252.
Pierce, P. L., & McWilliam, P. J. (1993). Emerging literacy and children with severe speech and physical impairments (SSPI): Issues and possible intervention strategies. Topics in Language Disorders, 13, 47-57.
Snow, C. E., & Ninio, A. (1986). The contracts of literacy: What children learn from learning to read books. In W. H. Teale & E. Sulzby (Eds.), Emergent literacy: Writing and Reading (pp. 116-138). Norwood, NJ: Ablex.
Sulzby, E. (1986). Writing and reading: Signs of oral and written language organization in the young child. In W. H. Teale & E. Sulzby (Eds.), Emergent literacy: Writing and reading (pp. 50-89). Norwood, NJ: Ablex.
Sulzby, E. (1989). Assessment of emergent writing and children's language while writing. In L. Morrow & J. Smith (Eds.), The role of assessment in early literacy instruction (pp. 83-109). Englewood Cliffs, NJ: Prentice-Hall.
Sulzby, E. (1994). Children's emergent reading of favorite storybooks: A developmental study. In R. B. Ruddell, M. R. Ruddell, & H. Singer (Eds.), Theoretical models and processes of reading. (4th ed., pp. 244-280). Newark, DE: International Reading Association.
Sulzby, E. (1996a). Roles of oral and written language as children approach conventional literacy. In C. Pontecorvo, M. Orsolini, B. Burge, & L. Resnick (Eds.), Early text construction in children (pp. 25-46). Hillsdale, NJ: Erlbaum.
Sulzby, E., & Barnhart, J. (1992). The development of academic competence: All our children emerge as writers and readers. In J. W. Irwin & M. A. Doyle (Eds.), Reading/writing connections: Learning from research (pp. 120-144). Newark, DE: International Reading Association.
Sulzby, E., & Edwards, P. A. (1993). The role of parents in supporting literacy development. In B. Spodek & O. N. Saracho (Eds.), Language and literacy in early childhood education (pp. 156-177). New York: Teacher's College Press.
Sulzby, E., & Teale, W. H. (1987, November). Young children's storybook reading: Longitudinal study of parent-child interaction and children's independent functioning. Final report to The Spencer Foundation, University of Michigan, Ann Arbor.
Sulzby, E., Teale, W. H., & Kamberelis, G. (1989). Emergent writing in the classroom: Home and school connections. In D. Strickland & L. Morrow (Eds.), Emerging literacy: Young children learn to read and write (pp. 63-79). Newark, DE: International Reading Association.
Tannen, D. (1983). Oral and literate strategies in spoken and written discourse. In R. W. Bailey & R. M. Fosheim (Eds.), Literacy for life: The demand for reading and writing (pp. 79-96). New York: Modern Language Association.
Teale, W. H., & Sulzby, E. (1986). Emergent literacy as a perspective for examining how young children become readers and writers. In W. H. Teale & E. Sulzby (Eds.), Emergent literacy: Writing and reading (pp.vii-xxv). Norwood, NJ: Ablex Publishing.
Teale, W. H., & Sulzby, E. (1987). Literacy acquisition in early childhood: The roles of access and mediation in storybook reading. In D. Wagner (Ed.), The future of literacy in a changing world (pp. 111-130). New York: Pergamon Press.
Teale, W. H., & Sulzby, E. (1989). Emergent literacy: New perspectives on young children's reading and writing development. In D. Strickland & L. Morrow (Eds.), Emerging literacy: Young children learn to read and write (pp. 1-15). Newark, DE: International Reading Association.
Whitehurst, G. J., Arnold,D. S., Epstein, J. N, Angell, A. L., Smith, M., & Fischel, J. E. (1994). A picture book reading intervention in day care and home for children from low-income families. Developmental Psychology, 30, 679-689.
Whitehurst, G. J., Falco, F. L., Lonigan, C. J., Fischel, J. E., DeBaryshe, B. D., Valdez-Menchaca, M. C., & Caulfield, M. (1988). Accelerating language development through picture book reading. Developmental Psychology, 24, 552-559.