Vocabulary and Beyond: Developing Language for School Achievement
In the first panel session, participants discussed research on how vocabulary develops; how vocabulary relates to other aspects of language, such as grammar; the effects of vocabulary on achievement and achievement gaps; and ways to develop vocabulary to reduce achievement gaps. The discussion went beyond vocabulary to explore research on academic language, focusing on current definitions of the concept and promising instructional practices for developing both language and knowledge in the disciplines students are expected to master in school.
VOCABULARY AND SOCIOECONOMIC STATUS
Erika Hoff summarized findings from her review of research on vocabulary conducted from a developmental and psychological perspective. She focused on one finding related to achievement gaps that is very well documented, that socioeconomic status (SES) predicts how well children will do in school. In this context, Hoff said, the evidence available is quite convincing that vocabulary helps to explain the observed achievement gap between children from lower and higher SES families, though at this point the evidence is “circumstantial” or indirect (see Hoff, 2009).
First, children with lower SES have lower school achievement. Second, children with smaller vocabularies have lower school achievement. Third, children with lower SES have smaller vocabularies. One interpretation of this evidence is that vocabulary differences help to explain at least some portion of SES-related differences in school achievement. Moreover, Hoff
speculated, since being a minority and English-language learner in the United States is confounded with SES, vocabulary probably also affects achievement gaps for those groups to some degree.
Another well-established finding is that vocabulary predicts how well all children will do in school, regardless of SES: after controlling statistically for SES and SES-related factors (such as nutrition), vocabulary still matters for school achievement. According to Hoff, this finding points to the vital importance of vocabulary because, theoretically, if SES and those other SES-related factors measured in those studies were somehow magically fixed, children with larger vocabularies would still perform better in school. Yet, Hoff was not aware of any studies that have included the analyses needed to test more directly whether vocabulary helps to explain the correlation between SES and achievement. Nor is there evidence about how much of the difference in achievement between lower and higher SES students can be attributed to vocabulary.
Experimental research also has not been conducted to test whether the SES-related achievement gap could be narrowed by increasing vocabulary. Nor has there been any research on how much the gap would be narrowed by addressing vocabulary directly—rather than other known SES-related influences on achievement, such as nutrition.
The literature does suggest, however, some sources of SES-related differences in vocabulary that might have implications for intervention, Hoff said. One possible focus for intervention is the amount of language input experienced by children of different socioeconomic backgrounds. This claim has sometimes been subject to controversy but it is derived from several findings.1
First, studies show that vocabulary size and growth are associated with the amount and complexity of language children hear in their everyday lives. Children from lower SES homes who, on average, have smaller vocabularies hear less language and less complex language at home than children from more economically advantaged backgrounds (Hart and Risley, 1995; Hoff, 2003, 2006, 2009). If children hear two languages at home, they have larger vocabularies in the language they hear most, suggesting again the importance of amount of language input to language learning.
VOCABULARY AS A PROCESS
Given what is known about how new words are learned, it is impossible, Hoff explained, to consider vocabulary separately from other aspects of language, and this fact has implications for intervention. As described in Hoff (2009), several aspects of linguistic knowledge converge to support learning a word. The ease with which new words are learned depends on being familiar with the phonology (sounds) of the word in the language system that the word belongs to; knowing the grammar of the language, which can be used to help interpret the meaning of new words in a sentence; and knowing the concept to which a new word refers. As a result of this research showing on the connectedness between vocabulary and the rest of a language system, Hoff infers that interventions that teach words in isolation from other aspects of a language are not likely to work very well. However, more research is needed to determine exactly what language interventions are needed, at what ages, and for how long to boost language sufficiently for narrowing achievement gaps.
Advances could be made, Hoff proposed, if researchers conceptualized vocabulary more as a dynamic process than a static skill and studied the cognitive processing aspects of vocabulary development. For instance, Anne Fernald and her colleagues have shown that 20- and 30-month-old children with more speech addressed to them not only had larger vocabularies, but also accessed the meaning of words from memory more quickly (Fernald, Perfors, and Marchman, 2006; Hurtado, Marchman, and Fernald, 2008). Likewise, the research of Cindy Fisher and colleagues with older children has shown that children retrieve a word faster if they hear it more often. Hearing a word has a long-lasting priming effect, such that a word heard within even the last week is more rapidly accessed than a word not heard during that time frame (Fisher, Church, and Chambers, 2004).
Since retrieving words and the meaning of words with little effort are both important for engaging in academic learning in the classroom, Hoff suggested that children who use larger vocabularies in school may not only know more words, but also retrieve those words and their meanings faster for both language comprehension and production. Thus, if teachers were to talk more often with students using new words relating to academic subjects and also encourage students to use those words, the words would be easier for children to remember. This process would be especially important for building vocabulary not typically used in children’s homes.
Most research on the development of vocabulary and oral language focuses on whether the vocabulary young children initially bring to school puts them on a trajectory of vocabulary learning and academic achievement. Research shows that intervening at early ages is important for early school achievement, but it is likely not to be sufficient, Hoff said. Vocabu-
lary continues to develop across the life span, and vocabulary learned at ages 10 or 12 may influence school achievement just as much as the vocabulary children bring to school initially. Building new vocabulary and more fluent processing of words and word meanings likely depends on the continuing opportunities children have to hear the specific words they need to be able to retrieve.
FUNCTIONAL USE OF LANGUAGE
Mary Schleppegrell explored other characteristics of “academic language,” the language associated with school. How might familiarity and facility with academic uses of language affect school learning? From the point of view of scholars, to what extent do differences in grammatical and lexical development contribute to the achievement gap observed between economically disadvantaged and advantaged students? What is known about the effectiveness of interventions to help teachers support language development in the classroom? (For a perspective on the controversies surrounding academic language, see Chapter 4; also see Valdés, MacSwan, and Alvarez, 2009.)
Schleppegrell addressed these questions in a review of research grounded in Michael Halliday’s theory of systemic functional linguistics (Halliday, 1985). A main tenet of the functional linguistics approach is that, from infancy, the desire to make and share meaning drives language development and use. The continual development of language depends on having on-going opportunities for social experience and interaction. Linguistic structures and grammatical forms are derived from the communicative meanings that are routinely created and shared in social interaction. This approach can be contrasted with two others: (1) nativist views in linguistics that propose innate and universal knowledge of linguistic structures, the specific application of which is simply triggered by children’s earliest language environment; and (2) cognitive views that propose cognitive constraints that channel attention toward social and other cues that support language learning (see Valdés, MacSwan, and Alvarez, 2009). Another tenet of the functional linguistics approach is that vocabulary and grammar interact to create meaning, hence the field’s use of the term lexico-grammar, which denotes the difficulty in establishing where vocabulary ends and grammar begins in language.
THE LANGUAGE OF SCHOOLING
People decide how to use language for meaning depending on the content and purpose of the communication, the relationship between speaker and listener or between speakers, the mode used for communicating (e.g.,
in person, by text, by e-mail), and their familiarity with the resources the language has to offer. The language of schooling, according to this perspective, has its own forms and functions created for academic purposes, or “academic register.” Register refers to specific features of language that vary according to the context and purpose of a communication. Speakers of all languages and of all dialects are assumed to engage in register variation. One linguistic feature of an academic register, used in academic writing, is “nominalization,” as in the sentence that follows from an 11th-grade history text: The destruction of the buffalo and removal of Native Americans to reservations emptied the land for grazing cattle. This type of structure packs noun phrases into sentences to condense information, and it can present challenges to readers unfamiliar with the writing style.
The academic language used to construe meaning and communicate in school, Schleppegrell argued, is made up of sets of academic linguistic registers that differ from language registers used in everyday conversation. And every academic subject has its own register. Learning academic registers is inseparable from learning school subjects, in her view, as children must master new language forms and functions for academic tasks and purposes. As explained further in Schleppegrell (2009), these tasks include reading and writing reports, articulating arguments, formulating hypotheses, building theories and developing explanations, and using the linguistic and discourse features typical of scientific, technical, and humanities fields.
According to research described by Schleppegrell (2009), teachers can be taught to recognize the linguistic challenges of their academic subjects and to create opportunities for the social interaction that helps to develop the needed language. Fostering teachers’ knowledge about language in the context of disciplinary teaching offers a concrete approach teachers can use to help children learn both language and content. In this approach, teachers learn to talk with students about language and apprentice them into academic uses of language while engaging with the school curriculum.
In ethnographic research, for example, Pauline Gibbons has illustrated how language development for learning can be supported in the context of teaching a unit of 5th-grade science instruction on magnetism (described in Schleppegrell, 2009). Starting with language that is more conversational, children eventually appear to learn not only science concepts, but also the language that is typically used and valued by scientists for communicating about magnetism. As students experiment with magnets, listen to the teacher’s explanations of magnetism, and then give oral and written reports, they move from using here-and-now language or narrative language for recounting personal experience to more presentational, scientifically exact, and decontextualized language for creating explanations of magnetism that do not rely on shared experi-
ence for understanding (Gibbons, reported in Schleppegrell, 2009). Their knowledge becomes less concrete and more abstract and theoretical for conveying information in an academic context. As students learn to speak and write using the language of the discipline, they also become more easily recognized by teachers and others as making progress in learning about science and about magnetism. Gibbons’ work also illustrates that although students bring many different backgrounds and life experiences to the classroom, it is possible to structure opportunities for students and teachers to interact around a shared experience that enables focusing students’ attention on new features of language used in the discipline along with conceptual, academic content.
Morphological or grammatical accuracy is not the main emphasis of this approach, Schleppegrell explained. Rather, a premium is placed on learning to use the language of a discipline so that students can develop content knowledge and perform school tasks, despite possible “infelicities” (errors) in the morphological marking or grammatical structure of their language.
Discussant Nonie Lesaux described her work with Spanish second-language learners. She began by emphasizing the particular challenge and opportunity that written text presents for both language learning and academic learning. School texts function as gate-keepers to academic achievement, according to Lesaux, because the language register used in texts is not accessible to all students. Yet, as she seeks to show in her intervention research (see, e.g., Lesaux et al., 2010b), texts have great potential to foster a language-rich environment that supports the development of language and learning in school.
Before describing the intervention research, Lesaux described findings from three studies to illustrate the challenge of closing the persistent achievement gaps of Spanish second-language learners, who have to both learn to decode text and develop vocabulary and reading comprehension skills.
First, in a longitudinal study, Lesaux followed a cohort of 100 U.S.-born children of immigrant Spanish speakers in grades 4 through 8 who had enrolled in U.S. schools as kindergartners. Significant differences emerged between decoding and comprehension scores obtained with standardized measures of receptive and expressive vocabulary and comprehension. As shown in Figure 2-1, average percentile scores were higher for decoding than for oral and print comprehension measures (Crosson and Lesaux, 2010; Lesaux et al., 2010a). While decoding-based skills matched national norms, vocabulary and comprehension scores
were significantly below national norms at both the 4th and 8th grade. Analysis of the growth trajectories for these children from ages 9 to 15 showed growth in all skills at a rate that is in line with the rate of growth that would be expected for English-only speakers (see Figure 2-2). Yet the Spanish-speakers’ percentile rank did not change over time since they would have needed to experience even more accelerated growth to make up the difference and match English-only norms.
A second study focused on Spanish-speakers’ vocabulary growth in both English and Spanish. Lesaux followed 200 English-only and Spanish-speaking children from the beginning of Head Start until almost the end of 6th grade. As in the longitudinal study, the rate of English vocabulary growth was consistent with what would be expected from English-only norms. The gap had began to close somewhat, but scores continued to trail English-only speakers (see Figure 2-3). In contrast to findings for English vocabulary, Spanish-speaking children lost ground in Spanish vocabulary compared with norms for monolingual Spanish speakers (see Figure 2-3) (Mancilla-Martinez and Lesaux, 2010, in press).
Finally, a cross-sectional study of 581 6th graders struggling with reading comprehension showed that average percentile scores for word reading accuracy (decoding) were 64 for native speakers of English and 57 for language-minority children, respectively (Lesaux and Kieffer, in press). In contrast, reading comprehension scores were much lower for both groups: native speakers of English, 18; language-minority students, 19. Similarly, vocabulary scores showed significant differences: native speakers of English, 33; language-minority students, 22.
In discussing these findings, Lesaux noted that results from the second longitudinal study showing loss of Spanish combined with lower comprehension of English is likely to be influenced by policy and instructional climates for these and other English-language learners in the United States that prevent the use of Spanish in school and that discourage the use of Spanish at home. As a result, Spanish proficiencies do not reach expected norms and cannot be used as a resource for thinking through the meaning of words and concepts to be learned in school. Given this
reality, what interventions might lead to the greatest gains in English oral language and reading comprehension in the shortest time possible?
Taken together, Lesaux said, the three studies illustrate the “daunting” challenge of reducing language and achievement gaps, and she encouraged thinking beyond specialist models and other sporadic approaches to intervention in use today. A universal design approach to instruction in which all children in the classroom receive the same curriculum and instruction for increasing vocabulary would be more feasible, she argued, in the face of rising immigration rates, increasing diversity in the languages spoken (in both rural and urban areas), and demands on classroom teachers to cope with this growing diversity. Lesaux has been studying the effectiveness of such a universal approach to developing language and reading comprehension in middle school classrooms. Academic Language Instruction for All Students (ALIAS) is a text-based curriculum designed as a 20-week intervention to be delivered for 45 minutes each day in all K-12 classrooms with high numbers of English-language learners.
ALIAS was developed by identifying “strong pieces” of curriculum-relevant text containing academic words and phrases likely to present difficulties to students who are struggling in school. The curriculum targets vocabulary by providing multiple opportunities to use words, posting visual resources for learning words, affirming the correct use of words, giving examples of word usage, and supporting students’ writing and facilitating their talk using the words. Morphology activities, such as analyzing parts of words, are a focus for the equivalent of 2 full days. Despite their research showing the importance of morphology awareness to reading comprehension (Kieffer and Lesaux, 2008), morphology instruction typically does not occur in classrooms, and teachers reported that this part of the curriculum was the most challenging for them (see Figure 2-4).
ALIAS has been evaluated using a quasi-experimental design involving 476 students (346 language-minority and 130 native English speakers) in 21 6th grade classrooms assigned to be either a treatment classroom that received the curriculum or a matched control (Lesaux et al., 2010a). Measures of word mastery for the targeted vocabulary showed large and statistically significant effects, though these translated into learning an average of three to five new words. Measures of word analysis using morphological decomposition showed 6 months of extra growth against what would be expected from test norms for the same period. In addition, the Gates-MacGinitie reading comprehension measure revealed 8-9 months of growth beyond expected norms. Lesaux characterized the results as promising, especially in light of more limited gains reported in earlier research. A subsequent randomized trial was conducted in 2008-2009 that
involved 14 middle schools (51 teachers and 2,500 language-minority and native English speakers). Preliminary results appear to replicate the findings from the first study.
Several questions remain about the meaning of some of the gains observed. For instance, what are the real implications of gaining three to five new words for school learning? Might individual differences be masked in reports of average group gains?: this possibility needs to be explored in these data and in other intervention research. What “dosage” of the intervention is required to result in a certain amount of gain? And what factors moderate the effectiveness of the intervention for certain students and under which conditions?
Moving forward, Lesaux asked what might be the best ways to systematically develop the type of explicit knowledge of language and language development that teachers likely would need to implement the ALIAS curriculum and other language-intensive approaches. Capacity building and paradigm shifting would both be required, she argued,
since in typical English-language acquisition instruction today very little instructional time is spent on vocabulary, and the techniques used to teach vocabulary, such as providing a single definition or example of a word, are too superficial for learning in light of research findings on how language develops.
Lesaux stressed that her own work, as well as that of Schleppegrell and others, makes clear that vocabulary is not the only essential focus for language instruction. Yet vocabulary might be a useful gateway for introducing knowledge about language to teachers. Teachers are familiar with vocabulary, and in Lesaux’s study they reported liking the opportunity to help students learn vocabulary while the program actually included much more about language relating to morphology, the construction of sentences, and other linguistic features.
Several possible constraints on the effectiveness of vocabulary and reading comprehension interventions need to be examined and resolved, Lesaux noted. One difficulty lies in finding the most beneficial words to teach. Often teachers use lists associated with a certain curriculum or standardized test. But what words should teachers really target to produce the greatest gains in language and achievement? Another constraint may be the nature of the measures used in research, which may affect the size of intervention effects. The National Reading Panel (National Institute of Child Health and Human Development, 2000), for instance, found little evidence for the effectiveness of vocabulary interventions in studies with commonly used standardized measures of reading comprehension, though a few effects were found with investigator-developed measures.
Increasing the amount of classroom talk beyond what typically occurs could support language growth, but it is not enough for teachers simply to talk more. The quality of teachers’ talk and students’ contributions to classroom conversation need to be considered along with the quantity of talk. According to Lesaux, teachers tend to do the vast majority of talking in classrooms today, but preliminary analyses from her research with colleague Perla Gamez points to a lack of strong relationship between the quantity and syntactic complexity of teachers’ talk in the classroom. The right balance needs to be found, Lesaux said, between teacher modeling of language and providing opportunities for students to participate in classroom talk in order to practice various uses of language for school. Structuring the linguistic environment to maximize learning would require teachers to have knowledge of language, the conditions that foster language development, and how to create these conditions in the context of school.
THE LINGUISTIC ENVIRONMENT
Respondent Aida Walqui said that the reviews by both Hoff (2009) and Schleppegrell (2009) supported that vocabulary is connected to a larger linguistic system and is only one of several aspects of language that need to be developed for academic purposes. In addition to vocabulary size, length and complexity of utterances also predict school achievement (Hoff, 2009), and composite measures of various aspects of children’s oral language skills predict reading achievement better than vocabulary alone (Hoff, 2009). In her view, the findings of both panelists support providing students with rich and legitimate opportunities to hear good academic language models with students participating as valued contributors to discussion. The presentations also support guiding students’ use of language while engaged in learning in the classroom to build specific aspects of language facility that are part of the school curricula.
Drawing on her own experience as an educator, Walqui noted several elements of language described in Schleppegrell (2009) that could be valuable to incorporate into academic instruction as the language encountered in school becomes more complex and comprehension demands increase. These include knowledge of genre (tasks and text types of different disciplines) and register variation (patterns of language that characterize various genres). Also needed is analysis of text to identify features that may affect comprehension, such as the authors’ interpretive perspective, use of logical connections and cohesive reference chains, and other attributes that—if made explicit—might help students understand and interpret the meaning of a text. Also, teachers might arrange opportunities to speak in everyday conversation about academic themes while progressively inviting students to begin using academic registers for talking and writing about academic material. Walqui urged providing teachers with better analysis and understanding of the density and abstraction that appears in academic texts so that students’ facility with comprehending these texts can be better supported.
But systemic changes in schools and teacher preparation would be required, Walqui argued, to make instructional practice consistent with findings in the research literature. In the United States, Walqui reported, 57 percent of English-language learners in middle schools and high schools are second- or third-generation immigrants who entered school as English-language learners in kindergarten but continue to be unable to perform complex academic tasks and to underperform in comparison with the native English-speaking population (Batalova, Fix, and Murray, 2007). As in the United States, first-generation language learners in Australia and Canada perform below the native-speaking population, according to data from the Program for International Student Assessment (Organisation for Economic Co-operation and Development, 2005). However, in
contrast to the United States, second-generation students in those countries begin to outperform native-speaking populations academically. In her view, one possible explanation for what appears to be greater progress for second-generation students in Australia could be that teacher education and professional development is grounded in a widespread approach to teaching language through text developed by Derewianka (1990) that focuses on developing academic language in a manner consistent with Schleppegrell (2009).
From her perspective as someone who offers professional development for teachers, Walqui concluded with several suggestions for translational research—studies to move basic research findings into practical applications, such as approaches to intervention, instruction, and assessment. First, teachers talk for most of the class time, leaving students little time to interact with one another around the academic material. Since this pedagogical approach has been ingrained in teachers’ practice, teachers would benefit from education and support for planning activities that are structured to help students learn language as part of learning academic content.
Second, teachers often lack a deep knowledge of subject matter, including awareness of how language is used to construe knowledge and communicate in the disciplines they teach. Those who educate teachers also lack such awareness, as well as how to offer this knowledge about language uses in disciplines and supporting pedagogies to classroom teachers.
Third, research could help to create a coherent system of on-going professional development opportunities to help teachers support language growth across the curriculum. Currently, teachers of different disciplines (e.g., mathematics and science) attend separate workshops such that a “cacophony of practices” emerges with little coordination or collaboration within a school.
Fourth, research could explore ways to augment assessments of “mastery” or accuracy of linguistic structures with measures of other linguistic outcomes that are valued in the context of school, such as fluency and complexity of communication. For instance, assessments currently evaluate mastery of vocabulary isolated from academic content, and emphasize accuracy of grammatical structures. Measures could be developed to also evaluate how students actually use vocabulary and other aspects of language to complete academic tasks: Are students able to use vocabulary appropriately to put forth coherent arguments, order and prioritize ideas, and discuss ideas with one another? In this context, Walqui suggested, students may show more progress if grammatical errors are attended to only after students begin to use, depend on, and recognize the value of certain grammatical structures for communicating academic content in school.
Then, errors can begin to be noted and revised with the goal of achieving more grammatically accurate versions of student communications.
Finally, intervening with parents and families is more likely to complement language development efforts in school if those who work with parents and families understand and appreciate their linguistic resources while helping them talk with children to foster language growth.
William Labov observed that progress with learning vocabulary is too slow, as Lesaux’s data show. On what basis, he asked, do teachers, textbook authors, and so on, select vocabulary to teach? Words tend to be chosen, Schleppegrell explained, on the basis of curriculum topics, such as the Declaration of Independence, and so any efforts to teach vocabulary would need to start from the curriculum that is to be taught. Although “controlled vocabulary” is not part of the educational system, said Labov, it would be possible to create texts that introduce vocabulary in more systematic ways that relate to the curriculum and that also support broader language growth for academic learning. Vocabulary would be selected and ordered in a logical learning progression, using knowledge from research about language structures and how these develop.
Kenji Hakuta added that theory would need to guide which words to use and set expectations for learning, such as described in Words Worth Teaching (Biemiller and Boote, 2006; see also, Biemiller, 2008) which participant Claude Goldenberg went on to characterize as an actuarial approach. For instance, based on data from thousands of children, at what age does 90 percent of the population know a certain word? At what age does 50 percent of the population know a given word? Goldenberg characterized the three approaches that emerged in discussion as a linguistic approach, a curriculum-based approach, and an actuarial approach. He said that, together, they could be used to design more systematic ways of teaching vocabulary in a logical progression with supportive text materials.
A question was raised about the nature of the evidence for academic language instruction in supporting students’ progress. The notion of academic registers, Otto Santa Ana observed, appears to be promising for supporting school learning in part because it focuses less attention on grammatical forms, at least initially, and so could more quickly develop
students’ capacities to understand academic concepts and their ability to cope with dense, abstract texts. But what is the evidence, he asked—from studies conducted either in the United States or in other countries—that justifies using the framework as a means to develop and measure academic growth. According to Schleppegrell and Walqui, most of the work to this point has been qualitative (see Schleppegrell, 2009).
This discussion led to a topic that recurred throughout the workshop: the need for a clearer definition—an operational definition—of academic language. One participant wondered if it is “just” curriculum-based vocabulary or if there are other aspects of language that are needed to understand the lesson at hand. Labov suggested that the nominal style, discussed and illustrated in Schleppegrell (2009), is a good first approximation because prose with dense propositions is what children encounter in subject-matter text and academic learning. He noted, however, that this style “is being fought by the best teachers of composition” who favor a more verbal or direct style of prose.
Despite agreeing with observations reported in Hoff (2009) for children from lower SES homes, John Rickford voiced concerns about characterizing those children and children from minority groups on the basis of those data as living in “linguistic deserts.” He said that his research and the work of Shirley Brice Heath and others have shown that children may be developing language valued in their communities, including rich narratives of personal experience. Lynne Vernon-Feagans said her work with narrative research with lower SES African American samples confirms such findings (e.g., Vernon-Feagans, 1996). Hoff responded that although children from disadvantaged or minority-language backgrounds may bring valuable language skills to school, “they know less of what works in the academic setting” in the United States. Several participants commented on how instruction might leverage some of the linguistic skills and practices that children bring to school to make academic learning easier. It could be useful, Rickford proposed, to study schools and teachers who have had at least some degree of success in building from the linguistic capabilities children already possess to develop the additional ones that would benefit learning in school.
Vernon-Feagans urged caution in interpreting past data on language input and the quality of children’s language, given findings in her narrative research with lower SES and African American samples (e.g., Vernon-Feagans, 1996). After accounting for SES, which Hart and Risley (1995) and others did not do, African American and non–African American families did not differ in the amount of language used at home. (Hers was a large representative sample of every child born in three low-income counties in North Carolina and Pennsylvania.) SES differences in input also were not as stark as found in other studies. In addition, African American boys told
more complex stories at school age than middle-class white boys though this did not translate into facility with school tasks. For instance, African American boys in the study did not re-tell a story accurately when asked to listen to a story and paraphrase it, but instead expanded and embellished it. Vernon-Feagans said this story-telling style tends to be valued in many African American homes, but teachers see these expansions as errors or as a failure to follow directions.
Standardized measures also may be “limiting” what is known about students, according to Lesaux. Though used in research and high-stakes testing, current standard measures may not capture the range of children’s language facility or identify certain linguistic strengths that could be leveraged for learning language related to school tasks. Similar to the findings of Vernon-Feagans, Lesaux noted that her research also shows that the complexity of ideas and language present in the oral narratives of children was not captured by standard vocabulary and comprehension measures that required forced choice. She suggested that the typical rubrics used to score narratives needed to be further refined and developed to capture variations associated with the influence of Spanish on narratives produced in English. Research on narrative production is underdeveloped; it could be advanced as a method for identifying existing linguistic strengths to inform the design of instruction.
David Dickinson stressed that to be valuable for narrowing achievement gaps, narrative research would need to target the particular speech and discourse skills relevant to school. Aida Walqui agreed and noted that not all aspects of narrative development that academic researchers tend to study are likely to matter for supporting the aspects of language used to engage with the rigorous school curricula that children must master.
Role of Families
A question was raised about how families should be engaged to support language for school learning. Hoff argued that, for families that do not provide rich language input, family intervention to increase language input can be helpful but it would not be enough to produce the magnitude of change in children’s language development to make a difference in school achievement. In addition, family interventions might not always be feasible because they could place an unreasonable burden on already busy and stressed families. So, while engaging families might contribute some additional linguistic input, more would be needed from other sources.
Goldenberg agreed with Hoff about the magnitude of change that could be expected from at least certain kinds of family interventions.
For instance, studies of dialogic reading (see, e.g., Valdez-Menchaca and Whitehurst, 1992; Whitehurst et al., 1994) in which parents are taught to engage in elaborated discussions around book-reading tend to show “a bump” in language and literacy outcomes, but the effects were very small relative to the size of observed language and achievement gaps. Schleppegrell agreed with Hoff and noted that schools vary in the opportunities and resources they offer children. Creating the school environments that support language, Dickinson pointed out, will require more than material resources. Teachers must become experts in using those resources to engage children in the kinds of conversations that develop language.
Hirsh-Pasek cautioned against completely disregarding the impact of families. The NICHD Study of Early Child Care and Youth Development (funded by the National Institute of Child Health and Human Development) showed that school quality accounted for, at most, 5 percent of variability in outcomes, which included language and other proficiencies such as math; in contrast, 25 percent of the variance in outcomes was accounted for by variations in parenting (National Institute of Child Health and Human Development Early Child Care Research Network, 2003). Also, with respect to communities, some promising efforts in the United States include revamping libraries to go beyond lending books to include language-enrichment programs that model reading and speaking for children. It could be promising, she argued, to study the effects of such programs, as well as other partnerships that might be created among families, schools, local agencies, and community organizations. In establishing partnerships and collaborations, Robert Bayley cautioned against what he described as false collaborations, such as those he observed while working with Latino and Mexican American families in South Texas and the Bay Area, where schools paid little attention to families’ desires to maintain the home language while children learn a new one, especially when working with English-language learners.
Given the wealth of evidence about the importance of early language input, several participants suggested that pediatricians and well-baby clinics are another potential community resource, though pediatricians probably need a better understanding of language development and better materials for helping parents to support language development, including bilingual development.
More needs to be understood about peer influences on language development, especially for children from lower SES backgrounds. Vernon-Feagans described classroom structures in Raleigh, North Carolina, where
the goal is to have no more than 40 percent of children from low-income and minority communities. Achievement is higher than would be expected according to national data on achievement for these groups. (For background about the Wake County district focus on integration and achievement, see Grant, 2009.) In addition, recent analyses of data on children living in poverty from the Early Childhood Longitudinal Study-Kindergarten Cohort (ECLS-K) study showed that, although teacher instruction and initial reading levels were strong predictors of reading during the first 3 years of school, minority segregation was the best predictor (Kainz and Vernon-Feagans, 2007). Research might explore whether peer influences could have contributed to these ECLS-K findings and, more generally, how peers affect language learning for and in the context of school. Schleppegrell agreed, and added that her work in classrooms in Dearborn, Michigan, supports the fruitfulness of studying peer influences on language. Children in her research were English-language learners from the Middle East who spoke Arabic as a first language and had varying levels of proficiency in English. They were not integrated into the community and spoke only Arabic at home, but opportunities for increased peer interaction in the classroom led to improvements in English oral language.
Schleppegrell emphasized that teachers’ expectations are known to influence student performance. In her experience working with teachers, low expectations for children from lower-SES and minority language backgrounds are widespread and are exacerbated by the lack of opportunities to talk with students in the classroom. When conversations do start to occur in classrooms, teachers report being surprised about the strengths of their students and they begin to have higher expectations for performance. As a teacher educator, Susana Dutro agreed and added that in her experience teacher beliefs and misconceptions about language that stem from lack of knowledge about language and how it develops can contribute to low expectations for students. This situation, in turn, is compounded by the few opportunities teachers have to listen to students’ thinking expressed in their own language and language varieties.