the transitional approach was the only one that supported language and literacy outcomes in both Spanish and English.

In closing, Hammer summarized several areas of research that would help to better understand dual-language learners and their progress with language, learning, and achievement. First, better descriptions of study populations and samples are needed to understand the heterogeneity of English-language learners. Better documentation of the characteristics of study samples, including information about proficiencies in all languages spoken and other characteristics, would help to interpret study findings.

Second, how early language experiences relate to reading and later achievement is not well understood: much more needs to be known about how uses of first and second languages at home affect first and second language growth. Thus, longitudinal studies of children’s language and literacy are needed to assess growth in first- and second-language and literacy proficiencies and to identify influences on their development. Third, most of the research to date has confounded language status with SES and has focused mainly on low-SES bilinguals: more research is needed that disentangles language experiences and SES.

Fourth, more second languages than Spanish need to be included in research studies to help interpret whether findings apply to the general experience of learning two languages or only to a particular language or population. Finally, intervention studies are needed to determine the effectiveness of various instructional models, curricula, and teaching techniques. Such studies should identify which approaches work best in terms of children’s language proficiencies and other characteristics. As part of this work, language facilitation strategies with evidence for their role in supporting the language development of monolinguals, such as modeling language, recasting, extending, and elaborating utterances, would be useful to incorporate into instructional approaches and to test.

Kathy Hirsh-Pasek and Roberta Golinkoff began their comments by noting two “big ideas” in understanding reading. First, reading is parasitic on language, a point first made by Gibson and Levin (1975) and later elaborated by Hollis Scarborough (2001) who compared learning to read to weaving a rope. Learning to read involves weaving together separate strands of skill; vocabulary and decoding, for instance, are but single strands of the rope. A broader view of language and its relation to reading development is essential to, first, understand the nature of each strand and, then, decipher how the strands of the rope become woven together. Second, language learning is malleable, as shown by both home and classroom intervention research.