Crossover of phonological processing skills: a study of Spanish-speaking students in two instructional settings.

ABSTRACT

This study investigated three questions: Do phonological processes show cross-linguistic transfer? How does the language of instruction influence the relationship between phonological processes and decoding? Does performance on Spanish and English phonological processing tasks similarly predict English decoding for the same English learners (ELs)? We studied first-grade ELs who had been enrolled for 2 years in two programs that differed by language of instruction (English only and bilingual). Phonological processing skills were examined following a theory of core phonological processing deficits that postulates that three related constructs--phonological awareness, phonological coding, and phonological recoding--are the major components of phonological processing. The results indicate that (a) phonological processes do exhibit cross-linguistic transfer in young ELs; (b) phonological awareness might best be conceptualized as comprising two developmentally overlapping components; (c) language of instruction influences English and Spanish word reading and Spanish pseudoword decoding, but not English pseudoword decoding; and (d) phonological awareness is the only theoretical phonological processing construct significantly related to all English and Spanish reading tasks.

**********

THERE IS GROWING CONSENSUS ABOUT THOSE ELEments of reading acquisition for young children that are most fundamental and susceptible to instructional intervention (Committee on the Prevention of Reading Difficulties in Young Children, 1998; National Reading Panel, 2000). However, significant questions remain about the growing number of young students who may not become proficient in reading English because English is their second language (L2; Hakuta & Garcia, 1989; Rumberger, 2000).

The primary purpose of this study is to frame some of the basic questions about the reading difficulties of English learners (ELs) in terms that relate to cognitive mechanisms that serve as the substrate for learning to read in English. We are particularly interested in what kinds of cognitive resources related to first language (L1) competencies are available for ELs via cross-linguistic transfer as they learn to read English. In selecting this approach, we do not believe that we are reducing a significant educational and social problem to a few cognitive constructs, nor are we denying the overarching and long-term importance of specific and larger instructional events and social conditions that influence and give meaning to long-term outcomes.

However, we do believe that form should follow function. Therefore, how young students function cognitively as they learn to read English ought not only to inform us but also to dictate to a large extent the forms of instruction and the programmatic arrangements that we mandate. Unfortunately, although there is considerable empirical literature on the cognitive aspects of L2 learning, there appears to be a significant dearth of literature on very young children who must learn to read in an L2. Fortunately, there is now a significant body of literature on individual cognitive differences between young monolingual readers that may inform us in pursuing better understanding of young ELs.

TWO THEORETICAL CONSIDERATIONS

We are informed by and attempt to bridge between two theoretical positions on L2 reading acquisition by young students. One is a "macro" theory of global language ability in the cultural and social context of its acquisition and use. The other is a "micro" theory of universal cognitive mechanisms underlying reading acquisition. We reason that broad language competencies emerge, in part, on a substrate of far simpler language-processing mechanisms and that individual differences in the development or functioning of these mechanisms may interact with broad environmental factors, including school instruction, to either increase or decrease the risk of failure to become academically proficient in English reading (e.g., see Hakuta, Butler, & Witt, 2000, & Rossell, 2000, for a discussion of time-to-proficiency issues).

The first theory to frame our work, a theory of common underlying proficiencies (CUP), states that a common underlying knowledge about language lies beneath the surface of bilingual or multilingual performance (Cummins, 1996). In other words, knowledge about reading in L1 is an available resource for assisting in L2 reading acquisition.

The second is a theory of core phonological deficits that represent a convergence of several lines of investigation showing that students with reading difficulties, particularly in word reading, have deficits in phonological processes (Liberman & Shankweiler, 1985; Morris et al., 1998; Shankweiler & Crain, 1986; Stanovich, 1988; Wagner & Torgesen, 1987; Wagner, Torgesen, & Rashotte, 1994).

This present research proceeds from a specific theory in which phonological processing is composed of three independent but highly related components: phonological awareness, phonological coding, and phonological recoding (Wagner & Torgesen, 1987). Phonological awareness refers to one's awareness of and access to the sound structure of oral language (Wagner, 1988). Phonological coding refers to coding of sound-based information in working memory. Phonological recoding refers to the ability to retrieve phonological information from long-term memory at a rapid pace.

It is reasonable to suppose that Cummins' (1996) bilingual theory of common underlying proficiency implies core cognitive and linguistic mechanisms that may also account for individual differences in young children's ability to learn to read words in a second language. Specifically, cross-linguistic transfer is conceptualized as the access and use of linguistic resources in L1 by students while learning other languages. Research on monolingual students in several languages provides considerable evidence that specific phonological skills improve students' chances of becoming strong readers, and, conversely, lack of these skills predicts later reading difficulties (Liberman & Shankweiler, 1985; Wagner, 1988). It is logical, therefore, that the ability to process phonological information is part of the common underlying proficiency posited by Cummins (Cisero & Royer, 1995; Durgunoglu, Nagy, & Hancin-Bhatt, 1993; Gerber, English, & Leafstedt, 2000; Gerber, English, & Leafstedt, 2002). The present study extends the existing research literature by examining relative influences of native language, language of instruction, and individual differences in phonological processes as these relate to early acquisition of word reading.

CROSSOVER OF PHONOLOGICAL PROCESSING SKILLS

Relatively little research has been conducted thus far on the cross-language transfer of phonological processes in early reading. Most of this research has focused primarily on one component of phonological processing: phonological awareness. Examination of the cross-linguistic transfer of phonological awareness skills has focused primarily on correlations among phonological measures in Spanish and English (August, Calderon, & Carlo, 2000; Cisero & Royer, 1995; Durgunoglu et al. 1993) and through examination of outcomes in English for bilingual students (Muter & Diethelm, 2001). Thus far, research findings have consistently revealed evidence of cross-linguistic transfer. Across comparable English and Spanish phonological awareness tasks, correlations have been high to moderate and statistically significant. These data can be used to argue that the measures are tapping the same underlying construct rather than a specific English or Spanish skill (Cisero & Royer, 1995). Preliminary factor analysis of comparable Spanish and English phonological awareness tasks administered to a large sample of Spanish-speaking kindergartners has suggested underlying constructs that are not specific to a particular language (Gerber, English, & Leafstedt, 2000). Additional research has shown that correlational relationships among English and Spanish phonological awareness tasks remain stable over time (Cisero & Royer, 1995). Some…