Adult first-time readers in a second language
by Martha Young-Scholten
Is there a critical period for learning to read?
Unschooled adult refugees are often resettled in countries where survival depends on oral proficiency and literacy in a second language (L2). Development of reading is slow; as Strucker and Davidson (in press) note for low-literate Spanish speakers, weak decoding skills persist even for those with an average of 6 ½ years’ native language (NL) schooling. Are adults are too old? While language develops with mere exposure to appropriate input for all young members of the species within a specific time span (Lenneberg’s 1967 critical period hypothesis), mere exposure to printed text is rarely sufficient for reading. If it does not emerge naturally, is there a critical period? A dearth of evidence precludes an answer to this question. While much is known about children’s reading development, closing the research gap in post-puberty first-time literacy should be a priority, given connections between low literacy and income for L2 adults (Burt 2003).
Phonological awareness and learning to read
Considerable research links children’s development of reading to awareness/manipulation of phonological units. Initial awareness emerges naturally, with pre-school children becoming aware of increasingly smaller units, from word to syllable to onset to rhyme. Only through learning to read in an alphabetic script does the child become aware of phonemes and grasp the (alphabetic) principle that graphemes correspond with segments (e.g. Goswami & Bryant 1990). If no critical period exists, adult first-time readers should show awareness of word, syllable, onset and rhyme prior to instruction, and awareness of phonemes as they learn to read. In groundbreaking work on adult Portuguese first-time NL readers, Morais et al. (1988) found that prior to instruction, awareness of units larger than the phoneme existed, but phonemic awareness emerged only with training. Robson (1982) arrives at similar conclusions for adult Hmong refugees: alphabetic literacy rather than education per se led to L2 reading progress. For learners with some schooling, language proficiency seems to be a factor determining at what point NL reading skills transfer (Bernhardt and Kamil’s 1995 language threshold).
Adult English learners with little or no schooling
Presented here are results from a study of the phonological awareness, L2 reading ability and oral proficiency of 10 Somali and 7 Vietnamese living in Seattle. US residence ranged from about a year to 20 years, and 16 had family members with NL and/or English literacy, with some siblings or children having university experience. English instruction varied as did NL schooling (see Table 1). Saville-Troike (1976) claims learning to read occurs once, so the schooled Somali and Vietnamese, whose languages use the Roman alphabet, may transfer reading skills. However, two Vietnamese learners were at least partly schooled in Chinese. Morais et al. ‘s claim that adults develop phonemic awareness only through alphabetic script training is mirrored by studies on adult L2 readers from non-alphabetic script backgrounds. Ng (2000) concluded that mere exposure to written English is insufficient for development of phoneme manipulation skills by Chinese readers (see also Read et al. 1986), and Ben-Dror et al. (1995) found that English readers surpassed unpointed Hebrew readers in phoneme manipulation.
Tests administered measured NL and English reading skills and phonological awareness, and English morphosyntactic proficiency and phonological competence. The awareness component used real words on tasks involving repeating words in a story read aloud, tapping out syllables, identifying a non-rhyming or non-alliterating word in a set of four, and removing first, last, middle sounds from words to create new words. Tasks were done in the NL and then English. Results from a sub-set are presented here.
Table 1. shows that English instruction and NL schooling is generally linked to reading level. But to measure the literacy skills of low-literate adults, children’s standardised tests were unusable due to such adults’ weak linguistic proficiency, and tests for low-literate L2 adults were of marginal use in their focus on what adults cannot do. Components of some tests recently used (see Condelli et al.) were adopted or simplified: (1) unordered/varied font letter identification; (2) common sign reading; (3) single-sentence cloze exercises (multiple choice); (4) word-pair discrimination; (4) single word reading (based on spoken lexicon); (5) paragraph reading (word awareness task story). Writing was measured through asking learners to write personal details. All learners scoring ‘1’ for reading were unschooled in their NL, though this is not a two-way relationship; not all unschooled learners scored a ‘1’ in reading, and there is no obvious relationship between English instruction, schooling and reading level. Somali subject S3 was unschooled, had two weeks of English instruction, yet scored ‘4’ in reading.
Table 1. Reading level for adults with variable amounts of schooling
Learners scoring ‘1’ could identify all letters in test (1), at least some street signs and write their names, but scored between 0 % and 25% on other reading tests. At the other end of the spectrum were learners scoring 100% on all sub-tests including single word decoding who read the paragraph fluently. None of the 17 either read all single words in isolation but not the paragraph or read the paragraph but not words in isolation. The ability to read words in isolation was then used as a measure of decoding ability and compared to phonemic awareness (see below). The two Vietnamese learners with Chinese schooling exhibited similar reading levels, but without a larger sample one cannot say whether V5’s year of Vietnamese schooling enabled him to reach the same level in six months of English classes that it took V2, with two years of Chinese schooling, a further 1 ½ years to reach.
We now turn to whether there a critical period exists for reading, and address this through a comparison of adults with little or no schooling with children (tests and scores from Karmiloff-Smith et al. 1996 for word awareness and Burt et al. 1999 for the rest). The figure below shows common patterns for pre-school and school-age children and the 17 adults. Onset and rhyme pattern differently depending on age, and onset and rhyme awareness slightly exceeds syllable awareness for Somalis, but both adults’ and childrens’ word awareness invariably exceeds syllable/onset and rhyme awareness, and both exceed phonemic awareness.
What correlations exist between linguistic proficiency, reading level and phonological awareness?There was a highly significant (p <.01) positive correlation between phonemic awareness and word reading, and a significant (p < .05) positive correlation between rhyme-onset awareness and word reading. Tested through picture naming, phonological competence scores were calculated as percentages of learners’ production of initial and final consonant clusters.(1) Morpho-syntactic competence was scored on a scale of 1-5 using criteria in Table 2. Stages reflect research carried out over the last 30 years revealing a uniform order of acquisition of verb-related inflectional morphology in L2 English independent of age, exposure type and generally, the learner’s NL (Hawkins 2001).
(1)The test of segmental competence yielded unrealiable results due to learners’ meagre lexicon not allowing determination of relevant English contrasts.
Table 2. Organic Grammar stages for verbal syntax (from Vainikka and Young-Scholten 1994)
Of the eight unschooled adults, two demonstrated phonemic awareness and read words in isolation (scoring highly on all other reading sub-tests). Unlike these two, whose level of oral proficiency was rated as high, the six non-readers were at a level of morpho-syntactic development resembling children’s one-word stage. This and statistical measures showing a significant positive correlation (at p < .01) for the Somalis for overall reading ability and both morpho-syntactic and phonological competence indicates existence of a language threshold for non-schooled adults, too. .
Implications for the classroom
If there is no critical period for learning to read, what accounts for the problems unschooled adults experience? That L2 exposure may be involved is implicit in Cunningham-Florez and Terrill’s (2003) claim that 500 - 1,000 hours of instruction is required for NL-literate adults to start functioning in their L2. Clearly non-literate adults require more exposure. Those who have addressed low-literate adults’ reading development recommend various solutions, from use of real life materials and basic skills work (Condelli et al.’s study of 500 low-literate adults in classrooms across the USA) to the teaching of NL reading in an alphabetic script prior to introduction of English (Burtoff 1985 on Haitian Kreyòl). Answers will come through closing the research gap on non-literate adult refugees resettled in literate societies.
Ben-Dror, I., R. Frost and S. Bentin. 1995. ‘Orthographic representation and phonemic segmentation in skilled readers|. A cross-linguistic comparision’ Psychological Science 6:176-181
Bernhardt, E. B. and M. L. Kamil. 1995. ‘Interpreting relationships between L1 and L2 reading: consolidating the linguistic threshold level and the interdependence hypotheses’ AppliedLinguistics. 16:15-34.
Burt, L., A. Holm and B. Dodd. 1999. ‘Phonological awareness skills of 4-year-old British children: an assessment and developmental data’ International Journal of Language and Communication Disorders. PAGES!!
Burt, M . 2003 ‘Issues in improving immigrant workers’ English language skills’ ERIC Brief. Center for Applied Linguistics, Washington, DC.
Burtoff, M . 1985 ‘Haitian Creole literacy evaluation study: final report’ ERICDigest 277273. Center for Applied Linguistics, Washington, DC.
Condelli, L ., H. Spruck Wrigley, K. Yoo, M. Seburn and S. Cronen. 2003. What Works. Study for Adult ESL Literacy Students. Volume II: Final Report. American Institutes for Research and Aguirre International. US Department of Education, Washington, DC.
Cunningham-Florez, M . and L. Terrill. 2003 ‘Working with literacy-level adult English language learners’ ERIC Q & A Center for Applied Linguistics, Washington, DC.
Goswami, U. and P. E. Bryant. 1990. Phonological Skills and Learning to Read. Hove: Psychology Press.
Hawkins, R. 2001. Second Language Syntax. A Generative Introduction. Oxford: Blackwell.
Karmiloff-Smith, A., J. Grant, K. Sims, M-C. Jones and P. Cuckle. 1996. ‘Rethinking metalinguistic awareness and accessing knowledge about what counts as a word’Cognition 58:197-219.
Lenneberg, E. H. 1967. The Biological Foundations of Language. New York: Wiley.
Morais, J., L. Cary, J. Alegria and P. Bertelson. 1979. ‘Does awareness of speech as a sequence of phones arise spontaneously?’ Cognition 7:323-331.
Ng, C . 2000. ‘English phonological skills in Chinese ELF learners with different L1 literacy experience’ Paper presented at the AAAL conference, Vancouver.
Read, C., Y. Zhang, H. Nie and B. Ding. 1986. ‘The ability to manipulate speech sounds depends on knowing alphabetic spelling’ Cognition 24:31-44.
Strucker, J . and R. Davidson. In press. Adult Reading Components Study (ARCS) Boston: NCSALL.
Robson, B. 1982. ‘Hmong literacy, formal education and their effects on performance in an ESL class’ In B. T. Downing and D. P. Olney (eds). The Hmong in the West: Observations and Reports. Minneapolis: U. of Minnesota.
Vainikka, A . and M. Young-Scholten. 1994. ‘Direct access to X'-Theory: Evidence from Korean and Turkish adults learning German’ In T. Hoekstra and B. D. Schwartz (eds.) Language Acquisition Studies in Generative Grammar: Amsterdam: Benjamins. pp. 265-316.
Copyright 2000-2014© Developing Teachers.com