ORTHOGRAPHIC REPRESENTATION AND PHONEMIC SEGMENTATION IN SKILLED READERS: A Cross-Language Comparison liana Ben-Dror,'

The long-lasting effect of reading experience in Hebrew and English on phonemic segmentation was examined in skilled readers. Hebrew and English orthographies differ in the way they represent phonological information. Whereas each phoneme in English is represented hy a discrete letter, in unpointed Hebrew most ofthe vowel information is not conveyed by the print, and, therefore, a letter often corresponds to a CV utterance (i.e., a consonant plus a vowel). Adult native speakers of Hebrew or English, presented with words consisting of a consonant, a vowel, and then another consonant, were required to delete the first "sound" of each word and to pronounce the remaining utterance as fast as possible. Hebrew speakers deleted the initial CV segment instead of the initial consonant more often than English speakers, for both Hebrew and English words. Moreover, Hebrew speakers were significantly slower than English speakers in correctly deleting the initial phoneme, and faster in deleting the whole syllable. These results suggest that the manner in which orthography represents phonology not only affects phonological awareness during reading acquisition, but also has a long-lasting effect on skilled readers' intuitions concerning the phonological structure of their spoken language. Phonological awareness is the ability to recognize and manipulate internal phonemic constituents of spoken words. Previous research has provided ample evidence that this ability is necessary for reading acquisition and is related to skilled reading performance (for recent reviews, see Bentin, 1992; Goswami & Bryant, 1990). For example, reliable correlations were found between children's ability to manipulate subword units and the rate and efficiency of learning to read (Goswami & Bryant, 1990; Liberman, Shankweiler, Liberman, Fowler, & Fisher, 1977; Mann & Liberman, 1984; Treiman, 1985). In addition, phonological awareness in kindergarten was found to be a good predictor of reading success in the early school years (Bradley, 1989; Bradley & Bryant, 1983; Lundberg, Olofsson, & Wall, 1980; Mann, 1984; Stanovich, Cunningham, & Cramer, 1984), A causal connection between phonological skills and reading acquisition has been supported by studies showing that intervention aimed at improving phonological skills facilitates reading acquisition (Ball & Blachman, 1988, 1991; Bentin & Leshem, 1993; Blachman, 1989; Bradley & Bryant, 1983; Lundberg, Frost, & Peterson, 1988). Other studies have shown, however, an inverse causal connection, that is, that exposure to literacy enhances phonological awareness (Bentin, Hammer, & Address correspondence to Ram Frost, Faculty of Social Sciences, The Hebrew University, Mount Scopus, Jerusalem 91905, Israel, Cahan, 1991; Morais, Bertelson, Cary, & Alegria, 1986), Together, these results suggest a strong bidirectional infiuence between reading acquisition and phonological awareness. Probably, the exposure to clearly defined orthographic segments triggers awareness of coarticulated phonemic segments, while at the same time this awareness fosters the acquisition of grapheme-to-phoneme correspondence rules. This interpretation is further supported by comparing the effects of different reading instruction methods on reading skills, Alegria, Pignot, and Morais (1982) reported that children who learned to read by analytic methods emphasizing letter-sound correspondences performed better on tests of phonemic segmentation than children who learned by hohstic methods. These results suggest that the manner in which the writing system represents the spoken language may infiuence phonological awareness. Support for this claim is gained from crosslinguistic studies, Mann (1986) compared phonological awareness of syllables and phonemes in Japanese and American first graders and found that Japanese children performed more poorly than American children on tests assessing awareness of phonemes but not of syllables, Mann argued that the Japanese children's performance was infiuenced by their reading experience with a syllabary orthography, whereas American children were affected by their reading experience with an alphabetic orthography. Her conclusions are supported by Read, Zhang, Nie, and Ding (1986), who showed that literate Chinese adults who learned to read the alphabetic (pinyin) orthographic system performed better in phonemic segmentation tests than literate Chinese adults who read only the logographic (kanji) system, A possible conclusion from these studies is that the size of the phonological unit that the beginning reader becomes aware of is affected by the size of the speech segment into which orthographic units are mapped. In the present study, we investigated this hypothesis by comparing phonological sensitivity of skilled adult readers trained initially to read either Hebrew or English, We sought to examine the infiuence of different orthography-to-phonology mapping rules on the ability of mature readers to manipulate the various segments of spoken words. The Hebrew writing system is characterized by several properties that make it interesting for comparison with other alphabetic orthographies, such as English orthography (e.g.. Frost & Bentin, 1992). In Hebrew, letters represent mostly consonants, and most ofthe vowels are represented by diacritic marks (dots and dashes). Some vowels, however, are represented by letters (which have dual function and represent either a consonant or a vowel). There are two modes of writing Hebrew; pointed and unpointed. The pointed writing system contains all the diacritic marks and is used mainly for children's books, holy scripts, and poetry. The unpointed print uses the same letter characters as 176 Copyright © 1995 American Psychological Society VOL, 6, NO, 3, MAY 1995 PSYCHOLOGICAL SCIENCE liana Ben-Dror, Ratn Frost, and Shlomo Bentin the pointed system (including the vowel letters) but omits the diacritic marks. The pointed system is taught in the early elementary grades. However, starting in the third grade, the vowel marks are gradually omitted from textbooks, and adult readers use the unpointed writing system almost exclusively. Hence, although Hebrew has an alphabetic orthography, its basic orthographic units usually represent more than single phonemes. Because the letters are mostly consonants onto which vowels are subsequently attached, these orthographic units usually correspond to consonant-plus-vowel (CV) utterances. In the present study, we examined whether the size of the unit represented by the orthography affects the size of the segments that mature readers are aware of. Specifically, we hypothesized that reading in Hebrew, in which alphabetic units represent mostly CV segments, should foster awareness of spoken word segments of that size (CV), In contrast, reading in English, in which most letters are mapped into phonemic segments, should enhance awareness of single phonemes. We were interested in the intuition of adult readers concerning the phonemic structure of their spoken language, as well as in their ability to manipulate single phonemes. Several studies have suggested that for skilled readers, orthographic and phonological representations interact so that orthographic knowledge affects the recognition of words in the auditory modality. It has been shown that lexical decisions to spoken words are facilitated if successive words share the same speUing (Jakimik, Cole, c& Rudnicky, 1980), Similarly, using the naming task, Tanenhaus, Flanigan, and Seidenberg (1980) demonstrated a visual-auditory interference in a Stroop paradigm. However, although it seems evident that reading and hstening could share one lexicon, allowing identical messages to be understood in the two modalities in the same way, it is possible that phonological awareness and the basic phonological skills of adult native speakers are independent of the special characteristics of their writing system. If orthographic knowledge affects phonological skills, such as phonemic segmentation, then English and Hebrew speakers should differ, for example, in their ability to omit the first consonant of a spoken word and pronounce the remaining phonemes (a phoneme deletion task). Whereas this task should be fairly simple for readers of English, the reading experience of Hebrew readers would cause them to delete the initial CV units of Hebrew words. Moreover, even a correct deletion of the initial phoneme instead of the initial CV unit would involve greater cognitive effort for Hebrew readers and, consequently, would result in slower deletion latencies for Hebrew than for English speakers. In contrast, similar performance of the two subject groups would support a view that basic phonological skills of literate adults are not affected by their reading experience, A second aim of the present study was to examine the effect a first language may have on phonemic segmentation in a second language, that is, how Hebrew bilingual readers segment words in English, and vice versa,