Phonological and orthographic visual word recognition in the two cerebral hemispheres: Evidence from Hebrew

Abstract

Studies on the cerebral mechanisms of reading have mostly used Latin-based writing systems and assume that the left, but not the right, cerebral hemisphere is capable of phonological processing. The present study used Hebrew as the test language to examine the effects of phonological and orthographic information in the two hemispheres. In unvoweled Hebrew script, words are read via consonant information alone. We used two naming tasks with an interference paradigm, where phonemically, orthographically, and figurally incorrect vowel information conflicted with the consonant information of words presented in the left, right, or central visual fields. Interference patterns indicated that the left hemisphere automatically transforms graphemes into phonemes (Experiments 1 and 2), whereas the right hemisphere processes vowel diacritics as visual objects (Experiment 1), although it possesses some phonological categories (Experiment 2). The significance of these findings for models of visual word recognition in the cerebral hemispheres is discussed.

We would like to thank Ruth Kimchi, David Plaut, and an anonymous reviewer for many helpful comments. The first experiment was presented at the XIth Conference of the European Society for Cognitive Psychology.

Notes

¹ Zaidel (1998) pointed out that “language in the disconnected right hemisphere following complete cerebral commissurotomy permits assessment of positive language competence, presumably free of callosally mediated inhibitory effects of left aphasiogenic lesions.”

² We thank Ram Frost for kindly providing us with the frequency counts.

³ The rather high error rates in Experiment 1 and Experiment 2 emerged due to a strict way of error count, including errors on both vowels and consonants. For example, the naming of the word ℵ (lake, /agam/) instead of the correct target word ℵ (pear, /agas/) was counted as error. Since this led to the exclusion of many participants from the data analyses, Appendices A and B provide the data analyses including all participants for Experiments 1 and 2, respectively.

⁴ The interaction between visual field and pointing was significant by items when all participants were included in the analyses, F ₂(2, 424) = 4.90, p < .0079.

⁵ The difference between figural and phonemic pointing interference in the RVF was significant by items when all participants were included in the analyses, F ₂(1, 211) = 6.72, p < .0102.

⁶ Phonemic interference was equally strong in both visual fields: Experiment 1, 32 ms in RVF vs. 25 ms in LVF, F ₁ < 1; F ₂(1, 209) = 3.06, p = .0820; Experiment 2, 54 ms in RVF vs. 44 ms in LVF, F ₁ < 1; F ₂(1, 182) = 2.63, p = .1069.

⁷ Figural distortion more strongly affected the LVF than the RVF: Experiment 1, 29 ms vs. 10 ms, F ₁(1, 31) = 5.39, p = .0270; F ₂ < 1; Experiment 2, 43 ms vs. 22 ms, F ₁(1, 27) = 3.14, p = .0874; F ₂ < 1.

⁸ We thank David Plaut for suggesting this interpretation.