Identification and classification of characters in two alphabets

Bilingual subjects were presented with pairs of characters drawn from either the English or the Hebrew alphabet. They were asked to indicate whether the two characters belonged to the same alphabet, or whether the two characters denoted the same sound, regardless of whether they belonged to the same alphabet. A model is proposed whereby the two characters are processed in parallel, with characters belonging to a nonnative alphabet requiring more processing time than characters belonging to one's native alphabet. The native-language effect appears to be localized peripherally. The data do not suppott models which categorically place identification before classification or vice versa.