Abstract

Four experiments to probe visual processing unit of Korean (Hangul) syllables were executed by introducing the postcueing, forced-choice task on briefly exposed stimuli. Positive or negative repetition effect (PRE or NRE) denotes that detection rate of a target is higher or lower when flanked by the same distractor than by a different one respectively. Of the six Hangul syllable types, <암> and <옴> types, contrasted in this study and most widely used in Hangul words, have a Consonant(C1; 'ㅇ')-Vowel(V; 'ㅏ' & 'ㅗ')-Consonant(C3; 'ㅁ') structure. Within each type, the two candidates of target at the C1 or C3 position were selected to make either a dimension stimulus set ('ㄴ' and 'ㄱ') or a feature stimulus set ('ㄱ' and 'ㅋ'). In the display of just C1 and C3 without V (not making a Hangul syllable), NREs were observed both in the dimension and feature set. But in the Hangul syllable context with one of the two Vs added, NREs disappeared in detecting C1 and C3 of the dimension set and C1 of the feature set, but not C3 of the feature set. In other non-syllable contexts replacing V by random dot bar or dashed line NREs appeared again. It was concluded that the disappearance of NRE in Hangul syllable context resulted from the grapheme processing of V, which was integrated with C1 and also with C3 in some conditions. From NRE pattern of this study was suggested a possibility there would be dual routes of Hangul syllable processing: wholistic and two-stage route. That is, all the three letters of Hangul syllables having a CVC structure can be processed as a whole when less complex visually, but in other cases C1 and V can be processed as a unit first and C3 later.