Lukatela G, Turvey M T, Todorović D
University of Belgrade, Yugoslavia.
J Exp Psychol Learn Mem Cogn. 1991 Jul;17(4):653-63. doi: 10.1037//0278-7393.17.4.653.
Phonologically ambiguous Serbo-Croatian words are identified more slowly and erroneously than their phonologically unique counterparts. Five experiments addressed the reduction of these ambiguity effects when Roman (Cyrillic) targets are preceded by consonants unique to the Roman (Cyrillic) alphabet. Alphabet-specific nonword contexts were presented briefly with masking. With forward masking, performance was better when the phonologically ambiguous target words and their preceding nonword contexts were alphabetically congruent. Similarly, where backward masked contexts acted themselves as backward masks for the target stimuli, identification was highest when the context masks were in the same alphabet as the targets. Results were discussed in terms of automatic, prelexical processes within a network model of visual word recognition in Serbo-Croatian.
与语音上独特的塞尔维亚-克罗地亚语单词相比，语音上有歧义的单词被识别得更慢且更容易出错。五项实验探讨了在罗马（西里尔）字母的目标词之前出现该字母表特有的辅音时，这些歧义效应的减少情况。特定字母的非单词语境在有掩蔽的情况下短暂呈现。在前向掩蔽中，当语音上有歧义的目标词及其前面的非单词语境在字母表上一致时，表现更好。同样，在反向掩蔽语境本身作为目标刺激的反向掩蔽的情况下，当语境掩蔽与目标词使用相同字母表时，识别率最高。研究结果在塞尔维亚-克罗地亚语视觉单词识别网络模型中的自动、词汇前加工过程方面进行了讨论。
