This paper analyzed word recognition in two patterns of Chinese characters, cross referenced with word frequency. The patterns were defined as uni-part (semantic radical/component only) and bi-part (including the phonetic radical/component and the semantic radical/component) characters. The interactions of semantic and phonological access in both patterns were inspected. It was observed that in the naming task and the pronunciation-matching task, the subject performance involving the uni-part characters showed longer RT than the bi-part characters. However, with the lexical decision and meaning-matching tasks the uni-part characters showed shorter RT than the bi-part characters. It was also observed that the frequency, which is regarded as a lexical variable, displayed a strong influence. This suggests that Chinese characters require lexical access in all tasks. This study also suggested that the phonological process is primary in visual word recognition; as there is a significant phonological effect in processing the Chinese bi-part characters, resulting in either the facilitation or inhibition of phonology due to the differing demands of the two tasks
This study devised a pronunciation computer program to examine whether mobile-assisted language learning (MALL) could facilitate college students’ acquisition of English phonemes and word stress. Thirty-eight participants enrolled in the remedial English class offered at the language center of a national technological university in central Taiwan. Before the class, they were asked to read a word list. In the following six weeks, they were taught to distinguish and articulate English phonemes and to predict word stress locations using the designed computer program. They were also instructed to review the learning materials using the smart-phone version of the devised program. After the teaching session, each participant was asked again to read the same word list and fill out an assessment questionnaire. The sound analyses show that their readings of English minimal pairs and word stress placement were more accurate than their performances before the instruction. Their responses to the questionnaire indicate that both the given instruction and the designed computer program were satisfactory. In the open-ended questions, some of them said that they have built up a better understanding of phonemes and word stress, and that they would try to predict polysyllabic word stress when reading English articles. The present findings can be further applied to research on MALL-based English pronunciation acquisition.