Research into kanji (Chinese characters used in Japan) learning and recognition has given rise to various theories on how kanji are learnt and identified by non-native learners of Japanese. However, an overall understanding of the underlying process by which they learn kanji remains unclear. The aim of this thesis was thus to produce a synthesis of kanji learning from cognitive outcomes and socio-cognitive behaviour to perceived strategies among learners of Japanese as a foreign language (JFL). This was examined through three separate but interrelated studies. The first study examined the use of kanji learning strategies and their perceived efficacy by JFL learners by means of a questionnaire. The second study examined the outcomes of identifying single kanji characters by means of a kanji identification task. Finally, the third study explored kanji learning behaviour in detail by examining affective factors, kanji attributes and the effect of instructional methods on six beginner JFL learners over a semester of kanji study. Analysis of Study 1 revealed three main categories of kanji learning strategies, namely, mnemonic, analytic and rote learning. On average, learners claimed that the strategies they used most were also most helpful. Despite their exposure to Chinese characters, the learners from Chinese backgrounds studying in Australia demonstrated similar preferences for kanji learning strategies to English first language (L1) learners, while learners from Sri Lanka showed preferences for different types of strategies to these two groups. Study 2 revealed that all learners performed better in matching kanji with their shapes than with their meanings or pronunciations (readings). Chinese L1 learners performed better than their alphabetic (English L1) or alphasyllabic (Sinhalese L1) counterparts in matching the meanings and shapes of kanji. Similarity, whether in shape, pronunciation or meaning, did impair the performances of all JFL learners. The findings of Study 3 underpin the need to develop individualised learning styles within the kanji classroom. Some learners collaborated with their study partner in finding solutions; some demonstrated abilities to recognise and assess their own learning behaviour, and others initiated and developed activities for learning kanji to varying degrees. In general, experiences of staying in Japan appeared to have produced negative impressions regarding kanji study. Moreover, asymmetries in reading and writing were prevalent among the learners. In spite of the varied instructional modes exercised on the learner groups, all learners were able to assess the strengths and weaknesses in each method and develop their own kanji-learning styles. Taken together, the three studies reported in this thesis all contributed to deepening our understanding of the kanji learning process of pre-intermediate and beginner JFL learners. These three studies constitute the starting point in the endeavour to propose a framework for kanji learning.

Research into kanji (Chinese characters used in Japan) learning and recognition has given rise to various theories on how kanji are learnt and identified by non-native learners of Japanese. However, an overall understanding of the underlying process by which they learn kanji remains unclear. The aim of this thesis was thus to produce a synthesis of kanji learning from cognitive outcomes and socio-cognitive behaviour to perceived strategies among learners of Japanese as a foreign language (JFL). This was examined through three separate but interrelated studies. The first study examined the use of kanji learning strategies and their perceived efficacy by JFL learners by means of a questionnaire. The second study examined the outcomes of identifying single kanji characters by means of a kanji identification task. Finally, the third study explored kanji learning behaviour in detail by examining affective factors, kanji attributes and the effect of instructional methods on six beginner JFL learners over a semester of kanji study. Analysis of Study 1 revealed three main categories of kanji learning strategies, namely, mnemonic, analytic and rote learning. On average, learners claimed that the strategies they used most were also most helpful. Despite their exposure to Chinese characters, the learners from Chinese backgrounds studying in Australia demonstrated similar preferences for kanji learning strategies to English first language (L1) learners, while learners from Sri Lanka showed preferences for different types of strategies to these two groups. Study 2 revealed that all learners performed better in matching kanji with their shapes than with their meanings or pronunciations (readings). Chinese L1 learners performed better than their alphabetic (English L1) or alphasyllabic (Sinhalese L1) counterparts in matching the meanings and shapes of kanji. Similarity, whether in shape, pronunciation or meaning, did impair the performances of all JFL learners. The findings of Study 3 underpin the need to develop individualised learning styles within the kanji classroom. Some learners collaborated with their study partner in finding solutions; some demonstrated abilities to recognise and assess their own learning behaviour, and others initiated and developed activities for learning kanji to varying degrees. In general, experiences of staying in Japan appeared to have produced negative impressions regarding kanji study. Moreover, asymmetries in reading and writing were prevalent among the learners. In spite of the varied instructional modes exercised on the learner groups, all learners were able to assess the strengths and weaknesses in each method and develop their own kanji-learning styles. Taken together, the three studies reported in this thesis all contributed to deepening our understanding of the kanji learning process of pre-intermediate and beginner JFL learners. These three studies constitute the starting point in the endeavour to propose a framework for kanji learning.