This thesis investigated (1) the reading comprehension and strategies by two orthographically different LI background subjects, Chinese- and non-Chinese backgrounds in reading two orthographically different Japanese passages, one with Kana only (Kana passage) and the other with mixture of Kanji and Kana (Kanji-Kana passage), (2) the relationship between comprehension and strategy use, and (3) the relationship between the perception of reading Japanese text and actual performance.
The study was prompted by the recent increase of pedagogical difficulties in a mixed population of two diverse LI backgrounds—Chinese- and non-Chinese-backgrounds— who have distinctly different needs and problems in learning to read Japanese text, which consists of logographic Kanji and syllabic Kana. The study aimed to identify particular needs, problems and characteristics specific to each group of learners so that findings can be utilised to promote better learning of written Japanese by both of them in a same classroom.
The study was qualitatively oriented and was conducted within an ethnographic perspective which focuses on people's behaviour in groups and on culturally characteristic patterns. Research started with questions and hypotheses posed on reading comprehension, reading strategies, relationship between reading comprehension and strategies, and relationship between perception of reading Japanese text and performance. Research data collected for analysis were from various sources, such as a Hiragana recognition test, questionnaires, protocols of the subjects' oral translation of the passages and retrospective accounts of their reading behaviour.
Chinese and non-Chinese background learners demonstrated different reading strategies and comprehension in reading Kana and Kanji-Kana passages. It was found that the differences were caused by the subjects' LI background and features of the passages—existence or non-existence of Kanji. In reading the Kana passages both groups demonstrated wider range of strategic reading behaviour from the word-level processing to global understanding. However, one of the notable findings was that even in reading the Kana passages where no Kanji were involved, the Chinese-backgrounds still heavily depended on the use of the knowledge of Kanji, often through the medium of their LI, Chinese characters. In reading the Kanji-Kana passage, use of visual access to meanings of Kanji, again often mediated by knowledge of Chinese characters, was a popular strategy among Chinese-backgrounds. While they showed exclusive use of graphic and semantic codes of Kanji/Chinese characters, Non-Chinese-backgrounds showed additional use of the phonological code of Kanji to retrieve the meanings. Non-Chinese-backgrounds had to learn all three aspects of Kanji and when the graphic shape did not give cues to meanings they were able to resort to the phonological code. On the other hand, since Chinese-backgrounds were already familiar with the graphic and semantic codes of Kanji thanks to knowledge of Chinese characters, they may not have taken the learning of readings to be as necessary as non-Chinese-backgrounds did. As result, when the graphic features of Kanji did not give cues to meaning, they were unable to resort to the readings of Kanji. This was reflected in the comprehension of the Kana passages where non-Chinese-backgrounds performed notably better than Chinese- backgrounds. It was thus argued that prior knowledge of Chinese characters had both positive and negative roles. It was positive and exceptionally useful when the reader could relate Japanese Kanji words to the relevant Chinese words. It also had a perceptionally favourable role in that the resemblance between Chinese characters and Kanji took pressure off the Chinese-backgrounds while reading Japanese text. Paradoxically, however, this favourable role often played out negatively, as when the search for meaning tended to get word-bound and other access tended to be uncultivated. Reliance of Kanji by Chinese-backgrounds was also reflected in their perceptions of reading Japanese text. They perceived that Kanji-Kana passages as easy because of the graphic and semantic similarities between Kanji and Chinese characters. However, such a perception, while it had a favourable element, tended to lead to a casual attitude to learning Kanji and Japanese in general.
The findings of the thesis suggest that the educators of Japanese as L2, while attending to the needs of non-Chinese-backgrounds as now, should pay more attention to those of Chinese-background learners.