A preliminary experiment on more efficient kanji learning for Japanese introductory level learners

漢字の構成要素

漢字構造

漢字部品

非漢字部品

Japanese kanji characters consist of many complex components. In Japanese kanji learning, especially for JSL（Japanese as a Second Language） children, one of the most difficult things is to recognize and remember these many complex forms. Hayakawa, Honda, & Iori（2019a） deconstructed 1,006 elementary school kanji into smaller components for JSL children, which we termed “kanji parts” and “non-kanji parts”. We suggested that if these children remember only 57 nonkanji parts and the most frequent 30 kanji parts, they should be able to recognize 475 kanji, comprising 47.2% of the school kanji to be learned. In this study, we conducted a kanji recognition task for Japanese learners in an introductory course at a university as a preliminary test to validate the previous classification, and to see if student kanji recognition could change through kanji learning. Participants were first asked to memorize a list of 36 kanji. Afterwards, they were shown additional kanji on a monitor, and they had to judge whether this kanji belonged to the list they had memorized. Their responses were timed. The data were analyzed with a three-way, repeated measures ANOVA design which compared learning time（beginning, end）, kanji type（kanji parts only, non-kanji parts only, mixed）, and kanji structure（left-right, top-bottom, holistic）. The results indicated that the kanji recognition abilities of introductory level Japanese learners are prone to be affected by kanji type and structure factors. Additionally, the learning time factor only affected speed of recognition. Based on these results, we suggest an efficient order for introductory level learners to learn kanji. We recommend that introductory level learners focus on non-kanji part only, and kanji part only characters ; first learning the holistic, then top-bottom and left-right characters.