Advances in Social Science, Education and Humanities Research, volume 222

2nd Social Sciences, Humanities, and Education Conference (SoSHEC 2018)

Orthographic Similarity Effects of Kanji on

Japanese Education Students’ Cognition

Mintarsih Bambang Yulianto

Universitas Negeri Surabaya Universitas Negeri Surabaya
Surabaya, Indonesia Surabaya, Indonesia
[email protected] [email protected]

Subandi
Universitas Negeri Surabaya Didik Nurhadi
Surabaya, Indonesia Universitas Negeri Surabaya
[email protected] Surabaya, Indonesia
[email protected]

Abstract— the current study explores the similarity effect of process of the logographic type. Logographic stimuli typically
kanji orthographic to 53 studentsat Japanese education display greater dependence on the various graphic features
departmentwhile reading and writing kanji. This research found in visual stimuli. The kanji processing is influenced by
employed qualitative method and error analysis technique. Kanji several factors, e.g. a given kanji contextual arrangement,
studied is a combined kanji of 2 kanji (nijijukugo). The similarity specific features of the familiarity, frequency, and complexity
effect of graphical kanji to reading and writing errors is classified of kanji [1].
into 2 categories, (1) misspelling of target kanji which then affect
meaning, (2) non-kanji errors type. This study indicates that
effect of orthographic similarity on process of understanding A. Kanji architectural design as orthographic
kanji on cognition of learners involves various information The architectural design of kanji as a letter consists of
contained in kanji i.e. graffiti, radical, kanji component, bushu (radical) components, kakushuu (number of strokes in
pronunciation, and meaning relationships. Existence of this kanji writing), hitsujun (strokes order in kanji writing),
relationship proves it is also important to conduct research from rikusho (kanji origins) and how to read kanji (onyomiis
effects of phonological, morphological, or semantic point of kanji
Chinese pronunciation and kunyomiis Japanese
lexical. This research has implications for kanji teaching
materials within more attention to effect of similarity between pronunciation). In Japanese is known four types of script, i.e.
kanji, especially when individual kanji formsmanycombined kana letter consisting of hiragana and katakana, kanji, and
kanji of 2 kanji or more called jukugo. romaji (Roman letters). Japanese characters are classified into
hyoonmoji (phonogram) and hyoimoji (ideogram) [2]. The
phonogram consists of alphabetical Roman letters
Keywords—effects, orthographic similarity, niji jukugo, reading
of kanji ) andsyllabogram of kana, while the ideogram in the form of
kanji, as shown in the following chart.
I. INTRODUCTION FIGURE 1. JAPANESE CHARACTERS
As the number of identifiable individual kanji increases,
more words will also be formed from combined kanji.
However, similarity effect of kanji visual form is often one of
obstacles affecting learning of kanji. Therefore, this study tries
to identify, analyze, and describe how kanji visual similarity
effect to learners‟ recognition.
The recognition of kanji as letters, words, and symbol in
Japanese activates information such as graphics,
pronunciation, or kanji meaning that takes place interactively
in the minds of learners. This activation process occurs
because of a visual stimulus kanji graphical configuration in
writing. Kanji processing is not essentially the same as kana or
alphabet processing, especially in the early stages of
processing. There is a difference in the pattern recognition

Copyright © 2018, the Authors. Published by Atlantis Press.

This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/). 400
 Advances in Social Science, Education and Humanities Research, volume 222

Hiragana and katakana itself each consist of five vowel pronunciation without alternative of onyomi and kunyomi
sounds, forty syllable sounds, and one consonant. Then add pronunciation. This is said to be consistent if every
some derivative sounds (e.g.dakuon は /ha/ is ば /ba/, constituent character has identical pronunciation in all
handakuonis ぱ /pa/, and you-onare び ゃ /bya/, ぴ ゃ /pya/). words containing that character in the same position,
Yokosuka (2002) explains kanji has a characteristic which can e.g.kanji 義 /gi/ (consultation), 題 /dai/ (title).
function as a letter, sound marker of object being referred, and b. Inconsistent (a character of kanji that has many ways of
word. Unlike alphabet that is a letter marker sound, kanji pronunciation in all words containing the same character)
symbolizes meaning as well as the sound. Each kanji can be a jukugoconsists of two kanji characters at least one of two
word because it kanji also called characters of kanji combined has an alternative
logogram/logograf„hyogomoji‟. When placing kanji as a letter,
pronunciation kunyomi. However, the correct
kanji-forming elements such as bushu (kanji root), tenkaku
(kanji character streak), senbun (line segment) need careful pronunciation of kanji is onyomi, although kunyomi may
attention, since the mistake of writing kanji elements will be used in other combined kanji.Typical inconsistency
result in kanji not being read or possibly forming other kanji jukugois said to be a typical inconsistency if every
characters that have similarities. In another context, a single constituent is an inconsistent character but pronunciation
kanji character can serve as a word„go‟, and in event of a of both characters is statistically typical. Atypical
merger of two or more kanji characters kanji serves as a inconsistency jukugois classified as atypical
combined kanji„jukugo‟. Because it has such a character, then inconsistencies if both characters have multiple
in general kanji can also be called a word [3]. pronunciationat a particular position and the
pronunciation of one or both kanji is statistically atypical
Kanji has root position of kanji (bushu) i.ehen, tsukuri,
e.g. kanji 神 when pronounced onyomi sounds /shin/,
kanmuri, ashi, tare, nyoo, and kamae, bushu or radical
/jin/, kunyomisounds /us/, /kan/. In corpus of 31,000
consists of two types of phonetic radicals and semantic
radicals; the procedure of writing kanji through sequence of words, this kanji appears in first character in 86
graffiti (kanji-forming line)and formation of kanji (rikusho). composite kanji, and is pronounced /shin/ in 71 words of
Fushimi, Ijuin, Patterson, and Tatsumi [4] explain many (but total number of occurrences of 86 words, e.g. 神 経
not all) kanji characters containing two components: semantic /shinkei/, pronunciation /shin/ for kanji 神 located in the
radicals that often (but not always) provide a clue to the first position is called a typical design, and for other
meaning of kanji, phonetics radicals that often (but not pronunciations are /jin/, /us/, /kan/ classifiedas atypical
always) provide a sound clue of kanji character. Saito, Inoue, [4].
and Nomura, Ito in Kess and Miyamoto [1] describe the views c. Shibahara, Zorzi, Hill, Wydell, Butterworth [5] added
of the basic formations of rikusho being classified into four inconsistent kunyomi consists of two kanji characters,
namely i.e shokeimoji (ideographic kanji) is a kanji formed this kanji pronunciation is kunyomi which each
from simplification of natural form, shijimoji(diagrammatic component have onyomi but used in other word. The
kanji) is a simplification of something abstract or a concept, whole word pronunciation is kunyomi, although the
kai-i moji (compound-semantic kanji) is a kanji derived from a typical pronunciation of each character is onyomi instead
composite of two previously created kanji that form a new of kunyomi.
meaning, keiseimoji (phonetic-semantic kanji) is a kanji which
is a combination of two kanji that express meaning and C. Kanji teaching materials in Japanese education
speech. Shokeimoji and shijimoji percentage occupy the department
smallest number of joyokanji, the largest percentage Kanji studied in Japanese education department of
proportion is kai-i moji, and the largest proportion of the UniversitasNegeri Surabaya for basic level refers to
number of kanji is keiseimoji which almost covers 80% of compulsory textbooks in learning Japanese in universities. The
total joyokanji[1]. instructional materials was conducted on textbooks of
Kanji orthographic similarities can be seen from kanji- Japanese elementary level i.e.Minna no Nihongo vol.1-2 [6],
forming components, i.e. radical, graffiti, kanji components, basic kanji textbooksKanjivol. 1-2 [7], and basic reading
e.g. in kanji 休 /kyu/ (rest) with 体 /tai/ /karada/ (body); 親 bookDokkai vol. 1 [8].Counting of kanji was focused on the
emergence of kanji in each chapter of textbook. The
/oya/ (parents) with 新 /shin/ (new), 顔 /kao/ (face) with頭 occurrence of kanji has been calculated in form of combined
/atama/ /to/ (head), etc. kanji, while in Kanji vol.1-2 in form of individual kanji.
Purpose of this descriptionis to try to make a general
B. Phonological aspects of kanji overview, what kanji and how many kanji appear in each of
Associated with kanji when viewed phonologically, kanji early Japanese learning books. Description can be seen in
has an onyomi and kunyomi pronunciation, and spelling in following table.
kana writing. The number of pronunciations either onyomi or
kunyomi on an individual kanji results inconsistent
pronunciations in the combined kanji„jukugo‟. Fushimi, Ijuin,
Patterson, and Tatsumi [4] explain the following.
a. Consistentjukugo is made up of two kanji characters each
character component of this type of kanji has one onyomi

401
 Advances in Social Science, Education and Humanities Research, volume 222

TABLE I. DESCRIPTIONOF LEARNING MATERIALS IN THE BASIC Participants more or less studied about 700 combined kanji
JAPANESE TEXTBOOK and 250 individual kanji.
Textbooks Total number of Number of individual
kanji composed of 2 simple and complex B. Instrument
kanji kanji
Minna no Nihongo Instrument used in form of a written test consisting of 30
524 -
Vol. 1 questions on kanji noyomikakikata (reading kanji as isolation
Minna no Nihongo word) and translate meaning, 10 questions on kanji no
321 -
Vol. 2 kakikata (writing kanji), and 5 questions on kanji noyomikata
Kanji Vol. 1 - 221 in a sentence. Participants wrote answers on answer sheets, the
Kanji Vol. 2 - 296 test lasts for 45 minutes.This tes was administered in class
time.
Dokkai Vol. 1 30 -
C. Data Analysis
Total number 875 518
Data analysis technique was referring to three stages of
error analysisthat were data collection, identification and error
The number of combined kanji on Minna no Nihongo vol.1-2
classification. And then, explain error by describing location
is 845, on Dokkai vol.1 is 30, with notes in this calculation
of error, explaining the cause of error, and giving a correct
there is no repeating of same combined kanji. This amount
example.
will be even more if kanji on combined kanji is counted as
individual kanji. There are 875 composite kanji consisting of 2 III. RESULT AND DISCUSSION
kanji characters in Minna no Nihongo vol.1-2, and Dokkai
Research results of error analysis on reading and writing
vol.1,and 518 individual kanji on Kanjivol 1-2. So it can be
kanji can be classified on 2 category type of errorsi.e. A
said that standard of kanji mastery ability based on basic level
category was indicated that misspelling of target kanji due to
of textbook is approaching an N3 JLPT standard i.e. 650 kanji,
similarity of kanji graphic form or one of kanji graphic forms
but the word mastery has not met an N3 level standard that is
of combined kanji, then affect a meaning of target kanji. B
3750 words. If predicting from 875 combined kanji on based
category was indicated the error of writing target kanji on the
level textbooks, vocabulary mastery is above N5 standard
test of kanji no kakikata (writing of kanji) raising non-kanji
level of 800 words. In other words, the results of kanji
errors. A meant by non-kanji errors is that letter is not a kanji
teaching material computingin the basic Japanese textbooks
even though it is written from parts of two kanji, as shown in
for the students if referenced to Japanese language proficiency
following tables.
standard issued by The Japan Foundation can be said to
approach N3 standard for mastery of kanji that is 518 of 650
TABLE II. PARTICIPANT‟S ERRORS ON READING AND WRITING
kanji and above N5 standard for mastery of kanji is 875 of 800 KANJI
words.
Mastery of kanji is very important in learning Japanese, Misspelling of target kanji (A Category)
because it can affect the fluency in reading, and smoothness is Target kanji Partisipants
Analysis
the foundation towards understanding. Pexman [9] states the answer
Orthographic similarity effects
general purpose of reading is to generate the meaning of the between kanji 調 /chou/ and 週 /shuu/
words of the text being read. Reading text involves a number 調子 /choushi/ /shuuko/ on a phonetic radical called tsukuri.
of component processes, one of which is the recognition of the But /shuuko/ has no meaning, because
word visual. That is, the recognition of isolated words is /shuuko/ is a non-word
assumed as one component of reading skill [9]. McClung, Orthographic similarity effects
between kanji 濯 /taku/ and 曜 /you/
O'Donnell, and Cunningham quotes Ouellette, Treiman, 洗濯 /sentaku/ /senyou/ on a phonetic radical called tsukuri.
Shahar-Yames [10] share most theories and research on the But /senyou/ has no meaning, because
acquisition of orthographic knowledge emphasizes the /senyou/ is a non-word
importance of phonological encoding, some studies paying Orthographic similarity effects
/zasshi//zangy
attention to its connection with the spelling process. It is a 確認/kakunin/ between kanji確認 /kakunin/, 雑誌
ou/
/zasshi/, 残業 /zangyou/
recurring act of linking oral and written form, through Orthographic similarity effects
decoding or spelling, which helps to produce a good between kanji 転勤 /tenkin/ with kanji
orthographic representation. This is a very important thing for 運動 /undou/, which is kanji 運 /un/
reading skills. looks similar to kanji 転 /ten/ then on
転勤 /tenkin/ /undou/
kanji 動 /dou/ looks similar to kanji 勤
II. METHOD /kin/, if note further kanji 動 /dou/ has
a similar combination between kanji 転
A. Participants /ten/ and 勤 /kin/
Participants in this study were 53 students in third semester The sound of/en/ indicates a stopped
in Japanese education departmentatUnesa in Surabaya. 速達 calling process when participant is
/en/
/sokutatsu/ going to spell kanji 速達 /sokutatsu/
being /ensoku/ because one of kanji on

402
 Advances in Social Science, Education and Humanities Research, volume 222

Misspelling of target kanji (A Category) 歌舞伎 /kabuki/ it should be spelled /bu/, the other
kanji 遠足 /ensoku/ has an participant‟s perception on kanji 歌舞伎 /kabuki/ is a
orthographic similarity with kanji 速 similarity component between kanji 伎 /ki/ and妓 /ko/, two of
/soku/, as shown速達 /sokutatsu/ and
kanji there are equally in phonetic radical but differences in
遠足 /ensoku/. For more details, it is a
graphic similarity effects between kanji semantic radical. Kanji 舞妓 /maiko/ has the similarity on
速 /soku/ and 遠 /en/, and the meaning with geisha, but 歌舞妓 /shimaiko/ has no meaning,
similarity of kanji pronunciation /soku/
because /shimaiko/ is a non-word; kanji 段階 /dankai/ and
but denoted by different graphic form
速 /soku/ on 速達/sokutatsu/ with 足 階段 /kaidan/ is formed from 2 kanji which have exactly same
/soku/ on 遠足 /ensoku/ kanji graphical form but different order as shown (1) 段 (2) 階
orthographic similarityeffects between /dankai/ and (2) 階 (1) 段 /kaidan/. Similarity of kanji graphic
室 /shitsu/ on kanji室 and 屋 that is an element kanji-
kanji 会議室 /ya/ formers position under kanmuri on form on both kanji affects kanji readings, whereas the two of
/kaigishitsu/ kanji 室 /shitsu/ and tare on kanji 屋 kanji have very different meanings 段階 /dankai/ (level,
/ya/ class), 階段 /kaidan/ (ladder). Finally, all of data was
orthographic similarities effects
料 /ryou/ on
classified in 2 types. Typeof misspelled error category is
between kanji 料 /ryou/ and科 /ka/. In
kanji 資料 /ka/ both kanji there are similarities in classified in part of A, and non-kanji errors in part of B.
/shiryou/ phonetic radicals but differences in The results of analysis can be said that in recognizing
semantic radicals. kanji, participants also need to pay attention to similarity of
There are 2 kanji which have the same kanji graphic apart from pronunciation. Cunningham, Perry, &
graphic form and pronunciation but
different order as shown(1) 段 (2) 階
Stanovich in McClung, O'Donnell, and Cunningham [10] state
段階 /dankai/ /kaidan/
/dankai/ and (2) 階 (1) 段 /kaidan/. not only phonological abilities that influence development of
speech recognition skills, but comprehensive reading models
Non-kanji writing’s errors (B Category) also need to reflect on role of other components orthographic
processing, because at the time of reading individual processes
Partisipants
Targetkanji
answer
Analysis and encodes certain sequence of letters in writing. Although
Combine two the theorists differ in the notes of skillful reading, but all
kanji which emphasize the fact that word recognition effectively requires a
have same substantial supply of orthographic representations closely
pronunciation
さがす/sagasu/ /sagasu/ there Pronunciation similarity effects linked to semantics and phonological information. In addition,
were探すand these experts suggest that weakness in ability to obtain detail,
捜すbecome print-specific, and knowledge of words can lead to reading
one kanji difficulties. Thus, developing a general understanding of how
Combined two orthographic representations are obtained and how later to
kanji which
have
support word recognition is a necessary part of reading study
orthographic [10].
similarity McClung, O'Donnell, and Cunningham [10] add
ねる/neru/ there Orthographic similarity effect
orthographic processes defined as ability to form, store, and
were寝る/ner access orthographic representations, which (1) prioritize
u/ and 帰る
/kaeru/become
possible sequence of letters in the orthographic field of
one kanji language in question; and (2) are closely connected with
phonological, semantic, morphological, and syntactical
The similarity effect of visual factors (orthographic) information in their language. That is, definition includes
indicated as seen in example of data, that is kanji 転勤 /tenkin/ procedural aspect and the declaration of orthographic
spelled /undou/, /tenkou/occurs because of similarity between knowledge.
kanji 転勤 and kanji 運動. Kanji 運 /un/ has the same kanji IV. CONCLUSION AND SUGGESTION
component with kanji 転 /ten/, as follows 運→転 and kanji 動
This research provesthat orthographic similarity leads to
/dou/ appears as a composed component on kanji 転 + 勤 → false recognition of nijijukugo, with type of
動 /dou/. Then it affects a meaning of kanji, as if there is a errorsi.e.misspelling and non-kanji. An inconsistent
related meaning between /tenkin/ (work mutation), with pronunciation on kanji also cause participants to misspell
/undou/ has been interpretedas motion, /tenkin/ has been kanji. It gave rise to different onyomi readings (kanji reading
interpretedas displacement. All of them have an equal in with Chinese pronunciation)and then bring a new or wrong
meaning that indicates movement; orthographic similarity meaning in combine of two kanji. The relation of orthographic
between two kanji 室 and 屋 also affects reading of kanji. form, pronunciation,and meaning of kanjishow, besides the
Kanji 室 on kanji 会議室 /kaigishitsu/ should be read /shitsu/ orthographic effect, it is also important to do research from
but read /ya/; kanji 歌舞伎 /kabuki/ spelled 歌舞妓 phonological, morphological, or semantic lexical point of view
/shimaiko/, kanji 舞 can be spelled /mai/ and /bu/, but in kanji of kanji. This study has implications for kanji teaching

403
 Advances in Social Science, Education and Humanities Research, volume 222

materials within more attention to the effect of similarity

between kanji, especially when kanji studied is more
numerous in form combined kanji (jukugo).

REFERENCES
[1] J.F. Kess, and T. Miyamoto, The Japanese Mental Lexicon:
Psycholinguistic Studies of Kana and Kanji Processing. Amsterdam:
John Benjamin Publishing Company, 1999.
[2] Y. Abe, and M. Nakamura, Moji Goi O Oshieru: Nihongo Kyojuho
series 3. Tokyo: Hituzi Library, 2007.
[3] R. Yokosuka,”Goi oyobi Kanji Gakushu Sutorateji no Kenkyu,” in
Nihongo Kyoiku to Nihongo Gakushu: Gakushu Sutoratejiron ni
mukete, S. Miyazaki, and J.V. Neustupny, Eds. Tokyo: Kurushio
Shuppansha, 2002, pp. 97-116.
[4] T. Fushimi, M. Ijuin, K. Patterson, and I.F. Tatsumi, “Consistency,
Frequency, and Lexicality Effects in Naming Japanese Kanji,” Journal
of Experimental Psychology: Human Perception and Performance, Vol.
25, No. 2, 1999, pp.382-407.
[5] N. Shibahara, M. Zorzi, M.P. Hill, T. Wydell, and B. Butterworth,
“Semantic Effects in Word Naming: Evidence from English and
Japanese Kanji,” the Quaterly Journal of Experiment Psychology, 56A
(2), 2003, pp. 263-286.
[6] Y. Tanaka, et.al., Minna no Nihongo Vol. I&II. Tokyo: 3A Corporation,
1999.
[7] K. Nishiguchi, et.al., Minna no Nihongo Vol I&II: Kanji I&II (English
edition). Tokyo: 3A Corporation, 2001.
[8] A. Masano, et.al., Minna no Nihongo Vol I&II: Shokyu de Yomeru
topikku 25 Vol I&II. Tokyo: 3A Corporation, 2001.
[9] P.M. Pexman, “Meaning-based Influences on Visual Word
Recognition,” inVisual Word Recognition, Vol. I: Models and Methods
Orthograpy & Phonologhy, J.S. Adelman, Eds. New York: Psychology
Press, 2012, pp. 24-43.
[10] N.A. McClung, C.R. O‟Donnell, and A.E. Cunningham,“Orthographic
Learning and the Development of Visual Word Recognition,” inVisual
Word Recognition, Vol. I: Models and Methods Orthograpy &
Phonologhy,J.S. Adelman, Eds. New York: Psychology Press, 2012,
pp.173-190.

404