Bachelor’s Programme in Japanese

Swedish students and the phonetic component

of Kanji

by

Jesper Gustafsson

JAPK11
Bachelor’s Thesis (15 credits)
June 2019
Supervisor: Axel Svahn
Examiner: Arthur Holmer
Abstract
This thesis was written to answer the questions How much do Swedish students of Japanese
know about the phonetic component of kanji? and Does awareness of the phonetic com-
ponent help students guess the reading of kanji?. The phonetic component of kanji gives
you the reading of kanji, and can give you an edge in correctly reading unfamiliar kanji.
By doing some background research into the curricula of Swedish universities and kanji
knowledge of their students, an online questionnaire fit to their presumed knowledge was
constructed. The questionnaire measured the participants’ ability to guess the reading of
unfamiliar kanji before and after being given information regarding the phonetic compo-
nent of kanji. The results of this questionnaire indicates that the participants in general
seem to know of the phonetic component, but are not utilising it very well. There was a
significant improvement in most of the participants, and as such, introducing, or focusing
more on this topic in school could prove to be beneficial.

Keywords: Japanese language, kanji, phonetic component

1
Acknowledgements
I would like to thank my supervisor for help and guidance when trying to decide on my
research topic. I also highly appreciate the help I received from the universities professors
from answering my questions and helping me find a suitable set of kanji for the ques-
tionnaire. In addition, I would also like to thank Leonardo Boiko from namakajiri.net for
the wonderful introduction to the phonetic component, which sparked my interest in this
topic.

2
Contents

1 Introduction 5
1.1 Research questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2 Thesis layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2 Background 7
2.1 The history of kanji . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Radicals: The building blocks of kanji . . . . . . . . . . . . . . . . . . . 8
2.3 Kanji categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.4 The phonetic component . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.5 How to utilise the phonetic component . . . . . . . . . . . . . . . . . . . 10
2.6 Frequency in JLPT & Japanese curriculum . . . . . . . . . . . . . . . . . 10
2.7 Curricula at Swedish universities . . . . . . . . . . . . . . . . . . . . . . 11
2.7.1 Lund University . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.7.2 Gothenburg University . . . . . . . . . . . . . . . . . . . . . . . 12
2.7.3 Dalarna University . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.7.4 Stockholm University . . . . . . . . . . . . . . . . . . . . . . . 14
2.7.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.8 Textbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.8.1 Genki . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.8.2 Tobira . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.8.3 Kanji Look and Learn . . . . . . . . . . . . . . . . . . . . . . . 16
2.8.4 Heisig’s Guide to Remembering the Kanji . . . . . . . . . . . . . 16

3 Method 18
3.1 The approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.2 Before the questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.3 The questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.4 After the questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.5 How the kanji were chosen . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.6 Questions & Information . . . . . . . . . . . . . . . . . . . . . . . . . . 20

4 Results 22
4.1 Kanji knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3
 4.2 Change in guesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.3 Individual kanji . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

5 Conclusion 25
5.1 Limitations & Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.2 Research answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.3 Pedagogical implications . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.4 Future research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

References 28

A Phonetic Series with KanjiVG kanji 30

B Genki Kanji 32

C Potential Phonetic Series 34

D Data 36
D.1 Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

E Questionnaire 44

4
1 Introduction
Japanese is a language with four scripts, hiragana, katakana, kanji and rōmaji. Kanji,
literally translated as Chinese characters, are logographic characters imported from China.
In Japanese they typically represent lexemes and can be read in multiple ways.
Hiragana and katakana, collectively known as kana, are a Japanese invention and func-
tion more similarly to the latin alphabet. Each character represents a mora, the smallest
unit of sound in Japanese. Any word in Japanese can be written using only hiragana,
but they are most commonly used as grammatical markers or are attached to a kanji and
change depending on the inflection of the word.
Rōmaji is the romanization of the Japanese sounds using the Latin alphabet as a base.
There are a few systems for romanization. Two of these systems are the Hepburn system
and kunrei-shiki.
Out of these four scripts, this thesis will focus on the kanji script. The subject of kanji
has been researched extensively, and the specific areas of kanji readings are no exception.
This thesis aims to investigate a method of deciphering the reading of kanji and test it on
a small scale to see if it can help students figure out the reading of kanji.
The aforementioned deciphering method utilises the kanji component which represents
the sound of the kanji, the so called phonetic component. It is not unheard of foreigners
assuming that every kanji are purely pictographic or drawings, with no rules of the associ-
ated sound. Because of this assumption among foreigners, this thesis will also investigate
if this component is known to Swedish students of Japanese.
In reality, if we look at the tōyō kanji set (See glossary), just over a tenth of all kanji are
pictographic while nearly two-thirds are of the phono-semantic type (Suzuki, 2007:59).
Ivarsson (2016:48) also provides some statistics for kanji classifications. She mentions
that in the more modern kanji set, the jōyō kanji set, two-thirds are of the phono-semantic
type. In addition, she also mentions as much as 80% of all kanji could be of this type.
Phono-semantic kanji consist of a phonetic component, which represents sound and a se-
mantic component, representing meaning. One thing to note is that as previously men-
tioned, a kanji can have multiple readings. There are two types of readings: kun-yomi
(Japanese reading) and on-yomi (Chinese reading). The phonetic component only plays a
role when it comes to on-yomi.

5
1.1 Research questions
This thesis will aim to answer two questions.

1. How much do Swedish students of Japanese know about the phonetic component
of kanji?

2. Does awareness of the phonetic component help students guess the reading of un-
familiar kanji?

While research into the phonetic component is not unheard of, there seems to be a lack
of examining the usage from a learning perspective. Ivarsson (2016:64) provides a meta-
analysis of a few studies which implies that Chinese children seem to develop an aware-
ness of the phonetic component somewhere between ages six and nine years. Japanese
children on the other hand, seem to develop this awareness a bit later, still having limited
radical awareness and difficulty reading words with on-yomi into the 7th grade (Ivarsson,
2016:79-81). The reason for choosing to investigate these topics is that it would be inter-
esting to know to what degree the phonetic component could help students when reading
kanji. If there is a noticeable change it could be of interest to spend some time on this in
the classroom as it could make it easier for students to learn kanji both inside and outside
of the curriculum.

1.2 Thesis layout

Excluding this introductory chapter the thesis contains five more chapters. The second
chapter, Background, provides a short history lesson on kanji, their readings and previous
research on the subject of kanji reading and the phonetic component. It also goes into the
method of teaching kanji that some textbooks use, as well as how Swedish universities
teach kanji. The third chapter, Method, explains the method of acquiring data to answer
the research questions. The fourth chapter, Results, presents the results gathered trough the
method in the form of tables. The fifth chapter, Analysis, provides analyses of the results.
The sixth chapter, Conclusion, provides a summary of the thesis as well as discusses the
method used, results and analysis as well as possible practical implications and possible
future research.

6
2 Background
This chapter examines some previous research into kanji, the phonetic component and the
curricula of Swedish universities. Firstly a short history lesson on the origin and develop-
ment of kanji and their readings through history will be given. It will then take a deeper
look into kanji, their categories and building blocks, including the phonetic component.
In addition, it will also provide some information regarding the curricula at the Swedish
universities and common textbooks. Finally, a short summary will be given.

2.1 The history of kanji

Ivarsson (2016:32-36) provides a short description of the history of kanji. As previously
mentioned, kanji are characters imported from China. However, Japanese kanji and Chi-
nese hanzi can be quite different. They can be different in two ways, the phonetic aspect
and the graphic aspect. In Chinese, the characters generally only have one reading while
kanji more often that not carry at least two (Ivarsson, 2016:34). The readings of kanji can
be divided into two categories: kun-yomi and on-yomi. Kun-yomi refers to the reading of
kanji for native Japanese words while on-yomi refers to the reading of Chinese origin, used
in Sino-Japanese words. This study focuses on the on-yomi reading and the kun-yomi is as
such sidelined. Apart from the fact that Chinese and Japanese are languages with different
sounds there are others reason for the difference in the on-yomi of kanji and the Chinese
reading of its hanzi counterpart. It is in part due to time itself as the pronunciation de-
veloped differently in the countries separated by sea. Another reason for the difference is
the manner of import. Not only did the process of adoptation span several centuries, but
also across many regions and thus, the dialect it was imported from also differed signifi-
cantly. In the beginning of the adoption process, imports came from the southern Chinese
dialect via Korea. Between the 7th and 9th century, the import had shifted to the cen-
tral Chinese dialect, and lastly from the northern dialect of Chinese. Due to this gradual
and inconsistent adoptation, the on-yomi of kanji does not necessarily correspond to their
hanzi counterpart. One example of these kanji is 行 (walk). 行 can be read as gyō as
in 行列 (gyō-retsu, “procession”), based on the southern dialect, kō as in 行進 (kō-shin,
“parade”), based on the central dialect as well as an as in 行脚 (an-gya, “walking tour as
pilgrimage”), based on the northern dialect. Kanji has also in time been modified, leading
to various forms of the same kanji. Suzuki (2007:60) include the examples 学 (learn),
simplified from 學 and 国 (country), the simplified form of 國.

7
2.2 Radicals: The building blocks of kanji
In the first chapters of Seely (2016), the origin and structure of kanji are explained. Kanji
can be quite complex characters, ranging from one single stroke to more than twenty-
five strokes. They are compounds of different smaller characters sometimes referred to
as radicals or components. There are a total of 214 radicals, with some having multiple
forms, such as the alternate form of 人 (person), 亻, found in 袋 (bag). There are also
variants of radicals, such as 儿 for 人 and 足 for 疋. There is also 月 which carries the
meaning of moon, but is at the same time an alternate form of 肉 (meat) and 舟 (boat)
(Seely, 2016:34-40). Variants can also be seen in kanji and hanzi. One old example is 鬽
and 魅, both meaning “malevolent spirit”. Variants are however nowadays less common,
most likely due to more widespread education and official script simplifications in both
countries (Seely, 2016:16).

2.3 Kanji categories

Seely (2016:12-14) mentions the four larger out of the six traditional categories of Chinese
characters which are classified based on their original formation. The first classification
are pictographic characters (象形 shōkei). These represents pictures of something con-
crete, such as 木 (tree), and 山 (mountain). However, an issue with this type of kanji is
that it can be quite hard to represent more abstract concepts such as ”above” or ”below”.
These more abstract ideas are represented with kanji classified as diagrammatic charac-
ters (指事 shiji). These include concepts such as direction and numbers: 上 (above), 下
(below), 一 (one), 二 (two) and 三 (three). Simply adding a horizontal line to the exist-
ing pictographic kanji for tree 木 can represents parts of a tree: 本 (root) and 末 (apex).
Combining these pictographic and diagrammatic characters leads to another classifica-
tion, semantic composites (会意 kaii). One obvious example is 林 (grove) and 森 (forest)
consisting of two and three 木 (tree). Another example is 峠 (mountain pass). The left
side represents a mountain, and the right side represents ”up” and ”down”. In ancient
China one would often borrow kanji that had the same reading to represent a different
word, eventually leading to a lot of kanji carrying multiple meanings. Let us take 莫 for
instance. 莫 originally had the meaning of ”sunset”, but was borrowed to also represent
“there is none, not any”due to its similar reading. This could lead to confusion and even-
tually semantic radicals were added to these kanji to differentiate the different words. For
our example, this resulted in 暮, with the added semantic radical 日 (sun) carrying the
meaning of ”sunset”, and 莫 keeping the meaning of “there is none, not any”. These kanji
came to be known as phono-semantic composites (形声 keisei). The characters consists of

8
one part representing a meaning (metal, tree, bird) and one part representing the pronun-
ciation. One more example of this is 憬 (yearning), where the left side 忄 is the semantic
part meaning ”heart” or ”mind”, and the right part 景 representing the pronunciation of
the kanji, kei. This phonetic component can be quite useful to be aware of for reasons
discussed below.

2.4 The phonetic component

According to Ivarsson (2016:48-49), the majority of kanji actually fall into the phono-
semantic composites (PSC) classification. Within the jōyō kanji, around two thirds are
considered phono-semantic composites. However, unfortunately not every PSC kanji can
be consistently read correctly. Only 57.6% are considered completely consistent, that is to
say that they all have the same on-yomi as the phonetic component dictates. Combining
these percentages one can accurately guess the pronunciation of a character by the phonetic
component 40% of the time. This inconsistency of PSC kanji can partly be explained
as previously said, to the modification or change of kanji and pronunciation throughout
history. One example of a modified PSC kanji is 仮. 仮 is read as ka even though its
phonetic component 反 is read as han. This inconsistency is due to the simplification of
the character. The character originally was written as 假, with its right side representing the
pronunciation ka. In other words, the simplified character 仮 retained the pronunciation of
叚 while replacing it graphically with 反. Another example of this happening is 読 derived
from the original character 讀 (Seely, 2016:96). These inconsistent kanji make up about
9.7% of the jōyō kanji. The other 32.7% are considered partially consistent, meaning the
pronunciation is similar but not exact. An example of a partially consistent kanji is 海
which is read as kai even though the phonetic component 毎 is read as mai.
Interestingly enough, Ivarsson (2016:64) mentions that while this component plays
such a huge role, awareness of the phonetic component seems to develop quite late for
both Chinese and Japanese students. This development can be observed even later among
Japanese students, likely due to kanji having several readings while hanzi typically only
has one reading per character. Due to this difference in kanji and hanzi, Japanese stu-
dents need to learn more kanji characters before being able to make connections between
readings of kanji and phonetic component. Ivarsson (2016:66-67) goes on to cover an-
other study comparing second-language learners and natives which concludes that second-
language learners depend more on phonological information compared to native speakers.

9
2.5 How to utilise the phonetic component
In theory, using the phonetic component could be a great tool when it comes to learning
or recalling kanji, but how can one actually use this kanji to determine kanji? Looking
at the various phonetic lists of kanji included in the appendix of this thesis, a pattern that
appears is that the phonetic component is a large majority of the time on the right side of
the kanji when the kanji can be divided vertically, such as 館, with the semantic component
食 (food, eat) on the left and the phonetic component 官 (kan) on the right. When kanji
can be divided horizontally, the phonetic component is generally on the bottom, as in the
case of 芥 with 艹 (grass) on the top and 介 (kai) on the bottom. There are quite a few
exceptions to these rules, for example 効 and 資, where the phonetic component is to
the left and top respectively. Once the phonetic component has been identified, recalling
readings of kanji with the same component should lead to a fairly high change of guessing
correctly, or at the very least lead to the reading that has the highest probability of being
correct. Let us see an example of this process for the kanji 鉸. Step one is identifying
the phonetic component, it is highly probable that it is the component on the right side of
the kanji, 交. Step two is trying to recall readings of kanji with this component. A very
common word, 高校 (high school, kōkō) includes 校 (kō). It is a quick two-step process,
which can further be reduced to a one-step process when getting more familiar with these
phonetic components.

2.6 Frequency in JLPT & Japanese curriculum

Suzuki (2007:59-60) provides the two tables 2.1 and 2.2 which include data of kanji cate-
gory frequency in an example JLPT (Japanese Language Proficiency Test), as well as fre-
quency of kanji categories in the Japanese education system. From these tables, one can
see that for both the easiest level of the example JLPT, N4 as well as the first grade kanji
mostly consist of pictographic kanji. However, with each level and grade, this amount
greatly reduces while other types becoming more frequent. The majority of the kanji for
N2, N1, fifth grade and sixth grade are phono-semantic kanji. This makes sense as the
phono-semantic and semantic compounds are made out of pictographic and diagrammatic
kanji and learning these first would make learning subsequent kanji easier. The informa-
tion here is based on the jōyō kanji list before it was updated in 2010.

10
 Table 2.1: Table of kanji frequency in an example JLPT test (Suzuki, 2007:59)
type of kanji N4 N3 N2 N1 jōyō
pictographic 35(42.5%) 32(17.7%) 98(13.0%) 75(8.1%) 216(11.7%)
diagrammatic 9(11.3%) – 1(0.0%) – 10(0.5%)
semantic compounds 20(25.0%) 65(35.9%) 218(29.0%) 190(20.5%) 413(22.3%)
phono-semantic 19(12.5%) 82(45.4%) 430(57.2%) 658(71.1%) 1211(65.5%)

Table 2.2: Table of kanji frequency in the Japanese education system’s first six grades
(Suzuki, 2007:60)
type of kanji 1st grade 2nd grade 3rd grade 4th grade 5th grade 6th grade
pictographic 44(57.9%) 47(32.4%) 24(12.3%) 17(8.7%) 13(6.7%) 25(13.2%)
diagrammatic 5(6.6%) – – – – –
semantic 15(19.7%) 38(26.2%) 53(27.2%) 68(34.9%) 43(2.1%) 48(25.3%)
compounds
phono- 12(15.8%) 60(41.4%) 116(59.5%) 109(55.9%) 139(71.3%) 117(61.6%)
semantic
total 76 145 195 195 195 190

2.7 Curricula at Swedish universities

According to the statistics, one could argue that utilising the phonetic component may be
an advantage when it comes to learning kanji, but to what extent do the Swedish univer-
sities and their textbooks discuss or utilise the phonetic component? This section of the
thesis will look at the kanji-teaching side of Swedish universities. I will base the teach-
ings of Lund University on my own experiences of the courses and the others on comments
from professors of the respective universities.

2.7.1 Lund University

Lund University offers a full-time bachelor’s program. As many students from Lund Uni-
versity spend their third year as exchange students in Japan with different textbooks and
curricula, only the courses for the first two years will be looked at.

11
 In the course plan from the first 20-week course JAPC01 (Japanese Beginners’ Course)1
it says that at the end of the course the students are expected to be able to describe the three
writing systems, including the systematic structure of kanji. They are expected to master
hiragana and katakana as well as 175 basic kanji whose systematic structure is taught out.
In the course plan for the following 20-week course JAPC02 (Japanese: Level 1) 2 it
says that 250 additional kanji are taught and that the students are expected to be able to
identify and write over 400 kanji, and being able to use them in phrases and sentences.
The courses of the second year are JAPC03 (Japanese: Level 2) and JAPK11 (Japanese
B.A. Course). As of the publication of this thesis, I am currently attending the JAPK11
course, and have passed the other three. I will base the information for these courses on
my own experiences.
In JAPC01 we were introduced to the concepts of hiragana, katakana and kanji. We
were also introduced to a few semantic radicals along with learning new kanji. The exis-
tence of the phonetic component was however mentioned briefly enough to forget or not
mentioned at all. That is not to say that there was no strategy behind the method of teach-
ing us kanji. For example, if a kanji was similar to one we had learned before, this was
mentioned. And when components looked like a kana character, this was also mentioned,
such as ム in 私 and 強 or カ in 動 and 助. The same strategy was utilised in all courses,
but we were no longer expected to learn new semantic radicals past JAPC01. As for learn-
ing the reading of kanji, this was generally done by learning vocabulary which included
the kanji. As a result, there were quite a few kanji which we only learned one reading of.
An example is 私. We had already learned the word kun-yomi reading, watashi, but the
on-yomi reading shi was not taught when we were introduced to the kanji.

2.7.2 Gothenburg University

The following questions were asked to a professor at Gothenburg University.

1. Are the students taught about the different categories of kanji? (keisei, shokei, kaii,
shiji moji)

2. Are the students taught about the phonetic component in keisei kanji?

3. Do you use the phonetic component when you teach new kanji?

4. Do you have the impression that your students actively use the phonetic component
when they read/learn new kanji?
1
http://kursplaner.lu.se/pdf/kurs/sv/JAPC01, accessed 05-05-2019
2
http://kursplaner.lu.se/pdf/kurs/sv/JAPC02, accessed 05-05-2019

12
The answer to all of these questions was ”yes”. The professor was also asked which kanji
the students are expected to know at the end of the first two courses, which was the kun-
yomi and on-yomi of all kanji in Genki I and Genki II.

2.7.3 Dalarna University

The following questions were asked to a professor at Dalarna University.

Question 1: ”Are the students taught about the different categories of kanji? (keisei,
shokei, kaii, shiji moji)”
Answer: ”Yes, very briefly and roughly, but yes.”

Question 2: ”Are the students taught about the phonetic component in keisei kanji?”
Answer: ”Same as above.”

Question 3: ”Do you use the phonetic component when you teach new kanji?”
Answer: ”We don’t teach kanji much during the lessons because of the lack of teaching
hours, so we have kanji worksheets with some instructions as well as one online
lecture with several online tests. We mention about the point a little bit both in the
online lecture and the worksheet instruction, but not very much.”

Question 4: ”Do you have the impression that your students actively use the phonetic
component when they read/learn new kanji?”
Answer: ”They are learning the first book of GENKI right now, so there are not so many
kanjis they can apply their knowledge to read new kanji, but as you suggest, we
should probably emphasize more on that point.”

Question 5: ”In the course plans for ”Japanska I: Grundläggande språkfärdighet” it says
that after the course, the student should be able to ”redogöra för kanji-tecknens ur-
sprung och grundläggande struktur”. Could you perhaps explain this a bit more?
Are they supposed to be able to break down each kanji into their components and/or
radicals, should they know the names and/or meaning of the radicals?
Answer: They need to answer/pass a small online test after watching a lecture about kanji.
The questions are about the history of Kanji, different categories, guessing the read-
ing of new kanji from the component.”

Dalarna University does mention the phonetic component but does not utilise it very much
in their teaching methods. They do have online test(s) to test the students on kanji history,
categories and ability to read unfamiliar kanji.

13
2.7.4 Stockholm University
The following questions were asked to a professor at Stockholm University.

Question 1: Are the students taught about the different categories of kanji? (keisei, shokei,
kaii, shiji moji)

Answer: In my kanji lessons (japanska II), I don’t teach the categories. I intentionally
avoided introducing them because I thought it will be too much information in this
level.

Question 2: Are the students taught about the phonetic component in keisei kanji?

Answer: No. Among 指示文字・会意文字・象形文字・形成文字, I feel that 形

成文字 is the most complicated one. With this reason, I don’t see any reason to
introduce the concept of 形成文字 to the students.

Question 3: Do you use the phonetic component when you teach new kanji?

Answer: Do you mean if I teach both kun- and on- reading? If so, yes.

Question 4: Do you have the impression that your students actively use the phonetic com-
ponent when they read/learn new kanji?

Answer: I don’t know. Since I speak Japanese as a mother tongue, I use the phonetic
component when I read unknown kanji. However, the students in Stockholm may
not necessarily see it in the same way. If they have never seen a kanji, they may
simply give up without associating or actively using phonetic component to identify
an unknown kanji.

Question 4: In the course plans for Kanji IA, IB, IIA, IIB it says that after the course,
the student should be able to ”uppvisa kännedom om tecknens struktur och uppbyg-
gnad”. Could you perhaps explain this a bit more? Are they suppose to be able to
break down each kanji into their components and/or radicals, should they know the
names and/or meaning of the radicals?

Answer: Both in Kanji I and II, students are introduced both the names and meaning of
radicals. In addition to that, students are explained kanji by emphasizing on 会意
文字 which is the combinations of two/three components in one kanji.

The third question was misunderstood, and a follow up question was asked.

14
Follow up question: From your other answers I would assume that you do not explicitly
use the term phonetic component, but maybe you would mention that component?
Would you also mention that they share the same reading?

Answer: Now I understand your question, and the answer is no. I don’t mention the term
phonetic component or sharing the same reading at all.

Stockholm University does not mention or utilise the phonetic component at all in their
teachings.

2.7.5 Summary
Gothenburg University seems to be the only university that uses the phonetic component
extensively in their teaching method. The other universities only briefly or do not mention
these categories at all. But they all seem to at least touch on the subject radicals. All the
universities but Lund University mention the different categories.

2.8 Textbooks

2.8.1 Genki
Genki (Banno et al., 2011a,b) is a two-part textbook series with a workbook for each part.
Genki is used by all the Swedish universities. The first textbook very briefly mentions
the phono-semantic kanji (In the book referred to as ”phonetic-ideographic compounds”)
when it lists up classifications of kanji (Banno et al., 2011b:31). The textbooks do however
not use this or any other classification in the chapters intended to teach the students to write
and read the kanji, and therefore 一 and 時 are not differentiated by classification. What
is listed together with the kanji is: stroke order, some vocabulary, some readings and their
meaning. Genki does in other words not employ the phonetic component when teaching
kanji. Genki I and Genki II include 145 and 172 kanji respectively for a total of 317
characters.

2.8.2 Tobira
Tobira (Oka, 2009) is another popular series, which is used at Lund, Gothenburg and
Dalarna for the second year. Tobira uses a very similar list of kanji. One key difference is
that Tobira also includes a space for the radical of kanji and the radical’s name.

15
2.8.3 Kanji Look and Learn
Kanji Look and Learn (Banno and Ikeda, 2009) is a textbook used at Gothenburg Uni-
versity in their second year together with Tobira. Similarily to Genki before it, it shows
the on-yomi and kun-yomi readings of the kanji, stroke order as well as some vocabulary.
What is unique to this book is that they use pictures to help readers remember how to draw
them and what their meanings are.

2.8.4 Heisig’s Guide to Remembering the Kanji

Heisig’s Guide to Remembering the Kanji (Heisig, 2011, 2012) is a book series that utilises
the phonetic component to teach 3000 kanji. Heisig (2011:10) argues that it is inefficient
to learn the writing and reading of kanji simultaneously, which may be why he split up
his guide into three parts. The first book advises the use of mnemonics and stories to
remember how to write kanji. You can often find a series of kanji in a row that have
similar components since the guide typically introduces one or a few components and
then the following kanji consists of this component. The guide does however only give
the meaning and a story as well as the stroke order of a kanji and does not list neither the
kun-yomi nor the on-yomi. However, in book 2 Heisig does focus on the reading of kanji
and introduce what he calls ”signal primitives” which is the same as phonetic components.
While book 1 and book 2 consists of the same 2200 kanji, the order is different. Heisig
groups the kanji into 10 groups:

1. The Kana and Their Kanji: Characters from which kana are derived from.

2. Pure Groups: Characters where the signal primitive dictates the reading with no
exceptions.

3. One-Time Chinese Readings: Characters whose on-yomi readings are unique.

4. Characters with No Chinese Readings: Characters with no on-yomi reading.

5. Semi-Pure Groups: Characters where the signal primitive dictates the reading with
a single exception.

6. Readings from Everyday Words: Characters from everyday words.

7. Mixed Groups: Characters where the signal primitive dictates the reading of at least
two characters and do not belong to the Pure or Semi-Pure Groups.

8. Readings from Useful Compounds: Characters that are frequent in non-conversation.

Such as newspapers, billboards and signs.

16
 9. A Potpourri of Readings: Characters that do not fit into the previous groups cate-
gories but are too common to be left to the last group.

10. Supplementary Readings: Characters that are less common and do not fit into the
previous groups.

The categories which utilise the signal primitive are the pure, semi-pure and mixed groups
and together they contain the majority of all the kanji in the book. There is also the third
part in the series which includes an additional 800 kanji, which expands upon the groups
found in the second book.

17
3 Method

3.1 The approach

In order to answer the two research questions (How much do Swedish students of Japanese
know about the phonetic component of kanji?, Does awareness of the phonetic component
help students guess the reading of kanji?), an online questionnaire was sent to four Swedish
universities: Lund University, Gothenburg University, Stockholm University and Dalarna
University. Before constructing the questionnaire, professors of each university were
asked to check a list of phonetic series of kanji to make sure that the questionnaire would
contain relevant kanji and questions.

3.2 Before the questionnaire

In order for the questionnaire to give something of substance, how the questions were
formulated had to be carefully thought over. Students may use the phonetic component
of kanji without realising it and because of this, the questions need to be formulated in a
manner that can access this knowledge without explicitly asking for it.

3.3 The questionnaire

The questionnaire can be split into three parts.

1. Asking the participant to guess the reading of a set of kanji.

2. Questions and information regarding kanji and the phonetic component.

3. Asking the participant to guess the reading of the same set of kanji again.

The reason for this ordering is because of the second research question (Does awareness
of the phonetic component help students guess the reading of kanji?). By analysing the
proposed readings of each kanji in the first part and comparing them to the third part, one
can get an idea of whether the information helped them guess correctly.

18
3.4 After the questionnaire
To answer the first question, an analysis of the answers from the second part of the ques-
tionnaire was made. In order to answer the second question, an analysis of the answers
from the first and third part was made. The analyses are explained in more detail in the
next chapter. Due to time constraints, all the parts were sent as a single questionnaire. One
could argue that it might be better to have a period of time between the second and third
part or maybe follow up with a fourth part later on because the results might be skewed if
they have just been primed to look for the phonetic component.

3.5 How the kanji were chosen

In order for a kanji to be chosen for this questionnaire it had to pass two criteria and
preferably a third criteria.

1. It should not be included in the students’ course curricula.

2. It should share a phonetic component with at least one kanji that the student knows.

3. The student should also preferably be confident in the on-yomi of the known kanji.

One kanji that satisfies all of these criteria is 郊. It is not included in any of the students’
curriculum. It shares the phonetic component 交 (kō) with 校, which is included in the
common word 高校 (high school, kōkō) which is present in all the first-years’ curriculum,
which all students should be comfortable with.
To find a set of kanji based on these criteria, two lists were compared. The first list
were of kanji that all students should know (See appendix B). As all universities used
the Genki books in the curricula for their first year, this list was a list of all kanji from
Genki I and Genki II. The second list was based on phonetic series found online in the
form of a spreadsheet1 created by Leonardo Boiko, author of the blog Namakajiri (Boiko,
2011). From this spreadsheet, only entries with a reading-accuracy above 50% as well as
a size of minimum 5 characters were considered. The two lists were then compared to
find entries of phonetic series which contained at least one Genki kanji to make a third list
(See appendix A). The entries of this third list contains kanji which satisfy the first two
criteria, as all kanji outside the kanji found in Genki are not included in the students’ course
curricula and all kanji share a phonetic component with the kanji from Genki. To satisfy
the third criteria, the professors of the universities were given words using the on-yomi
reading of the kanji were searched for in the vocabulary found in Genki2 and asked if their
1
https://namakajiri.net/data/kanji/components_phonetic.kanjivg.tsv, accessed 05-05-2019
2
http://genki.japantimes.co.jp/resources/saku_tango, accessed 05-05-2019

19
students should be comfortable in the reading of the word, further reducing the number
of potential kanji a bit. From the potential kanji, a final set of 36 kanji were chosen from
18 different phonetic series, with two from each series. This final set can be seen below,
with kanji from the same series next to each other. For the questionnaire, the order of
appearance was randomized using an online randomizer3 but the randomized order was
the same for all participants.

Table 3.1: Chosen kanji

衙 梧 按 鮟 藁 犒
跣 筅 絆 袢 舘 棺
岱 玳 剱 鹸 抹 茉
廂 孀 鮫 鵁 撼 緘
錵 囮 峙 塒 漣 蓮
价 芥 柞 搾 澗 癇

3.6 Questions & Information

To answer the first research question (How much do Swedish students of Japanese know
about the phonetic component of Kanji?), a number of questions for the questionnaire were
produced. They had to be framed and ordered in a way that does not prime the participants
to the degree that their answers are influenced. Since the participants may unknowingly
make use of the phonetic component while not knowing of its existence or purpose, ques-
tions about their experience with reading kanji were put first, while questions explicitly
asking about the phonetic component were put afterwards. To see the questions in detail,
see appendix E. As introductory questions the students were asked which university they
attend, which year and how many kanji they estimate that they know. Next they were
asked to guess the reading of 36 different kanji. After this, they were asked questions
about their thought process when guessing as well as questions about kanji categories and
the phonetic component. They were also given information about how to utilise compo-
nent and then asked to guess the reading of the kanji again. As an ending question they
were asked if they feel that they had an easier time guessing the second time around. The
information given in the questionnaire concerning kanji and the phonetic component is
given in full below.

Kanji can be distributed into four different categories (pictographic / pictograms, diagrammatic / simple
ideograms, semantic / compound ideograms and phono-semantic / phonetic-ideographic). Pictographic
kanji are depictions of something concrete such as a tree/mountain (木/山), diagrammatic are more abstract
such as up/down (上/下), semantic kanji combines characters from the previous categories to represent
words such as forest/mountain pass (森/峠). The fourth category is the phono-semantic category. Phono-
semantic kanji can be broken into two components, the phonetic and the semantic component. An example
are the kanji for different types of metal. 銅 (bronze), 鉛 (lead), 銑 (pig iron), 鋼 (steel), 鉱 (ore) all share
the semantic component 金 (metal) on their left side, but what does the right side mean? The right side (the
phonetic component) does not convey any meaning but rather the on-yomi of the kanji. Unfortunately, due
3
https://www.browserling.com/tools/random-letters, accessed 04-20-2019

20
to both the writing of some kanji and their reading changing throughout history, it is not 100% sure that
the reading of kanji is the same as the phonetic component. 銀 (silver) is read as gin and not gon like 艮.
However, using this information, it is possible to correctly guess the reading of kanji that you have never
seen before. 広, 昿, 砿, 絋 and 鉱 are all part of the phonetic-series of 広, and are all read the same as 広,
kō. Out of the jōyō kanji, around two thirds are phono-semantic characters and out of these, slightly over
half are completely consistent, meaning that around one third of the kanji has their reading decided by their
phonetic component. The phonetic component is usually to the right or at the bottom of kanji, but they can
appear in other positions.

21
4 Results
This chapter will reference the tables found in Appendix D. These tables contain the an-
swers from all participants and if one wishes to disregard the answers from some par-
ticipants for reasons discussed below, the numbers might change. In the appendix, each
participant has an ID number which are also referred to when specific participants are dis-
cussed. The first and third part of the questionnaire where the participants were asked to
guess the readings of the kanji are differentiated as ”the first kanji test” and ”the second
kanji test”.
A total of 27 students participated in the questionnaire. Out of the 27 participants,
14 of them guessed correctly more often during the second kanji test. Since 16 of the
participants are from Stockholm University and only five from Gothenburg University,
three from Lund University and two from Dalarna University, there probably is not much
to say about the difference of the participants’ answers depending on their university, this
coupled with only three second year students participating, with the rest being first year
students, no difference will be made between participants from different universities or
year groups.
One reason why one may want to disregard some answers is that some participants
opted not to answer in one or both of the tests, some participants explicitly said that the
reason for not answering on the second kanji test was that they thought they would just an-
swer the same. Some of the participants showed an incredible difference between the first
and second kanji test, which could possibly have to do with participants searching for the
correct answers in between the two tests. Regardless of the reasons behind their answers,
their entries were still kept, but should be taken into consideration when answering the
second research question, whether awareness of the phonetic component helps students
guess the reading of unfamiliar kanji.

4.1 Kanji knowledge

From Table D.2 we can see that most of the participants use the components of kanji in
some way or another when trying to guess the reading of the presented kanji. Only one
participant thought that it is impossible to guess the reading from just the kanji. As for how
many of the jōyō kanji are keisei kanji, 14 of them answered half or less than half (or that
they did not know). Seven participants answered within the range of 50% - 59%. Nine
participants answered within the 60-70% range (including the participant who just said
”Majority”). Only two participants answered above 70% at 80%. Taking the estimation
from Ivarsson (2016:48) of 66% in mind, we can see that only 9 participants were correct

22
within 10 percentage points. Regarding where they have heard of the phonetic compo-
nent, most participants seem to have some knowledge of it, and classes or the internet are
the most common sources. 10 people have heard of it from class and eight people have
heard of it on the internet. Nine participants have not heard of the phonetic component
at all. However, this does not mean that they do not know what it is, they could simply
be unfamiliar with the specific term ”phonetic component”. This also applies to the terms
used for kanji categorization. Six of the participants found the given information new, 18
of the participants found it partly new, while only three participants did not consider any
of the information new. The three participants who did not find the information new had
a slight percental change at -3%, 0% and 3% in the second test.
If we look at Table D.1, there does not seem to be an obvious correlation between
how many kanji they are familiar with and their general knowledge of kanji. The biggest
correlation seems to be that the more kanji you can read, the more likely you are to use
the phonetic component actively, but general knowledge of kanji seem irregular. We can
also see that participants with knowledge of the phono-semantic category have all but one
heard of the phonetic component.
From Table D.3 we can also see that most participants had some knowledge of the
information given, but to most, at least some of the information was new. Some changes
are dramatic, up to 1000%, while others are not as dramatic. This does however not nec-
essarily mean that the information was less helpful in general for the participants with a
lower swing. The higher one scores on the first kanji test, the potential swing is lower.
The participants with over 10 correct answers in the first kanji test could be very diligent
students, confident in the reading of the kanji from Genki and had an easier time seeing
patterns with the kanji already in the first kanji test.

4.2 Change in guesses

From Table D.3 we can see that for 14 participants there was a positive shift toward getting
more correct answers after hey were given the information about the phonetic component.
For three of the participants there was a slight decrease in correct answers. For the re-
maining 10 participants, there was no difference in correct amount of guesses before and
after they were given information. 11 of the participants felt that they had an easier time
guessing. Whether the participant felt that it was easier to guess correctly during the sec-
ond kanji test seems to correlate with the change in correct answers, with the exception
of three participants. One of these participants thought it easier the second kanji test and
yet had no increase in correct answers. The other participants did not find it easier and yet
had a increase from three to five and zero to seven respectively.
One thing to note is that only answers which are 100% correct were considered correct
guesses. For instance, one participant answered ”kyō” instead of kō for the kanji 藁, which

23
one could say is a pretty close guess. There are also some answers which seem to indicate
that the participant did use the phonetic component, but for whatever reason decided to
spell out every component in the character. For example ”igo” for instead of just go for
the kanji 衙. Finding the phonetic component, but using the kun-yomi of that component
is also common, for example ”saki” for 跣, a kun-yomi reading for 先, instead of its on-
yomi, sen. Looking at one participant (ID 16) specifically and their comment, it seems that
just the possibility of being able to guess the reading increased their correct answers from
4 to 20, an increase of 400%. This participant did however already say that they used the
phonetic component when guessing in the first kanji test, which could indicate that just the
confidence boost or getting reminded of this information, can give a substantial increase
in being able to guess correctly.
Three participants (ID 10, 24, 27) in particular are interesting. They all had 0 correct
guesses in the first kanji test but had 21, 9 and 7 correct guesses in the second kanji test
respectively. This may seem like an unlikely increase and is further discussed in the next
chapter.
One participant (ID 21) only had one mistake in the first kanji test, but managed to
get it correct during the second kanji test. This kanji was 緘 kan, which they mistakenly
answered ”dou” in the first kanji test. This participant said that they knew the reading of
800 kanji, which is on the higher end of the participants. However, it is still amazing that
one participant could score this high. This could also be the case of a participant looking
up the reading of the kanji during the questionnaire.

4.3 Individual kanji

Looking at Table D.4 there does seem to be a pattern of a participants guessing more
correctly if the phonetic component is on the right hand side of the kanji, as is the most
common. 剱, 鵁, 藁, 衙 and 囮 were the only kanji which had a negative or neutral swing.
In both 剱 and 鵁 the phonetic component is on the left hand side of the kanji, while in 藁
and 囮, the phonetic component is in the middle. When the phonetic component was on
the right side, as in the other kanji of the pair, the swing was positive. One thing to keep
in mind is that unfortunately one kanji, 廂 was mistakenly included twice, instead of the
kanji 孀 in the first kanji test. This could be the reason why two participants (ID 1, 21)
had a change of one correct answer and could also partially contribute to the change in
several of the participants answers.

24
5 Conclusion

5.1 Limitations & Problems

As mentioned in the Methods chapter, one issue of this thesis was time constraint. Due to
time constraints, all the parts were sent as a single questionnaire. One could argue that it
might be better to have a period of time between the second and third part or maybe follow
up with a fourth part later on because the results might be skewed if they have just been
primed to look for the phonetic component. Another issue was the choice of kanji. Which
kanji do the students know? Which phonetic series would be good to choose from? How
many kanji should be included in the questionnaire?
Luckily, all the universities used the Genki books in their first year courses, and I
was able to find the set of kanji that the students should be comfortable with quite easily.
The professors of the universities were also very helpful in confirming which kanji they
should be able to read. It was quite hard to find a good list of phonetic series for kanji.
I did find a Wiktionary article1 with an extensive listing of Chinese phonetic series, but
it could be hard to remove all the hanzi without kanji counterparts. It also did not list
the japanese reading, which would mean I would need to find the reading for a huge
number of kanji. Eventually the spreadsheet2 from Boiko (2011) was decided upon as
it had information regarding predicted reading, size of phonetic series and how consistent
the series were. But this spreadsheet is not without faults. The phonetic series are not
based on actual phonetic components, but rather components in general. This meant for
example that small components such as ” 丶” would have their own ”phonetic series”.
Luckily these series had a very low consistency (some lower than 1%), as they are not
actual phonetic components. One could also argue that whether a component is officially
a phonetic component does not matter, as long as it seems to be a consistency of reading
between the kanji it is in.
Another concern is the balance of how consistent and how big the series should be. I
ultimately decided on a size of at least 5 kanji and a consistency of 50%. 50% may seem
low, but the list of the phonetic series included all on-yomi readings of kanji, not just the
most common one. For example, the phonetic series of 半 has 9 kanji in its series, all of
which can be read as han, but three of the characters also have extra readings, which gives
the series a potentially misleading 75% reading coverage.
Regarding the three participants (ID 10, 24 and 27) pointed out in the last chapter, there
1
https://en.wiktionary.org/wiki/Wiktionary:About_Chinese/phonetic_series, accessed 05-05-2019
2
https://namakajiri.net/data/kanji/components_phonetic.kanjivg.tsv, accessed 05-05-2019

25
may have been an issue of them looking up answers in the middle of the questionnaire.
Participant 10 entered nothing at all in their first round of guessing, and said ”I can’t guess.
They mean nothing to me.” when asked what they were thinking about when guessing
the reading of the kanji. The information given was partly new, but an increase from
0 to 21 correct guesses is quite unlikely. Perhaps this could be a case of the participant
looking up the answers. This participant did think that it was easier the second time around.
Participant 24 did also not answer anything the first time around but unfortunately they
did not leave a comment regarding how they base their guesses. They also interestingly
did not think it was easier the second time around. Participant 27 did answer the first time
around, but they also unfortunately did not leave a comment regarding how they based
their guesses. This participant did think that it was easier the second time around. While
the drastic change for participant 10 is a cause for alarm, the change for these two may
not be.
These issues could perhaps have been avoided by requiring the participants to answer
every answer as well as trying to make sure that they would not look up the answers in
between the two tests. The reasoning behind not requiring an answer for every single ques-
tion was that some participants may have quit the questionnaire as there were a total of 72
times that they had to guess the reading of kanji. It was unfortunately not mentioned in the
questionnaire that the participants should not look up the reading before they finished the
questionnaire, and it could have easily been mentioned. For future studies, a physical test
rather than an online one would most likely reduce the chance of these issues occurring.

5.2 Research answers

The two research questions for this thesis were How much do Swedish students of Japanese
know about the phonetic component of kanji? and Does awareness of the phonetic com-
ponent help students guess the reading of kanji?. Did the research help answering these
questions? For the first question, there definitely seems to be some knowledge of the pho-
netic component among most of the participants, but to which degree is very scattered and
less than half heard about it during class. The information about the phonetic component
was new or at least partly new to most of the participants of the questionnaire. As for the
answer to the second question, half of the participants had a noticeable increase in guess-
ing the readings correctly, some benefiting greatly. Only three participants had a slight
decrease while most of the stagnant ones did not appear to try especially hard to guess,
which could imply that even more participants could benefit from this. These results indi-
cates that yes, awareness of the phonetic component does indeed help students guess the
reading of kanji.

26
5.3 Pedagogical implications
Given the fact that there seems to be a noticeable improvement for most of the participants
after being given a short explanation of the phonetic kanji, maybe it would be a good idea to
take advantage of this during lectures. Gothenburg University and Stockholm University
does mention the component in their curricula, but it is possible that they would see an
improvement if they focused a bit more on it. It is also important to note that just because
you can read the on-yomi of a kanji, that does not mean that you can understand the word.
While being able to read an additional 50% of kanji sounds great, this does not mean that
you will increase your vocabulary very much. Most words using the on-yomi reading are
also compounds of several kanji, meaning that you will be less likely to be able to read all
of the kanji of a word. One more thing to keep in mind is that almost all participants of
this questionnaire were first year students. It is possible that there is a higher occurrence
of second year students who already know all of this and they may not benefit as much.
As an anecdote I could mention that I did not know about the phonetic component before
my second year, but I did notice that there seemed to be a orthographic similarity between
kanji that shared readings. When I first learned of the phonetic component during the
start of my second year while browsing the internet, I definitely seemed to gain more
confidence and could guess readings of unfamiliar kanji more often. I would also imagine
that if you put your mind to it and try to remember the most consistent series’ phonetic
components and their readings, you will be able to read many unfamiliar kanji. Given the
small investment and high reward, I do not see why this could not be implemented, or at
least taken into consideration when designing the courses. Another issue was the number
of participants. While I do think that I managed to answer my research questions even with
the number of participants I got, more would be even better. Specifically, more from other
universities. More than half my the responses came from Stockholm University. Perhaps
one could observe a difference based on the participants’ universities if there were more
from the other universities. I think I could have gotten a lot more responses from Lund
University had I went to one of their classes, but unfortunately the questionnaire was not
ready before their last class.

5.4 Future research

As always, an experiment should be repeated before making too many assumptions, if one
were to repeat this experiment, I would recommend doing it in a classroom setting with
paper and pen rather than an online test. This could avoid previously mentioned issues
of the participants not even trying to answer the questions or looking up answers in the
middle of the questionnaire. I would also recommend meeting the students in person, or
at least join them in the classroom via a live video-chat using a program such as Skype

27
or Zoom. The professor from Stockholm University and I arranged a Zoom-meeting and
it is likely the reason for why there was a great number of responses from Stockholm
University.
What would probably be the best, most accurate and also fairly simple to perform
experiment, could be to hold a short lecture (the information in this questionnaire could
easily be held in less than 15 minutes, including potential questions from students) for one
class and see if there is a desirable increase in the students’ ability to read kanji compared
to previous classes.
Another thing to research could be when the best time is to introduce this aspect of
kanji. Introducing it at the end of the first year or the start of the second year seems to
be a good point, as the majority of the participants of this questionnaire were first year
students. It could also be interesting to do a similar test on native Japanese speakers. I
have had some (unrecorded) conversations about this topic with native Japanese speakers,
albeit a very small set, and interestingly, they do not all seem to know about the phonetic
component, and when giving them similar (unrecorded) tests, they scored about the same
as the highest-performing participants of this questionnaire. Another thing to pursue could
be how much the position of the phonetic component matters, the results of this thesis’
questionnaire showed that students with knowledge of the phonetic component in general
may think that the phonetic component is always on the right hand side of the kanji.

28
Bibliography
Banno, E. and Ikeda, Y. (2009). Kanji Look and Learn, Tokyo: Japan Times.

Banno, E., Ikeda, Y. and Ohno, Y. (2011a). Genki: An Integrated Course in Elementary
Japanese II), 2nd edn, Tokyo: Japan Times.

Banno, E., Ikeda, Y. and Ohno, Y. (2011b). GENKI I: An Integrated Course in Elementary
Japanese), 2nd edn, Tokyo: Japan Times.

Boiko, L. (2011). Testing the predictive power of phonetic components in Japanese kanji,
viewed 2 March 2019, <https://namakajiri.net/nikki/testing-the-power-of-phonetic-
components-in-japanese-kanji/>.

Heisig, J. W. (2011). Remembering the Kanji (Book 1), 6th edn, Hawaii: University of
Hawaii Press.

Heisig, J. W. (2012). Remembering the Kanji (Book 2), 4th edn, Hawaii: University of
Hawaii Press.

Ivarsson, F. (2016). A Study of the L2 Kanji Learning Process: Analysis of reading

and writing errors of Swedish learners in comparison with level-matched Japanese
schoolchildren., PhD thesis, Gothenburg University.

Oka, M. (2009). Tobira: Gateway to Advanced Japanese Learning Through Content and
Multimedia, Tokyo: Kuroshio Shuppan.

Seely, C. (2016). The Complete Guide to Japanese Kanji: Remembering and Understand-
ing the 2,136 Standard Characters, Tokyo: TUTTLE Publishing.

Suzuki, Y. (2007). Kanji Kyōiku Ni Kansuru Oboegaki (A Note on Kanji-Education),

Departmental Bulletin Paper, Waseda University, viewed 26 December 2018,
<hdl.handle.net/2065/26475>.

29
 A Phonetic Series with KanjiVG kanji
Phonetic series with at least 50% reading coverage and a series size of at least 5, with at
least one kanji from the Genki book series in each series. The kanji set KanjiVG (6300
kanji). The series are sorted by reading overage and then size. Reading coverage takes all
readings of all kanji into account, which is why the percentage is lower than 100% even
though every kanji in some series can be read the same.

Component Kanji in phonetic Predicted Size of Readings Kanji with extra

series readings phonetic coverage % readings
series
安 安按晏案鞍鮟 アン 6 100
介 介价堺界畍疥芥 カイ 7 87.5 芥
広 広拡昿砿絋鉱 コウ 6 85.714 拡
連 嗹漣縺蓮連鏈 レン 6 85.714 漣
知 智痴知蜘踟 チ 5 83.333 聟
五 五伍吾唔圄寤悟 ゴ 14 82.353 圄衙
晤梧牾珸衙語齬
吾 吾唔圄寤悟晤梧 ゴ 12 80 圄衙
牾珸衙語齬
半 伴判半叛拌畔絆 ハン 9 75 伴判叛
胖袢
官 官棺管舘菅館 カン 6 75 綰菅
秋 啾愀愁楸湫甃秋 シュウ 11 73.333 愀湫鍬鰍
萩鍬鞦鰍
代 代垈岱玳袋貸黛 タイ 7 70 代垈袋
化 化囮花訛貨錵靴 カ 7 70 化囮花
次 咨姿恣次瓷粢茨 シ 9 69.231 懿次盗茨
諮資
高 嚆塙敲槁犒稾稿 コウ 13 68.421 亭塙嵩髞
縞膏蒿藁鎬高
先 先洗濳筅跣銑 セン 6 66.667 讚贊鑽
交 交佼傚効咬效校 コウ 17 60.714 佼咬校狡皎蛟較
狡皎絞纐蛟較郊 餃駮
餃鮫鵁
相 孀廂想相箱霜 ソウ 6 60 廂想湘相

30
僉 倹儉剣剱劍劒劔 ケン 17 58.621 僉匳斂歛瀲簽臉
嶮検檢瞼臉険險 験驗鹸
験驗鹸
間 嫺澗燗癇簡繝間 カン 7 58.333 澗燗簡繝間
奴 努呶奴孥帑弩怒 ド 8 57.143 呶孥帑怒拏
駑
咸 咸喊感憾撼緘轗 カン 9 56.25 咸喊減箴轗鍼鹹
鰔鹹
無 嘸廡憮撫無蕪 ブ 6 54.545 嘸憮撫無蕪
韋 偉圍幃緯葦違韋 イ 7 53.846 幃衛諱韓韜
弟 剃弟悌梯涕睇鵜 テイ 7 53.846 弟悌梯睇鵜
周 凋彫稠蜩調雕鯛 チョウ 7 53.846 周惆稠綢週
父 交佼傚効咬效校 コウ 17 53.125 佼咬斧校父爺狡
狡皎絞纐蛟較郊 皎蛟較釜餃駮
餃鮫鵁
古 估個倨凅古固姑 コ 34 50.746 個倨做克兢凅剋
居怙故枯楜沽涸 嫡尅居据摘敵涸
湖瑚痼箇糊罟胡 滴瑚箇糊胡苦裾
葫蛄蝴裾詁踞辜 謫踞適醐鋸鏑
醐鈷鋸錮餬鴣
乍 乍作咋搾昨柞炸 サク 11 50 乍作咋怎炸祚胙
窄筰胙酢 詐鮓
寺 侍塒寺峙恃持時 ジ 10 50 侍塒待恃特畤痔
畤痔蒔 等蒔詩
弔 剃弟悌梯涕睇第 テイ 8 50 弔弟悌梯睇第鵜
鵜
同 同恫桐洞胴銅 ドウ 6 50 恫桐筒粡興
末 抹末沫秣茉靺 マツ 6 50 末沫秣茉靺

31
 B Genki Kanji

一 二 三 四 五 六 七 八 九 十 百 千 万 円 時
L3:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
日 本 人 月 火 水 木 金 土 曜 上 下 中 半
L4:
16 17 18 19 20 21 22 23 24 25 26 27 28 29
山 川 元 気 天 私 今 田 女 男 見 行 食 飲
L5:
30 31 32 33 34 35 36 37 38 39 40 41 42 43
東 西 南 北 口 出 右 左 分 先 生 大 学 外 国
L6:
44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
京 子 小 会 社 父 母 高 校 毎 語 文 帰 入
L7:
59 60 61 62 63 64 65 66 67 68 69 70 71 72
員 新 聞 作 仕 事 電 車 休 言 読 思 次 何
L8:
73 74 75 76 77 78 79 80 81 82 83 84 85 86
午 後 前 名 白 雨 書 友 間 家 話 少 古 知 来
L9:
87 88 89 90 91 92 93 94 95 96 97 98 99 100 101
住 正 年 売 買 町 長 道 雪 立 自 夜 朝 持
L10:
102 103 104 105 106 107 108 109 110 111 112 113 114 115
手 紙 好 近 明 病 院 映 画 歌 市 所 勉 強 有 旅
L11:
116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131
昔 々 神 早 起 牛 使 働 連 別 度 赤 青 色
L12:
132 133 134 135 136 137 138 139 140 141 142 143 144 145
物 鳥 料 理 特 安 飯 肉 悪 体 空 港 着 同 海 昼
L13:
146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161
彼 代 留 族 親 切 英 店 去 急 乗 当 音 楽 医 者
L14:
162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177
死 意 味 注 夏 魚 寺 広 転 借 走 建 地 場 足 通
L15:
178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193
供 世 界 全 部 始 週 以 考 開 屋 方 運 動 教 室
L16:
194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209
歳 習 主 結 婚 集 発 表 品 字 活 写 真 歩 野
L17:
210 211 212 213 214 215 216 217 218 219 220 221 222 223 224
目 的 力 洋 服 堂 授 業 試 験 貸 図 館 終 宿 題
L18:
225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240
春 秋 冬 花 様 不 姉 兄 漢 卒 工 研 究 質 問 多

32
 Table B.1 continued from previous page
L19:
241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256
皿 声 茶 止 枚 両 無 払 心 笑 絶 対 痛 最 続
L20:
257 258 259 260 261 262 263 264 265 266 267 268 269 270 271
信 経 台 風 犬 重 初 若 送 幸 計 遅 配 弟 妹
L21:
272 273 274 275 276 277 278 279 280 281 282 283 284 285 286
記 銀 回 夕 黒 用 守 末 待 残 番 駅 説 案 内 忘
L22:
287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302
顔 情 怒 変 相 横 比 化 違 悲 調 査 果 感 答
L23:
303 304 305 306 307 308 309 310 311 312 313 314 315 316 317

33
C Potential Phonetic Series

Component Predicted Kanji from Word from Phonetic

reading Genki (chapter) Genki (chapter) series
安 アン 安 (13) 不安 (13) 安按 晏案
鞍鮟　
介 カイ 界 (16) 世界 (13) 介价 堺界
畍疥芥
広 コウ 広 (15) 広告 (13) 広拡昿砿絋
鉱
連 レン 連 (12) 連絡 (21) 嗹漣 縺蓮
連鏈
五 ゴ 語 (7) 日本語 (1) 五伍吾唔圄
寤悟晤梧
牾珸衙 語
齬
半 ハン 半 (4) 半 (1) 伴判半叛拌
畔絆 胖袢
官 カン 館 (18) 映画館 (15) 官棺 管舘
菅館
代 タイ 代 (14) 時代 (14) 代垈 岱玳
袋貸黛
化 カ 化 (23) 文化 (11) 化囮 花訛
貨錵 靴
高 コウ 高 (7) 高校 (1) 嚆塙敲槁犒
稾 稿縞膏
蒿藁鎬高
先 セン 先 (6) 先生 (1) 先洗濳筅
跣銑
交 コウ 交 (7) 高校 (1) 佼傚効咬效
校狡皎絞纐
蛟較郊餃鮫
鵁
相 ソウ 相 (23) 相談 (14) 孀 廂 想相
箱霜

34
僉 ケン 験 (18) 試験 (9) 倹儉剣剱
劍劒劔嶮検
檢瞼臉険險
験驗鹸
間 カン 間 (9) 時間 (4) 嫺澗 燗癇
簡繝間
咸 カン 感 (23) 感動 (13) 咸喊感憾撼
緘 轗鰔鹹
乍 サク 作 (8) 作文 (9) 乍作咋搾
昨柞 炸窄
筰胙酢
寺 ジ 時 (3) 時間 (4) 侍塒 寺峙
恃持時畤痔
蒔
末 マツ 末 (22) 週末 (13) 抹 末沫秣
茉靺

35
D Data
The results are divided into four tables. The first table shows the responses to some of
the short answers. The second table shows paraphrased longer answers. The third table
shows correct answers for the two kanji tests, as well as the difference in percentage. It
also includes answers to question 10 and 13, which are the most relevant to the potential
change in reading accuracy. The fourth table shows the collective correct answers for
each individual kanji. For each table, each row represents a single participant. The same
participant is represented by the same row in each table. In other words, the participant
represented by row 5 in table 1 is also represented in row 5 in the other tables. Each
participant has an ID to make it easier to traverse the different tables. The participants are
ordered by university and then by year. As the questions are sometimes quite long, they
cannot fit neatly into the table and for that reason, all the columns that have the header
Q1, Q2, Q3, etc. corresponds to Question 1, Question 2, Question 3, etc. The relevant
questions can be seen below each table. Comments from participants are also included at
the end of this appendix.

36
 Table D.1: Short answers
ID University Year Q1 Q2 Q4 Q5 Q6 Q7 Q11 Q12
1 Dalarna 1 A few hundreds Yes Yes Yes Yes Yes Yes Yes
2 Dalarna 1 2000 Yes Yes Yes Yes Yes Yes Yes
3 Gothenburg 1 300 Yes No Yes No Yes No No
4 Gothenburg 1 200 Yes Yes Yes No Yes No No
5 Gothenburg 1 100-150 Yes No Yes Yes Yes No No
6 Gothenburg 2 200 Yes Yes No No Yes Yes No
7 Gothenburg 2 700 Yes Yes No No Yes Yes Yes
8 Lund 1 30 Yes No No No No No No
9 Lund 1 About 340-350 Yes No No Yes Yes Yes No
10 Lund 1 ∼200 No No Yes No No No No
11 Stockholm 1 1000 Yes Yes No No Yes Yes Yes
12 Stockholm 1 100 No No Yes No Yes No Yes
13 Stockholm 1 400 No No No No No No No
14 Stockholm 1 100 Yes Yes No Yes Yes Yes No
15 Stockholm 1 350 Yes No Yes No No No No
16 Stockholm 1 350 Yes No No No Yes Yes Yes
17 Stockholm 1 400 No Yes Yes Yes Yes Yes Yes
18 Stockholm 1 150 Yes No Yes No No No Yes
19 Stockholm 1 150 Yes No Yes Yes Yes No No
20 Stockholm 1 approximately 400 Yes No Yes Yes Yes Yes No
21 Stockholm 1 800 Yes Yes Yes Yes Yes Yes Yes
22 Stockholm 1 150 Yes No Yes Yes Yes No No
23 Stockholm 1 100 Yes No Yes No No No No
24 Stockholm 1 200 No No No Yes No No No
25 Stockholm 1 200 Yes Yes Yes Yes Yes Yes Yes
26 Stockholm 2 Like over 100 No No Yes No No No No
27 Stockholm 2 40 Yes No Yes Yes Yes No No

Q1: How many kanji do you think that you can read? (on-yomi) Q2: Do you practice
writing or reading Japanese outside of the class-material? Q4: Do you find that you can
sometimes correctly guess the reading of kanji that you have never seen before? Q5: Did
you know that kanji can be grouped into different categories? Q6: Have you heard about
the phono-semantic/phonetic-ideographic category of kanji? Q7: Have you ever heard of
the phonetic component of kanji? Q11: Do you consciously use the phonetic component
when trying to guess the reading of kanji? Q12: Do you consciously use the phonetic
component when trying to remember the reading of kanji?

37
 Table D.2: Longer answers
ID Q3 Q8 Q9
1 2nd part of a kanji On the Internet 20%
2 Radicals On the Internet 20%
3 Nothing On the Internet Minority
4 Trees and fish etc The Kanjibook (genki) 50%
5 Phonetic component In class 50% - 60%
6 Nowhere Majority
7 Radicals In class 70%
8 Particles I have not heard about it No idea
9 Radicles On the Internet 60%
10 I can’t guess. They mean nothing to me. I have not heard about it I don’t know.
11 Parts, ”feeling” In class 50%
12 Nothing, impossible to guess In class 80%
13 Parts I know from other kanji I have not heard about it 30%
14 Components, wild guesses In class 60%
15 I have not heard about it No idea
16 Radicals, common on-yomi From classmates, On the Internet 30%
17 Parts I know From classmates 60%
18 Radicals, wild guesses I have not heard about it 70%
19 Elements, ”general feel” In class, On the Internet 15%
20 Phonetic component In class, On the Internet 70%
21 Phonetic component On the Internet 65%
22 Similar kanji, wild guesses In class, From classmates 50%
23 Parts I have not heard about it 60%
24 I have not heard about it 50%
25 Right hand side of kanji In class, From classmates 80%
26 Instinct and radicals I have not heard about it 33%
27 In class 60%

Q3: When guessing the readings of the kanji, what did you think about? Q8: Where have
you heard about the phonetic component? Q9: How many percentage of jouyou kanji do
you think are phono-semantic/phonetic-ideographic?
The full, un-edited answers for Question 3 can be found on the next page.

38
 1. ”I have learned that the 2nd part of a kanji is often the on reading, while the first
denote meaning”

2. ”radical”

3. ”Nothing, it’s very hard to think of a lead that would make me come up with a
possible reading. I’ve heard radicals can help you but I have no idea in what way.”

4. ”Trees and fish etc, I really don’t get all the coincidal multi-meaningbearing short
sounds yet. Haven’t figuered out how you” would talk about a spec kanji without
writing it down and showing it. But guessing you express it WITH the trail of
hiragana but still, then its also several different words.. so wierd haha.”

5. ”Phonetic component”

6. ””

7. ”I look at the radicals and if it has a radical I know from another kanji I often assume
it has the same reading. It can be difficult however if it has 2 different radicals that
I believe is the radical containing the reading.”

8. ”The what are they called, particles they consist of”

9. ”The radicals (just a little bit though, as I don’t know how, when or where in the kanji
they ever effect the pronounciation or if radicals that by themselves also are kanji
have a separate pronounciation as radicals). For some I simply used ”feeling”.”

10. ”I can’t guess. They mean nothing to me.”

11. ”On reading of the parts they consist of (or in the case of some of them I already
new the on reading)”

12. ”I could not think of anything, in my point it is impossible to figure out the onyomi
by just randomly guessing.”

13. ”Parts that I knew how to read/pronounce”

14. ”Some are completely wild guesses, others I guessed based on components that I
know the on-yomi of.”

15. ””

16. ”What common on-yomi exist, and also the radicals as they sometimes contribute
to readings.”

17. ”Identifying one part of the kanji with an on-yomi reading I know, and guessing that
reading.”

39
18. ”I try to look at the radicals of the Kanji, or take a wild guess. ”

19. ”Their elements and general feel”

20. ”I thought about which kanji can be found as components, and guess the reading
based on what I think is the phonetic component. ”

21. ”I tried seeing if there were any obvious phonetic component(s) in the kanji, or parts
that were similar to other kanji”

22. ”1.If there was anything i recognised from other kanji i knew, maybe some with sim-
ilar symbols have the same reading.” 2. Straight up guessing, but with the ””rules””
of kanji in mind. I have so far not seen a single kanji with more than 3 (maybe
4) mora/moura (don’t know how that is spelled). So I made my guesses with that
””rule”” in mind. I have also seen A LOT of しゅ、しゃ、ぎょ、じゅ and so on,
so I made sure to include som of those in there.”

23. ”Guessed basing on kanjis that they had parts of”

24. ””

25. ”I mostly guessed the reading by comparing the right side of the kanji to kanji that
I know.”

26. ”Instinct and radicals”

27. ””

40
 Table D.3: Before and After
ID Correct answers Correct answers % change Q10 Q13
before explanation after explanation
1 29 28 -3% No No
2 6 6 0% Yes Partly
3 1 11 +1000% Yes Partly
4 3 3 0% No Partly
5 21 18 -14% No Partly
6 15 19 +27% Yes Partly
7 26 26 0% No Partly
8 0 0 0% No Yes
9 3 17 +467% Yes Partly
10 0 21 ∞% Yes Partly
11 22 20 -9% No Partly
12 0 0 0% No Yes
13 3 5 +67% No Yes
14 14 21 +50% Yes Partly
15 0 0 0% No Yes
16 4 20 +400% Yes Partly
17 21 23 +10% Yes Partly
18 1 5 +400% Yes Yes
19 0 0 0% No Partly
20 13 19 +46% Yes Partly
21 35 36 +3% Yes No
22 1 11 +1000% Yes Partly
23 0 0 0% No Yes
24 0 9 ∞% Yes Partly
25 25 25 0% No No
26 1 1 0% No Partly
27 0 7 ∞% No Partly

Q10: Was the information about keisei kanji and the phonetic component new to you?
Q13: Do you feel that you had an easier time guessing the reading of kanji the second
time around?

41
 Table D.4: Change for individual kanji
衙 梧 按 鮟 藁 犒
8/8 7 / 11 8 / 14 6 / 14 8/7 9 / 16
(0%) (57,1%) (75%) (133,3%) (-12,5%) (77,8%)
跣 筅 絆 袢 舘 棺
11 / 16 8 / 16 10 / 14 9 / 13 11 / 16 11 / 15
(45,5%) (100%) (40%) (44,4%) (45,5%) (36,4%)
岱 玳 剱 鹸 抹 茉
6/7 11 / 15 6/4 7 / 10 5 / 10 7/9
(16,7%) (36,4%) (-33,3%) (42,9%) (100%) (28,6%)
廂 孀* 鮫 鵁 撼 緘
1/2 ? /2 7 / 11 5/4 11 / 15 8 / 12
(100%) (? %) (57,1%) (-20%) (36,4%) (50%)
錵 囮 峙 塒 漣 蓮
4/6 5/5 4/7 4/7 5/7 6/7
(50%) (0%) (75%) (75%) (40%) (16,7%)
价 芥 柞 搾 澗 癇
5/9 5/8 6/8 5/8 9 / 13 5/6
(80%) (60%) (33,3%) (60%) (44,4%) (20%)

This table shows the correct amount of guesses for each kanji in the first kanji test and
the second kanji test. It also shows the swing in percentage points. Unfortunately, in the
first kanji test, the kanji 廂 was mistakenly included twice, instead of the kanji 孀, which
means that there was no data for it during the first test.

42
D.1 Comments
4. ”I did the best I could the first time, sry. ”

7. ”I’m sorry I just skipped it the second time because I feel that I would have put in
the exact same answers since I pretty much used this method already. This was fun
though and I learned some new stuff, good luck on your report!”

13. ”This is really good information to know, but I felt that I couldn’t really use it on
many of the kanji because I don’t know/ remember the onyoumi readings for many
of the kanji. ”

16. ”I feel like without knowing about the phonetic components, you feel a lot less con-
fident in guessing the readings. The other time around, despite not really knowing
any more, just knowing that I had a certain chance of being correct was better than
just guessing blindly, so it gives you more confidence to guess.”

43
E Questionnaire

44
Reading of kanji https://docs.google.com/forms/d/1xRUA-HajK6FtR9HrHcIjfNlhqCYd...

Reading of kanji
This questionnaire aims to measure your ability to guess the on-yomi (Chinese reading) reading of
kanji that you have not learned.

*Obligatorisk

1. Which University do you study at? *

Markera endast en oval.

Gothenburg University
Lund University
Stockholm University
Dalarna University

Övrigt:

2. Which year are you in? *

Markera endast en oval.

1st year
2nd year
Higher

3. How many kanji do you think that you can

read (on-yomi)? *

4. Do you practice writing or reading Japanese outside of the class-material? *

Markera endast en oval.

Yes
No

Guess the reading! - Part 1

Keep in mind that I assume that you have not yet learned these 20 kanji. Even if you don't have a clue
on how to read the kanji, just try to guess the on-yomi. Please use either hiragana (あ), katakana (ア)
or roomaji (a/A) when filling in.

5. 衙

6. 按

1 of 10 4/27/2019, 8:02 PM

45
Reading of kanji https://docs.google.com/forms/d/1xRUA-HajK6FtR9HrHcIjfNlhqCYd...

7. 澗

8. 藁

9. 跣

10. 絆

11. 梧

12. 舘

13. 岱

14. 剱

15. 玳

16. 棺

17. 抹

18. 袢

2 of 10 4/27/2019, 8:02 PM

46
Reading of kanji https://docs.google.com/forms/d/1xRUA-HajK6FtR9HrHcIjfNlhqCYd...

19. 廂

20. 鮫

21. 撼

22. 錵

23. 緘

24. 鹸

25. 峙

26. 漣

27. 价

28. 囮

29. 柞

30. 癇

3 of 10 4/27/2019, 8:02 PM

47
Reading of kanji https://docs.google.com/forms/d/1xRUA-HajK6FtR9HrHcIjfNlhqCYd...

31. 蓮

32. 搾

33. 筅

34. 犒

35. 廂

36. 鵁

37. 茉

38. 塒

39. 芥

40. 鮟

Questions about kanji

41. When guessing the readings of the kanji, what did you think about?

4 of 10 4/27/2019, 8:02 PM

48
Reading of kanji https://docs.google.com/forms/d/1xRUA-HajK6FtR9HrHcIjfNlhqCYd...

42. Do you find that you can sometimes correctly guess the reading of kanji that you have
never seen before? *
Markera endast en oval.

Yes
No

43. Did you know that kanji can be grouped into different categories? *
Markera endast en oval.

Yes
No

44. Have you heard about the phono-semantic/phonetic-ideographic category of kanji? *

Markera endast en oval.

Yes
No

45. Have you ever heard of the phonetic component of kanji? *

Markera endast en oval.

Yes
No

The phonetic component & jouyou kanji

The phonetic component is one of the components that can make up a kanji. There are several types
of kanji, the ones making use of this component are the phono-semantic/phonetic-ideographic (keisei)
kanji. The jouyou kanji are the 2136 kanji that are taught in Japanese schools from primary until high-
school and are the ones that allowed to be used in official governmental documents.

46. Where have you heard about the phonetic component? *

Markera alla som gäller.

In class
From classmates
On the Internet
I have not heard about it

Övrigt:

47. How many percentage of jouyou kanji do you

think are phono-semantic/phonetic-
ideographic? *

The Phonetic Component & Keisei kanji

Kanji can be distributed into four different categories (pictographic/pictograms, diagrammatic/simple

5 of 10 4/27/2019, 8:02 PM

49
Reading of kanji https://docs.google.com/forms/d/1xRUA-HajK6FtR9HrHcIjfNlhqCYd...

ideograms, semantic/compound ideograms and phono-semantic/phonetic-ideographic). Pictographic

kanji are depictions of something concrete such as a tree/mountain (⽊/⼭), diagrammatic are more
abstract such as up/down (上/下), semantic kanji combines characters from the previous categories to
represent words such as forest/mountain pass (森/峠). The fourth category is the phono-semantic
category. Phono-semantic kanji can be broken into two components, the phonetic and the semantic
component. An example are the kanji for different types of metal. 銅 (bronze), 鉛 (lead), 銑 (pig iron),
鋼 (steel), 鉱 (ore) all share the semantic component ⾦ (metal) on their left side, but what does the
right side mean? The right side (the phonetic component) does not convey any meaning but rather the
on-yomi of the kanji. Unfortunately, due to both the writing of some kanji and their reading changing
throughout history, it is not 100% sure that the reading of kanji is the same as the phonetic component.
銀 (silver) is pronounced gin and not gon like ⾉. However, using this information, it is possible to
correctly guess the reading of kanji that you've never seen before. 広, 昿, 砿, 絋 and 鉱 are all part of
the phonetic-series of 広, and are all pronounced as 広 (kou). Out of the jouyou kanji, around two
thirds are phono-semantic characters and out of these, slightly over half are completely consistent,
meaning that around one third of the kanji has their reading decided by their phonetic component. The
phonetic component is usually to the right or at the bottom of kanji, but they can appear in other
positions.

48. Is this information new to you? *

Markera endast en oval.

Yes
No
Partly

49. Do you consciously use the phonetic component when trying to guess the reading of
kanji? *
Markera endast en oval.

Yes
No

50. Do you consciously use the phonetic component when trying to remember the reading of
kanji? *
Markera endast en oval.

Yes
No

Guess the reading! - Part 1

Keep in mind that I assume that you have not yet learned these 20 kanji. Even if you don't have a clue
on how to read the kanji, just try to guess the on-yomi. Please use either hiragana (あ), katakana (ア)
or roomaji (a/A) when filling in.

51. 衙

52. 按

6 of 10 4/27/2019, 8:02 PM

50
Reading of kanji https://docs.google.com/forms/d/1xRUA-HajK6FtR9HrHcIjfNlhqCYd...

53. 澗

54. 藁

55. 跣

56. 絆

57. 梧

58. 舘

59. 岱

60. 剱

61. 玳

62. 棺

63. 抹

64. 袢

7 of 10 4/27/2019, 8:02 PM

51
Reading of kanji https://docs.google.com/forms/d/1xRUA-HajK6FtR9HrHcIjfNlhqCYd...

65. 孀

66. 鮫

67. 撼

68. 錵

69. 緘

70. 鹸

71. 峙

72. 漣

73. 价

74. 囮

75. 柞

76. 癇

8 of 10 4/27/2019, 8:02 PM

52
Reading of kanji https://docs.google.com/forms/d/1xRUA-HajK6FtR9HrHcIjfNlhqCYd...

77. 蓮

78. 搾

79. 筅

80. 犒

81. 廂

82. 鵁

83. 茉

84. 塒

85. 芥

86. 鮟

End of Questionnaire questions

Thank you so much for participating in this questionnaire.

87. Do you feel that you had an easier time guessing the reading of kanji the second time
around? *
Markera endast en oval.

Yes
No

9 of 10 4/27/2019, 8:02 PM

53
Reading of kanji https://docs.google.com/forms/d/1xRUA-HajK6FtR9HrHcIjfNlhqCYd...

88. Any other comments?

Tillhandahålls av

10 of 10 4/27/2019, 8:02 PM

54
55