Semantics in Japanese Kanji
description
Transcript of Semantics in Japanese Kanji
Semantics in Japanese KanjiSemantics in Japanese Kanji
山
yama
やま
登山[tozan]
climbing
火山[kazan]
volcano
KanjiKanji• Around 50 000 kanji
• 5th century from China
• 3000 in daily usage• 200 kanji account 50 %• 1000 kanji 90 %• 2000 kanji 99%
• Varying pronunciations due to historical reasons: – on-yomi (chinese reading)– kun-yomi (japanese reading)
下 o-, kuda-, sa-, shita, shimo (kun)ka, ge (on)
= under, lower, down, inner, inferior, latter half下 shita/shimo =under, down, below
下 moto = under
… 下 ...ka = under..., below...(-> 影響下 = under the influence)
下りる oriru = descend, get down下ろす orosu = take/put down
下さい Kudasai = give to me, grant to me
下さる kudasaru = give, grant, bestow
下がる sagaru = hang down(vi), come down
下げる sageru = hang(vt), lower, move back
下町 shitamachi = downtown (+town)
下着 shitagi = underwear (+wear)
靴下 kutsushita = socks (shoe+)
下院 kain = Lower House (+House)
下記 kaki = below-mentioned (+describe)
下部 kabu = lower part (+part)
下落 geraku = fall, degradation (+fall)
下品 gehin = vulgar, unrefined (+quality)
下半期 shitahanki = latter half year
(+half+period)
下旬 gejun = last ten days (+ tendays)
* 下手 heta = unskillful, poor at (+hand)
日本語
日本
本日 語
Nihon-goJapanese
[nichi], [jitsu]
[-ka], [hi]
[-go]
language
word
sun, day
hon (-bon, -pon)
moto
月曜日日曜日休日
* 今日
book,
origin, root, base,
true, real, this,
slender obj.
Japan
言
i-
gen, kon
語る
[nihon]
[kataru]
to talk
吾
吾がI, my me
言う言葉言語
伝言…
Two sets of syllabary charactersTwo sets of syllabary characters
Hiragana ひらがな- Transformed originally from Kanji
in the 9th century
• Used:– to represent readings of Kanji – to teach new Kanji– to represent grammatical features
(particles, auxiliary verbs and the inflections of words)
Katakana カタカナ• Correspondent to hiragana (46
moras)
• Used– in foreign names– in loan words– in onomatopoemic words
このテキストは漢字とひらがなと
カタカナで書いてある。
[ko-no]this
[te-ki-su-to]text
[kan-ji]
[hi-ra-ga-na]
[ka-ta-ka-na][ka-i-te a-ru]
is written
[wa]topic particle
[de]by
[to]and
山
やま yama
Previously held believe in reading Japanese
Naming Kanji slower than HiraganaNaming Kanji slower than HiraganaFeldman & Turvey (1980)Feldman & Turvey (1980)
• Color words usually are written in Kanji
frequency effect predicts that Kanji naming faster
黒緑
茶色灰色朱色栗色
くろみどり
ちゃいろはいいろしゅいろくりいろ
kuro
midori
chairo
haiiro
shuiro
kuriiro
Kana was named fasterKanji -> lexicon -> phonology?
Saito (1981)Saito (1981)- abstract -- abstract -
• Reading-out task: • Kana words were read out faster than Kanji words
• Silent reading:• Kanji were judged faster than Kana.
Kanji words easier to identify than Kanji words easier to identify than Katakana wordsKatakana words
(Yamada, Mitarai & Yoshida, 1991)(Yamada, Mitarai & Yoshida, 1991)
**** ****
新聞配達 エアコン
****
1s
30ms
response
Tachistoscopic study
1-4 characters
kanji/katakana
word/non-wordエアコン
****
Kanji Katakana
2 4 2 4
Word .80 .53 .38 .41
Non-word
.02 .00 .09 .00
Kanji Katakana
2 4 2 4
Word 82.7 64.2 46.1 45.5
Non-word
31.0 13.3 30.4 15.1
Whole word identification Whole word identification Constituent word identificationConstituent word identification
- Kanji words and their constituent characters were identified faster than katakana words of same lenght
- The effect of script type disappeared in non-word condition
Not complexity but inferability
Vocal inference only in KanaVocal inference only in KanaKimura (1984)Kimura (1984)
Synonymity judgement of pairs of words in
1) Kanji
2) Hiragana,
with or without concurrent articulation (nro 1-5)
ichi, ni, san, yon, go, ichi, ni, san,
yon, go, ichi, ni, san, yon, go, ...
結果 ー 実験けっか ー じっけん
CRITICS (Leong & Tamaoka, 1995)- Unfamiliarity of seeing words in
Hiragana - Homonyms in Hiragana (7 of 30
pairs) -esim. kyoukai =church, association or border
Semantics and phonological Semantics and phonological access access
Yamada (1998)Yamada (1998)
かわ -> kawa 川 -> kawa
かわ -> river川 -> river
1) Naming
2) Translation
faster than
faster than
Reading Kanji in lexical and semantic level:Reading Kanji in lexical and semantic level:semantics, ortography and phonologysemantics, ortography and phonology
(Morita & Tamaoka)(Morita & Tamaoka)
1) Lexical decision task "To respond as quickly and as accurately as possible deciding if the item was
correct japanese two-kanji compound word
2) Proofreading: " to decide as accurately as possible if the words in the sentence were correct"
in phonology: "if you find nonword in the sentence please press the button as soon as possible"
3) Semantic decision at the sentence level"To read by meaning and decide whether the words in the sentence were
correct/misspelling (in phonology)"
Ortography of Kanji:Morita and Tamaoka (2001)
Lexical decision Proofreading Semantic decision
RT (ms) Error (%) RT (ms) Error (%) RT (ms) Error (%)
Correct "Yes" 720 3.2 2506 4.6 2465 6.0
Correct "No"
similar
919 43.1 2618 55.6 2447 50.0
Correct "No"
dissimilar
821 6.5 2088 10.0 2228 4.2
予約 booking, reservation矛約 orthographically similar nonword 吐約 orthographically dissimilar nonword
p<.005 p<.005 p<.01 p<.001 p<.005 p<.001
Semantics of Kanji:Morita and Tamaoka (2001)
Lexical decision Proofreading Semantic decision
RT (ms) Error (%) RT (ms) Error (%) RT (ms) Error (%)
Correct "Yes" 710 4.2 2385 1.4 2443 6.9
Correct "No"
similar
859 18.5 2292 38.0 2500 14.8
Correct "No"
dissimilar
792 6.0 2402 9.8 2195 1.9
残額 [zan-gaku] = The money left over (real word)
余額 semantically similar nonword ( 余 ~ 残 = left over)
乱額 semantically dissimilar nonword ( 乱 = disorder)
p<.005 p<.001 p<.05 p<.001 p<.001 p<.001
Phonology of Kanji:Morita and Tamaoka (2002)
Lexical decision Proofreading Semantic decision
"no" RT (ms) Error (%) RT (ms) Error (%)
RT (ms) Error (%)
Pseudo-homophones
842 9.4 2066 8.3 1955 7.7
Nonwords 802 6.2 2130 12.0 2167 3.6
規則 [kisoku] = rule regulation (real word)
基則 pseudo-homophone (nonword)
想則 [sou-soku] phonetically dissimilar (nonword)
p <.10 p<.05 non non p < .01 p<.05
Reaction times:Comparison of 3 studies
600
650
700
750
800
850
900
950
Ortography Semantics Phonology
Correct "Yes"
Correct "No" (similar)
Correct "No" (dissimilar)
1900
2000
2100
2200
2300
2400
2500
2600
Ortography Semantics Phonology
190020002100220023002400250026002700
Ortography Semantics Phonology
Experiment 1: Lexical decision
Experiment 2: Proofreading Experiment 3: Semantic decision
Error rates (%)Comparison of 3 studies
0,0
10,0
20,0
30,0
40,0
50,0
Ortography Semantics Phonology
Correct "Yes"
Correct "No" (similar)
Correct "No" (dissimilar)
0,0
10,0
20,0
30,0
40,0
50,0
60,0
Ortography Semantics Phonology
0,0
10,0
20,0
30,0
40,0
50,0
60,0
Ortography Semantics Phonology
Experiment 1: Lexical decision
Experiment 2: Proofreading Experiment 3: Semantic decision
Not only semanticsNot only semantics
• Leong & Tamaoka (1995)
” kanji are not always accessed directly through the visual or lexical route
• difficult/rare kanji may be processed via phonetic recoding
• chlidren and less skilled readers– more influence of concurrent articulation
Another storyAnother story
• Parallel processing of semantics and phonology
• For example:Phonologically mediated access to meaning for Kanji: Is a rows still a rose in Japanese Kanji?
(Wydell, Patterson and Humphreys, 1993)
Parafoveal viewParafoveal viewWhat can be extracted from paravovea in natural reading?
• Word length information (Rayner, Fischer, & Pollatsek, 1998)
• Orthographic features (Beauvillain & Doré, 1998)
• Frequency (Kennedy, 1998; 2000; Hyönä & Bertram, 2004)
• Phonology• Semantics
Text features Readers skillsScript type
Semantics and parafoveal view in Semantics and parafoveal view in JapaneseJapanese
- The role of semantics in reading Kanji- Meaning entities
- High information density
Semantic influence from parafoveal kanji?
Experimental settingExperimental settingSemantics Ortography REeading and
meaning
Identical 赤 + + [aka]
red
semantically similar
血色
+ - [chi] blood
[iro]color
Ortographically similar
歩 - + [aru(ku)]
[ho] walk
control word 気 - - [ki] spirit
Boundary techniqueBoundary technique
1) そしてその花は赤になった。 red
2) そしてその花は血になった。 blood
3) そしてその花は歩になった。 walk
4) そしてその花は気になった。 spirit
Then that flower became red.
赤になった。 red
赤になった。 red
赤になった。 red
赤になった。 red
Important featuresImportant features
– Natural reading condition– Frequency (word/character?) – Complexity– Number of strokes– Number of constituents– Familiarity and frequency – Age of Acquisition– Imageability– Reader skills
ReferencesReferences
• Feldman, L.B. & Turvey, M.T. (1980). Words written in kana are named faster than the same words written in kanji. Language and Speed 23, 141-147 (abstract only)
• Flores d'Arcais, G. B. & Saito, H. (1993). Lexical decomposition of complex Kanji characters in Japanese readers. Psychological Research, 55, 52-63.
• Flores d'Arcais, G. B., Saito, H., & Kawakami, M. (1995). Phonological and semantic activation in reading kanji characters. Journal of Experimental Psychology: Learning, Memory and Cognition, 21, 34-42.
• Kess, J. F. & Miyamoto, T. (1997). Accessing the japanese mental dictionary through the japanese writing system. ???
• Kimura, Y. (1984). Concurrent vocal interference: Its effect on kana and kanji, Quarterly Journal of Experimental Psychology, 36 A, 117-131.
• Leong, C. K. & Tamaoka, K. (1995). Use of phonological information in processingkanji and katakana by skilled and less skilled Japanese readers. Reading and Writing: An Interdisciplinary Journal, 7, 377-393.
• Saito, H. (1981). Japanese Journal of Psychology, 52, 266-273, in japanese (abstract only).
• Yamada (1998). The time course of semantic and phonological access in naming kanji and kana words. Reading and Writing: an Interdisciplinary journal, 10, 425-437.
• Yamada, J. Japanese kanji as a semantically based orthography. Psychological reports, 84, 637-642)