The Japan News by The Yomiuri Shimbun

Home > Opinion > Culture and Education


Culture and Education

Familiarity of Kana Words and Kanji Words

Yasushi Hino
Professor at the Faculty of Letters, Arts and Sciences, Waseda University

The Japanese language consists of two kinds of characters, kana (Japanese syllabary) and kanji (characters imported from China), which are different in nature. What characteristics do kana and kanji words have? When one reads or hears them, how are the words being processed in the person’s brain? In this article, I will present new findings on the nature of kana and kanji words based on my recent research.

Do Kana Words Have Higher Familiarity than Kanji Words?

Several pairs of katakana and kanji words are listed below. Decide which word in each pair sounds more familiar to you.

ジャム(Jam)-剣術(Fencing)  オリーブ(Olive)-属国(Vassal state)  ゴリラ(Gorilla)-動悸(Palpitation)
クレープ(Crepe)-訳本(Translation)  ランプ(Lamp)-誤植(Misprint)  リンス(Rinse)-暗唱(Recitation)
ビーフ(Beef)-海女(Woman diver)  ミルク(Milk)-鈍器(Blunt weapon)

I believe that in most, if not all, pairs of words, katakana words are more familiar to you than kanji words. However, each pair of katakana and kanji words is actually made of words with equal frequency counts. The frequency with which the words appear is obtained by checking and calculating how many times they are being used in sentences in publications such as newspapers which people read in daily life. Therefore, if the two words in each pair have equal frequency counts, it is assumed that we have almost equal opportunities of reading these words in daily life. Nonetheless, why do people feel more familiar with katakana words than kanji words? Our experience in hearing words seems to greatly affect our evaluation of how familiar each word is even when they are being written.

Effects of Experience in Hearing Words

By repeating the process of reading words in a sentence, we will not only learn the spelling but also increase the degree of familiarity of these words. On the other hand, when we hear a word in a conversation, information of its corresponding written characters is not provided. Nonetheless, there are many studies indicating evidence that when we hear a word, our brains search and utilize its corresponding character information.

For example, phonologically-similar words of the word "密着(mit-chaku)" include "決着(ket-chaku)," "発着(hat-chaku)," and "密告(mik-koku) where only one of the sounds (mora: units of sound that are equivalent to what we call Japanese syllabary) is being replaced. Each of these phonologically-similar words hold one of the kanji characters included in the original word "密着(mitchaku)." In other words, since these words consists of the same character with the similar sounds, there is a consistent relationship between sounds and characters. In this paper, this type of words is called a "consistent word."

Meanwhile, phonologically-similar words to the word "庭園(tei-en)" include "定員(tei-in)," "提案(tei-an)," "永遠(ei-en)," and "声援(sei-en)." Although all these words are made up of completely different kanji characters, each word is phonologically similar (i.e. involving kanji characters with similar sounds) to the word "庭園(tei-en)." In other words, while these words share the same sounds with the character in the original word, they are all spelled differently and, hence, these words do not have a consistent relationship between sounds and characters (inconsistent words). By comparing the hearing performance for the consistent and inconsistent words, my recent research suggests that it takes longer to comprehend the inconsistent words than the consistent words in their spoken form (Hino, Kusunose, Miyamura & Lupker, 2017). This is called "consistency effects." The reason why these effects take place is due to an attempt of the brain to search and use information of the written form of a word upon hearing it. When a person’s brain lookup for characters that correspond to a particular sequence of sounds, unnecessary character information is also being lookup, obstructing the whole search process. As a result, the process of hearing the word takes longer if there is no consistent relationship between sounds and characters. Therefore, consistency effects are observed.

What about kana words? Are consistency effects due to a consistent relationship between sounds and characters observed in the results of a hearing experiment using kana words? According to my recent research, a consistency effect was not observed using kana words (Hino & Lupker, submitted). By nature, auditory stimuli are temporal. Hence, they are serially processed and disappear with the passage of time. Furthermore, kana, which is a phonetic symbol, has a systematic relationship between sounds and characters. For this reason, if it is known to the listener that a particular word is written in katakana or hiragana, correct character information can be retrieved easily even before the process of hearing the word is completed. As a result, a consistency effect does not arise for kana words.

Experience in Hearing a Word and the Familiarity of its Spelling

As described above, while correct character information can almost always be searched when hearing kana words, in the case of kanji words, it is more likely that the correct character information can be retrieved from the sounds only if it has a consistent relationship from sounds to characters like "密着(mitchaku)". These differences between kana and kanji words also affect the familiarity ratings of words. An analysis of the familiarity ratings (the familiarity ratings of written words in Amano and Kondo, 2003) shows that only in the case of kanji words, such data depend on the degree of consistency. We classified 32,990 kanji words and 3,405 kana words into six groups based on their familiarity ratings and plotted the average degree of consistency expressed in numerical terms in Figure 1. As clearly demonstrated by this figure, kana words show an almost constant average degree of consistency regardless of the familiarity ratings. The graph on the other hand shows that the higher the familiarity ratings are for the kanji words, the higher it is for the average degree of consistency.

As shown above, it seems that the familiarity ratings of kana words depend greatly on not only the experience in reading the words, but also in hearing them. In the case of kanji words, on the other hand, the degree of contribution of experience in hearing them to their familiarity ratings seems to depend on the consistency of relationships from sounds to characters. These reasons seem to be behind the higher evaluations of the familiarity of kana words than that of kanji words even if the number of times they are seen (the frequency of their appearance) is equal.

Figure 1: Average Sound-Character Consistency in Each of the Kana and Kanji Word Groups Based on the Orthographic Familiarity Ratings.
The words' familiarity ratings are taken from Amano and Kondo (2003). Group 1 consists of words with the lowest familiarity ratings while Group 6 consisted of those with the highest ratings. The sound-character consistency is close to 1.0 if words possess more consistent sound-character relationships, while it is closer to 0.0 if they have inconsistent relationships from sounds to characters. This figure shows that while kana words have almost the same level of consistency irrespective of how high the familiarity ratings of their spellings are, the value of kanji words' consistency tends to become greater as the familiarity ratings of their spelling increases.

  • Amano, N. & Kondo, K. (2003). NTT Detabesu Shirizu: Nihongo No Goi-tokusei Dai 1-ki CD-ROM-ban. [NTT database series: Lexical Properties of Japanese, Vol.1, CD-ROM Version.] Tokyo: Sanseido.
  • Hino, Y., Kusunose, Y., Miyamura, S., & Lupker, S. J. (2017). Phonological-orthographic consistency for Japanese words and its impact on visual and auditory word recognition. Journal of Experimental Psychology: Human Perception and Performance, 43, 126-146. doi:
  • Hino, Y., & Lupker, S. J. (submitted). The impact of phonological-orthographic consistency on orthographic familiarity ratings and lexical decision performance for Japanese words. Paper submitted for publication.

Yasushi Hino
Professor at the Faculty of Letters, Arts and Sciences, Waseda University

Yasushi Hino, Professor at the Faculty of Letters, Arts and Sciences, Waseda University After graduating from Ritsumeikan University with a bachelor's degree in literature, Yasushi completed the master's course at the Graduate School of Chukyo University. In 1993, he completed his Ph.D. program at the Graduate School of the University of Western Ontario in Canada. Since 2004, he has been affiliated with the Faculty of Letters, Arts and Sciences of Waseda University. His field of expertise is cognitive psychology and psycholinguistics. He has contributed several papers to the Journal of Experimental Psychology: Human Perception and Performance, Journal of Experimental Psychology: Learning, Memory and Cognition, Journal of Memory and Language, and other publications.
Currently, he is a professor at the Faculty of Letters, Arts and Sciences at Waseda University. He is also a member of the Japanese Psychological Association, American Psychological Association, and Psychonomic Society as well as a member of the steering committee at the Japanese Cognitive Neuropsychology Society.