CMC and sound information:

A case study of Vowels and language plays in Japanese mobile phone e-mails

Noboru Sakai, Tamagawa University [About | Email]

Volume 21, Issue 2 (Discussion Paper 3 in 2021). First published in ejcjs on 16 August 2021.


This study scrutinises how people utilise language plays (LP) in mobile phone e-mail (Keitai-mail) based on the five vowels as the criterion. The analysed data corpus is a collection of 43293 Keitai-mail by 60 Japanese people who use it regularly and naturally in daily life. The result shows that people tend to use LP to create the impression of a milder sound (such as ungrammatical use of small letters, or long vowel symbols) where /a/あ, /i/い,/e/え, /o/お, vowel sounds in their messages have a slightly weaker sound than their English equivalent. I postulate that this suggests that Japanese as a language naturally tends to a preference for milder sounds. On the other hand, /u/う as a rounded vowel tends to be omitted in keitai-mail, since /u/ mostly stops air flow and seems not to be suitable to longer pronunciation, or people have a preference to circumvent that sound in their articulations. Keitai-mail texting of course reflects the manner of articulation; this is an interesting instance of representations of oral communication, like text messaging, naturally being affected by real oral communication. 

Keywords: CMC; Japanese mobile phone; e-mail; language plays; vowels

1. Introduction

Keitai, mobile phones, are a tool that people always have with them, anywhere, at any time (Hayashi, 2007). However, a dynamic view of Keitai-medium communication, in contrast to studies of different ways of communicating utilising Keitai, has yet to be developed. To challenge this matter, this study analyses how the mode of Keitai-mail itself influences Keitai-mail composition, namely, in the sense of capturing or recreating a mode of oral communication. Some researchers have mentioned the nature of Keita-mail as something similar to oral communication, to suggest that Keitai-mail is like oral communication in nature (Tsuji, 2003), or is a form of textualised oral communication (Katayama, 2003). Miyamoto and Kotera (2004) also suggest that communication through Keitai-mail has an oral-communication nature; it can be said to be a conversation which takes place over time. In other words, Keitai-mail are an easy and convenient method of communication which further provides users with opportunities to write and read Japanese. At the same time, however, as Katayama (2003) argues, Keitai-mail communication is more a matter of textualised oral communication than the pure written communication found in letters, with the result that the content of e-mail is rather plain and is not necessarily the result of deep and logical thinking in composition.

Language plays (LP), unorthodox usage of basic scripts to include some extra-textual effects, are often used in textualised oral communication, and these unique patterns have been gaining attention in the study of CMC including Japanese Keitai-mail. In addition, Japanese is a language which heavily relies on vowels (あ/a/、い/i/、う/u/、え/e/、お/o/; further details in the next section), and this will influence text and oral communication including violating uses of these features for interesting effects. Therefore, this study investigates quantitatively how vowels are utilised in Japanese Keitai-mail in order to analyse the extent to which people utilise their language system intentionally or unintentionally.

2. Background: Japanese language and LP patterns

The Japanese writing system consists of hiragana, katakana, kanji, Arabic numerals and Romaji (or the alphabet). Only the first three of these are official scripts but the other two are widely used as well. Hiragana and katakana are phonetic syllabaries. They have 46 symbols, representing five vowels and 41 consonants. The five vowels in Japanese are あ(a), い(i), う(u), え(e) and お(o). Their pronunciations are shown in Figure 1.

Except for the vowels themselves and the single ‘n’, syllables are combinations of a consonant ‘k’, ‘s’, ‘t’, ‘n’, ‘h’, ‘m’, ‘y’, ‘r’, ‘w’ and one of the five vowels: for example, か(ka) or す(su). Syllables with voiced consonants are represented by adding diacritics to the unvoiced version (e.g., た ta → だ da).

Figure 1. IPA chart for vowels (The International Phonetic Association, 2005) Note. Japanese vowels are circled by the author.

In the Japanese writing system, long vowels are expressed by symbols such as ‘ー’, ‘~’, indicating that the pronunciation of the vowel embedded in the last moji preceding them should be prolonged. For CMC, in addition to the basic input system of the Japanese word processing system, picturial emoticons, or emoji, have symbols seeming like long vowel symbols. Further, the sounds of the Japanese writing system can also be transcribed in the English alphabet, Romaji. One letter is used to represent each consonant and each vowel, and a Japanese syllable can thus be written in two letters (some compound syllables consist of three letters).

About the LP and Japanese writing system, Miyake (2007) summarises the usual patterns that occur in Japanese Keitai-mail:

1.  Kanji instead of Katakana: 怒気怒気←ドキドキ (ドキドキ, dokidoki’ means ‘The sound of a heartbeat’, and should be written in Katakana.)

2.  Hiragana instead of Katakana: はっぴい←ハッピー (ハッピ’, happi-’ happy’ should be written in Katakana.)

3.  Katakana instead of Kanji: リョーカイ←了解 (了解, ‘ryo¯kai’, ‘all right’ in Katakana is not conventional writing.)

4.  Romaji instead of Hiragana: いいyo←いい (よ, ‘yo’, sentence-ending particle, should be written in Hiragana.)

5.  English instead of Katakana: Thank you←サンキュ (normally written in Katakana.)

6.  English acronym instead of Japanese: NG←ダメ(‘NG’ is an acronym for ‘no good’.)

Non-Standard Letter Choices

7.  Small letters 1: ありがと←ありがとReducing the size of normal letters.

8.  Small letters 2: おやすみ←おやすみAdding a small letter to lengthen the preceding vowel.

9.  Long vowel: . . .するよ Lengthening the vowel sound of the preceding letter.

10.  Wrong spelling: ま←ま, そ←そConsciously taking up a wrong spelling when the pronunciation is unaffected by the mistaken spelling.

11.  Additional aspiration mark (Soku-on): よし←よし, Expressing a desire to show determination, desperation, etc.

12.  Specialised signs: いちゃ②←いちゃいちゃUsing mathematical, scientific and other specialised signs. In this case, ②playfully applies the concept of mathematical squaring to avoid having to repeat a whole word.

13.  Omitting punctuation:ありがとう ←ありがとうReplacing punctuation with graphic characters.

(pp. 64-65, numbering is modified for the discussion)

Among the list, vowels are involved as a major source of LP (patterns 1, 2, 7, 8, 9, 10), and people seem familiar with the patterns as commonly used features.

As a possible reason for various LP embedded in the vowels, the Japanese sound system is CV or V structures, and vowels are combined with consonants in most Japanese syllables. Therefore, focusing on vowels in Japanese discourse seems a meaningful method of analysis to describe the use of CMC as a method of generally time-tight text communication.

From the discussion above, the research question is: Are language plays influenced by manners of articulation based on the 5 Japanese vowels あ、い、う、え、お?

3. Methods

This study investigates the corpus from raw data of communication practices that appeared in Keitai-mail the researchers gained through field work. In detail, this study analyses 43,295 Keitai-mail for communication purposes collected from 60 Japanese young people (aged 18 to 30, the ratio of male to female is 1:1. i.e, 30 male participants and 30 female participants) during May 2009 to January 2010.[1]1 The participants use mobile phones sold by the three major Japanese Keitai companies: NTT Docomo, Softbank and au, and the numbers of participants using each Keitai company are as follows: Docomo 28, SoftBank 11, and au 21. Keitai-mail are collected by Keitai-mail backup software Keitai-master MX (ver 4.5) (Jungle Inc., 2009), with the selection of participants which Keitai-mail could offer. Because of conflicting specifications, Docomo emoji stored in au mobiles cannot be collected.

This analysis further counts how many times each vowel has been changed, replaced, or even omitted based on the manner of articulation, and discusses how oral sound production affects Keitai mails text properties, to give a sense of immediate communication between users, which further discloses a characteristic of this particular CMC.

4 Results and discussion 

Vowels play a big role in the application of Language Plays (LP: unique use of language beyond standard formal rules) in Japanese. The importance of vowels in English is also discussed by Crystal (2008) in that they are omitted as contractions (in the middle) and clippings (at the end), particularly in the language used on the Internet. The importance of vowels in Japanese with its consonant-vowel syllable structure may easily be suspected to be crucial in analysing the language found in Keitai-mail as a type of CMC. This section analyses, in Fairclough’s terms, the presentation of phonological information as a part of the style of Keitai-mail.

All simple vowels have a large and small moji form; as well, they can be replaced by long vowel symbols such as ‘ー’ or ‘~’, even though in standard Japanese these symbols are not normally used in hiragana words. In addition, each Japanese syllable ends in a vowel, which, when the word is written in hiragana or katakana, can be extended by this means. LP are commonly applied at the end of words and therefore it is understandable that vowels have a central function in LP. For these reasons, this section further analyses the relationship between vowels and LP.

Table 1 summarises how each type of LP is applied to each vowel and Figure 1 illustrates how each LP occurs in terms of each vowel based on the manner of articulation. This figure includes four types of LP: addition (adding large moji and small moji), subtraction, replacement of long vowel symbols, and replacement of large moji with small moji (e.g., あ with ぁ). Figures 2 and 3 represent Table 1 in visual form.

Table 1.

Figure 2. Language plays and vowels based on the manner of articulation

Abbreviations. ADD: Addition, SUB: Subtraction, REP: Replacement
L to S: Large moji (L) to small moji (S)

Figure 3.  Each vowel and language play.

Figure 4. Total number of applications of language plays for each vowel.

Abbreviations. ADDL: Addition of large size moji, ADDS: Addition of small size moji
SUB: Subtraction; Symbol ー: Replacement of long vowel symbol ー
Symbol ~: Replacement of long vowel symbol ~
L to S: Large moji (L) to small moji (S)

Looking at Figure 3, it is apparent that each form of LP tends to affect a certain vowel most. For instance, small moji addition occurs heavily with /a/, whereas subtraction can mostly be seen with /u/. Figure 4 depicts LP application for each vowel: the figure shows this tendency to apply a particular form of LP more clearly. In addition, Figure 4 also shows that LP are largely applied to /a/ and /u/, while on the other hand, /e/ and /o/ exhibit relatively small application of LP. As seen in Figure 2, LP application relates to the manner of articulation to a certain extent.

LP and manner of articulation

The figures shown in the last section indicate the relationship between LP and the manner of articulation. Even though Keitai-mail are written-based communications, this shows that sound information is a significant factor in written expression as well, as the following discussion shows.

Firstly, /e/ and /o/ have a relatively small number of LP applied to them. Both of these are close-mid vowels (see Figure 2). Close-mid position is close to the central position (where the /ə/ sound is generated) and is relatively easy for pronunciation. Because of this, /e/ and /o/ are not heavily influenced by LP.

As the opposite phenomenon to this, large to small replacements have a strong tendency to occur in high or low positions (i.e., /i/, /u/, and /a/). It may be the case that /a/, /i/, and /u/ require a firmer tongue position than /e/ and /o/, and small vowel replacement decreases the impression of burden in the production of the sound since small moji indicate a slight blurring of the sound. This means that changing into a small vowel gives the impression of decreasing the effort needed to pronounce these sounds through moving the tongue to a more decided position (not in the middle) and is preferentially applied to those vowels which have an associated burden in articulation.

Among these three vowels, this replacement is applied to /i/ and /u/ more than to /a/. The pronunciation of the close sounds of /i/ and /u/ requires a certain extent of interference. Kubozono (1999) explains /i/ and /u/ as vowels that have high consonant properties, meaning that they have more interference in the air current. Therefore this interference of the air current involves a certain lag in smooth articulation.

When texting, this tendency to slow down is more discernible than it is with speaking, since people cannot input as fast as they speak and unconsciously try to avoid a lack of smoothness through the platform available to them. This phenomenon of the reflection of unconscious phonological sense corresponds with other linguistic phenomena in adopting new forms of language which go beyond the existing Japanese language framework. For example, Peperkamp, Vendelin, and Nakamura (2008) argue that loanword pronunciations unconsciously reflect their manner of articulation in the original languages and this differentiates how each loanword is spelled even though the orthography of the original words in different languages is quite similar.

The impression of smoothness in pronunciation seems to be reflected in /u/, a close back rounded vowel. For the most part, /u/ is omitted or replaced by long vowel symbols. For reasons of subtraction, to pronounce /u/ would need the most effort compared to other vowels because of the position of the tongue, and it will be avoided. Replacement with long vowel symbols is explained by the same mechanism. When /u/ is replaced by long vowel symbols, pronunciation is taken to be a continuation of the sound of the vowel in the last moji, and it is not necessary to pronounce /u/. Moreover, Japanese vowels are pronounced ambiguously compared with other alphabetical languages. In other words, their position of tongue to pronounce vowels is deeper inside in the mouth. This seems to influence the increased number of omissions of /u/ since the Japanese /u/ sound is less clear than other languages and Japanese people can understand /u/ with a more ambiguous articulation, or even in case of its not being heard. Therefore, when applicable, long vowel symbol replacements are preferred in order to indicate eased pronunciation, or simply when it does not present a problem for intelligibility.

As for addition, this pattern is applied most to /a/, followed by /e/. This can be explained by the fact that sounds generated with an open mouth are easily added to; since these sounds are naturally prolonged, there is little or no air blockage in articulation. The small number of applications to /i/ and /e/ can be explained by the same reason as that pertaining to long vowel symbol replacements.

5. Conclusion

To sum up, the application of LP to vowels has a close relationship to the manner of articulation when we observe the type of LP applied to each vowel. Keitai-mail are written-based media: information about sound, in theory, plays no role in their composition. However, sound information in fact has a significant impact on written composition as well as on speech and the strategies used in pronunciation are reflected in Keitai-mail through the available input system. Moreover, this phenomenon also provides strong evidence that the language of Keitai-mail is quite similar to spoken language and that people use Keitai-mail in this ‘spoken’ mode, by applying a typical phonological style which can be expressed using the framework of the Keitai input system.


This research project was supported by several scholarships and a research grant: The University of Queensland, the Faculty of Arts International Scholarship covered the tuition fees. A living allowance was provided by Nanette Gottlieb’s Australian Professorial Fellowship, funded by the Australian Research Council. In addition, Tokyo Foundation also supported my data collection in Japan. I am thankful for these forms of financial support. In addition, I would like to extend my thanks to Emeritus Professor Nanette Gottlieb, the University of Queensland, for proofreading this work.


Crystal, D., 2008. Txtng: The Gr8 Db8. Oxford: Oxford University Press.

Jungle. Inc., 2009. Keitai-master MX ver4.5 [Computer software].

Hayashi, R., 2007. Industrial design changes and social trends in view of mobile phone history. Sharp Technical Journal, 95, 24-28.

Katayama, A., 2003. A study on digitization of making sentences by using information instruments. Kyōiku Jōhō Kenkyū, 18(4): 13-20

Kubozono, H., 1999. 日本語の音声 [Japanese phonetics and phonology]. Tokyo: Iwanami Shoten.

Miyake, K., 2007. How young Japanese express their emotions visually in mobile phone messages: A sociolinguistic analysis. Japanese Studies, 27(1): 53-72.

Miyamoto, S., & Kotera, M., 2004. Gender consciousness in mail communication by cellular phone : From the perspective of politeness strategies . Research Reports, 6: 127-137

Peperkamp, S., Vendelin, I., & Nakamura, K., 2008. On the perceptual origin of loanword adaptations: Experimental evidence from Japanese. Phonology, 25: 129-164

The International Phonetic Association., 2005. The international phonetic alphabet. Retrieved 4 August, 2011, from

Tsuji, D., 2003. Summary report on a questionnaire survey of young people's communications and relationships with friends and parents. Bulletin of the Faculty of Sociology, Kansai University, 34(3): 373-389.

About the Author

Noboru Sakai is an adjunct lecturer at the Center for English as a Lingua Franca, Tamagawa University/the College of Libera Arts and Sciences, Kitasato University, Japan. He holds a PhD (Language studies) from the University of Queensland, an MA (Applied Linguistics) from the University of Queensland, and BS (Information systems) from Soka University, Japan. His research interest is communication in society from a holistic view, including its related multidisciplinary research fields. He in particular studies computer mediated communication (particularly among young people) based on sociolinguistic perspectives. He also researches in applied linguistics, emphasising computer assisted language learning and Japanese translation.

Email the author

Back to top