Japanese Native Speakers’ Perception on Learners’ Pronunciation of Double Consonant Sounds in Japanese Adverbs ( Fukushi )

The purpose of this study is to determine the perception of Japanese native speakers on the pronunciation of the double consonant 「促音」 ( sokuon ) in Japanese adverbs or 「副詞」 ( fukushi ) namely 「ゆっくり 」 ( yukkuri ), 「ずっと」 ( zutto ), and 「やっと」 ( yatto ) by ten Japanese language students in a state university in East Java, Indonesia. This study uses both quantitative and qualitative approach with conducting assessments of students’ pronunciation of Japanese adverbs ( fukushi ) by three Japanese Native Speakers (JNS). The results showed that eleven data were pronounced perfectly and naturally, sixteen data were pronounced naturally, four data were pronounced a little naturally, and one data was pronounced a little unnaturally and unnaturally. However, there were seven data misunderstood by the JNS. The results also show that differences in JNS perception of the respondent’s pronunciation have an impact on the assessment, and the double consonants also affect the duration of pronunciations which affects the assessment from JNS. In addition, the Japanese language students participated in this study were all considered able to pronounce the three 「副詞」 ( fukushi ) or Japanese adverbs correctly according to JNS. Double Consonant;


INTRODUCTION
Language learning includes four aspects of language skills, namely listening, speaking, reading, and writing. Matsuzaki and Kawano in Pratiwi, Dahidi, and Haristiani (2016) argue that Japanese language learners are generally weak in pronunciation and many students have not realized the importance of understanding the pronunciation of foreign languages correctly which is included in speaking skills. However, the ability to speak correctly is considered the most sought-after skill for an individual to be accepted in the field of a foreign language (Kurum, 2016), and the "speaking-listening" language activity is very important part (Soepardjo, 2012). The ability to speak in learning Japanese is related to the pronunciation of the sounds of the letters. Judiasri (2017) argues that to be able to speak one language well, the speaker must master the pronunciation, structure, and vocabulary. The elements of Japanese sounds that are a problem for the learner are accents, intonations, long sounds, double consonant sounds, and others (Najoan, 2019) caused by difficulties in mastering special beats such as long vowels, "n" sounds, and double consonant sounds (Hirata, n.d.). Therefore, this is also a cause of many e-ISSN 2528-5548 | p-ISSN 27764478 mispronunciations of Japanese sokuon sounds by Japanese language learners.
The sound of a double consonant or called a double consonant sound or called 「促音」 (sokuon) in learning Japanese letters is one of the characteristics of a distinctive sound. The double consonant sound (sokuon) is a closed sound or a clogged sound, which in Indonesian can be called a double, which is the use of the same consonant sound as the consonant in a syllable in the next section (Mael, 2021) and written using the letter 「つ」 (tsu) small (Kawahara & Braver, 2014). In the katakana letters use small 「ツ」 (tsu) while in the Roman letters it is written using the double "t". Table 1 below is a series of hiragana letters and romaji that are sounded in duplicate (Kawarazaki, 1980).

□っか
In Japanese letters, not all series of hiragana and katakana letters can be sounded in pairs. Even so, the double consonant sound (sokuon) is one of the typical Japanese phonemes that is difficult to hear by foreign speakers (Fujimoto, 2014). Furthermore, Hirata conveyed that the double sound (sokuon) remains difficult for second language learners, even for advanced level learners because these double consonants tend to be pronounced as two separate words (Kubozono, 2017). As a foreign language learner, awareness of the sound of double consonants (sokuon) is felt to be lacking even though there is a basic difference between single and double consonants, namely the duration of the consonant (Arai, Iwagami, & Yanagisawa, 2017) that double consonant sounds (sokuon) in Japanese undergo a longer sound narrowing process than non -double words (Kawahara, 2015). Therefore, the pronunciation of Japanese sounds, especially double consonant sounds (sokuon) is difficult to pronounce because not all languages have the same pronunciation system as Japanese.
In addition, double consonant differences can also affect the duration of the articulation. For the comparison of differences in Japanese double consonants, for example, double consonants [t] has a ratio of 2.24 while double consonant [k] has a ratio of 1.91 (Kawahara, 2015).
On the other hand, Indonesian adult Japanese learners generally have more difficulty following pronunciation like native Japanese speakers. The imitation ability, voiceless pronunciation practice, and foreign language acceptance are depending on their language ability (Ohta, 2001). Hence, most adult learners are consciously aware that awareness factors (cognition of adult second language learners), interfere with language acquisition. And adult language learners are more aware of the social values and attitudes related to language, etc. which is exaggerated to hinder the acquisition of a second language (Bainbridge, 2002). Therefore, it is important to equalize perceptions between Japanese language learners in Indonesia and the Japanese native speakers.
Adverbs are a class of words that often appear before verbs. Japanese adverb or 「副詞」(fukushi) is a non-conjugated word that modifies other words (Kamiya, 2012). The function of Japanese fukushi is to modify verbs, adjectives, other adverbials, and certain nouns such as「ゆっくり」(yukkuri), 「とても」 (totemo), and「もっと」(motto); the second function is used to indicate special expressions such as「けっ し て 」 (kesshite) and 「 た ぶ ん 」 (tabun). Japanese adverbs that contain double consonant sounds (sokuon) are often unaware of their existence by Japanese language learners. The double consonant sound (sokuon) is rarely pronounced correctly compared to the double consonant sound (sokuon) in verbs. This can also give the impression that the low awareness of students of the sound of double consonants (sokuon) in a word.
Research on double consonants (sokuon) in Japanese was also conducted by Wahyuni and Sutedi (2020) to determine the language errors and causal factors of Japanese language learners on the ability to read and write hiragana containing sokuon and choon. This study found that most of the errors in reading and writing words that contain double consonant sounds (sokuon) are because learners do not know the meaning of the word in question and do not know the actual rules of Japanese. In addition, the results were obtained due to the influence of the level of understanding of hiragana letters and the neglect of the rules of writing hiragana letters. This study only focuses on the variety of written language so there are no research results on pronunciation errors sokuon and choon. Next is a study by Mael (2021) that focuses on sokuon in conversational language in comics, who found that oral language writing in comics if transcribed into written language will affect sokuon which inconsistent with Japanese grammar. In addition, the sokuon varies because it is spoken in a conversational language. However, this study does not discuss the pronunciation of sokuon in spoken language so there is a need for the development of research on the pronunciation of sokuon. e-ISSN 2528-5548 | p-ISSN 27764478 In contrast to above studies, this study aims to determine how the perceptions of Japanese Native Speakers (hence JNS) on the pronunciation of three adverb 「 副 詞 」 (fukushi) namely 「 ゆ っ く り 」 (yukkuri),「ずっと」(zutto), and「やっと」(yatto) by Japanese language learners in a state university in East Java, Indonesia. The results about JNSs' perceptions of learners' pronunciation of the three「 副詞」(fukushi) is expected to be useful as a guide on how to pronounce 「 副 詞 」 (fukushi) a word that contains double consonants in Japanese speaking activities.

Research Approach
This research approach used qualitative research. Then for the presentation of data in this study in the form of a description. Furthermore, the validation technique used to determine how the perception of native Japanese speakers of the pronunciation that is often pronounced by learners is to use triangulation. There are various ways to use triangulation validation done in this study are: 1. Triangulation of data collection, which is collecting data from various sources of informants; 2. Triangulation of methods, i.e. data collection is done with various methods. (Budiastuti & Bandur, 2018)

Data and Research Subjects
The data used is an audio recording containing the pronunciation of 10 students practicing Program Pengenalan dan Pengelolaan Manajemen or Learning Introduction and Management Program (hence LIM Program), which are members of class of 2018 in Japanese Language Education Study Program in a state university in East Java who have passed JLPT N4. The total data collected about the pronunciation of the three fukushi or Japanese adverbs that contain the double consonant sound (sokuon) 1 word「ゆっ くり」(yukkuri), 2 words「ずっと」(zutto), and 1 word「やっと」(yatto) was 40 data.

Data Collection
Students as respondents were given a speech script and then invited to read and ask if there are things that are not understood. Furthermore, students are given the opportunity to read in the heart first, when ready to start reading and record it in .wav format and upload it to the specified Google drive.

Data Analysis
The first step in the analysis is the assessment of audio recording data by Japanese Native Speakers (JNS). The assessment is done by listening to the audio recordings of the students that contain the pronunciation of three 「副詞」(fukushi) and then choosing whether the pronunciation done by the students is determined as「自然」(shizen) 'natural', 「やや自然」(yaya shizen) 'slightly natural', 「やや不 自然」(yaya fushizen) 'slightly unnatural', or 「不自然 」(fushizen) 'unnatural'. Three JNS have participated in this study as the assessors. Furthermore, of the 3 JNS, 2 native speakers (JNS1 and JNS2) are already understand Indonesian and Javanese while 1 native speaker (JNS3) only understands Japanese. This may give different perceptions of the pronunciation uttered by students due to language experience, and differences in knowledge about the linguistics of the language being listened to (Williams & Escudero, 2014).

JNS Assessments on Each Learner's Pronunciation of Adverbs 「 ゆ っ く り 」 Yukkuri, 「ずっと」Zutto, and 「やっと」 Yatto
The data taken in this study were「副詞」(fukushi) or Japanese adverbs pronunciations consisting of the words「ゆっくり」(yukkuri), 「ずっと」(zutto), and 「やっと」(yatto) with 40 total amounts of data. The assessments from JNS were divided into four categories namely 「 不 自 然 」 (fushizen) which is 'unnatural', 「やや不自然」(yaya fushizen) or 'slightly unnatural', 「やや自然」(yaya shizen) means 'slightly natural', and「自然」(shizen) that is 'natural'. The following are the data findings based on the type of assessment data from JNS for each respondent (below abbreviated as 'R', hence Respondent 1 is 'R1' and so on). Table 2 contains the pronunciation assessment of R1. with a high accent of 「 ゆ 」 (yu) and the double consonant was not clearly heard. Although it sounds a little unnatural but can still be understood by all JNS. When pronouncing the word「やっと」(yatto), R1 gets an assessment from JNS as 'slightly natural' because there is an inequality of perception from the JNS-3. However, the opinions of JNS-1 and JNS-2 were very different because JNS-1 assessed as 'unnatural' while JNS-2 assessed it as 'natural'. In addition, the opinion of JNS-1 says that R1's intonation of「や」(ya) was high, and JNS-3 thinks that if the intonation of「と」(to) is lower it will sound more natural. While R1's pronunciation of the first word「ずっと」(zutto) assessed to be 'slightly natural' and the second「ずっと」(zutto) assessed as 'natural', both「ずっと」(zutto) pronunciation by R1 was commented to have an accent in the pronunciation of the「ず」(zu) sound by JNS-1.

JNS Assessment of R2's pronunciation
R2 pronouncing the word「ゆっくり」(yukkuri) get an assessment that tends to be the same from all JNS which is 'unnatural'. The accent of the word「 り」(ri) was high and the double consonant was not heard which is why all of JNS gives the result as in Table 3. When pronouncing the word「やっと」(yatto), R2 gets an assessment from JNS as 'natural'. But JNS-3 says a little more might sound natural. When R2 pronounces the first and the second「ずっと」 (zutto), R2 also gets an assessment from JNS as 'natural'. But the first and the second「ずっと」 (zutto) by R2 commented by JNS-1 that there is an accent in the pronunciation of the「ず」(zu) sound.

JNS Assessment of R4's pronunciation
R4 pronunciation of「ゆっくり」(yukkuri) gets an assessment from all JNS as 'natural'. Meanwhile, when pronouncing the word「やっと」(yatto), R4 got a comment from JNS-1 that the duration of pronunciation「と」(to) was too long. However, the pronunciation of 「 や っ と 」 (yatto) by R4 gets a 'slightly natural' assessment from JNS-3, as presented in Table 5. R4 when pronouncing「ゆっくり」(yukkuri) gets the same assessments from all JNS which is 'natural'. Meanwhile, when pronunciation of the word「やっ と 」 (yatto) got a comment from JNS-1 that the duration of the pronunciation「と」(to) was too long, although the pronunciation of「やっと」(yatto) by R4 got a 'little natural' assessment from all JNS.
R4 when pronouncing the first and the second「ず っと」(zutto) gets the same assessment as 'natural' from JNS-2 and JNS-3. However, in both words「ず っ と 」 (zutto) by R4 there was an accent in the pronunciation of the sound of 「 ず 」 (zu) which assessed as 'unnatural' by JNS-1. Table 6 contains JNS assessments for R5. R5 when pronouncing 「 ゆ っ く り 」 (yukkuri) gets the same assessments all JNS which is 'natural'. Meanwhile, when pronouncing the word「やっと」(yatto), R5 get an assessment that tends to be different from all JNS.