Errors in Learning Japanese through Listening-Misheard Cases-Gede

This paper aims to study the error that happens when students learning Japanese through listening. This paper describes misheard cases by students during listening class. The data in this research collected from students’ quiz and test results. Students participated in this study were first-year and second-year students, including 37 first-years students and 24 secondyear students, with total participants 81 students. The data collected in this study then categorized based on the type of errors. The results showed that the errors occurred include confusion between two sounds, reduction of sound, and mis-guessing long vowel. Confusing of two sounds happened when the students misheard two different sounds such as alveolar nasal consonant /n/ in [hinan] with liquid consonant /r/ as in [hiran]. Furthermore, reduction of sound is occurred when students confused the same vowel at particle with front or back vowel sound of the word, such as yamagaafureru which misheard with yamagafureru. This error occurred because the vowel sound /a/ on particle /ga/ which covering up the vowel sound /a/ in the front of the word afureru. Lastly, there are errors that happened because thin overlapping borderline between error or mistake, where students mostly misheard or mistaken short vowels sound such as [ba∫o] with long vowels such as [ba∫o:].


INTRODUCTION
During language learning, errors can happen. The errors in learning process are more complex than the acquisition process. This complex situation caused by the interference from L1 to L2 (Nemati and Taghizadeh, 2013). Brown (2007), identified that speech sound is one of the language acquisition devices. Identifying speech sound is the first step for learning. The processes include hearing, imitating, and understanding. In the pedagogical context, repeating the word from what the teacher said, or reading with a loud voice, is one of them.
On the other hand, the term 'misheard' in the tittle less related to the term 'mondegreen' or 'soramimi' in Japanese. As we know mondegreen means mishearing L1 to become L2 because has similarity sound or homophone. Even though both terms (mondegreen and soramimi) has different level meaning. Otake (2007) stated that mondegreen can occur in one language only (in L1 only or L2 only).
Like, 'bring the phone' misheard with 'microphone'. Different from mondegreen, soramimi occurs cross-linguistically. Sometimes when we hear foreign song, we misheard them with our native language. In Indonesian, these phenomenon become meme or joke, like in the Rihanna song "work" has a line 'he said me haffi' which misheard by Indonesian listener as 'Sumiati'. Or the other example is in the BTOB song "movie" has the Korean line 'nugubodado' which misheard by the foreign listener as 'nugu potato'.
Interestingly, Otake (2007) on his research found 194 misheard song lyrics which broadcast on TV between 1992 and 2007. The results showed that the phonetic distance between the actual lyrics and perceived songs varied from good preserved to be very little maintained. Playing using soramimi on song lyrics can make more attention to listeners. This strategy calls ambiguous code. Otake (2007), noted that soramimi can analysis with two phonological types processing. The first is at the segmental level. In the segmental level, we can found additions, omissions, and replacements of specific speech segments. The second is at the phrasal level. Besides from segmental errors, misperceptions of time often lead to errors in this level. Then in the case misheard song lyrics occurred at both the segmental level and the phrasal level.
Nakata (2016), argues that song lyrics have a complex analysis of mondegreen because has rhythm or accent pattern as a musical instrument. Background sound can occur misinterpretation of meaning. The same situation in real conversation, while speaker are talking with a noisy voice as background sound and that's noisy voice can caused misinterpretation. However, Nakata (2016) noted that's the situation set in his study was different from recorded material in language class. The conversation was recorded in a studio, which made them noise-free and the speed of audio can be set up, and made easy to understand. However, the focus in this study is different from soramimi or mondegreen phenomenon, but the term 'misheard' is used because it has similar concept with both phenomenon.

ERROR ANALYSIS
The basic theory of error analysis is came from contrastive analysis to divergence from one sociolanguage to another socio-language. Lennon (2008), said contrastive analysis is stated by Robert Lado (1975). The beginning concept is how to compare one language and culture to another language culture. These processes identify the characteristics of a native language. By comparing the process, we can know whether languages have similarities or differences, or how to learn a language. Not only the language, but the culture must also be include while learning, because language itself can not be compared without the knowledge about its' culture and people as language environment and user.
Then, in practice, contrastive analysis can be used to predict errors. This implies that errors and difficulties can be predicted. However, this might not be certainly true, because language difficulties are connected to psycholinguistic concept, whereas mistakes are part of the product of language (Lennon, 2008).
Error analysis used interdisciplinary approach, such as linguistic, cognitive, and educational theory. Error analysis has two main problems: first, students can pay attention on aspects of the language that they find difficult so they don't make mistakes; and second, students actually make mistakes in areas whithout realizing that they made a mistake (Lennon, 2008).
Then, the contrastive analysis assumes that mistakes only have one reason, namely maternal influencetongue (interference). However, after discovering that intralingual and interlingual factors often participated to produce errors, the reason for errors is not just one reason (Lennon, 2008).
Castillejos (2009) sees error analysis is part of language teaching. In the evolution of language teaching methods, error analysis has given a very important role. The first formal method assumes that mistakes are evidence of misleading a language. This method only looks at the structure of language, and perfect oral and written production is the goal for students. However, there are developments, the latest trends of communicative methods focus on developing communicative skills, in the presence of mistakes or not is the main thing as long as it does not hindercommunication. Seeing from the formal method and its development up to now, the error analysis shows good development. Lennon (2008), belives the analysis contrastive assume the errors originate from the first language disorder. In the easiest ways, we can say if the problem of error analysis is more related to the intrinsic difficulties than the from cross-language influences.

| P a g e e-ISSN 2528-5548
According to Castillejos (2009), the study of errors offers great advantages for improving language pedagogy. The reason is error analysis results can't go out of date because they prove areas that need language teachers focus on areas like grammar, lexicon, discourse, and others. It is means that is by analyzing students' error we can have an important suggestion for the design of language methods that can be made, these involve all pedagogical design fields, from the syllabus to teaching material.
In the easiest ways, the error analysis is an analysis of the mistake of students L1 when learning foreign language L2. Then the method used is comparative analysis. It is means the benefits of this analysis are to provide advice for teachers on the subsequent learning process, curriculum development, selection of teaching materials, and others.

The Border of Error and Mistake
Error and mistake have a grey area, and both sometimes overlapping (Brown, 2007). However, there are several ways to distinguish between mistakes and errors. The first is associated with L2 students' consistency in the performance. If students sometimes use the right form of certain structures or rules and then they used it wrong, then it is a mistake and can be corrected by themselves. However, if the student always uses it wrong, then it is an error. The second way is to determine error and mistakes is by asking an L2 students to correct their utterance. If they can not, then it is an error, and if the students succeeded to correct it, it is a mistake (Al-Khresheh, 2016).
Analyzing students' error has beneficial point for teacher or lecturer, and for language researcher. Selinker (1969( , as cited in Touchie, 1986 giving three points of beneficial doing error analysis as follows 1. The error can indicate the students progress in the learning process; 2. The error can answer how language is learned; 3. Error is significant to the student himself/herself as the students get involved in hypothesis testing. On the other hand, a mistake refers to a performance error that is a random guess or 'slip' because of a failure to use a properly known system (Brown, 2007). It is means that the students have the knowledge, but inadvertently make mistakes (maybe because of lack of concentration and others).
Afterall, a mistake can be self-corrected because the person who did mistake know the right language system. On the other hand, an error can not be self-corrected, because the person who did an error not realized if he/she did an error, since they think they are doing right or using the proper language system.

RESEARCH DATA
The data in this study collected from a quiz and test. The type of the questions are 'fill in the blank' questions for second-year student. The students listen to the news, and they fill in the blank with a word or a sentence from what they listened. The audio were repeated three times. By repeating the audio three times, the student who could not understand the news they are listening can guess the answer from what they heard. The resource of audio is NHK Easy News. Figure 1 The example of NHK Easy News contents Figure 1 shows the NHK Easy News that will be used as a test. The first step is to choose words or clauses that will be omitted. The chosen words or clauses are N4 level words and clauses which the students never learn before, with a purpose to give the students new vocabularies. Figure 2 is an example of fill in the blank test. The example shows an NHK Easy News title taifuu ga chikaku ni kita toki ni ki o tsukeru kotowith five questions. The learner must fill the blank with a single word or short sentence. The answer number 1 is hinan shite kudasai, number 2 is kanban ga tondari, number 3 is denchuu ga taoretari, number 4 is uekibachi ya jitensha, and number 5 is kuzurete ie. From the answer, have a words that learners never learn before, such us hinan (evacuation), kanban (signboard), uekibachi (plant pot), and kuzurete(to crumble). The test for second-year students who's

JAPANEDU: Jurnal Pendidikan dan Pengajaran Bahasa Jepang
Vol. 04, No. 02, December 2019, pp. 126-132 129 | P a g e e-ISSN 2528-5548 have learned Japanese language before, and their listening skill can be appear when they are can answer the question. The student of the Japanese language at Undiksha using Bahasa as L1, and Japanese as L2. The first-year students, are never learn Japanese before. The second-year students have learned Japanese in the first year. The results showed that the first-year students misheard every single sound because they are using bottom-up a concept in their mind. They just guessing the sound without knowing the meaning of a word they have heard. On the other hand, the second-year students using top-down the concept in their minds, using their one-year experience to guessing the next sound or a word they have learned. If the second-year students only heard '-pitsu', they can easily guess the word as 'enpitsu' because they learned the word and understood the meaning (enpitsu means 'pencil').
In listening class (choukai), the students focus on understanding the Japanese language from what they watched or heard. I usually use video or audio tape from Minna no Nihongo, Dailymotion, or NHK easy news for learning media. During learning process, sometime the students did misheard, eventhough the audio repeated twice or three times. This problem is the first background to initiate this study, which is to find the pattern of the error. Finding the pattern expected to make easier to evaluate the learning process and help the student to reduce the errors.
The procedure of this study is instructing the first-year students to do a listen-writing test. They will hear a word, and then write the word they have heard. Then, the will answer the question with hiragana and romaji, because not all student can write hiragana easily, to reduce the factor which can be a barrier in this study.

RESEARCH METHOD
This research used a descriptive method. The research participants were 37 first-years students and 24 second-year students, with total participants were 81 students. The research object is the student error in the listening test. Test answers then identified and classified by the type of the error and analysed based on Corder's (1974, cited in Lennon 2008 five-step error analysis procedures as follows. a. Doing election of a language corpus. b. Doing identification of errors in the corpus. c. Doing classification of the identified errors. d. Write the explanation of the psycholinguistic causes of the errors e. Doing evaluation of the errors.
However, since there are limitation in this study, this study only followed three steps from five steps mentioned above.

RESULT AND DISCUSSION
The problem of misheard found when teaching Japanese listening of the beginner level. Before learning a sentence pattern and practice listening, the students doing exercise by hearing sound from the Japanese speaker using Minna no Nihongo audio for Japanese pronunciation (nihongo hatsuon). From exercise and test results, there are errors found. The results from one-time quiz and test are described as follows.

Confusing Two Sounds
The first error pattern is confusing two sounds. The two sounds are not quite similar, like plosives sounds /b/ and /p/, with difference in voice and voiceless. The students errors in using these two sounds are as follow. These case happen maybe because there is difference in Japanese liquid consonant with Indonesian liquid consonant. Indonesian liquid consonant voices are trill and lateral. On the other hand, Japanese liquid consonant voices is the flap. Indonesian students have difficulty to produce Japanese flap voices, so they usually used Indonesia lateral voice for Japanese flap voices (Tjandra, 2004).
This means that, Indonesian students guessing Japanese flap voices because in Indonesian voices don't have similar voice like Japanese flap voice. The students tried to recognize the Japanese flap voices, yet they are failed to recognize the voices.

c. Consonant Fricative Glottal and Vowel
The third case is mishearing the fricative glottal voice [ha∫i] with [a∫i]. The reason of this mishearing is the voiceless of consonant fricative glottal, so the students only heard the last voices. This case is similar with the first case where the students only heard [..a∫i]. Table 1 shows data of error which occurred because students confusing two sounds. From the three examples above, there are interesting data about consonant fricative glottal and vowel. These errors occurred both on second-years students and first-year students. The second-year students mishearing fricative glottal to vowel (omission), while the first-students from the vowel to fricative glottal (addition). The complete data are as shown in Table 1.
The alveolar affricate and alveolar fricative mostly misheard by students. Because in the L1 of the students doesn't have alveolar affricate voice [ts]. Eventhough the students learning Japanese voices, but that's not enough to make the students realize that Japanese has different sound between affricate and fricative.
Confusing two sound is an error caused by similar two sounds. It is caused because L1 doesn't have similar voice as in L2, or because the students only able to hear the half part of the word. firstyear studen t

Reduction of Sounds
The second error pattern is reduction of sound. This pattern occurs when the students filling the blank with a sentence or phrase. The student confused by the same vowel on a particle with front or back vowel sound of the word. The example are as follows.
1. Correct answer: suugaku no no-to o gakkou ni wasureta Students answer: suugaku no no-to gakkou ni wasureta The confusion pattern: 'suugaku no no-to o gakkou ni wasureta' 2. Correct answer: mizu ga afureru basho Students answer: mizu ga fureru basho The confusion pattern: 'mizugaafurerubasho' Example 1 showed that the student reducted the vowel [o] as particle because of the back vowel sound of the front word similar to the following vowel sound. Different with example 1, example 2 shows that the students reducted the vowel [a] in the front word [afureru] because particle 'ga' has similar vowel [a] in the back of particle. As the listening aspect, example 1 still has meaning even though the reduction occurs. But for example 2, misinterpretation can happen because verb fureru has meaning in Japanese. The verb fureru means 'to touch', while the verb afureru means 'overflow'.
The reduction of sounds has similar pattern with confusing two sounds. The difference is that confusing two sound only occurs in word level, while reduction happens in clause level.

Long Vowel
Long vowel sounds in Japanese are double vowel sounds such as /aa/, /ii/, /uu/, /ee/ or /ei/, and /oo/ or /ou/. Long vowels are counted as one syllable or two mora. Long vowels in Japanese have a role in distinguishing meaning. Wrongly putting a short vowel with a long vowel will cause different meanings. Therefore it is very important to distinguish long and short vowels.

Example:
/obasaN/(aunt) VS /oba:saN/(grandma) /ojisaN/ (uncle) VS /oji:saN/(grandpa) /koko/(here) VS /ko:ko:/(high school) The student has knowledge about long vowel, so they know the word with a long vowel or not. While write a word with long vowel, some student might forgot to write the long vowel, but some student could be have the wrong knowledge. Some of second-year students believe that 'ishoni' (means 'together') is using long vowel, becomes ishouni. Even if they reminded repeatedly, they still use the wrong word and did the 'mistake'. The error in using long vowel is as shown in Table 2. Table 2 shows an error about a long vowel. The interesting is in the second year students still have students who's made this error. The error in word kaichuudentou and denchuu can be understood because those two words are a new word. But for word basho, this should not happen, because they are as usually using this word.