Mispronunciations in Graduate Students’ Presentation Projects

This paper explored the mispronunciations produced by Indonesian graduate students in their presentation projects. The data were collected from presentation projects of 11 graduate students of the English Education Master’s Program of Sanata Dharma University. The researchers used a participant observation technique to collect data and a document analysis to analyze the data. The theory from Bonaventura, Herron and Menzel’s (2000) was employed to classify the data. Results showed that 89 mispronunciations existed in the presentation projects. The mispronunciations were categorized into three types of errors, namely: 26 mispronunciations belonged to the first type, the problems in the pronunciation of non-native sound; 36 mispronunciations belonged to the second type, the carry-over of pronunciation regularities from the mother tongue (L1); and 27 mispronunciations belonged to the third type, the over generalizations of target language (L2) regularities. It is expected that the findings can become a stepping stone for graduate students to improve their English pronunciation.


Introduction
Every teacher has their own way in using certain approach and/or method in order to achieve specific learning goal(s) set by the government, school, or even themselves (Richards & Rodgers, 2001). These approaches and methods are usually implemented in the form of learning activities. For example, some teachers design weekly quizzes to improve students' performance in the intended subjects because students tend to spend their time studying before class and they may get feedback on their weekly performance (Ruscio, 2001;Sporer, 2001). On the other hand, other teachers prefer to combine their learning activities by using games because gamebased learning approaches can increase students' learning motivation and encourage students to continue learning (Escudeiro & Carvalho, 2013;Johnson, 2010;cf. Indriani, 2019). Escudeiro and Carvalho (2013) state the two factors are needed in achieving learning goals. Meanwhile, some groups of teachers decide to put students' presentation projects in their syllabus.
The implementation of students' presentation projects can vary from one teacher to another. Some teachers ask the students to present a topic individually while the others ask the students to present it in a group. Some teachers ask the students to present predetermined topics related to the subject of study, while the others ask the students to present students' progress on certain tasks. Haber and Lingard (2001) relate the word 'presentation' to a personal ability in communicating with others. By presenting something, people actually try to deliver a message or information to the interlocutors or audiences. Besides, sometimes a presentation aims to persuade someone to do something (Johlke, 2006).
Based on the concepts above, many experts agree that getting engaged in presentation projects can give students many benefits. In this stage, the researchers only focus on two benefits which influence pre-service teachers related to their future careers. First, it enables students to train their ability or skills in speaking or communicating in front of the public (public speaking) because in presentation projects, students should present or deliver ideas in front of interlocutors or audiences (Behnke & Sawyer, 2000;Collins, 2004; see also Asmaruddin, 2018;Angelina, 2019). This ability or skill is needed by pre-service teachers because they will deliver many topics of discussion to their students in the future. Second, the implementation of students' presentation projects helps pre-service teachers in preparing themselves when they become teachers after their graduation. Asmaruddin (2018) states in a presentation, students can deliver their field of study that might be able to increase the students' interest to speak. Therefore, this activity is relevant to improve speaking skills since it will help them to perfume better in sentence structure and vocabulary choice. Further, Yigit (2009) states that presentation projects can be used as a simulation for pre-service teachers to deliver learning materials. It functions as a teaching training that enables teachers to develop their "knowledge, skills, attitudes and effective teaching in classrooms" (p. 56). It means pre-service teachers do not only learn how to speak in front of many people, but they also learn how the audiences or future students understand the material or topic they are talking about in a presentation.
According to Collins (2004), Gelula (1997), Grez, Valcke andRoozen (2009), andCapecce (2011), there are many aspects that should be noticed when speaking in front of others, such as content and organization, interaction with the audiences, body movement and gesture, eye contact, time management, facial expression, voice (volume, pace, intonation), articulation, pitch and inflection, and pronunciation. In this study, the researchers focus on how Indonesian graduate students pronounce English words. Non-native speakers often make lexical, syntactic, and phonological errors because their control in using the target language is not as big as when they speak in their mother tongue (Boxer & Pickering, 1995). Therefore, this study was conducted to explore mispronunciations in presentation projects of Indonesian graduate students so that they can eliminate the mistakes in the future and to examine the linguistic factors contributing to the mispronunciations in the presentation projects of the Indonesian graduate students. In this study, the researchers formulated the following two research questions to resolve. First, what are the mispronunciations in presentation project among Indonesian graduate students? Second, what are the linguistics factors contributing to the mispronunciations in presentation project among Indonesian graduate students?

Mispronunciation in spoken English
To understand better the mispronunciations that will be discussed in this study, it is important for the researchers to review some related theories. Pronouncing words correctly is crucial because Hoffmann and Siebers (2009) state that mispronunciation may lead to unintelligibility. The statement means that a mispronunciation can cause misunderstanding. The listeners may not recognize what the speaker means because of the mispronunciation. An error in pronunciation or mispronunciation is a situation when the deviation arises as a result of lack of knowledge or competence.
As cited in Ellis (2008), Corder classifies three types of errors in pronunciation according to their systematicity. First, pre-systematic errors happen when the speakers are aware of the existence of a certain rule in the L2. Second, systematic errors happen when the speakers have discovered a rule but it is the wrong one. Third, post-systematic errors occur when the speakers have already known the correct L2 rule but use it inconsistently. Meanwhile, Dulay and Burt classify errors into three categories (as cited from Ellis (2008), namely developmental, interference, and unique. Developmental refers to errors that are similar to L1 acquisition. Interference is for the errors that reflect the structure of L1. Last, unique is for the errors that are neither developmental nor interference.
However, the researchers would like to employ Bonaventura, Herron and Menzel's (2000) theory to classify the findings of this study since the description of the errors in pronunciation is more relatable to the data and findings of this study. According to Bonaventura et al. (2000), there are three types of mispronunciations that are categorized based on distinguished problem areas. The first type is problems in the pronunciation of non-native sounds. Based on the theory, in this type, the speakers will replace phonemes that do not exist in their original mother tongue (L1) with the closest-sounding phoneme from their native language. For example, in pronouncing the word "seven", rather than pronouncing /ˈsɛv(ə)n/, the speaker tends to pronounce /ˈsɛfən/. They change the medial consonant /v/ with the consonant sound /f/. This problem may be caused by the sound /v/ and /f/ which have the same manner and place of articulation, but the sound [f] is voiceless, whereas the sound /v/ is voiced. To simplify the pronunciation, the speaker substitutes the sound /v/ with the sound /f/ (Habibi, 2016).
The second type is carry-over of pronunciation regularities from the mother tongue (L1). It involves mostly phonological changes, such as the devoicing of final voiced stop consonants or the spelling-to-sound mapping of the mother tongue. For example, Indonesians sometimes devoice English final voiced stop consonants, as in the word "six", which should be pronounced /sɪks/. However, Indonesians pronounce it /sɪk/. The final stop consonant /s/ is omitted.
The last type is overgeneralizations of the target language (L2). In relation to this type, the speakers apply a possible pronunciation in an unsuitable word. Coarticulation errors also belong to this type. This type of error consists of deletions or assimilations of initial and final consonants or vowels when the adjacent phones influence each other. For example, when pronouncing the initial vowels of the words "cognitive" and "compare", the initial vowels of those words should be /ɒ/or /ɑ/ for "cognitive", and /ə/ for "compare". However, the speaker tends to pronounce Acces article distributed under the terms of the Creative Commons Attribution license (https://creativecommons.org/licenses/by-sa/4.0/) the initial vowels of both words similar to /ɒ/. They pronounced /ˈkɒɡ.nə.tɪv/ for "cognitive" and /kɒmˈpeə/ or /kɒmˈper/ for "compare".

The influence of L1 on L2
Rahmat (2019) states the speakers' mother tongue can be a barrier for them to learn the target language since it is easier and sounds more natural for them to use with their mother tongue when they communicate with others. They are afraid of getting a lot of attention when they use foreign language to communicate with others. Besides, the researchers explore whether the speakers' mother tongue (L1) has an effect on the speakers' second language (L2) performance. Therefore, we consider it as an important aspect for non-native speakers to grasp the phonological system of the second language fluently. Van Weeren and Theunissen (1987) state that the phonological system includes the sounds and intonation of the language. The effect of Bahasa Indonesia on their English oral proficiency is discussed in this paper because it is a factor that is non-negligible in the occurrence of the mispronunciations. The phonetic differences between the speakers' L1 and L2 can be the main reason of the occurrence of mispronunciations among L2 speakers. Flege (1993) explains that when there are sounds that do not exist in one's L1, it is replaced with similar sounds that do exist in one's L1 when using the L2.
Furthermore, the findings of previous studies in the field of the mispronunciation problems of Indonesian speakers (Rosyidah, 2014;Habibi, 2016)  . As a result, /lʊk/ and /ˈpʌblɪk/ were pronounced /luːk/ and "public" /ˈpablɪk/. These examples indicate that the participants tended to replace some English vowel sounds with other similar sounds. Furthermore, the studies also reported that most of Indonesian advanced students often pronounced vowel sounds based on the orthographic writing. This problem occurred when they mispronounced the schwa /ə/. Besides, previous studies (Mustikareni, 2013;Habibi, 2016)

Method
To answer the two research questions, the researchers employed a qualitative study, defined as a study of facts in a narration with words to describe a connection between parts of a whole and depth of understanding of a phenomenon (Ary et al, 2010;Gall & Borg, 2007). Besides, Merriam et al. (2002) state the result of a qualitative study would be in a form of words or pictures to carry all data that have been collected through interviews, observations, or document analyses to identify the patterns of the data. The type of qualitative study that employed in this study was ethnographic study. Ary et al (2010) define it as an extensive study of people in their natural environment in order to obtain understanding into how people interact within their natural environment. Its methods are such as prolonged participant observations, face-to-face interviews, and documents or artefacts observations. The researchers were convinced that this research design was suitable for this study since they collected the data through participant observation during the presentation which occur in their natural setting, video recording, and video transcription. Moreover, the goal of this study was to examine the linguistics factors-the influence of L1 to L2-contributing to the mispronunciations in presentation project conducted by the participants.
The data of this study were the mispronunciations produced by graduate students of batch 2017 of the English Education Master's Program of Sanata Dharma University. The 2017 batch consisted of 21 students; however, the researchers randomly selected approximately half of the population, 11 students, as the participants of the study. Furthermore, the mispronunciation data from the sample were used to draw conclusions of this study. The data were taken from group presentation projects that the students did in one of their classes. The participants belonged to two different groups: group 1 consisted of five members and group 2 consisted of six members. The group was decided by the students themselves; the researchers did not contribute to the group selection process. Each group did one presentation so that the researchers analyzed two presentations as the data of this study. The researchers conducted a study on the two groups since the members of the groups had different experiences in using their spoken English. Kretzschmar (2008) states this phenomenon is influenced by students' social environment. Here, a student who came from X region might pronounce a word in a different way compared to other students from Y region. In a multicultural country like Indonesia, Bahasa Indonesia might not be the participants' first language or mother tongue (L1) since this country has many regions with its local languages. As a result, this variation of cultural backgrounds among the participants would be a good consideration for the result of this study.
The researchers used a document analysis as a technique to analyze the data. Leedy and Ormrod (2005) define a document analysis as a detailed and systematic examination of the contents of a particular body of material by identifying purposes, themes or biases. They further state that a document analysis examines human communication, including books, newspaper, films, television, art, music, videotapes of human interaction, and transcripts conversation. In this study, the presentations were recorded and transcribed to ease the researchers in detecting the mispronunciations. The researchers focused on the presenters' speech. If there were some question and answer sessions with the audiences of those presentations, the Acces article distributed under the terms of the Creative Commons Attribution license (https://creativecommons.org/licenses/by-sa/4.0/) researchers did not transcribe them since the audiences were not the participants of this study. After completing the transcriptions, the researchers re-watched the videos and marked the mispronunciations while watching those videos. To avoid mistakes, first, each of the researchers was in charge to watch one video, make transcriptions, and find mispronunciations. Then, they listed the mispronunciations that they had found. Next, they exchanged their results and double checked them. To check that the participants mispronounced words, the researchers consulted both offline and online dictionaries to find the correct pronunciations, both American English and British English.
After identifying all mispronunciations, the researchers categorized them based on the three types of mispronunciations proposed by Bonaventura et al. (2000). As mentioned earlier, those are the problems in pronunciation of non-native sounds, carry-over of pronunciation regularities from the mother tongue (L1), and overgeneralizations of target language (L2) regularities. Afterwards, the researchers compared the linguistic features of Indonesian (L1) and English (L2).

Findings and discussions
This section discusses the findings of mispronunciations which occurred in the presentation projects of Indonesian graduate students. The results show that there are 89 mispronounced words which were produced by the participants of this study because of some reasons. In this study, the mispronounced words were classified into three categories based on a theory from Bonaventura et al. (2000), namely problems in the pronunciation of non-native sound, carry-over of pronunciation regularities from the mother tongue (L1), and overgeneralizations of target language (L2) regularities. There were 26 mistakes for the first type of mispronounced in which the problems in the pronunciation of non-native sound were caused by the replacement of phonemes that do not exist in their original mother tongue with the closest-sounding phoneme from their native language. The total number of this type of mistakes reached 29.21%. Another type of error happened because the participants tended to carry over the pronunciation regularities from the mother tongue. There were 36 mistakes for this kind of error or it was equivalent to 40.45%. Lastly, the errors happened because the participants used a possible pronunciation to unsuitable words. There were 27 mistakes for this error, equaling 30.34%.

Problems in the pronunciation of non-native sound
The following table shows some of the words that belong to the first type of mispronunciations that were classified based on the theory of Bonaventura et al. (2000). As shown in Chart 1, type 1 errors reached 29.21% from the total mispronunciations. It indicated that from 89 mispronunciations that were identified, there were 26 words that belonged to the first type. Based on the theory of Bonaventura et al. (2000), the mispronunciations of this type happened when the participants tended to replace some correct phonemes in English with the closestsounding phonemes in Bahasa Indonesia. Basically, the participants were aware of the correct pronunciations in English but the phonemes that appeared in English words did not exist in their original mother tongue (L1), namely Indonesian.
The examples of this mispronunciation can be seen when the participants pronounced the words "overall" and "think". The two words were mispronounced by the participants (see Table 1, numbers 11 and 24). The word "overall" should be pronounced /ˌəʊ.vəˈrɔːl/ in British English or /ˌoʊ.vɚˈɑːl/ in American English but some participants pronounced it /ˌo.vərˈɔːl/. It revealed that the participants tended to change the initial vowels which were pronounced by using a diphthong sound /əʊ/ with the phoneme /o/. As explained before, it happened because there was no diphthong sound in Bahasa Indonesia so the participants tried to replace it using the closest-sounding phoneme in Bahasa Indonesia. The participants also made the same mistake in pronouncing the words "emotion" and "open". On the other hand, the word "think" should be pronounced /θɪŋk/ both in British English and American English but some participants pronounced it /tɪŋk/. It revealed that the participants tended to replace the phoneme /θ/ with /t/ because that phoneme did not exist in Bahasa Indonesia. The participants also made similar mistakes in pronouncing the words "theory" and "three".

Carry-over of pronunciation regularities from the mother tongue (L1)
The following table shows some of the words that belong to the second type of mispronunciations that were classified based on the theory of Bonaventura et al. (2000). As shown earlier in Chart 1, errors of type 2 were the most frequent type which occurred in the data, reaching 40.45% of the total mispronunciations. It indicates there are 36 words from 89 mispronunciations that have been found by the researchers that belong to the second type. Based on the theory that has been discussed in the review of related literature (see p. 3), this type carries meaning that when the participants pronounce the English words, they have a tendency to carryover of pronunciation regularities from Bahasa Indonesia. According to the theory, the carry-over of pronunciation regularities mostly involves phonological changes, such as the devoicing of final voiced stop consonants or the spelling-to-sound mapping of the mother tongue.
Furthermore, this type can be exemplified as when they had to pronounce the words "focus" and "ask". Those words have been found by the researchers to be mispronounced words by the participants (see Table 2, numbers 2 and 15). In the case of pronouncing the word "focus", instead of pronouncing it /ˈfəʊ.kəs/ as in British English or /ˈfoʊ.kəs/ as in American English, the participants tend to pronounce /'fo.kus/, just similar to the written word. The participants substitute the medial consonant /c/ with the similar sound consonant that exist in Bahasa Indonesia, which is /k/. They also replace the diphthong sound /əʊ/ with the simple vowel sound /o/. It may happen because Bahasa Indonesia has the word "fokus" which have the same meaning with the English word "focus". In relevant to the second type of mispronunciations, the phenomena happens because the carry-over of pronunciation regularities from Bahasa Indonesia. Since they have known that "focus" in English and "fokus" in Bahasa Indonesia have the same meaning, they simply pronounce the word "focus", that should be pronounce /ˈfəʊ.kəs/ or /ˈfoʊ.kəs/, with the pronunciation of the word "fokus", which is /'fo.kus/.
Moreover, the same type of mispronunciations happens when the participants pronounced the word "ask". In English, that word should be pronounced /ɑːsk/ in British English or /aesk/ in American English. However, most of the participants pronounced it /ɑːs/. They devoiced the sound of final stop consonant /k/ as suggested by their mother tongue's orthography. This may happen because Bahasa Indonesia rarely had a word which has a consonant cluster at the end of a word, like /sk/ in the word /ask/. Therefore, they simply omitted the final sound /k/ and pronounced it /ɑːs/.