How First and Second Languages Influence Indonesian Students’ English as a Third Language: Transfer and Interference Analysis

In the face of the scarcity of linguistic studies on Batak Toba language, this study tries to contribute by investigating how Batak Toba students’ local language and mother tongue’s elements are transferred to their English as a foreign language output. Specifically, this study aims to describe the transfer and interference of Batak Toba and Indonesian as first and/or second language and how they influence learners’ English as a third language. The study observed 15 English Education programme students who were selected via snowball sampling and analysed the way the students speak and translate a set of phrases to their respective second and third languages. Results indicated that the students’ first and second languages interferes with their English language performance at the phonological, morphological, and syntactic levels. The novelty of this study is that it offers a detailed investigation of how students whose first language, Batak Toba , and second language, Indonesian , interfere with the foreign language that is a mandatory subject in the country’s formal education system, namely English . Recommendations for teachers, particularly teachers of Indonesian learners whose students are primarily bilingual and multilingual, are discussed.


Introduction
Throughout the history of linguistics, many theories have been developed by experts in the context of learning and acquiring language, both in learning and acquiring the first language, as well as the second language as the target language. However, in the application of theories of acquisition and learning of this language, many learners, understandably, tend to be ignorant and do not realize or are not aware of these theories (Gao, 2013). This is where the role of the language teacher explains how a learner should choose an effective learning strategy when learning the target language (Al-Rawahi & Al-Balushi, 2015;Shi, 2017).
From various theories of second language acquisition, it seems that transfer and interference theory are two theories that simply transfer and make comparisons between the first language and the second/third/additional language. The process of acquiring an additional language cannot be separated from the process of transfer and interference (Derakhshan & Karimi, 2015). Both of these are related to the influence of the first language or the previous language on someone's acquisition of an additional language. Learners tend to transfer elements of their first language when carrying out the use of an additional language, and this process happens especially at the starting weeks of the acquisition process (Rahayu, 2012). This transfer from a person's mother tongue to their additional language can be considered positive if there are similarities between the two languages, and deemed as a negative transfer or interference if the structure between the two are too different. Due to the frequency of this interference, many students, teachers and language professionals alike have considered first language a "nuisance" in mastering a new language because, consciously or not, the learner transfers the structure, phone, lexical item of the language they know to the language that they are learning in the classroom, which often leads to errors, code-mixing, and other linguistic forms considered as mistakes by teachers in the classroom (Iswandi, 2021;Marwa, 2014;Purandina et al., 2021).
The theory of transfer and interference is born from the decades of scientific attempt to describe the differences and similarities between a pair of languages, which is known on the field as contrastive linguistics. Majority of these studies, as part of the larger field of second language acquisition, has been focused on contrasting students' first or native language and English as their second language. Little research has been done in the context of when the learners already had a first and a second language, and how these influence their third language. The novelty of this study is that it will investigate how students whose first language, Batak Toba, and second language, Indonesian, interfere with the foreign language that is a mandatory subject in the country's formal education system, namely English. This study is interested to apply the transferization and interference linguistic theories to understand the third language output of Indonesian students from the Batak Tribe.
Indonesia as a multicultural country boasts approximately 1,001 local languages shared by 714 tribes, according to the 2010 census of the Central Bureau of Statistics (Azanella, 2019). A closer investigation on one of the local languages of the most linguistically diverse countries in the world will contribute to the world's ongoing, strong trend of multiculturalism (Augoustinos, 2022;Fisher et al., 2020). The languages of Batak Tribe in particular is highly unique with its distinct sounds that almost all Indonesians can immediately recognize them just by hearing them talk. The Batak languages have two main branches, Northern and Southern, and the researcher, as part of the Batak Toba people, is interested to study the Batak Toba dialect and how natives of this language transfer its elements to English as a third language. The results of this study will contribute to the body of research on Indonesia's local languages and the growing interest on learners' third languages.

Transfer and Interference Theory
The transferization theory is concerned on the case when individuals are learning an additional language, in which they tend to transfer the rules of the language that they have already masters or previously acquired to the new language that they are learning. Due to the commonness of this transferization process, linguist experts agree that an individual's first language is an inevitable significant influence on the process of their second language acquisition. The event of an element of the first language entering the target language output is the interference process, and they often come in the form of errors because the language learners may not recognise the appropriate rule or structure in the target language. The forms of these errors can vary depending on the first and target language. For instance, Sabbah (2015), who traced the literature on Arabic and English studies, found and summarized seventeen differences that causes Arab learners of English to make mistakes, which include misuse of singularity, duality and plurality of nouns (Salim, 2013), countable and uncountable nouns, definite article, prepositions, adjectives and adverbs, errors in using some English modal verbs, word order, tenses (Aoun et al., 2009), relative clauses (Ali, 2007, pronouns, punctuation, pronunciation (Grami & Alzughaibi, 2012), spelling (Al-zuoud & Kabilan, 2013), lexical errors, lexicosemantic usage, redundancy in writing essays, and differences between Arabic and English vowel system (Al-Badawi, 2012). So many aspects in language can be produced as errors simply because of the students' first language's differences with their target language, therefore, the literature body of cross-language transfer has the mission to discover and reveal the necessary conditions which would enhance language learning for educators and researchers alike.
Various researchers have looked into how prior mastery of one language can interfere with the acquisition of another, usually foreign, language. Jabbari & Samavarchi (2011) were interested in the different phonotactic constraints in both English and Persian to investigate elementary level Persian learners' syllabification of English consonant clusters. They discovered a negative transfer based on the students' pronunciation of English primary consonants; because the equivalents do not exist in Persian, students tended to add a vowel or exclude a consonant to make the words easier to pronounce. A look into the how Thailand natives' first language interference to their English writing showed that there are three levels of interference (words, sentences and discourse) and that the most dominant type of error is literal translation, which is a chronic problem among Thai EFL students (Bennui, 2016). Manyasa (2021) analysed Tanzanian learners' morpho-syntactic interference errors using the interlanguage theory and the error analysis approach, revealing that students' dominant error type is the absence of French determiners, followed by nouns originated in English and omission of prepositions due their previously acquired or learned languages of Swahili, English and ethnic languages. A case study on a 5-year-old Japanese child whose target language is Indonesian showed that the most common type of language interference that occurred to the child was phonological, then lexical and grammatical errors (Adnyani et al., 2021).
Multiple other studies studied language transfer and interference in a more theoretical manner. Al-Khresheh (2016) reviewed the theory of error analysis; Bahr et al. (2015) argued that bilingual spelling patterns are more than simple language transfers; Deng & Zou (2015) used the perspective of children's native language acquisition to figure out the level of difficulty of adults' second language acquisition; and Chung et al. (2019) researched the theoretical frameworks of cross-language transfer processes (e.g., Lado's (1964) contrastive-typological framework, Cummin's (1981) linguistic interdependence framework, Geva and Ryan's (1994) common underlying cognitive process, and Koda's (2008) transfer facilitation model) to create their own framework to comprehensively describe the complex linguistic and cognitive processes involved in cross-language transfer. In the search for studies on transfer and interference, the researcher only located one citation that explicitly considered the influence of first language on learners' third language instead of the majority's second language, but this citation is a book chapter rather than an original study. Therefore, this paper will contribute more information on how students' first and second language can interfere with their third language.

Current Linguistic Studies on Batak Toba
Batak is a collective term used to identify a number of closely related Austronesian ethnic groups predominantly found in North Sumatra, Indonesia, who speak Batak languages (Claudia, 2018;Siahaan, 2015). The Batak languages is generally divided into two main branches, namely Northern Batak and Southern Batak.

Figure 1. Classification of Batak Languages
This study is concerned on the particular dialect of Batak Toba, which is simultaneously used as a noun for the language and an adjective for the speakers. There is currently very few published linguistic research on Batak Toba, and even fewer in the context of this dialect and English language classroom. A number of studies focused on carrying out analysis on the tribe's folklore and literary work. Simanjuntak (2021) analysed the meaning of metamorphic sentences in a commercially popular Batak Toba folk song called "Aut Boi Nian" written by Wervin Panggabean, concluding that it contains cultural and moral values of Batak Toba Tribe, particularly Bataknese men in general. Simanjuntak et al. (2022) also extended the previously mentioned study to publish an analysis of seven Batak Toba folk songs about Bataknese men (i.e., Tilhang Gultom's "Mardalan Ahu Marsada-sada", Dompak Sinaga's "Dang Hasuhatan Burjumi Inang", Charles Hutagalung's "Host of Pangintubu written", Korem Sihombing's "Unang Manarita Ho Inang", Tagor Tampubolon's "Tangiang Ni Dainang", Diara Jatoman Simare-mare's "Ingkon Borhat", and Abidin Simamora's "Amang Na Burju". They tried to find out if the English translations of the songs would still reflect the values of the tribe. On the flipside, Saragih (2021) chose to analyse eight Batak Toba songs that contain the theme of mother's love (i.e., Trio Elexcis' " Tangiang  Besides the tribe's lore, other studies focused on peeling the values embedded in Batak Toba's communication patterns. A linguistic anthropology study found out that Batak Toba society's local wisdom considers Indonesia's gotong royong (mutual cooperation) value as marsirimpa or marsirumpa (cohesive, in unison, and together). Novelita et al. (2019) observed the process of traditional Batak Toba marriage communication to promote knowledge of the tribe's uniqueness. Meanwhile, Lubis & Asnawi (2021) detailed 20 distinct unique addressee or pronouns that Batak Toba people have for 20 different family members, and according to their informants, if a person uses the wrong addressee when they are greeting a specific person, they would be called dang maradat (uncivilised).

Method
The design of this study was qualitative descriptive. The data sources were 15 students of the English Education Study Program, Faculty of Teacher Training and Education, Universitas Jambi whose native languages are the Batak Toba language and Indonesian language. Data were collected with the snowball technique, in which the first data source being an informant for the second data source, the second referring to a third student as data source and so on until the researcher reached the point of data saturation, which is fifteen.
The students are divided the students into two groups of informants to reflect two different patterns of what first language will influence the target language, in this case, English as a third language. The first pattern or group is coded as L1/BT, which include 8 students whose first language is Batak Toba and then they obtained Indonesian as their second language from parental and schooling encouragement and encouragement. The second pattern or group (code: L1/IND) are 7 students whose first language is Indonesian and their second language is Batak Toba. This pattern existed commonly because the learners had been born and grown up with Indonesian as their home language environment. Later on, in their daily social interactions, they hang out with friends and communicate with Batak Toba speakers, thus acquiring it as their second language.
This study's source of data relied on observation of the way the students speak and translate a set of phrases to their respective second and third languages. Data were analysed through data collection flow techniques, data reduction, and drawing conclusions. For the validity of the data, the theoretical triangulation technique was used such as confirming the data findings with the linguistic theory of transfer and interference in target language learning.

Results
Indonesian learners whose mother tongue is Batak Toba in general when learning the target language (English) often transfers at the level of the linguistic system in the field of phonology. The transferization process can be seen in Data 1.

Data 1_L1/BT
Data 1 is the observational data of the target language learning group whose first language is Batak Toba language. It should be noted that in fact Batak Toba language does not recognize the 'senteral' lexeme. This lexeme is a lexeme that comes from Indonesian, the 'sentral' lexeme into the Batak Toba language. In the pronunciation for Batak Toba speakers, it tends to be /senteral/. Then, the researcher compared it with Data 2 which comes from from a student whose first language is Indonesian, and Batak Toba is their second language.

Data 2_L1/IND
In Data 1, Batak Toba language learners pronounce (BT)/centRral/ direct phonological transfer from Batak Toba language (B1)/senteral/ to (B2)/sentral/. From /sentral/ to (BT)/centRal/. While in Data 2, target language learners whose first language is Indonesian pronounce (BT)/central/ directly from (B1)/sentral/ to (BT)/centRal/, without the influence of (B2)/senteral/. Thus, transfer theory is very helpful in the learning process (BT), English for speakers of the first language is Indonesian. Meanwhile, for speakers whose first language is Batak Toba, actually there is chaos in learning the target language (English). The chaos occurred in the transferization process from (B1) to (B2) then to (BT). Thus, there is an influence of the first language, Batak Toba language /senteral/ to /centeral/, which should be /centRal/.

Data 3_L1/BT
Batak Toba language learners pronounce (BT)/coRneR/ direct phonological transfer from Batak Toba language (B1)/korner/ to (B2)/korneer/ and then from /korner/ to (BT)/coRneR/. However, target language learners whose first language is Indonesian pronounce (BT)/coRneR/ directly from (B1)/korner/ to (BT)/coRneR/, without the influence of (B2)/korneer/. Thus, transfer theory is very helpful in the English learning process of speakers of the first language is Indonesian, but it is more "chaotic" for speakers whose first language is Batak Toba. The chaos occurred in the transferization process from (B1) to (B2) then to (BT). So, there was an influence of the first language, Batak Toba language /korneer/ into /corneer/, which should be /coRneR/.

Data 4_L1/BT
Data 4 shows that Batak Toba first language speakers, in pronouncing (BT)/bank/, immediately transfer phonologically (B1)/beng/ from Batak Toba language into English (BT)/beng/, thus facilitating the learning of the target language. In this case, the target language learners whose first language is Batak Toba, it seems that there is no influence of the second language (Indonesian). This pronunciation (BT)/beng/, which is almost the same as the pronunciation in Batak Toba language as the first language of the target language learner, is different from the pronunciation of the second language of the target language learner (Indonesian). This is because they are learners of the target language whose first language is Indonesian, then they acquire Batak Toba language as a second language. For more clarity, consider the following Data 5.

Data 5_L1/IND
For target language learners whose first language is Indonesian, there is no effect. Even though they already know that in their second language, Batak Toba, the pronunciation of /bank/ is (B2)/bankK/, while in (B1)/benK/, it makes the transfer process easier. They just follow the phonological pattern of the Batak Toba language as their second language (Data 5).
In addition to transferization at the phonological level, transferization also occurs at the morphological level. In the field of morphology, Indonesian learners whose first language is Batak Toba often transfer the structure of the language or the morphological system of the first language into the target language. This is especially common in

Data 6_L1/BT
In the case of Data 6, (BT) English 'pen I', a structural error was explained explaining (DM), which should have been 'My pen', (MD). Indonesian learners whose first language is Batak Toba language immediately transfer the structure or morphological system of Batak Toba language into English, which is supposed to represent pena saya, is (BT) 'my pen' (MD), they tend to immediately make 'pen'. I' (DM) is the same as the structure of both Indonesian and Batak Toba languages. It turns out that their transferization is not compatible with the English morphological system. In this case, they make a mistake or error. However, this transfer theory actually makes learning easier. It is just that an explanation from the instructor or teacher is needed. This is where the role of teachers is to provide training so that such cases do not occur.

Data 7_L1/IND
In the case of Data 7, it can be seen that what the target language learners (B1) Indonesian and (B2) Batak Toba language do, the case is the same as in the data case (6). Thus, even though (B1) Indonesian and (B2) Batak Toba, they did not show a significant effect in the use of language transfer. In this case, the target language learners, both in the Data 6 and 7 both made mistakes. This happens because the structure of the Indonesian language with the structure of the Batak Toba language, especially in terms of the morphological system is the same. In the structure of the Indonesian language, 'my pen' (DM), as well as in the Batak Toba language structure it is also 'my pen' (DM).
In the case of Indonesian learners whose first language is Batak Toba, the linguistic subsystem is that Indonesian learners often transfer, such as depicted in Data 8.

Data 8_L1/BT
Indonesian learners whose first language is Batak Toba immediately transfers the linguistic subsystem from Batak Toba language into English (Data 8). In this case, the target language learners directly transfer the linguistic subsystem of Batak Toba language and Indonesian language without adding the phoneme /s/ as one of the characteristics of the English linguistic subsystem at the end of the word /pen/, which should be /pens/. The addition of the 's' phoneme absolutely must exist in the English subsystem.

Data 9_L1/IND
Data 9 shows the case of when the target language learner has Indonesian as their first language. It does not seem different from what happened to Data 8. In this case, target language learners whose first language is Indonesian or Batak Toba are the same. They both carry out the same pattern of transfer of linguistic subsystems. This happens clearly because the linguistic subsystem of Batak Toba language with the Indonesian language subsystem has the same characteristics.
Taking into account of Data 8, it turns out to be the same as in Data 6 and 7. The transferization that they do causes errors in the morphological system of the target language (English). In the case of Data 6 and 7, an error occurred at the morphological level. What they did was actually something natural, because the morphological system of Batak Toba language with Indonesian tends to show the same characteristics. Meanwhile, the morphological structure of English is clearly different. It is the difference that causes errors when implementing transferization. However, the error can be corrected through an explanation from the instructor or supervising teacher.
In another case when the transfer theory is applied, at the morphological level it happens easily, although sometimes it is true, but it can also be wrong. Consider Data 10.

Data 10_L1/BT
Indonesian learners whose first language is Batak Toba language easily transfer morphologically from Batak Toba language into English. (B1) 'sada table', (B2) 'satu meja' and (BT)'one table', but actually they made a mistake. It can be seen that if in daily oral communication, without affixing 'a' to 'a table', the interlocutor must have understood and understood his purpose, that he was stating a table (Data 10). However, it needs to be emphasized, if the communication that takes place is not speech, but writing, of course without the 'a' in 'a table' is not justified, or in other words it violates the grammar or structure of the English language.
In the case of target language learners whose first language is Indonesian, it turns out to be the same when compared to those whose first language is Toba Batak, that is, they both make mistakes.
(B1) Indonesian : /satu meja/(a Observing Data 11, it is clear that what a target language learner whose first language is Batak Toba, is the same as what a target language learner whose first language is Indonesian. They both made mistakes. However, the error is actually not a very significant, because the meaning that arises from the structural error does not change, or does not give rise to new meanings or other meanings. The meaning remains the same that there are three pens.
At the level of linguistic construction, Indonesian learners often transfer to the formation of command sentences. In this case, Batak Toba language learners are no exception. For clarity, consider the following Data 12.

Data 12_L1/BT
The data indicates that the transferization process carried out by Indonesian learners whose first language is Batak Toba language and second language is Indonesian, they are very easy to do. However, in spoken language there seems to be no problem, but in written sentences, it should be noted that in English it is not acceptable as in Indonesian /Jangan datang terlambat/ atau /Jangan terlambat datang/. In English it is always certain that the form of the structure is explained to explain. The form of 'Don't come late', is not accepted as in Indonesian and Batak Toba language, 'Jangan terlambat datang', or in Batak Toba language 'Unang tarlalang ro'.

Data 13_L1/IND
To see whether the previous pattern is reflected or not in the case of language learners whose first language is Indonesian and the second language is Batak Toba language, the answer is shown in Data 13: There is no difference in what the target language learners do whose first language is Batak Toba language. This happens because the structure of Batak Toba language has similarities with the structure of the Indonesian language.
The type of interference that often occurs in learning the target language is in the syntactic order. The structure of the Toba Batak language and Indonesian is explained (DM). However, the English structure is the opposite, explainingexplaining (MD). This form of (DM) is often directly entered or interfered in learning English. Consider the following data (14)

Data 14_L1/BT
Subsequent interference also often occurs at the syntactic level. We note how the interference from the first language and the second language of the target language learners directly transfers the syntactic structure as in the following Data 15.

Data 15_L1/BT
Observing the data (15), it can be seen that Batak Toba language learners directly interfere with such as Batak Toba grammar and Indonesian without adjusting it to English grammar. This also seems to happen to target language learners whose first language is Indonesian, and the second language is Batak Toba language. (B1)'lomo hian rohana banana', (B2) 'Dia sangat suka pisang', is directly transferred to the target language form (BT) 'He very likes banana'. This form still uses the structure of Batak Toba language and Indonesian language. Learners should follow the structure of the English language. The form of the sentence should be; 'He likes banana very much'.

Discussion
This study offers a unique insight on how EFL students' language output is influenced by their first and second languages. This paper compared two patterns: how students' Indonesian as their first language and Batak Toba as their second language influence their English as their third language's performance, and how students' Batak Toba as their first language and Indonesian as their second language influence their English as their third language's performance. The results of this study will not only contribute to the research on third languages and shedding light on one of Indonesia's local languages.
Interference and transferization generally occur in phonological, morphological and syntactic settings. As in the data; 'senteral', 'sentral' to become 'central'. Likewise for data; 'korner', 'korner' to become 'corneR'. The same is true for data; 'beng', 'bang' to become 'benK'. With all of these data, learners, both whose first language is Batak Toba and Indonesian, both transfer and interfere in learning the target language at the phonological level. This type of influence also occurs on the process of Persian as a first language's influence on English as a second language (Jabbari & Samavarchi, 2011).
Then at the morphological level, transferization and interference of the first and second languages are also carried out. On data; 'pen hu', 'pen me', to become 'pen I', which should be 'My pen'. In this case, the target language learners, both whose first language is Batak Toba language, or Indonesian language, both transfer and interfere. Likewise in the case of data; 'tolu pen', 'tiga pena', 'three pens', and became 'three pens'. These cases of literal translation process are similarly echoed by Thai EFL students, as explained in the study conducted by Bennui (2016). Literal translation still frequently occurs and not limited to the classroom context (Farghal & Bazzi, 2017;Nabokov, 2021;Poudel, 2019). However, it, as well as code-mixing, can be problematic in the classroom if done too frequently and can negatively affect the students' academic performance (Akan et al., 2019;Sedina & Wan, 2018).
The process of transferization and interference in the linguistic construction of the target language, Indonesian learners who have a mother tongue of Batak Toba and Indonesian, also occur at the level of syntax. This can be seen in the data; 'lomo hian rohana banana', 'Dia sangat suka pisang, 'He very likes banana', then became 'He likes banana very much'. Further experimentation to investigate the syntactic errors of students may replicate Manyasa (2021) who studied how Swahili or English as a first language interfered with Tanzanian learners' French language output.
Considering the classic view that errors are often considered seen as mistakes, this study agrees with Sabbah (2015) who made a distinction between errors and mistakes-mistakes are lapses in performances, but errors are signals of a learner's gap of knowledge. In the process of this research, the author realised that discussion of the transfer and interference theory can help learners in realising their subconscious linguistic construction at a phonological, morphological and syntactic level. These value of these theories for students who are not aware of their how their second language can also interfere with their target language acquisition is much more emphasised for students in Indonesia who are dominantly bilingual and multilingual.
This study recommends that in teaching English to Indonesian learners, it is better to prioritize vocabulary and pronunciation, not language structure. Students are generally able to arrange the structure after they mastered the vocabulary and pronunciation. Previous studies have also showed that students often perform in poor terms due to crippling or low self-confidence of their lexical repertoire (Batubara et al., 2020;Liu & Ni, 2015;Mukminin et al., 2015).
Moreover, previous studies have mostly focused on translating Batak Toba's literary works (folklore stories and songs) into English to figure out whether the tribe's values are still clear in spite of the change of language (Saragih, 2021;Simanjuntak, 2021;Simanjuntak et al., 2022) and converting Batak Toba folklore stories into English lesson materials (Ginting et al., 2019;Ginting et al., 2021;Ritonga et al., 2021;Sihombing et al., 2021;Sinamo et al., 2021). Therefore, this study also offers a strong contribution as a study on Batak Toba language in the context of English language classroom, which heavily focused on analysing the linguistic differences of the local language, the national language, and the most popular and mandatory foreign language.

Conclusion
Transfer and interference theory are two theories that were previously considered to make it difficult for learners to acquire the target language. However, this study found that these two theories are very helpful for Indonesian learners whose English performance is influenced by their first and second languages, namely Indonesian and Batak Toba. Their first and second languages have been identified to influence their English as a third language on three levels: phonological, morphological and syntactic. This study concludes that in learning English for bilingual/multilingual Indonesian students, it is better not to put too much emphasis on language structure, but prioritise vocabulary and pronunciation mastery. Ultimately, the results of this study showcased the linguistic system of Batak Toba language and Indonesian language and how they influence the English linguistic system, which can be valuable contribution to the burgeoning field of third language studies.