NATIVE AND TARGET LANGUAGE INFLUENCE ON THE STUDENTS ’ INTERLANGUAGE PRODUCTION : A CASE OF INDONESIAN EFL

English compositions written by Indonesian EFL students contain erroneous sentences which portray learner language. The errors are neither of their native language nor the target language, but containing linguistic system from both. This is called an interlanguage. This study focuses on one of interlanguage features, that is, permeability, meaning the susceptibility of interlanguages to infiltration by first language and target language rules or forms. It aims to provide empirical evidence of the permeability of the students’ interlanguage production by describing the types and degree of the native and target language influence and explaining the possible causes of the influences. The data were 264 ill-formed sentences elicited from their English free compositions. Error analysis and interlanguage analysis were used as framework for collecting, identifying, describing, and explaining the data. The results indicate that their interlanguage production was influenced by their native language and the target language at both lexical and syntactical level. The dominant native language influence was on vocabulary (i.e. Indonesian borrowings) and the target language influence was on grammar (i.e. verb tenses). The native language influence had a little lower frequency compared with that of the target language. The main source of the influence was their possession of two language systems in their mind was activated regardless of their intention to use one language only. The native language influence was due to the good mastery of the native language and the limited knowledge of the target language. The target language influence was due to the learning strategy used.

The observable phenomenon indicates that English compositions written by Indonesian EFL learners contain erroneous sentences (Mardiono, 2003;Halimi, 2008;Faisal, Mulya, & Syamsul, 2016).These sentences which represent learner language are neither those of native language or NL (Indonesian) nor the target language or TL (English).They contain linguistic elements of both Indonesian and English as in the following sentences: "After pass the exams I will job at the police; "I really proud become citizen Indonesian".In second language acquisition (SLA) study, the learners' second language (L2) which deviates from the TL is termed by Selinker (1972) an interlanguage (IL) or learner language.It describes the type of language produced by second or foreign language learners who are in the process of learning a new language.It represents the learners' journey from their NL or first language (L1) to acquisition of the TL or L2.It is thought of as a third language that is unique to particular learners which represents a system that has a structurally intermediate status between the NL and TL.It is neither the system of the NL nor that of the TL, but instead falls between the two.Sharwood-Smith (1994, p.7) defined IL as "the systematic linguistic behavior of learners of a second or other language; in other words, learners of non-native languages."The world "language" suggests that IL is a separate linguistic system which has specific characteristics different from both the learner's NL and the TL being learned but linked to both.Meanwhile, the word "inter" suggests that this version is viewed to be an intermediate stage in the learner's linguistic development.The alternative term for IL is "learner language" which refers to "the oral and written language by L2 learners" (Ellis and Barkhuizen, 2005, p.4). Saville-Troike (2012) suggested that learner language has empirically been found to be systematic, dynamic, variable and simplified, both formally and functionally, relative to the TL and the learner's NL".
The IL features can be summarized from Ellis & Barkhuizen (2005) and Saville-Troike (2012) as follows: (1) IL has its own linguistic system; (2) IL is dynamic or approximative in nature in the sense that the system changes frequently; (3) IL is variable in a sense that at any stage of development the learner employs different forms for the same grammatical structure; (4) IL is the product of various learning strategies such as native language transfer, overgeneralization or simplification; (5) IL may fossilize, i.e. the learner's IL system stops to develop to achieve a full native-like grammar; and (6) IL is permeable or is open to influence from doi: dx.doi.org/10.17509/ijal.v7i1.6858learners' NL and TL system.Of these features, IL permeability becomes the focus of this study.
The permeability of IL was first noted by Brown (1994) and also Connor (1996) who categorized learner language errors into two types: interlingual and intralingual.Interlingual errors are those resulting from the influence the learner's NL or external factors, whereas intralingual errors are the ones resulting from the TL system itself or internal factors.Brown also states that beginning learners are commonly familiar merely with their NL system.Therefore, many learner errors in this stage are due to the influence from their NL system.
Likewise, O'Grady, Dobrovolsky, & Katamba ( 2002) discussed the dual nature of IL in the ontogeny model for SLA in which they categorized errors into two: transfer errors and developmental errors.Transfer error reflects the influence of NL, whereas developmental errors involve the sort of errors as a result of overgeneralization or simplification of TL rules.In this view, developmental errors are in fact similar to intralingual errors.It is predicted that the kind of errors made by L2 learners is dependent upon their level of proficiency.Beginning learners may rely much on their NL linguistic knowledge, while advanced learners have acquired a certain amount of TL knowledge, and this becomes a potential source of errors.
There have been efforts to investigate language errors which demonstrate the learner's NL influence in English writing.Two other terms used include language transfer and language interference.Choroleeva (2009) who studied Bulgarian students found that their NL influenced their English writing at the level of phonology, orthography, vocabulary and grammar.Watcharapunyawong and Usaha's (2012) study classified NL interference errors into 16 categories.The frequency of errors varied due to the different text types assigned to the students.The most frequent errors were on the use of verb tense, word choice, and sentence structure.Solano et al. (2014) studied the influence of Spanish on the Ecuadorian students' English writing.The results indicate that English grammar and vocabulary were the linguistic areas that suffered the highest level of NL interference, covering verb tenses, pronouns, prepositions, articles, and word order.Pudin et al. (2015) studied NL interference of Malaysian students' English writing and found that the most common errors were on pluralization and sentence structure.Finally, Owu-Ewie & Lomotey (2015) who studied NL (Akan) influence on junior high school students' English writing found that the types of errors include transliteration, word choice, spelling, and pronoun and the most frequent errors were on transliteration and omission.
Relevant to TL influence, some of the following studies on EFL writings are worth reviewing.Falhasiri et al. (2011) shed light on the most occurring interlingual and intralingual grammatical and lexical errors in low intermediate students' English compositions.It was found that 71% of the learner errors were categorized as interlingual errors and the rest were intralingual.Likewise, Qaid (2011) who studied intralingual errors made by Yemeni's students found six types of errors, consisting of omission as the dominant error, addition, fragment, word choice, and simplification.Along the same lines, Kaweera's (2013) study found that the interlingual interference includes lexical, syntactic, and discourse interference and seven intralingual interference includes false analogy, misanalysis, incomplete rule application, exploiting redundancy, overlooking co-occurrence restrictions, hypercorrection, and overgeneralization.Finally, Na-Pukhet and Normah's (2015) study found that the most frequent types of errors were word choice, verb tense, and preposition.The sources of errors were derived from interlingual and intralingual, and the interlingual interference was the dominant one.
Most of the previous studies reviewed have been devoted to NL influence (NL interference or interlingual errors) and TL influence (intralingual errors) of EFL learners with French, Spanish, Thai, Malay, Ghanan, Yemeni, and Persian language background.None of these studies used EFL learners with Indonesian background as research subjects.This is one of the reasons why the current study was worth conducting.
These works were error analysis-based research.So far, there has not been much systematic research carried out on these issues using both error analysis and IL analysis perspectives as research framework.This is another reason why the current study was carried out.As previously discussed, one of the IL features is permeability; its system is influenced by both the learners' NL and TL being learned.The current study would like to address both NL and TL influence using the perspectives of error analysis and IL analysis.Furthermore, very few studies on Indonesian EFL learners were carried out; hence, it would be crucial to share research findings within this area.It is expected that this study gives contribution to SLA research on Indonesian EFL learners and encourages further research in this area.
The main aim of the present study is to identify, describe, and explain the nature of NL and TL influence on the IL production of Indonesian EFL Learners.NL and TL influence on the learners' IL development is an agreeable issue.On the one hand, despite the rebuttals of the contrastive analysis hypothesis stating that NL influence is the prime cause of errors in L2 acquisition, researchers find the fact that NL plays a significant role in IL development.This continuing interest can be attributed to the fact that the NL influence on the learners' L2 production is quite apparent.On the other hand, TL influence is due to the complexity of the TL being learned.The nature of the influence cannot be traced to the learners' NL.It is "the confusion a language learner experiences when confronting patterns within the structure of a newly acquired language, irrespective of how the TL patterns might contrast with the learner's mother tongue" (Scovel, 2001, p. 51).Thus, both NL and TL influences areinevitable processes in IL development.Hence, the problem raised is how the nature of the influence of NL (Indonesian) and the TL (English) on the learners' IL production is.
The specific objectives of the current study are as follows: (1) to describe the types of NL (Indonesian) influence on the students' IL production; (2) to describe the types of TL (English) influence on the students' IL production; (3) to portray the degree of NL and TL influence on the students' IL production; and (4) to clarify the causes of the influences.

METHOD
This study is qualitative research of SLA in a classroom context or instructed SLA (Ellis, 2006).The framework used in this study is error analysis and IL analysis.This was carried out on students' English composition to identify, collect, describe, and explain the data.
The research subjects were 45 eleventh grade students of Muhammadiyah Senior High School of Surakarta Indonesia.The students speak Indonesian as their NL and have learned English as a foreign language for at least five years through formal schooling.Their average age was 17 to 18 years old.They were homogeneous in terms of nationality, language background, level of education, level of English proficiency, and age.
Data were elicited through free compositions with topics: about myself, my daily activities, and my last experience.The basic reason for the use of free composition as data elicitation was that it is one of the best forms of closely observed language production.The learners could use freely any patterns or rules they had learned or acquired for their own communication purposes.Thus, their free written composition was a free faithful record of their L2 production.The collection of free composition was a task directly performed by the subjects in the classroom.The students were supervised by the researcher when writing their free composition and submitted them in class.Each was around 300-350 words.
Data on learners' IL production were collected and identified on the basis of error analysis framework by James (2014).Only infelicitous sentences were taken into account and classified as instances of the students' IL production.Sentences were considered deviant when they were (1) appropriate but unacceptable, (2) acceptable but inappropriate, and (3) inappropriate as well as unacceptable.In other words, sentences were considered felicitous when they are appropriate and acceptable.
The collected data were also classified on the basis of James's (2014) error analysis framework, particularly linguistic categories and comparative taxonomy.Based on linguistic category, data were classified according to the deviation in terms of syntax and morphology.On the basis of the comparative taxonomy, the data were compared with the structures of the NL and TL.The comparison could yield the structures which could be traced to the NL and those to the TL.From this taxonomy, the data on NL and TL influence on the students' IL production could be collected.Finally, causes of the influences were explainable through different perspectives as discussed by Selinker (1997), Scovel (2001), Ellis and Barkhuizen (2005), James (2014), Kroll et al. (2015), etc.

FINDINGS Types of NL Influence on the Students' IL production
There are two major types of influence: lexical and grammatical.
The lexical influence was characterized by the use of Indonesian lexical items, Indonesian cognate, and Indonesian acronym.Lexical influence often provides for the borrowings and converts them to sound and spell more natural.The grammatical influence was characterized by the use of Indonesian collocation, passive construction, negative construction, and conjunction.

Lexical Influence from NL on the Students' IL Production
The first type of lexical influence is the use of Indonesian words in the students' IL production.They used Indonesian such as in "Our activities on Sundays are singing, dancing, and doing karawitan" and "Name of the dalang is Ki Narto Sapdo".The words "karawitan" and "dalang" are Indonesian words, referring to Javanese music orchestra and the puppet player in Javanese puppet show.These words can be categorized as cultural bound words or specific expression.Other Indonesian words were related to Indonesian food as in "Before going to school I usually breakfast with lontong sayur" and "Then I breakfast with rice uduk"."Lontong sayur" represents Indonesian dish made of rice in the form of a cylinder wrapped in banana leaf and served with vegetables, whereas "nasi uduk" is Indonesian steam rice cooked with coconut milk and other various herbs and spices.It is truly understandable that these kinds of term cannot be easily rendered into English as the TL being learned.The students had problems to cope with the true dilemmas as these words may be strongly rooted in their culture and difficult to translate into English.
The second type is the use of Indonesian cognate, "a word in one language which is similar in form and meaning to a word in another language because both languages are related" (Richards & Schmidt, 2002, p. 829), such as restoran (restaurant), pinalti (penalty), tiket (ticket), gitar (guitar), voli (volley), and sofenir (souvenir).They believed that these cognates were totally identical both, in Indonesian and English, in terms of form and meaning.For examples, they wrote "My mother bought some sofenir for the wedding party; "I and my friend bought tiket in the loket for the football match"; "I also played football and played voly"; I can play gitar and piano very well"; and "We went to a restoran near the school for lunch".In fact, these cognates share merely in meaning and not in spelling.The learners had taken some advantage of the similarity in order to cope with the problem in expressing their ideas in English.This was an obvious bridge to their English expression.
The third type is the use of Indonesian naturalized Arabic words, such as sholat (prayer), shubuh (morning prayer), dhuhur (noon prayer), wudhu (ablution) Romadhon (the fasting month), and Idul fitri (celebration of breaking the fast that marks the end of the fasting month).These words are naturalized borrowings from Arabic which represent culture-specific terms related to Islamic rituals.They wrote "In the morning I usually pray shubuh at 04.30 a.m."; "We pray dhuhur in school together with my or teachers"; "During Idul Fitri my parents and I visit my grand mother at Kampung; "Before sholat we take wudhu with clen water".The reason behind the use of such terms lies in the constant contact between the EFL students with their culture.As most students under this study were Muslims, the use of these terms was inevitable.In fact, for these students, the culture-bound words were one of the most difficult topics to express in English.
The fourth type is the use of Indonesian acronym, such as SMA (senior high school), SMP (junior high school), IPA (natural science), IPS (social science), SDN (state elementary school), and PPKN (civics education).These words belong to education-related terms which are intrinsically and specifically bound to Indonesian culture.The students were in the situations where they did not understand their English equivalence, hence, wrote "My sister is a student of SMP I now"; "I study at SMA II because it is command of my father"; "My activities daily are study and get lessons in IPA II", and "My brother is 8 years and he is school in SDN Kleco".They used these acronyms to fill in the existing gaps in their knowledge of English vocabulary.

Grammatical Influence from NL on the Students' IL Production
Grammatical influence from NL on the students' IL production was quite noticeable due to the inadequate linguistic knowledge of the TL.They employed the linguistic system they knew best from NL, covering the use of Indonesian collocation, passive construction, negative construction, and conjunction.
Some Indonesian collocations-two or more words that go together and sound right to the native speaker Indonesian but not to the native speaker of English such as the weather felt cold-were found in the data.This sentence is a translation from Indonesian cuaca (weather) terasa (felt) dingin (cold) and such collocation just sounds right to the native speaker of Indonesian.In English we can say I felt cold or the weather was cold.The students' tendency to have word for word translation when expressing themselves in the TL resulted in the creation of IL production such as, "My brother was sick and his body was hot"; "The Garuda Sport Store is completed store in my town"; "We went to Bandung climb on a car".The clause his body was hot is a translation from Indonesian Badannya (his body) panas (hot), which means "He was suffering from fever".In English, hot in this sentence may mean sexy.The phrase a completed store is a literal translation from Indonesian toko (store) yang komplit (provides various goods).They used the word completed as a translation of Indonesian komplit; a naturalized borrowed from Dutch kompleet.In Indonesian, tokonya komplit means it provides a variety of goods, whereas in English the store is completed may mean the store is over, and it is an unnatural collocation.The phrase climb on a car is a plain translation of Indonesia naik (climb) mobil (car) (English: to go by car).They used the word climb which does not collocate with word car.In English, we can say to climb a wall, to climb a tree but not to climb a car.All the examples above illustrate how Indonesian collocation has resulted in IL production.
Other grammatical influence was in the form of passive construction.The students produced "Hide and seek game can be played by hiding"; "Tuti is very sad.Her cat is lost"; "The game can be done indoors and outdoors"; "I was angry by my teacher because I didn't do home work".These illustrate that they regularly use passive construction in their TL expression, due to the fact that in Indonesian culture people tend to talk about things and people around them rather than about themselves.Therefore, they put forward the object (thing or person) being affected by an action and put the agent later.This was also due to the modesty virtue that Indonesian people tend to get rid of making blunt negative statement by hiding the subject or doer and prefer expressing it in passive construction.This tendency was reflected in the students' IL production.This style certainly does not belong to English version.The above examples will sound natural in active voice such as: "We play hide and seek game by hiding"; "Tuti is very sad.She lost her cat"; and "My teacher was angry with me because I didn't do my homework".
The next type was negative construction.The students assumed that English negative construction was similar to that of Indonesian, by adding negation word tidak which means no for formal usage or enggak for informal one.Hence, they wrote "My mother not work.She is a house wife"; "I feel sad because not can watch JKT 48 performance"; "My father not work again, he is already pension"; "In the morning I usually not have breakfast".In the examples, My mother not work was a literal translation from Indonesian Ibu (mother) saya (I, my) tidak (not) bekerja (work); I not breakfast from saya (I) tidak (not) makan pagi (have breakfast) and I not can watch from saya (I) tidak (not) dapat (can) nonton (watch).
The last grammatical influence was the use of conjunction, due to the fact that the students did not generally learn to use these cohesive elements as they did with other aspects of language, such as tenses.Their current understanding of English conjunction basically came from that of Indonesian.For example, they wrote "My name Ita Saraswati but people call me Ita"; "The children were playing and some others only watched them".Example one represents the common way Indonesian people introduce themselves, "Nama (name) saya (I, my) Ita Saraswati tetapi (but) orang (people) memanggilku (call me) Ira", mentioning the complete name first then the nickname, combining the two clauses with the conjunction but.Example two represents an acceptable familiar Indonesian compound sentence, using and as a connector as in Anak-anak (children) sedang bermain (were playing) dan (and) yang lain (others) menonton (watching).

Types of TL Influence on the Students' IL Production
There were two types of influence: lexical and grammatical influence.The lexical influence was characterized by the use English false friends (lexical similarity in meaning and in form), whereas the grammatical influence was characterized by the use verb tenses, pronoun, and preposition.

Lexical Influence from TL on the Students' IL Production
The lexical influence from TL was due to the students' confusion of English words which are similar in meaning.For examples, the words to follow and to join have one Indonesian equivalence mengikuti in "I usually follow English course in the afternoon".The words wide and large can be translated into Indonesian besar or luas and they wrote "My grandmother's garden is not so wide but clean".The words wage and salary were presumed the same as they wrote "This year your wage is Rp 500.000per month".Such words commonly appeared in the students' IL production.
Another lexical influence was due to the students' confusion with words which are similar in form.For examples, the words beside and besides orthographically look very similar and presumed identical as in "Besides that, I could write to some pen friends in English".Likewise, the words stationary and stationery are false friends as in "We can find books, pens, and erasers in the stationary".
In addition, they appeared to have problem with English part of speech, a category of word based on its syntactic function.There was a confusion in the use of verb and noun as in "Everybody communication in English and in Indonesian"; "At evening I dinner with my family"; "At noon I pray dhuhur and then lunch"; "My sister Sahara always attention to me"; "In the morning I breakfast with fried rice and egg".These nouns (communication, attention, breakfast, lunch, dinner) were commonly acquired first by the learners; therefore, they were more familiar and ready to use than the verb counterparts.

Grammatical Influence from TL on the Students' IL Production
The grammatical influence from the TL was due to the complexity of the TL (English) itself.As the learners had limited knowledge of English, they created approximative grammar, involving the use of verb tenses, pronoun, and preposition.
The first type was the use of English verb tense.In English, verb denotes an action, and the action has the time relationship with the doer of the action.To the students, learning verb tense was the most difficult area of grammar to acquire.This appeared to be a pervasive problem that results in the creation of IL since Indonesian does not have the concept of verb tense.The first type was the use of BE, as BE concept does not exist in Indonesian system.Addition of BE in the present and past tense appeared in the data such as "My parents is very love me, and me too"; "I love my parents because they are understand me"; "I don't know my mother; She was died"; "My father was died and my mom is single parent now".The omission of copula BE in nominal sentence also occurred in the data, as in "I really proud become citizen of Indonesian"; "I don't like Mathematics because it so difficult for me".The learners neglected the occurrence of copula BE in a sentence since it has no semantic contribution in the sentence.The second type was the addition of TO to present verb as they believed that TO and the verb were one word.For examples, "Sometimes, my father to help my mother when in market is crowded"; "I always to study hard as s long as my parent to hard work for my study".Third, the use of present tense in the place of past tense commonly occurred in the students' IL production.For examples, "I have an unforgettable experience last year"; "I have unforgettable experience when I date someone special in SMP"; "He break my heart when I know he love other girl".Fourth, using two verbs together to denote two activities committed at once was a real problem for the learners.They did not have any knowledge of such patterns, especially the form of the second verb.They just took it for granted as shown in the data, "I like sing the Manca songs"; "I want look for a job as a teacher; "I enjoy read books".They had no idea that the word like must be followed by gerund, want by infinitive with to, and enjoy by gerund.Next, irregular verb commonly occurred in their' IL production since to their current knowledge past tense form is made by adding -ed to all verbs, such as "I was choosed to deputy my school in Raimuna Nasional selection"; "20 people were sended to deputy my city to Raimuna Nasional selection"; "I was gived the chance to join Raimuna Nasional".Finally, the use of Verb-ing instead of Verb infinitive in the future tense also characterized the learners' IL production as shown in the data, "When finish SMA I will studing at university"; "After graduate I will working"; "After graduate my sister will searching a job".
Similarly, learning English personal pronoun was another complicated area for the learners due to the gender and number differences which do not exist in Indonesian language.They generally used subjective pronoun properly but not the other types of pronouns since it was the most familiar type they had in their mind currently.Subjective pronoun was used in different syntactic functions, i.e. as a possessive pronoun as in "I have one sister she name is Annisa" and as objective pronoun as in "I have two sisters, I love they so much".
Next, preposition appeared to be one of the most problematic categories the learners encountered in learning English due to the differences between the prepositional systems of English and Indonesian.English has much more number of prepositions denoting various kinds of relationships compared to Indonesian.For example, Indonesian preposition di to denote place and time may have several equivalents in English i.e. in, at, on.Overwhelmed by the various prepositions to choose from, they used one they were most familiar with as in "In night I studying and then I go to bed"; "My mother always in home with me and my brother"; "In 5 A.M. sometimes I listen to the Radio Prambos".
The students' IL production was also colored by the omission of {-s} in plurality, as in "I have one brother and two sister"; "I have two brother and three sister"; "My father and mother are hard worker".The students omitted {-s} due to the lack of training or incomprehensibility of the rules since it had no significant semantic contribution.Plurality also does not exist in the students' NL.Finally, article system appeared to be problematic for the students as well and was one of the sources of difficulty in EFL learning because the students' NL does not have articles or article-like morphemes.For example, "I love my mom because she is good parent"; "my mother is house wife".The absence of indefinite article "a" represents their incomplete understanding of the rule.

Degree of Influence from NL and TL on the Students' IL production
The results indicate that NL influence had a little bit lower percentage than that of the TL influence, that is, 122 cases or 46% and 142 cases or 54%.The NL influence consisted of 70 cases or 26% lexical influence and 52 cases or 20% grammatical influence, whereas the TL influence consisted of 53 cases or 20% lexical influence and 89 cases or 34% grammatical influence.To sum up, the data show that the students relied less on the NL rather than the TL as language resources when they wrote in English as the TL being learned.However, they relied more of their NT as vocabulary resources especially when they had problem to deal with cultural expression and to the TL as grammatical resources to deal with the complexity of TL grammar (see Table 1).

Factors which Contributes to NL and TL Influence on the Students' IL Production
One of the factors which contribute to the NL influence was the students' good mastery of Indonesian since this system provided the high chance of usage, especially when they found problems in expressing their ideas in English.Indonesian system became linguistic resources when writing in English.In fact, they used it as vocabulary resources, covering Indonesian words (kupat sayur), Indonesian cognate words (pinalti, restoran), Indonesian naturalized Arabic words (sholat, dhuhur), and Indonesian acronym (SD, SMP).Likewise, they used NL as grammar resources, involving Indonesian collocation, passive construction, negative construction, and Englishized conjunction from Indonesian.These resulted in IL production.The high mastery in Indonesian as their NL appeared to be decisive for NL influence in the students' IL production, both at lexical and as grammatical level.
Another factor that contributes to NL influence was the students' limited knowledge of the English as the TL.In order to be able to express their ideas in English, the students were required to have a good mastery of the language.In fact, their English vocabulary was quite insufficient and this forced them to slide back to their NL.They used Indonesian vocabulary in their English writing.In the same way, a lack of substantial knowledge of grammatical structures of the TL led them to grammatical confusion, resulting in the falling back to their NL source for grammar sources.These created an IL production, neither Indonesian nor English, which represents the students' current knowledge of English.The data also indicate that learning strategy was the major crucial factor which contributes to the occurrence of the TL influence.The students commonly used different learning strategies to deal with the complexity of the English as the TL being learned.They used their insufficiently acquired linguistic knowledge of the TL as the resources for vocabulary and grammar.
One of the learning strategies was false analogy.The students did not fully understand the distinction of verb forms in the TL; for example, regular and irregular past tenses.Their current knowledge of forming past tense was by adding -ED to the base verb form and applying it to any verb, including the irregular ones.They wrote "I loved her in SMP.I holded she hand and she holded my hand"; "When my old is ten I falled in the ditch in front of my house".The examples indicate that did not yet have the knowledge that the verb hold and fall were irregular verbs that must be treated differently.
Another strategy was the incomplete rule application.The learners, for example, used an incomplete rule of present tense in which the verb should be in the present tense form without BE.The learners currently assumed that the subject + BE (they are) were one word and they used them in the present tense as in "My parents is very love me, and me too"; "I love my parents because they are understand me".
The next strategy was the simplification of the TL system.It was noticeable that they omitted copula BE in the sentence, as in "I like Math because that subject really funny"; "I don't like physics because that subject really difficult".The omission of the copula BE in these cases may be due to the learners' inability to grasp minor details in the TL structure because of their limited knowledge of the TL.Hence, such copula BE also had no significant semantic contribution to the sentence and left unnoticed by the learners.This resulted in the creation of an IL.
The next strategy used by the learners was the overlooking co-occurrence restrictions.The learners appeared to observe the restrictions of TL existing structure, as in "I enjoy study English because my teacher is fun".Based on James (2014), such an example reflects the overlooking co-occurrence restrictions because the word "enjoy" co-occurs a gerundial complement.
Another strategy was overgeneralization, that is, the learners had learned a certain rule or pattern in the TL; they then, assumed that the rule or pattern could be operated without exception (Scovel, 2001).In other words, the learners overgeneralized this rule or pattern to other constructions inappropriately.For examples, they used the personal pronoun she at any syntactic function, as in "I have one sister she name is Annisa"; "I have one sister, she name is Nela"; and objective as in "I have two sisters, I love they so much".
Finally, the main factor that contributes to the NL and TL influence was because the students had two language systems (Indonesian and English) in their mind.They tended to activate the two language systems.It was obvious that they used both systems regardless of their intention to use one language only, that is, English as the TL being learned.They used both Indonesian and English when writing in English.In a particular situation, as they had problems to find appropriate English words and grammar, they grabbed the Indonesian resources for the compensation.In another situation, they took the resources from their incomplete knowledge of English.These resulted in the production of an IL.It was appealing to find out that the students' IL was open for influence from both NL and TL.

DISCUSSION
The result of this study provides empirical evidence of the NL and TL influence on the students' IL production.This supports the IL theory discussed by Selinker (1997), Ellis & Barkhuizen (2005), Gass and Selinker (2008) and Saville-Troike (2012) that learner language is open to the infiltration from both learners' NL and TL systems.NL influence found in the current study includes both lexical and grammatical aspects, which are in line with the findings of Choroleeva (2009), Watcharapunyawong and Usaha (2013), Solano et al. (2014), Pudin et al. (2015), andOwu-Ewie &Lomotey (2016).They found that students' NL influenced their English writing.TL influence found in this study also includes both lexical and grammatical aspects, and this corresponds to the findings of Falhasiri et al. (2011), Qaid (2011), Kaweera (2013), andNa-Phuket (2015).They found various instances of intralingual error, ill-forms of the TL use which were not traceable to their NL.However, there were differences among the findings regarding the number of types or forms of linguistic components, the frequency, and the dominant types.These I believe are due to the different data, the number of instances or cases, and topics assigned to the participants.However, most of the studies reviewed shared common findings of NL influence, particularly, on the use of NL borrowings, word order, and collocation and of TL influence on the use of verb tenses, preposition, article, pluralization, and negative construction.One thing is apparent from the studies that the learners' IL production is colored by the influence of the NL and TL regardless of their NL background.
A look at types of NL influence in the students' IL production shows that most NL influence occurred on vocabulary, particularly the use of NL words, while most TL influence on grammar, especially on the use of verb tenses.This means that vocabulary and grammar were the most troublesome that Indonesian EFL learners faced when learning English.This is one of the main sources for the IL production in this study.A similar problem was found in the studies of Kaweera (2013) and Na-Phuket (2015) in which the students' English writings were influenced by their NL, resulting in interlingual errors, whereas the learning strategies used to deal with the complexity of English resulted in intralingual errors.This is convincing that the students' reliance on the acquired language knowledge of NL and some of the TL provides us with information of their linguistic development and learning process.In sum, it can also be inferred that when there is a gap in the student's knowledge in the TL, they tend to borrow their NL words in order to fill in the existing gaps of English vocabulary (Choroleeva, 2009), and at the same time they apply grammar rules they already acquired from the TL by overgeneralizing or simplifying the rules (Kaweera, 2013).
This study found that the NL influence amounted to 46%, consisting of 26% lexical and 20% grammatical.Meanwhile, the TL influence was 54%, covering 20% lexical and 34% grammatical.This means that the students who were intermediate EFL learners found recourse to both NL and TL more or less equally when they found problems to express their ideas in English.However, when dealing with vocabulary and grammar they did differently.It is apparent that they relied more on their NL rather than TL when having problems with vocabulary and on their TL rather than NL when facing difficulties in grammar.
The students' reliance on NL in their IL production usually occurred when they dealt with culture-bound expressions and cognates.In fact, the students had problems to find the English equivalence for such culture-bound expressions.This is in line with Baker (1992), stating that the main problems to deal with such expressions lie on the ability to interpret them correctly and to render the various aspects of meaning they convey into the TL.An expression may have a similar counterpart in the TL, but its context of use may be different.As a result, such words appeared in the students' IL production.In addition, the students identified the cognates (gitar, pinalti) as identical to those in English (guitar, penalty), but the similarity of form is actually superficial.These pairs of words are known as false friends false cognates.The students wrongly assumed them as identical since they originate from a common root (Richards & Schmidt, 2002).
The students' reliance on TL in their IL production commonly occurred when they dealt with English verb tenses.This finding may explain that the use of English verbs was a major learning difficulty for the students.They found problem especially on when and how to use the tense and the form of the verb as well as the form of irregular verbs.This might happen because Indonesian has no tenses.This is in line with most of the studies reviewed that the researchers found tenses as one of the error types in EFL students' writing.It is an acceptable fact that tenses in English are considered one of the most difficult grammatical aspects faced by most EFL learners, including Indonesian students, as Bardovi-Harlig (2000) maintained that tense and aspect are apparently two major issues in language teaching.The present study found that the sources of influence includes the students' good mastery of Indonesian as their NL, the limited knowledge of the English as the TL, the learning strategies used, and the possession of two language systems in one mind.This corresponds with Selinker (1997) and Saville-Troike (2012) that IL is easily influenced by both learners' NL and TL systems.As the students' knowledge of the TL was very limited, they took recourse to the existing linguistic knowledge, both from NL and TL, resulting in IL production.Using the analogy of bilingualism, there is evidence that having two language systems in one brain means that both languages are active regardless of a bilingual intension to use one language only (Kroll et al., 2015).In fact, the students' IL production represents some language mixing in the spontaneous utterances due to the intention to express complex ideas but lacking the necessary vocabulary and grammar in the TL.It is apparent that the students are not aware of the differences in the language systems.

CONCLUSION
There are several conclusions which can be drawn from the current study.Firstly, this study reconfirmed that both NL and TL influenced the students' IL production.The influence was generally due to the students' ownership of two language systems (Indonesian and English) in one mind.As they intended to express ideas in English but lacking the necessary vocabulary and grammar in the TL, they found help from both NL and TL system.The NL influence appeared to be slightly lower than that of TL because the students were at the intermediate level of English proficiency; this is in consonant with O' Grady's et al. (2002) view.Secondly, the major influence from their NL was dealing with vocabulary in the form of Indonesian borrowings, including cultural bound expression, cognates, and acronym.The sources of NL influence were the students' good mastery of NL and their limited knowledge of the English as the TL.Thirdly, the major influence from the TL was related to grammar, particular dealing with verb tenses.The TL influence stemmed from the learning strategy (i.e., false analogy, incomplete rule application, simplification, overlooking co-occurrence restrictions, and overgeneralization) used by the students in coping with the difficulties in the TL.Lastly, in fact, the students' IL production represents some language mixing in the spontaneous utterances due to the intention to express complex ideas but lacking the necessary vocabulary and grammar in the TL.These conclusions were attributed merely for the students under the study and not intended for generalization.

Table 1 .
The degree of NL and TL on the students' IL production