THE EFFECTS OF ERROR TREATMENT ON INTERLANGUAGE OF INDONESIAN LEARNERS LEARNING ENGLISH AS A FOREIGN LANGUAGE

Interlanguage errors always exist in foreign language learning. They become the source for studying the system of the learners’ interlanguage (IL). Han (2004) reviews hundreds of IL studies and concludes that there are two competing views can be identified. One view suggests that error treatment (ET) has unconvinced value in classroom second language acquisition or SLA (Krashen 1982). Adults do not require constant correction in useful ways; thus, the role of the teacher is to provide comprehensible inputs for the learners to move to the next stage of IL. ET has little value in SLA because IL is fossilized (Mukkatash 1987; Thep-Ackrapong 1990). This view corresponds with Patkowsky (1980), and Johnson and Newport (1989) who believe fossilization is due to CP. The opposite view comes from White (1991), Spada and Lightbown (1993), and Muranoi (2000) who believe that that ET is very important in foreign language (FL) learning. Learners can take a lot of benefits from ET as they can develop their IL system to a higher level of accuracy. This view corresponds with Scovel (1988), White and Genesee (1996), and Bialystok (1997) who deny the existence of CP in SLA; CP may applicable for the acquisition of phonology but not for syntax. Thus, grammar is learnable at any age. This study investigates the effects of a short-term error treatment (ET) on IL errors, with specific attention to the learners’ ungrammatical items. The problem states “what are the effects of a ET on the learners’ ungrammatical items?” Are their ungrammatical items fossilized (in a sense that they are static in nature) or dynamic after the learners have been exposed to the ET? The data were the learners’ free compositions collected four times: prior and after the ET and two months afterwards. They were analyzed quantitatively and qualitatively. The result indicates that the ET changed the state of the learners’ ungrammatical items. They became so dynamic. At a certain period, some appeared; then due to the ET, some were destabilized, some were fluctuating, and others were still stabilized. New errors appeared as they started learning to use new grammatical items. The conclusion drawn from this study is that ET can change the state of the learners’ IL errors; ET contributes to the destabilization process. Errors may persist momentarily but they can be destabilized. The ET still works on the learners who are at their post puberty. Thus, there is a great possibility for the learners to acquire complete TL grammar since their ungrammatical items are dynamic.


Introduction
All learners make errors in learning a new language.Their TL always contains errors.In general, such errors are considered as "an inevitable sign of human fallibility" (Corder 1981: 65).Errors are inevitable in any learning situation which requires creativity such as in learning a FL.They are no longer viewed as mere deviations but rather as a source for studying the processes/strategies used by the learner in learning the TL.They are "evidence about the nature of the process and of the rules used by the learner at a certain stage in the course" (Corder 1977: 167).Therefore, if we want to study the learners' IL system, we should find clues to the systems by analyzing the errors they make.
This study deals with the effects of an ET on the learners' IL errors and the related error fossilization issue.Han (2004) reviews hundreds of studies of fossilization that have emerged over the past three decades and comes to a conclusion that there are two competing views can be identified.One view suggests that ET has unconvinced value for classroom SLA.Krashen (1982) believes that there are possible parallels between children's acquisition of their first language and adult's SLA and this led him to suggest that ET has dubious value in the classroom.Adults do not get much benefit from error correction; thus, the role of the teacher is to provide comprehensible inputs which learners can work on in order to move to the next stage of IL.Mukkatash (1987) and Thep-Ackrapong (1990) also believe that there is not much value in explicit and systematic ET in the case of adult FL learning since their IL errors are fossilized.This view corresponds with Patkowsky (1980), Johnson andNewport (1989), andLong (1990) who believe that a CP indeed exists for SLA and consequently FL learners cannot attain TL grammar since their IL errors are fossilized.
The opposite view comes from White (1991), Spada andLightbown (1993), andMuranoi (2000) who believe that ET is very important in FL learning.It gives positive effects on FL learning; learners can take a lot of benefits from the ET provided by the teachers; they can develop their IL system to a higher level of accuracy.This view corresponds with Scovel (1988), White and Genesee (1996), Bialystok (1997), Steinberg et al. (2001), andBirdsong (2004) who deny the existence of CP in SLA.They claim that CP may applicable for the acquisition of phonology but not for syntax.They believe that grammar is learnable at any age and consequently there is a possibility for FL learners to attain TL grammar.
The main research question says "what are the effects of an ET on the learners' ungrammatical items?Are they static or dynamic after the learners have been exposed to the ET?" To answer this question, five subsidiary research questions (Srq.) were raised, namely: (1) What kinds of ungrammatical item do the learners produce before the ET? (2) What are the effects of a ET on the learners' persistent and non-persistent ungrammatical items?(3) What is the nature or behavior of the learners' ungrammatical items after they have been exposed to the ET? (4) What cognitive factors contribute to the stabilization of the learners' ungrammatical items after they have been exposed to the ET? (5) What classroom aspects of the ET can contribute to the destabilization of the learners' ungrammatical items?
This study is very significant as it can give teachers and researchers clear pictures of the common phenomena usually occur in FL learning (i.e. the committing of errors, the L2 learning processes, and the phenomena of error stabilization and destabilization).The insights derived from this study can contribute to the development of the theory of applied linguistics, especially to the existing theorization of IL error and fossilization in SLA.In general, it can give insights into several aspects of adult FL learning (i.e. the processes and the constraints).

Research Method
The subjects of the present study were 30 Indonesian secondary school students grade three, (average age was 17) who learned English as a FL.They had been learning English for 7 years through formal instruction.This study used a hybrid method, a combination of a quantitative (error treatment) and a qualitative method.The short-term error treatment that was conducted to collect the needed data constitutes three stages, namely: pre ET, ET, and post ET.
The research was initiated by assigning the research subjects to write a free composition (C1) of about 150 to 200 words.To get a similar result, they were given pointers to write such as their study, parents, daily activities, past experience, and future ideas.An error analysis or EA (Corder 1982;James 1998) was carried out on their C1 to identify the grammatical errors shared mostly by all the students.The result of EA indicates that the learners produced a significant number (422 cases) of ungrammatical items which can be classified into 8 types: verb, to BE, bound morpheme {-s}, syntactic structure, noun, preposition, pronoun, and article.
The ET (as a method used to eliminate the learners' IL errors) then was conducted on these 8 grammatical items for one semester.This was intended to see its effects on their ungrammatical items; are they fossilized (in a sense that they are static) or dynamic?There were two main classroom activities carried out during the ET: error correction and explicit grammar instruction.Each session was dedicated for the discussion of one grammatical item.These two play a critical role in FL classroom, particularly in grammar acquisition since they create the conditions needed for grammar acquisition to occur.
At the post ET, the learners were as-signed to rewrite their C1 to produce composition two (C2).This was intended to investigate the effects of the ET as an attempt to eliminate their ungrammatical items.Two months after the ET, again they were assigned to rewrite their C1 to produce composition three (C3).In addition, they were also asked to write another free composition with different topic in order to produce composition four (C4).It was assumed that they would produce new error types (different from those they previously made) as they started learning to use new grammatical items.These four compositions (C1, C2, C3, and C4) constituted the primary data of this study.
Finally, a qualitative study (through observation, debriefing, and interview) was conduced to collect data or information needed to answer the Srq. 4 and 5, that is, the cognitive causal factors of error stabilization and the classroom aspects of the ET which contributed to error destabilization.This was carried out throughout the ET sessions.

Before the Error Treatment
The result of EA on C1 indicates that the learners produced a significant number (422 cases) of ungrammatical items, which were classified into: verb (119 cases), to BE (69 cases), bound morpheme {-s} (68 cases), sentence structure (65 cases), noun used as verb, (37 cases) preposition (36 cases), pronoun (16 cases), and article (12 cases).Each of the learners contributed different number of ungrammatical items.The highest number (29 cases) was made by student No. 30 and the lowest number (5 cases) was made by learner No. 26 and 27.Each learner produced 14 cases in average.The frequency of the learners' ungrammatical items before the ET is shown in the chart below.
Based on these, the writer concludes that the learners' English is considered as an IL.Their language system is neither that of English nor Indonesian; it contains the elements of both.Their IL is idiosyncratic in nature; it is distinct from both their NL and the TL.Their IL system proved to be systematic (Saville-Troike 2006: 41).The sentences they produced, though grammatically unacceptable, were not just a random collection of entities.They appeared to obey certain linguistic constrains (of their own).Thus, this study supports the theory that IL is systematic, as proven by researchers such as Dickerson's 1975;Beebe 1980;and Tarone 1988.The permeability, the susceptible to the infiltration by the NL and the TL rules or forms (Yip 1995: 12), of the learners' IL is also clearly noticeable in the learners' IL.On the one hand, the learners' IL structures were invaded by their NL as in All subjects I very like.But that I very like is biology.On the other hand, their IL followed the TL rules but they distorted them as in I telled her that I loved her.She holded my hand and I holded her hand.Their IL was idiosyncratic in nature; it was distinct from both their NL (Indonesia) and the target language (English), as shown in the figure below.
Both of these processes, permeation from the learners' NL known as NL transfer and infiltration from the TL known as overgeneralization, reflect the basic permeability of the learners' IL.The present research also supports the theory that IL is permeable or easily infiltrated by both the NL and the TL linguistic rules, as proven by researchers in the 1970s to 1980s such as Dulay 1974;LoCoco 1976;Grauberg 1977; and Wode 1986.

After the Error Treatment
This study basically tried to investigate the effects of an ET on the learners' grammatical errors and to determine whether they were fossilized (in a sense that they were static) or dynamic after the learners have been exposed to the ET.Result of the descriptive and statistical analysis of the tests is shown in table below.

The Output Sheet for t-test: Paired Two Sample Means
The table above shows that the t-value is much smaller than the t-Stat.That is to say, there is a significant difference between the two scores of C1 and C2.This demonstrates that the ET gives significant effects on the learners' Thus, the statistical analysis reveals that the ET gave significant effects on the learners' ungrammatical items.It changed their state (nature) and stimulated their dynamicity.They became so dynamic and not static.It effectively prevented some ungrammatical items (i.e. the future tense, deletion of BE as auxiliary, word order, negative construction, subject omission, preposition omission, and the conflation of the objective with the possessive pronoun) to reappear.In general, most of the learners produced fewer ungrammatical items; their IL system developed closer to the TL.It means that the ET gave the learners a usable feedback, providing them with both the positive input and the negative input which were useful for the learners.They found error correction combined with explicit grammar instruction of a great value.This outcome supports much of the previous literature on ET by White (1991), Spada and Lightbown (1993), and Muranoi (2000), claiming that L2 learners gained benefit from ET provided by the teacher.ET contributed to the development of the learners' IL system.This finding also supports Bley Vroman's (1990) fundamental difference hypothesis (FDH).
The quantitative analysis also reveals that some (142 cases or 33.64%) of the learners' ungrammatical items persisted regardless of the ET.With regards to this, to a certain degree, this finding corresponds with Mukkatash (1986) and Thep-Ackrapong (1990) who confirm that even with systematic ET, IL errors persist.The qualitative analysis on these persistent errors indicates that they were the results of cognitive mechanism such as NL transfer, overgeneralization, and simplification (Selinker, 1977(Selinker, , 1997)).These illustrate how the learners activated their interlangual unit with these cognitive processes in their attempt to produce the TL of which their knowledge was still quite limited.They relied on the linguistic knowledge they already acquired either from their NL or the TL.
The result of the interview (the learners were asked to comment on the errors) indicates that most of their comments showed their reliance to linguistic knowledge they already acquired (either the NL or the TL).In other words, the learners' IL as the product of cognitive process appeared to be much dependent on NL and TL rules.One thing is clear.
Having fewer resources at their disposal in the TL, they relied on the knowledge they already knew, either from their NL or the TL to help them cope with the problem.In one situation, they relied extensively on their NL; and in another situation, they relied on the TL grammar they already acquired but did it wrongly by over generalizing or simplifying the TL rules.They are common processes in FL learning or SLA.
The writer, however, believes that getting stuck at a certain stage in the learning course is common in FL learning and it is a temporary condition.This is caused by the learners' cognitive constraint (the determined latent psychological structure which was activated whenever they attempt to produce the TL).To prove this assumption, 2 months after the ET, the learners were asked to rewrite their C1 to produce C3.The result indicates that they produced fewer ungrammatical items (142 cases in C2; 94 cases in C3).The graph below shows the comparison of scores of C1,

Compositions
Freq uen cy C2, and C3, indicating the effects and the development of the ungrammatical items after the ET.
Thus, the writer concludes that there is a possibility to destabilize persistent errors if the learners gain further exposure and input of the TL.The learners' stabilized errors could be eliminated (destabilized) as shown in the chart below.
The qualitative analysis reveals that there were observable classroom activities of the ET which contributed to the error destabilization.
It was observed that some classroom events of the ET contributed to the error destabilization, since they provided the learners with the language acquisition opportunities.In the classroom, the learners got adequate input, feedback, frequent exposure, explicit grammar explanation, and they had the opportunity to practice the target language.These five classroom aspects could improve their TL linguistic knowledge and gave contribution to the error destabilization process.Such classroom ingredients of the ET could promote the learners' acquisition of grammatical items.
In addition to C3, the learners were also asked to write another composition (C4) with a new topic, because it was assumed that the learners would make new types of errors (different from the above mention) as they started learning to use new grammatical items.The assumption was true.The learners produced new ungrammatical items (i.e.adverb of manner, preposition -with, The-deletion in superlative adjective, that-Clause, the conflation of the past tense with the past continuous form, and pseudo passive) in their C4.
Further analysis indicates that as a result of the external pedagogical intervention, the persistent errors changed their state: some were still persistent, others became non-persistent (appeared only once within one composition); and the rest were eradicated.The non persistent errors were finally eradicated.New ungrammatical items appeared as they used new grammatical items.The pedagogical intervention could change their states as shown in the diagram below.

Conclusion
Several conclusions can be drawn from this study.Firstly, the ET (a combination of error correction and explicit grammar instruction) can change the state of the learners' ungrammatical items.It contributes to the destabilization process since it provides the learners with input, feedback, grammar explanation, and the opportunity to practice.All these classroom events of the ET are facilitative for the destabilization process to take place.ET is critical for the learners who mostly acquire English merely through classroom instruction.
The importance of ET in FL learning is derived from the fact that IL errors must and always exist in FL learning.They are inevitable part of learning process.We cannot avoid or prevent their existence since the making of error is human nature.The ET is proved to have contribution to the destabilization of the learners' ungrammatical items; the persistent errors are merely a temporary plateau and not a permanent condition.It is feasible that these errors finally can be destabilized at some point and under certain conditions (i.e. the learners still get further input and exposure to the TL).The error destabilization takes place when learners can incorporate new learning items into their developing language system or IL system.
Secondly, the ET stimulates the dynamicity of the learners' ungrammatical items.At a particular stage of FL learning, ungrammatical items appear.As a result of the ET, some of the ungrammatical items tend to stabilize; some tend to destabilize; and others fluctuate.The fluctuating ungrammatical items are likely to destabilize and the stabilized errors are likely to be destabilized.Other new IL errors are likely to appear when the learners start learning to use new rules.The learners' ungrammatical items remain dynamic as they continue learning the language.They keep evolving naturally as learning or ET provision continues.Han (2004) theorized that stabilized er-rors can be good candidates for fossilization.Nevertheless, this can only happen under the condition that learners stop learning or having inadequate input and exposure to the TL.Language exposure and input are very critical for interim grammar to develop.When learners stop learning, the destabilization process stops and the IL errors become fixed.On the contrary, when learners continue learning the language, the destabilization process keeps on going; IL errors change their nature and finally become part of TL system.Thus, due to the pedagogical intervention, the learners' ungrammatical items evolve naturally, developing towards complete TL grammar.Thirdly, the ET still works for the learners who are at their post puberty (post CP); in other words, grammatical items are learnable at their post puberty.The learners' capability of learning syntax does not decline at their post CP.It is not impossible to destabilize the learners' persistent ungrammatical items when requirements for language acquisition are fulfilled.This is an accord with the hypothesis which states that there is no CP for the acquisition of syntax of a foreign language.They may get stuck temporarily due to cognitive constraint and due to the learners' individual differences or the nature of the grammatical features themselves.Stabilization and destabilization commonly occur in SLA as long as the learners have not yet reached the TL system.Such a natural persistence to the new system (stabilization) can be overcome by further exposure to and hours of practice of the grammatical items involved.
Finally, the writer concludes that the learners' ungrammatical items are dynamic.At a particular point of learning course, the learners' ungrammatical items may get stabilized temporarily; but they are not fossilized.The learners' persistent errors are just a temporary and not a permanent condition.There is a possibility to destabilize the learners' persistent errors at some point provided that the learners are still learning the language.Their ungrammatical items may be temporarily stabilized since stabilization is a natural learning process.Thus, following Selinker and Lakshamanan's (1992) distinction between the terms fossilization and stabilization, the writer is of the view that the term stabilization rather than fossilization is more appropriate to describe the learning condition of FL learners who cease to develop their IL system in a particular stage of their learning course.
There are several limitations of the present study.One particular limitation of this study is that there is no separation between error correction and explicit grammar explanation.The ET conducted in this research is a combination of error correction and explicit grammar instruction.It is concluded that the combination of the two was beneficial for the learners SLA.In other words, the effect of error correction is not investigated separately from the effect of explicit grammar explanation.It is not clear whether the error correction or explicit grammar instruction or both lead to the results obtained.Future researchers, therefore, are suggested to investigate the two variables separately to make clear how each variable is beneficial for classroom SLA.
The second limitation is that this study deals with IL errors within a group of learners (macro-analysis), that is to say, it does not take into account the individual differences of the learners.It is a kind of macro-analysis which reflects condition of learners in general.Fu-ture researchers are recommended to conduct a detailed micro-analysis by considering individual differences of the learners.Microscopic analysis of individual learners will provide very unique insights into the complexity and multi factors which involve in classroom SLA.
Finally, a limitation is placed on the generalizability of the results achieved in this study.This study used a relatively small number of grammatical features (8 error types) as well as a small number of learners (30 students); therefore, it is too early to claim that the results can be generalized to all grammatical features and to all L2 learners.The results can be generalized only to the same grammatical features and to FL learners with more or less the same characteristics with the subjects.Any how, this is a case study and the findings of a case study cannot necessarily be generalized to other learners.Rather, it is useful to compare the findings with other case studies in order to search for useful general principles.
No research is without limitation.Future research, therefore, should consider the above research limitations in order to gain more satisfactory results.There are still other topics, on similar area which seem quite significant to investigate.Such studies will be quite useful for the improvement for English teaching and learning, especially in Indonesian context and at Junior and senior High School levels.The knowledge and insights derived from such studies will certainly help improve the quality of the teachers, researchers, and textbooks writers.

The
Effects of STET on the Errors and the Development of Errors after Adjemian, C. 1976."On the Nature of Interlanguage System."Language Learning.26: 297-320.Beebe, Leslie M. (Ed.) 1990.Issues in Second Language Acquisition: Multiple Perspectives.London: Newbury House.