A Review Study of Interlanguage Theory

The influence of first language (L1 hereinafter) in the learning of second language (L2 hereinafter) has witnessed an intense debate during the past years, resulting in the prevalence of Error Analysis (EA hereinafter) over Contrastive Analysis (CA hereinafter). A great number of empirical studies indicated that neither L1 nor L2 was always responsible for learners' errors (Bailey et al. 1974; Krashen et al. 1978, 2003; Larsen-Freeman, 2002; 2003). CA and EA paved the way for Interlanguage theory (IL hereinafter) in describing L2 learners' errors in the acquisition process of L2. IL, which has been in vogue for the last years, has witnessed huge criticism by different researchers and linguists from various L1 backgrounds. They all came to state that there are many points in this theory which are not clear. In light of this, this study aims at reviewing and discussing the role of IL in describing and explaining learners' errors in the process of acquisition of L2. Therefore, theoretical foundations, theoretical assumptions, limitations and significance of this theory are discussed in detail. This review reveals that IL theory is almost no longer valid in the field of second language acquisition (SLA hereinafter) for numerous reasons.


Introduction
Both CA and EA were criticized as being insufficient for describing L2 errors.From the one hand, CA was questioned by many scholars working in applied linguistics.The main criticism was that interlingual interference from L1 is not the only reason for the occurrence of errors in SLA.Another claim was that CA is most predicative at the phonological level and least predictive at the syntactic level because no language has been adequately compared yet to another language.From the other hand, EA was very popular in the 1960s and 1970s.It occupied the mainstream in the field of SLA research for its great contributions to this field.Despite its popularity, it was not without any limitations or criticisms.It has also been criticized by some researchers for its "poor statistical inference, the subjectivity of its interpretation of errors, and its lack of predictive power" (Bell cited in Abdel Qader, 2000:15).
Having said that, the way was paved for IL theory to take place in the field of SLA.Hence, it is good for the scope of this review to discuss the IL theory as another approach exploring its role in describing and explaining learners' errors in the process of acquisition of L2.Therefore, theoretical foundations, theoretical assumptions, limitations and significance of this theory are discussed in this review.

Theoretical foundations
The role of L1 in the acquisition of L2 has witnessed an intense debate during the past 50 years, resulting in the prevalence of EA over CA.A great number of empirical studies indicated that neither L1 nor L2 were always responsible for learners' errors (Bailey et al. 1974;Krashen et al. 1978;Larsen-Freeman, 1991, 2002).Thus, CA and EA paved the way for IL theory to take place in describing L2 learners' errors from its own perspective.
As learners' errors are not always considered undesirable, they can be devices foreign language (FL hereinafter) learners use to test their hypotheses.The occurrence of Corder's concept of idiosyncratic dialect (1967), Cooper's hypothesis testing theory (1976), Nemser's approximate language (1971;1974) and Selinker's Interlanguage (1972) propose the existence of "a separate linguistic system based on the observable output which results from a learner's attempted production of a target language norm" (Selinker, 1972:35).In this respect, Othman (2004:81) asserts that "these items refer to the same phenomenon but they emphasise different aspects of it".Tarone (1988;1990), declares that IL should not only be viewed through the lens of the FL/TL system, but they should also be treated as language systems with their own internal consistency.More so, L2 learning involves a gradual development or progression from the learners' L1/ mother tongue (MT hereinafter)/native language (NT hereinafter) towards the FL/ target language (TL hereinafter).During the language learning process, the learner naturally builds up a new language which does not refer either to his/her L1 or L2.It is a separate language having its own linguistic system.This new separate linguistic system is known as IL.
The term 'Interlanguage' was first introduced by Selinker (1972 who referred it to as L2 systematic knowledge independent of both L1 and L2.According to Richards et al. (1996), IL is the type of language which can be produced by FL/L2 learners who are in the process of acquiring or learning a new language.Within the cognitive perspective, IL 124 refers to the separateness of an L2 learner's system, a system which includes a structural status between the NL and TL (Brown, 1994).It reflects L2 learners' attempts at building up a linguistic system which progressively and gradually approaches the TL system (Fauziati, 2011).
As opposed to CA and EA, IL is not viewed as a process in SLA influenced by L1 or L2.In other words, it is neither the system of L1, nor the system of L2 or TL, but rather, as an independent linguistic system that exists independently.IL is found where L2 learners express the knowledge that they already have in the new language that they are trying to learn.Selinker (1972) suggests that IL, as the transitional processes between L1 and L2, is observable in a learner's language and can be explored.He considers IL as "a dialect whose rules share characteristics of two social dialects of languages, whether these languages themselves share rules or not" (Selinker, 1972cited in Corder, 1981:17).According to him, the notion of IL is illustrated in Figure 1.

Language A
Target Language Figure 1.The notion of the IL.(Adopted from Corder, 1981:17).In this diagram, Language A represents the learner's L1.Corder (1981) states that the learner's language could be considered as a dialect in the linguistic sense.He means that two languages which share some rules of grammar become dialects.Based on this claim, he states that language A and language B as illustrated in Figure 2 are in a dialect relation which leads to IL.

Theoretical assumptions
The appearance of the IL theory substantiated the shift in 'psychological perspectives' of L2 learning from a 'behaviourist approach' to a 'mentalist one'.Therefore, some of the IL's major assumptions were borrowed from the mentalist theories (Tarone, 2001).
During the process of L2 learning, the learner prepares the hypotheses about the rules of his/her TL.These rules can be viewed as mental grammars which create the IL system.Such grammars are exposed to some influences that might be external to the learner and/or internal derived from the learner's internal processing.In this case, it could be inferred that the learner's performance is variable.So the learner changes his/her grammar from time to time by deleting rules, adding rules and reconstructing the complete system.This indicates the role of IL in every stage of L2 learning.
Checking and rechecking hypotheses take place through the gradual process of L2 learning.The learner keeps changing his/her own IL until the TL system is fully shaped.This process is called 'Interlanguage Continuum'. Figure 3 illustrates this process.
Set of rules of language B Set of rules of language A The main theoretical assumption of IL theory is that when a FL/L2 learner is trying to communicate in the TL, he/she uses a new linguistic system different from the NL and the TL.This assumption has been supported by Selinker (1974:35) who states that IL can be "a separate linguistic system based on the observable output which results from a learners' attempted production of a TL norm".He claimed that whenever a L2/FL learner attempts to create a sentence, He/she activates 'the latent psychological structures' (LPS hereinafter), which he defines as an 'already formulating arrangement in the brain'.Additional assumptions were given by Mitchell andMyles, 1998 andLarsen-Freeman, 2003.They are summarised as follows: (1) SLA is a "process of increasing conformity to a uniform TL" (590-591).
(2) L2 learning is a gradual development from MT towards the TL.
(3) A system of certain rules which is neither the system of NL or MT nor the system of FL or TL can be developed by the learner at every stage of the language learning process.This system is a separate linguistic one.
(4) The language learning process includes hypothesis-testing or rule formation.
(6) Many language learners do not achieve the TL competence.
As was mentioned earlier, Selinker (1972) assumes the existence of IL as a result of dissimilar utterances, as observable data, of the same sentence created by L2 learners and NL learners of the TL.Accordingly, Selinker (1972:214) made a group of three sets of utterances, which can be psychologically related data of L2 learning, and theoretical predication in an important psychology of L2 learning in order to be the surface structure of an IL produced sentence.These sets are as follows: (1) utterances in the learner's MT produced by the learners.
(2) IL utterances produced by the learners.
(3) TL/FL utterances produced by NL speakers of that TL.
By determining these three sets of utterances, an investigation into the psychology of L2 learning can reveal the psycholinguistic process which might establish the knowledge that triggers IL behaviour.Within the LPS, there are some important notions: fossilisation and psycholinguistic processes.

Fossilisation
Fossilisation has received great interest among SLA researchers and has stimulated important differences of opinion.
Fossilisation is an important component of the IL process that appears at a particular point in the development of IL.
According to Selinker (1974), fossilisation is considered as one of the most important mechanisms of the LPS.In SLA research, the concept of fossilisation is basically related to the theory of IL that is considered by Selinker (1972) to be an essential phenomenon of all SLA.Selinker's concept of fossilisation is not too different from Tarone (1976), Nemser (1971), andSridhars' (1980) concepts; all of them had some attempts to explore the sources of fossilisation in the L2 learners' IL.His explanation was that "fossilisable linguistic phenomena are linguistic items, rules, and subsystems which speakers of a particular NL will tend to keep in their IL relative to a particular TL, no matter what the age of the learner or amount of explanation or instruction he/she receives in the TL" (Selinker, 1972:215).He supposes that a lot of L2 learners do not achieve the competence of the TL, because they stop somewhere in the middle of their language learning process affected by errors.He states that fossilisation occurs at different stages of the language learning process.In the process of IL continuum, only 5% of L2 learners reach the end of their IL continuum successfully whilst the majority of them do not do so (Birdsong, 2004).In other words, they cannot overcome fossilisation.When a learner of L2 stops progressing any further, his/her IL will be fossilised.In this case, only the successful learners will not fossilise as they constantly move along the IL continuum.Towell and Hawkins ( cited in Han 2005:13) claim that "even after many years of exposure to an L2, in a situation where the speaker might use that L2 every day for normal language, it is not uncommon to find that the speaker still has a strong foreign accent, uses non-native grammatical constructions, and has non-native intuitions about the interpretation of certain types of sentences".Selinker and Lakshamanan (1992) declare that the first or the preliminary sign of fossilisation is stabilisation.However, Fauziati (2011:25) says that the main difference between the two terms can be defined in terms of 'permanence'.He explains that L2 fossilised errors are permanent and continue to occur in the learner's performance despite any further exposure to the TL.In contrast, "stabilised errors are permanent; they are maintained in the learners' L2 production at a given level of IL development.It is just a momentary halt".
So on one hand, stabilised errors can be defined as those errors which eventually subside whilst the language learners make progress.On the other hand, fossilised errors refer to those which occur in spite of the given input and exposure provided for the learner.

Psycholinguistic Processes according to Interlanguage
There are several factors that lead to fossilisation.As suggested by Selinker (1972:56), they are: "language transfer, transfer-of-training, strategies of L2 learning, strategies of L2 communication, and overgeneralisation of TL linguistic material".These five essential processes of IL are involved in the latent psychological structures.
First, if it is experimentally confirmed that fossilisable rules, items and subsystems which appear in IL performance result from the L1, then we are dealing with the process of 'language transfer'.FL learners use their own MT as a resource.It is obvious that all learners fall back on their MT, mostly in the early stages of SLA.Second, when these fossilisable rules, subsystems and items come as a result of particular items in training procedures, then a process known as 'transfer-of-training' appears.
Third, if they appear as a result of a certain or identifiable approach by the learners to the input that is being presented, then a process known as 'strategies of L2 learning' takes place.L2 learning strategies are conscious based on problemsolving, directed by purpose, and seek to increase the efficiency in study.The learners of L2 might realise their errors and make progress when a proper learning strategy is adopted or followed.Only successful learners can make full use of learning strategies to accomplish their goals, whilst others cannot.
Fourth, if these fossilisable rules, items and subsystems result from identifying a certain approach by the learners for the sake of communication with native speakers of the TL, then we are dealing with 'strategies of L2 communication'.These communication strategies are associated with the skills that L2 learners use to overcome difficulties that they may encounter when they are unable to express themselves due to limited or partial language resources.Such strategies are considered very important aspects of communication skills.
Finally, if they result from a clear overgeneralisation of the rules of TL and semantic features, then we are dealing with the 'overgeneralisation of TL linguistic material' (Richards, 1974).These psycholinguistic processes are illustrated in Figure 4 below.Having explained these psycholinguistic factors that might lead to fossilisation, it is important to shed some light on the issue of variability in IL.

Variability in Interlanguage
Othman (2003:81) claims that "for the past three decades, there has been a growing number of empirical and theoretical works dealing with variation in interlanguage".She goes on saying that due to the interest of SLA research in IL, attention has been recently directed to the concepts of variability and systematicity in SLA theories.Ellis (2004) declares that variability in IL is exhibited by the nature of the task in which the learners of L2 are involved.He argues that the performance of learners is varied.They do not perform in the same way.Their performance is affected by the situation in which the learners are involved.Consequently, they produce different styles in accordance with different situations.Variability in IL was defined by Foulkes and Docherty (2005:1) as follows: Variability is one of the defining characteristics of human speech.No two voices are identical, no two utterances the same.Variability in speech is not, however, wholly random or chaotic.Rather, it results from a number of specific sources and may form rule-governed patterns.
As asserted by Ellis (2004), variability in IL can be noticed in the performance of NL and L2 learners.It can be classified into two different types, namely; systematic and non-systematic.Figure 5 below explains these types of variability.
Figure 5. Types of Variability in IL (Adapted From Ellis, 2004:76) As shown in Figure 5 above, variability of IL is characterised as being systematic and non-systematic.The systematic type can be either contextual or individuals and contextual variability might occur in a linguistic or situational context.In simple words, when the learners' performance is different from a linguistic or a situational context, this difference occurs due to contextual variability.However, the non-systematic type is divided into free variability and performance variability.In accordance with the linguistic context, Othman (2003) reveals that such a linguistic context refers to the linguistic environments which include various linguistic elements such as: pronunciation, as well as grammatical, syntactic and semantic categories.Situational context "includes factors such as the field and mode of discourse, interlocutor(s) with whom a speaker is interacting and the task in which a speaker is engaged" (Othman, 2003: 92).Hence, variability of IL across situational context(s) appears when the same language learner makes two different structures of the same form of the TL under the influence of such factors.

Investigating L2 errors
As IL theory believes that L2 learning is a gradual development from MT towards the TL, then, L2 learners' errors are natural.In other words, IL looks at the errors as devices FL learners use to test their hypotheses.They are not always considered undesirable.IL theory believes that the errors that the learners make in the rules of the TL are often correct by the rules of an IL created by the learners as "provisional and sufficiently workable substitute" (Frith, 1977).IL hypotheses deal with errors as evidence of L2 language learning strategies used by FL learners, rather than as signs of transfer or interference which should be eradicated.In accordance with the previous point, even CA and EA look at the errors from the same perspective but they indicate the importance of studying such errors in order to explore their causes and eliminate them to the minimum (Al-khresheh, 2010;2011).IL, however, assumes that making errors helps the learners in testing their hypotheses about the system of TL.Selinker (1992) argues that the evidence of IL is found in what he names as 'fossilisation', which has been discussed widely in the previous subsections.

Characteristics of Interlanguage
Tarone et al. (2001), mention that there are four characteristics or observable facts of IL theory.These characteristics are discussed and examined below.

Stability
Stability is to show consistency for using a certain rule or form over time in the field of IL learning.In other words, stability can be seen in using the same form twice by L2 learners.Henderson (1985) surprisingly argues that it is not clear for us that a new language hypothesis is needed for more explanation about the human propensity to keep making the same errors or mistakes, and to learn things gradually.This feature of IL becomes less interesting when we find that Tarone et al. (1976) decided to differentiate between two types of IL users.Type one is associated with those whose IL is distinguished by stability, whilst the other type is characterised by instability.The main problem comes from deciding which type a student is.Deciding is based on the stability of a learner's IL.IL is characterised as being systematic and not a random collection of rules or items.IL follows a particular system of rules which makes it systematic.Although the rules are not essentially the same as the rules of the TL, IL however, has a specific set of rules.Despite the variability of IL, it is probable to detect the rule-based nature of a learner's use of an L2/FL (Kasper, 2001;Kasper & Dahl, 1991).Tarone et al. (1976:97) state that L2 speech can be called systematic "when it evidences an internal consistency in the use of forms at a single point in time".
As a point of criticism, it is not very clear as to how internally consistent FL/L2 speech should be before considering it systematic.More specifically, when a student learns a FL, he/she uses a smaller range of styles than the styles being used by the native speakers of the language.This is due to the native speakers' ranges of styles which are much more, compared to the L2 learner who has just acquired a few styles and command at his/her reach.Such a person is said to have more styles in his/her NL, which is the L1 they acquired and are familiar with (House, 2000(House, , 2005(House, & 2009;;Cheng, 2005).

Mutual Intelligibility
Adjemian (1976:300) claims that ILs can, by and large, be used for the sake of communication among their speakers.They can share different functions of communication with natural languages.Linguistically, mutual intelligibility is regarded as a relationship between dialects or languages in which speakers of different languages can to some extent understand each other without extraordinary effort.Intelligibility among languages could be asymmetric (Bent and Bradlow, 2003;Bent et. al, 2008).
According to Henderson (1985), mutual intelligibility is the inherent property of the ILs which makes them to become members of the human language.The need to establish whether the learners of FL can communicate verbally with other languages other than their NL is put into consideration here.If this is found to be so, then the students are found to be able to share an IL and can be considered to be efficient.If the students are found not able to communicate with other languages, then they will be considered to possess the non-native grammar that causes them not to have the ability to make the emergence of an IL native (Bent &Bradlow, 2003).
The need to establish whether a learner in one FL class is able to communicate with and understand another student from another class being taught FL is crucial.If this is found to be so, then the students will have reached the IL.And if not, then they will be assumed not to having it (Darling-Hammond & Youngs, 2002).Accordingly, if students can understand each other, then they should have an IL.In case that they cannot, it is obvious that they will not have one.If they had one, they might be able to understand each other.

Backsliding
Backsliding is another feature of IL.All FL teachers are familiar with the concept of backsliding.It means the linguistic mastery of certain form in the TL, followed by loss, nonuse or misuse of the form (Butler-Tanaka, 2000).Selinker (1974) argues that backsliding is neither haphazard nor towards the speaker's NL but is toward an IL norm.He states that backsliding may happen when a L2 learner focuses on meaning and makes or produces a formerly learnt IL form.Fossilised forms or structures continue in spite of error correction, explicit grammatical instruction or explanation; and even if they are eradicated, they might occur again in spontaneous production.Such a phenomenon is called backsliding (Ellis, 1994).
In fossilisation, no alternative rule of the TL can be available to the learner, whereas in backsliding, there is always an alternative rule, but because of some contextual and emotional factors, the learners fail to use the right alternative rule.Therefore, Adjemian (1976:317) asserts that: The speaker should have intuitions about the correct rule or form, whereas in the case of fossilisation he may not ... This seems to me to imply that backsliding is evidence of a function in IL which has almost lost its permeability.

Conclusion
The importance of this approach in the field of SLA lies in the fact that it can be the first attempt which takes into consideration the possibility of L2 learners' conscious attempts to have control of their learning (Sercombe, 2000).Therefore, IL theory is important for some reasons.First, it looks at the L2 learner as an active participant because he/she has the ability to form rules from the data he/she might encounter.This leads to saying that the study of IL is universal and systematic by nature, which makes it similar to the innateness theory.Second, the study of IL theory might help us in determining what a FL leaner knows at a particular point in time and what he/she should be taught.Third, IL theory helps in understanding the L2 learners' problems better and in providing timely help to such learners; hence, they can achieve competence in their TL.Fourth, IL theory has also brought about significant changes in the teaching methodology, with communicative teaching making its way into the teaching system.IL also brought in the acknowledgement of the fact that errors are a part and parcel of the learning process, thus reducing the need for continuous supervision by the teachers (Richards, 1996;Rustipa, 2011;Ellis, 2008).
Despite the importance of this approach in certain aspects, some of its assumptions were criticised for their weaknesses.Thus, IL theory has some problems which are summarised below.
First, a major criticism of this approach might relate to its limited explanatory power.Compared to EA, IL assumes that the linguistic stage that learners are at could be predicted by analysing their errors.Henderson (1985:26) states that, "There is nothing that we can ever hope to observe which would disconfirm the IL hypothesis.It therefore tells us nothing about SLA.It makes no prediction which could ever turn out not to be the case".He adds that, "we are far more likely to be able to find a way to account for L2 competence (or incompetence) by looking for systems, and that in getting us away from our preoccupation with errors, IL has done the field a service.If professionals in language teaching now want to use the term to mean communicating in a foreign language, that's fine.But let's stop calling it a hypothesis and waiting for it to explain something" (ibid: 26).
Second, it is sometimes difficult to identify which of the five essential processes of IL is the noticeable data to be attributable to (Yuksel, 2007& Jordan, 2004).In accordance with this point, Richards (1974:42) surprisingly raises a question, that is, "can we always unambiguously identify which of these processes our observable data is to be attributable to?" and he answers "most probably not".He explains his claim by saying that we often do not know if we are in fact studying 'storage' or 'retrieval'.Accordingly, we may not be able to decide as to whether a certain constituent IL concatenation is due to language transfer or transfer-of-training, or both.
Third, predicting which linguistic items in which interlingual situations can be fossilised is a difficult task (Richards, 1974).Fourth, IL theory does not have the ability to determine how the exact position of a FL learner in between MT and TL can be interpreted.Henderson (1985:26) argues whether IL hypothesis is able to predict observable events and if it can 'be disconfirmed by looking at observable data'.He concludes that to disconfirm it, One would have to show that there is in fact no system separate from the NL and the TL which develops in the mind of the learner.One would have to demonstrate either: (a) that the learner has simply grafted some TL forms onto the internalised NL system, or (b) that he/she has in fact internalised the TL system in its entirety but also has a set of rules which mess it up when he/she tries to use it.Neither of these positions is provable, nor could any data be found to support them.
Fifth, regarding the IL continuum, the majority of L2 learners do not reach the end of this continuum.They might stop at a certain stage where they think that they do not need any additional or further development in order to communicate efficiently.Another reason is related to changes in their brains which restrict the learning mechanisms, and therefore stop any further learning.This fossilisation is not liable to remedy by instruction (Sulaiman, 2006).The premise which claims that IL is fossilised is still debatable (Fauziati, 2011).Kharma and Hajjaj (1989) state that IL is permeable as its rules are not fixed.It means that they are open to modification.Moreover, IL is a gradual process; consequently, any new rules or changes the structure goes through are slowly established which provide the structure with energy.Also, the students' utilisation of L2, despite being flawed and shifting, is systematic in some way (i.e., supported on the present 'rule system' at a definite phase).Yet, the volatility of IL brings about a mixture of troubles of explanation, given that it is hard to gather sufficient facts from one condition of a student's IL with the aim of rebuilding its structure.
Lastly, much correction by EFL/ESL teachers might lead to lack of motivation.Thus, many well-formed utterances will be unnoticed and neglected.In this case, L2 learners need to be restricted to the most important errors only.
This study has attempted to review the IL theory by providing some definitions, theoretical foundations and assumptions.It is quite obvious from the given discussion above that learners acquire a L2 through several processes, and in each process many errors can be made.The way in which these errors are made is attributed to a variety of causes.As discussed above, the processes which learners can go through are still a point of debate.Consequently, IL has witnessed huge criticism by different researchers and linguists from various L1 backgrounds.They all came to state that there are many points in this theory which are not clear.